id
stringlengths 20
20
| year
int64 2.01k
2.02k
| title
stringlengths 21
249
| sections
stringlengths 4.88k
32.5k
| headings
stringclasses 217
values | abstract
stringlengths 32
4.12k
| summary
stringlengths 227
2.45k
| keywords
stringlengths 7
1.31k
⌀ | toc
stringclasses 267
values |
|---|---|---|---|---|---|---|---|---|
journal.pcbi.1006227 | 2,018 | General differential Hebbian learning: Capturing temporal relations between events in neural networks and the brain | Most learning rules used in bio-inspired or bio-constrained neural-network models of brain derive from Hebb’s idea 1 , 2 for which “cells that fire together , wire together” 3 ., The core of the mathematical implementations of this idea is multiplication ., This captures the correlation between the pre- and post-synaptic neuron activation independently of the timing of their firing ., Time is however very important for brain processing and its learning processes 4 ., Differential Hebbian learning ( DHL ) rules 5 , 6 are learning rules that change the synapse in different ways depending on the specific timing of the events involving the pre- and post-synaptic neurons ., For example , the synapse might tend to increase if the pre-synaptic neuron activates before the post-synaptic neuron , and decrease if it activates after it ., As suggested by their name , DHL rules use derivatives to detect the temporal relations between neural events ., Here we will use the term event to refer to a relatively short portion of a signal that first monotonically increases and then monotonically decreases ., Events might for example involve the activation of a firing-rate unit in an artificial neural network , or the membrane potential of a real neuron , or a neurotransmitter concentration change ., DHL rules use the positive part of the first derivative of signals to detect the initial part of events , and its negative part to detect their final part ., By suitably multiplying the positive/negative parts of the derivative of events related to different signals , DHL rules can modify the synapse in different ways depending on how their initial/final parts overlap in time ., To the best of our knowledge , current DHL rules are basically two: one proposed by Kosko 5 and one proposed by Porr , Wörgötter and colleagues 6 , 7 ., These rules modify the synapse in specific ways based on the temporal relation between the pre- and post-synaptic events ., Formulating other ways to modify synapses based on event timing is the first open problem that we face here ., The development of dynamical neural-network models and learning mechanisms that , as DHL , are able to take time into consideration is very important ., Indeed , the brain is an exquisitely dynamical machine processing the continuous flow of information from sensors and issuing a continuous flow of commands to actuators so its understanding needs such types of models 8–11 ., In this respect , neuroscientific research on spike timing dependent plasticity ( STDP; 12 ) clearly shows how synaptic changes strongly depend on the temporal relation between the spikes of the pre- and post-synaptic neurons ., Given the typical shape of spikes , an important class of STDP models , called phenomenological models 13 , abstracts over the features of the spike signals and directly links the synaptic strengthening , Δw , to the time interval separating the pre-synaptic and post-synaptic spikes , Δt , on the basis of a function of the type Δw = f ( Δt ) 12 , 14 ., Such a function is usually designed by hand and reflects the synaptic changes observed in experimental data ., 15 ., The function f ( Δt ) generates a typical learning kernel that when plotted shows a curve where each Δt causes a certain Δw ., Phenomenological models are simple but are applicable only to spike events ., In comparison , DHL rules are more complex but have the advantage of computing the synaptic update as the step-by-step interaction ( based on multiplication ) between the pre-synaptic and post-synaptic events ., Therefore they are applicable to any complex signal that might exhibit events with variable time courses ., When applied to the study of STDP , the property of DHL rules just mentioned also opens up the interesting possibility of using them to investigate the actual biophysical neural events following and caused by the spikes that actually lead to the synaptic change , as first done in 16 ., The chain of processes changing the synapse is also captured by biophysical models ( e . g . , see 14 , 17 ) ., These models can capture those processes in much biological detail ( mimicking specific neurons , neuromodulators , receptors , etc . ) but at the cost of being tied to specific phenomena ., Because the level of abstraction of DHL rules lies between that of phenomenological models and that of biophysical models , DHL represents an important additional research tool ., Experimental study of STDP 18 , 19 shows that different types of neurons , for example excitatory/inhibitory neurons in different parts of the brain , implement a surprisingly rich repertoire of learning kernels ., It is reasonable to assume that the brain employs such learning mechanisms to implement different computational functions ., In this respect , an interesting fourth class of models appropriate for studying STDP , which might be called functional models , aims to derive , or to justify , specific STDP learning kernels based on normative computational principles 20–23 ., Investigating the functions of different STDP kernels is not in the scope of this work ., However , assuming that the variety of learning kernels discovered through STDP experiments supports different functions relevant to neural processing and that analogous functions might be needed in artificial neural networks , it is important to understand the computational mechanisms that might generate such a variety of learning kernels ., In this respect , an important question is this: is there a DHL learning rule , or a set of them , that can generate the complete variety of learning kernels found in the brain ?, Some existing research shows how different STDP learning kernels can arise from the same biophysical mechanisms 17 , or from the same DHL-based model 24 ., However , these studies propose specific mechanisms to address a sub-set of STDP data sets rather than proposing a general way to systematically reproduce STDP learning kernels ., Understanding the extent to which DHL can capture the known STDP phenomena , and how this can be done , is thus a second important open problem that we address here ., The rest of the paper addresses the two open problems indicated above in the following ways ., As a first contribution of the paper , the Section ‘G-DHL and the systematisation of DHL’ considers the first open problem—how different DHL rules can be generated in a systematic fashion—by proposing a general framework to produce DHL rules ., In particular , the section first reviews the DHL rules proposed so far in the literature; then it presents the G-DHL rule and shows how it is able to generate the existing DHL rules and many others; and finally it shows how one can filter the neural signals to generate events that correspond to the features of interest and can use memory traces to apply the G-DHL rule to events separated by time gaps ., As a second contribution of the paper , the Section ‘Using G-DHL to fit STDP data sets’ deals with the second open problem—understanding if and how G-DHL can be used to capture known STDP phenomena ., To this end , the section first illustrates how the G-DHL synapse update caused by a pre- and post-synaptic spike pair can be computing analytically rather than numerically , and then it presents a collection of computational tools to automatically search the rule components and parameters to fit a given STDP data set ., Addressing the same second open problem , and as a third contribution of the paper , the Section ‘Using G-DHL to fit STDP data sets’ uses those computational tools to show how the G-DHL rule is able to reproduce several learning kernels from the STDP literature ., To this end , the section first uses G-DHL to fit the classic STDP data set of Bi and Poo 25; then it illustrates how the G-DHL components found by the fitting procedure can be heuristically useful to search for the biophysical mechanisms underlying a given STDP data set; and finally it shows how to apply the G-DHL rule to systematically capture different aspects of all the STDP data sets reviewed by Caporale and Dan 18 ( such as their temporal span , long-term potentiation/depression , and variability around zero inter-spike intervals—e . g . sharp depression-potentiation passages , non-learning plateaus , Hebbian/anti-Hebbian learning ) ., The Section ‘Discussion’ closes the paper by analysing the main features of G-DHL and its possible development ., All software used for this research is available for download from internet ( https://github . com/GOAL-Robots/CNR_140618_GDHL ) ., As discussed in the introduction , different types of neurons exhibit surprisingly different STDP learning kernels ., For this reason we tested the flexibility of G-DHL by using it to capture several different STDP learning kernels involving pairs of pre- and post-synaptic spikes ., In the future G-DHL could be extended to capture STDP processes involving spike triplets or quadruplets ( 41; see 42 for a model ) by considering three or more multiplication elements rather than only two as done here ., To apply G-DHL to spike pairs , we first outline the procedure used to derive the formulas to compute G-DHL analytically , rather than numerically as done so far ., The procedure is illustrated in detail in Section 2 . 1 in S1 Supporting Information in the case in which one assumes that spikes and traces are described with some commonly used formulas ., Sections 2 . 7 and 2 . 8 in S1 Supporting Information show a method that leverages these formulas to use G-DHL to fit STDP data sets; examples of this fitting are shown in the Section ‘Results’ ., Before presenting the formulas , we discuss two important points ., The closed-form formulas for synaptic updates by the G-DHL rule have two main advantages ., First , they allow the mathematical study of the G-DHL rule ( see Sections 2 . 2 and 2 . 6 in S1 Supporting Information ) ., Second , the formulas allow a computationally fast application of G-DHL by computing the synaptic update through a single formula rather than as a sum of many step-by-step synaptic updates as done in its numerical application , an advantage exploited in the computationally intensive simulations of the Section ‘Results’ ., A second observation concerns the relation between the G-DHL explicit formulas and phenomenological models discussed in the introduction ., The G-DHL explicit formulas have the form Δw = f ( Δt ) typical of phenomenological models ., This shortcut is possible because spikes have a fixed shape: this implies that Δt is the only information relevant for computing G-DHL ., The resulting synaptic update is however the same as the one that would be obtained by numerically simulating the step-by-step interaction between the pre- and post-synaptic neural events mimicking more closely what happens in the real brain ., Therefore , the possibility of computing Δw = f ( Δt ) formulas for DHL rules does not violate what we said in the introduction , namely that G-DHL captures the mechanisms causing the synaptic update at a deeper level with respect to phenomenological models ., The procedure for the automatic fit of STDP data sets was first employed to fit the classic STDP data set of Bi and Poo from rat hippocampal neurons 25 ., Fig 8a summarises the results ( for ease of reference , henceforth we will refer to synapse strengthening/weakening as ‘LTP—long term potentiation’ and ‘LTD—long term depression’ ) ., The model comparison technique selected two G-DHL components: an LTP component ( σpp = 0 . 73 ) and an LTD component ( ηps = −0 . 025 ) ., The parameters σ and η differ in scale as they refer to differential and mixed G-DHL components involving signal-derivative or derivative-derivative multiplications ., Fig 8b shows the target data and their fit obtained with the G-DHL components and parameters shown in Fig 8a ., The G-DHL regression fits the data accurately ( FVU = 0 . 2725 ) ., While the original paper performed the fit with the usual exponential function for both positive and negative Δt , the G-DHL regression captures the LTP with the σpp ‘sharp’ component ( Fig 6 ) , concentrated on small positive inter-spike intervals , and the LTD with the ηps = −0 . 025 ‘softer’ component ( Fig 7 ) , concentrated on negative intervals ., We now illustrate with an example the idea of using the components found by the G-DHL regression to heuristically search for biophysical mechanisms possibly underlying a target STDP data set ., This example involves the Bi and Poo’s data set 25 analysed in the previous section ., The idea relies on the observation that each multiplication factor of the G-DHL components identified by the regression procedure has a temporal profile that might correspond to the temporal profile of the pre-/post-synaptic neuron electrochemical processes causing the synaptic change ., The steps of the procedure used to search the biophysical mechanisms are as follows:, ( a ) identify with an automatic procedure the G-DHL components and parameters fitting the target STDP data set;, ( b ) define the temporal profile of the two pre-/post-synaptic factors of each found component , and the LTP/LTD effects caused by the component;, ( c ) identify possible biophysical processes having a temporal profile similar to the one of the identified factors;, ( d ) design experiments to verify if the hypothesised biophysical processes actually underlie the target STDP phenomenon in the brain ., We now give an example of how to apply the steps ‘a’ and ‘b’ , and some initial indications on the step ‘c’ , in relation to the Bi and Poo’s data set 25 ., The example aims to only furnish an illustration of the procedure , not to propose an in-depth analysis of this STDP data set ., Regarding step ‘a’ , Fig 8 shows that the G-DHL regression identified two LTP and LTD components ., Regarding step ‘b’ , Fig 9 shows the temporal profile of the factors of the two components ., The first component is a ‘positive-derivative/positive-derivative’ component ( u ˙ 1 + u ˙ 2 +; Fig 9a , left graph ) with two factors ( Fig 9b , left graph ) :, ( a ) a relatively long pre-synaptic factor ( u ˙ 1 + ) lasting about 30 ms;, ( b ) a shorter post-synaptic factor ( u ˙ 2 + ) lasting about 7 ms . These two factors , amplified by a positive coefficient ( σpp = + 0 . 73 ) , produce LTP concentrated on small positive inter-spike intervals ( 0 ms < Δt < 30 ms; Fig 9a , left graph ) ., The second component is a ‘positive-derivative/signal’ component ( u ˙ 1 + u 2; Fig 9a , right graph ) with other two factors ( Fig 9b , right graph ) :, ( a ) a relatively long pre-synaptic factor ( u ˙ 1 + ) lasting about 30 ms;, ( b ) a longer post-synaptic factor ( u2 ) lasting about 50 ms . The two factors , amplified by a negative coefficient ( ηps = −0 . 025 ) , produce LTD covering negative-positive inter-spike intervals ( −30ms < Δt < 20ms; see Fig 9a , right graph ) ., When the two components are summed , LTP more than cancels out LTD for positive delays ( 0ms < Δt < 20ms ) ., This causes the sharp passage from LTD to LTP around the critical Δt values close to zero , which characterise the target kernel ( Fig 8 ) ., Regarding step ‘c’ of the procedure , directed to identify possible biological correspondents of the component factors identified in step ‘b’ , we now discuss some possible candidate mechanisms that might underlie the factors identified for the Bi and Poo’s data set ., Note that these brief indications are only intended to show the possible application of the procedure , not to make any strong claim on the possible specific mechanisms underlying such STDP data set ., Pioneering studies on hippocampus have shown that a repeated stimulation of the perforant path fibres enhances the population response of downstream dentate granulate cells ( long-term potentiation–LTP; 47–49 ) ., LTP also takes place in other parts of brain such as the cortex 50 , amygdala 51 , and the midbrain reward circuit 52 ., Other studies have shown the existence of long-term depression ( LTD ) , complementary to LTP , in various parts of brain , for example hippocampus 53 , 54 and motoneurons 55 ., More recent research has shown that LTP and LTD , and their intensity , depend on the duration of the temporal gap separating the pre- and post-synaptic spikes ( spike time-dependent plasticity—STDP; e . g . 56 , see 18 for a review ) ., The relation between the time-delay and the synaptic change depends on the types of neurons involved ( e . g . , glutamatergic vs . GABAergic neurons 57 , 58 ) , the position of the synapse ( e . g . , 59 ) , and the experimental protocols used ( e . g . , 60 ) ., Early findings that blocking NMDA receptors ( NMDARs ) can prevent both LTP and LTD , while a partial blocking can turn an LTP effect into an LTD , has led to the proposal of several calcium-based models of synaptic plasticity ( e . g . , 61–64 ) ., One view proposes that two independent mechanisms can account for the classic STDP learning kernel 19 , 65 ., This is in line with the two components , and their factors , found by our G-DHL based regression of Bi and Poo data set ., The first component was an LTP ‘positive-derivative/positive-derivative’ component ( u ˙ 1 + u ˙ 2 + ) formed by two factors ., The first factor was a pre-synaptic factor ( u ˙ 1 + ) lasting about 30 ms , compatible with a short-lived effect involving the pre-synaptic glutamatergic neuron spike and affecting the post-synaptic NMDARs 66 ., The second factor was a post-synaptic factor ( u ˙ 2 + ) lasting about 7 ms , compatible with a back-propagating action potential ( BAP; 67 ) ., The second component was a ‘positive-derivative/signal’ LTD component ( u ˙ 1 + u 2 ) formed by two factors: a relatively slow pre-synaptic element , ( u ˙ 1 + ) , lasting about 30 ms , and a slow post-synaptic element , ( u2 ) , lasting about 50 ms . Different biological mechanisms might underlie these two factors ., In this respect , there is evidence that post-synaptic NMDARs might not be necessary for spike-timing-dependent LTD 68 , while this might be caused by metabotropic glutamate receptors ( mGluR; 69 ) , voltage gated calcium channels ( VGCC; 25 , 69 ) , pre-synaptic NMDAR 70 , or cannabinoid receptors 68 , 69 ., We tested the generality of G-DHL by fitting all STDP kernels reported in the review of Caporale and Dan 18 ., The data sets addressed in this review encompass many different STDP experiments reported in the literature and proposes a taxonomy to group them into distinct , and possibly exhaustive , classes ., The taxonomy is first based on the excitatory or inhibitory nature of the pre- and post-synaptic neurons , giving the classes:, ( a ) excitatory-excitatory;, ( b ) excitatory-inhibitory;, ( c ) inhibitory-excitatory;, ( d ) inhibitory-inhibitory ., Some neurons in different parts of brain belong to the same class but exhibit different STDP learning kernels: in 18 , these have been grouped in ‘subtypes’ ( sub-classes ) called ‘Type I’ , ‘Type II’ , etc ., For the G-DHL regressions we used the original data when the authors of the experiments could furnish them ., When this was not possible , we used the data extracted from graphs in the publications ., Figs 10 and 11 summarise the outcome of the G-DHL-based regressions for the different data sets ., For each data set , the figures report this information:, ( a ) left graph: original data and , when available , regression curve of the original paper;, ( b ) right graph: regression curve based on G-DHL;, ( c ) top-center small graph: function with which the review 18 proposed to represent the STDP class of the data set ., In the following , we illustrate the salient features of these regressions ., Section 3 in S1 Supporting Information presents more detailed data on all the regressions as those presented in Fig 8 for the data set of Bi and Poo ., Understanding the functioning and learning in dynamical neural networks is challenging but also very important for advancing our theories and models of the brain—an exquisitely dynamical machine ., Differential Hebbian Learning ( DHL ) might become a fundamental means to do so ., Existing DHL rules are few , basically two 5 , 7 , and are not able to model most spike-timing dependent plasticity ( STDP ) phenomena found so far in the brain ., Building on previous pioneering research , this work addresses these limitations in multiple ways ., First , it proposes a framework to understand , use , and further develop DHL rules ., In particular , it proposes a general DHL ( G-DHL ) rule encompassing existing DHL rules and generating many others , and highlights key issues related to the pre-processing of neural signals before the application of DHL rules ., Second , it proposes procedures and formulas for applying DHL to model STDP in the brain ., Third , it shows how the proposed G-DHL rule can model many classes of STDP observed in the brain and reviewed in 18 ., With respect to other approaches for modelling STDP , DHL represents a complementary tool in the toolbox of the modeller and neuroscientist ., First , DHL differs from ‘phenomenological models’ ., Although simple and elegant , these models update the synapse based on mathematical functions directly mimicking the synaptic changes observed in empirical experiments in correspondence to different inter-spike intervals 14 , 15 ., Instead , DHL rules compute the synaptic update on the basis of the step-by-step interactions between levels of and changes in the neural variables of interest ., DHL rules also differ from ‘biophysical models’ ., These models can reproduce many biological details but have high complexity and rely on phenomenon-specific mechanisms ( e . g . , 14 , 17 ) ., Instead , DHL rules reproduce fewer empirical details but at the same time , after the systematisation proposed here , they represent ‘universal mechanisms’ able to capture many STDP phenomena ., G-DHL relies on two main ideas ., The first idea , elaborated starting from previous proposals 5 ( see also 29 ) , is that the derivative of an ‘event’ , intended as a monotonic increase followed by a monototic decrease of a signal , gives information on when the event starts and terminates ., This information is used by G-DHL to update the connection weight depending on the time interval separating the pre- and post-synaptic neural events ., The second idea is that the actual synaptic update can rely on different combinations of the possible interactions between the pre-/post-synaptic events and their derivatives , thus leading to a whole family of DHL rules ., Mathematically , this gives rise to a compound structure of the G-DHL rule which is formed by a linear combination of multiple components ., In this respect , the capacity of G-DHL to capture different STDP phenomena is linked to the power of kernel methods used in machine learning 34 , 35 ., The linear form of the rule facilitates its application through manual tuning of its parameters , as shown here and in some previous neural-network models of animal behaviour using some components of the rule 80–82 ., The linear form of the rule also facilitates the automatic estimation of its coefficients when used to capture STDP data sets , as also shown here ., G-DHL has a high expressiveness , as shown here by the fact that we could use it to accurately fit multiple STDP data sets ., In particular , the G-DHL components form basis functions that are well suited to model key aspects of STDP , in particular its long-term potentiation/depression features , its time span , and its variability around the zero inter-spike interval ( e . g . , sharp depression-potentiation passages , non-learning plateau , Hebbian/anti-Hebbian learning ) ., The regressions of the data sets targeted here employed seven out of eight components of the rule ., The regressions are particularly reliable because the optimisation procedure used here is highly robust with respect to local minima , so they show the utility of most G-DHL components for modelling different STDP data sets ., Future empirical experiments might search for STDP processes corresponding to the eighth non-used G-DHL component ( encompassing a multiplication between the pre-synaptic stimulus and the post-synaptic derivative negative part ) : this corresponds to a relatively long LTD peaking at a negative inter-spike interval but also involving low-value positive intervals ., The results of our regression based on G-DHL of the classic STDP kernel , represented by the classic Bi and Poo data set 25 , suggests the possible existence of two distinct mechanisms underlying LTP and LTD involved in such STDP learning kernel , so it is interesting to compare this result with different views in the literature ., A specific hypothesis on calcium control of plasticity was formulated in 83 and was followed by significant experimental evidence ., According to this hypothesis , post-synaptic calcium transients above a lower threshold cause LTD whereas calcium transients above a second higher threshold produce LTP ., In a detail model 84 , this phenomenon is captured with a single mechanism for which the synaptic change is caused by calcium concentrations at the post-synaptic neuron modulated by the temporal relation between the current at the pre-synaptic neuron ( causing NMDAR opening ) and the back-propagating action potential ( BAP ) at the post-synaptic neuron 67: low levels of post-synaptic calcium cause the synapse depression whereas high levels cause its enhancement ., Models of such type have been criticised on the basis of empirical evidence ., According to 65 , calcium models require a long-fading BAP-induced transients to account for LTD when the BAP occurs before the pre-synaptic action potential 12 ., Moreover , calcium models also predict a pre-post form of LTD even when the BAP occurs beyond a given time from the pre-synaptic action potential ., While this pre-post form of LTD has been registered in hippocampal slices 74 , other data 25 indicate that it is not a general feature of STDP ., In this respect , our findings agree with other proposals for which two independent mechanisms account for LTP and LTD in the classic STDP learning kernel 19 , 65 ., Future work might extend these preliminary results ., In particular , it could aim to understand in detail how some of the mechanisms mentioned above implement change detectors and these lead to STDP , as predicted by the G-DHL core functioning mechanisms based on derivatives ., Moreover , G-DHL could be used to heuristically guide the identification of the biophysical mechanisms underlying different STDP data sets beyond the classic kernel ., Future work might also investigate , both computationally and empirically , DHL rules different from G-DHL , namely:, ( a ) DHL rules formed by three or more components ( useful to model STDP involving more than two spikes 41 ) ;, ( b ) DHL rules using orders of derivatives higher than the first one used in G-DHL 32 , 33;, ( c ) DHL rules generated by other types of filters , rather than u ˙ + and u ˙ - used in G-DHL , to detect the increasing and decreasing parts of events ., Another line of research might aim to investigate the possible computational and behavioural functions of the different G-DHL components ., In this respect , the analysis presented here on the computational mechanisms underlying STDP might contribute to the current research on the possible functions of such plasticity 20–23 ., Indeed , this research mainly focuses on the computational function of the classic STDP learning kernel 25 , whereas the research presented here , by stressing how the brain uses different DHL rules , calls for the investigation of their different possible functions ., A different approach to understand the functions of different DHL rules and STDP kernels might use embodied neural models to understand their utility to support adaptive behaviour ., The development of G-DHL was in fact inspired by the need to implement specific learning processes in neural-network models able to autonomously acquire adaptive behaviours 80–82 ., Thus , it could for example be possible to establish a particular target computation or behaviour and then automatically search ( e . g . with genetic algorithms or other optimisation techniques ) the rule components and coefficients that are best suited for them ., For example , previous work 85 used a learning rule based on Kosco’s DHL rule 5 to obtain interesting/surprising emergent behaviours in physical simulated agents ., This approach might test other G-DHL components to produce different behaviours . | Introduction, Methods, Results, Discussion | Learning in biologically relevant neural-network models usually relies on Hebb learning rules ., The typical implementations of these rules change the synaptic strength on the basis of the co-occurrence of the neural events taking place at a certain time in the pre- and post-synaptic neurons ., Differential Hebbian learning ( DHL ) rules , instead , are able to update the synapse by taking into account the temporal relation , captured with derivatives , between the neural events happening in the recent past ., The few DHL rules proposed so far can update the synaptic weights only in few ways: this is a limitation for the study of dynamical neurons and neural-network models ., Moreover , empirical evidence on brain spike-timing-dependent plasticity ( STDP ) shows that different neurons express a surprisingly rich repertoire of different learning processes going far beyond existing DHL rules ., This opens up a second problem of how capturing such processes with DHL rules ., Here we propose a general DHL ( G-DHL ) rule generating the existing rules and many others ., The rule has a high expressiveness as it combines in different ways the pre- and post-synaptic neuron signals and derivatives ., The rule flexibility is shown by applying it to various signals of artificial neurons and by fitting several different STDP experimental data sets ., To these purposes , we propose techniques to pre-process the neural signals and capture the temporal relations between the neural events of interest ., We also propose a procedure to automatically identify the rule components and parameters that best fit different STDP data sets , and show how the identified components might be used to heuristically guide the search of the biophysical mechanisms underlying STDP ., Overall , the results show that the G-DHL rule represents a useful means to study time-sensitive learning processes in both artificial neural networks and brain . | Which learning rules can be used to capture the temporal relations between activation events involving pairs of neurons in artificial neural networks ?, Previous computational research proposed various differential Hebbian learning ( DHL ) rules that rely on the activation of neurons and time derivatives of their activations to capture specific temporal relations between neural events ., However , empirical research of brain plasticity , in particular plasticity depending on sequences of pairs of spikes involving the pre- and the post-synaptic neurons , i . e . , spike-timing-dependent plasticity ( STDP ) , shows that the brain uses a surprisingly wide variety of different learning mechanisms that cannot be captured by the DHL rules proposed so far ., Here we propose a general differential Hebbian learning ( G-DHL ) rule able to generate all existing DHL rules and many others ., We show various examples of how the rule can be used to update the synapse in many different ways based on the temporal relation between neural events in pairs of artificial neurons ., Moreover , we show the flexibility of the G-DHL rule by applying it to successfully fit several different STDP processes recorded in the brain ., Overall , the G-DHL rule represents a new tool for conducting research on learning processes that depend on the timing of signal events . | learning, medicine and health sciences, action potentials, engineering and technology, nervous system, signal processing, membrane potential, social sciences, electrophysiology, neuroscience, learning and memory, signal filtering, kernel functions, cognitive psychology, mathematics, neuronal plasticity, operator theory, animal cells, biophysics, physics, cellular neuroscience, psychology, cell biology, anatomy, synapses, physiology, neurons, biology and life sciences, cellular types, physical sciences, cognitive science, neurophysiology | null |
journal.ppat.1003555 | 2,013 | Quantitative Models of the Dose-Response and Time Course of Inhalational Anthrax in Humans | The causative microorganism of anthrax , Bacillus anthracis ( B . anthracis ) , is classified by the US Centers for Disease Control and Prevention ( CDC ) as a Category A ( highest priority ) bioterrorism pathogen , with the potential for causing a large number of infections and deaths after an effective aerosol release in a community 1 , 2 ., Reports of natural infections 3–6 and large scale accidental or intentional releases causing infections 7 , 8 provide limited insight into the risk ., To evaluate the threat posed by potential release scenarios , risk assessors , public health analysts , biodefense modelers , and other researchers require robust quantitative dose-response analyses to estimate the magnitude and timeline of potential consequences and the effect of public health intervention strategies 9–11 , such as the administration of prophylactic antibiotic regimens to potentially exposed cases 12 , and to interpret the significance of sampling results for detecting B . anthracis spores in indoor environments 13 , 14 ., For these analyses , it is particularly important to estimate the probability of infection after low dose exposures , which could cause the majority of cases after a large-scale release 15 , 16 ., Due to the deadly nature of the disease , there are no experimental data on exposure and response of humans to aerosolized B . anthracis ., Analyses of quantitative information from natural and accidental exposures and infections of humans 15 , 17 and experimental infections of non-human primates 18 , 19 are scattered in the literature , poorly understood , and often contradictory ., Mathematical dose-response modeling is useful when experimental data on the effects of low dose inhalational exposures are scarce or non-existent ., These models utilize information about the height and shape of a dose-response curve at higher doses where data or estimates are available and use an assumed functional form to extend the curve to lower doses where data are not available , but where risk estimates are required ., Different model forms can lead to very different extrapolated estimates from the same set of data ., This creates substantial uncertainty regarding the minimum dose required to cause infection in humans 20 and the dose-dependent time from exposure to appearance of illness ( incubation period ) , key parameters required for sound risk assessment by public health and emergency preparedness authorities 12 , 21 ., In this study , we critically evaluate the available published literature and identify candidate raw data sets to develop refined quantitative dose-response models for B . anthracis infection in humans with an emphasis on the low-dose effect ., We use the resulting models to estimate the incubation period as a function of the exposure and the relationships between duration of antimicrobial treatment after exposure and the probability of infection ., Three outbreaks of inhalational anthrax in humans having information to estimate dose-response are the 2001 letter attacks through the U . S . Postal Service , industrial workers handling contaminated animal products in the early-mid 1900s , and an accidental airborne release of spores from a facility in Sverdlovsk , Russia in 1979 ., The doses to which victims of the 2001 letters were exposed are not known , and it is a challenge to estimate exposure amounts without knowing the means by which spores were released from the envelopes , aerosolized , and inhaled ., Therefore , despite modeling efforts 16 , 22 , these incidents shed limited information on quantitative dose-response ., Some quantitative data exist for exposure of non-vaccinated industrial workers handling animal products contaminated with B . anthracis ., This evidence suggests that the infection rate for humans exposed in this setting is very low , especially for inhalational anthrax , as most of the infections that did occur were cutaneous 3 ., Workers in one mill were thought to have been inhaling hundreds of spores on a daily basis with not a single infection documented 23 ., A recent analysis of this case concluded that 600 spores or fewer would not be expected to cause disease in healthy humans and advocated the use of 600 spores as a threshold to use in risk assessments 17 ., However , it is possible that the industrial workers were immune to clinical infection from repeated low-level exposure , that there were undiagnosed cases , or that infections would result from low-dose exposures of individuals with unusual susceptibility 24 ., B . anthracis spores were accidentally released from a facility in Sverdlovsk ( Russia ) in the former Soviet Union in 1979 , causing infections in both humans and animals downwind of the facility 7 ., Doses inhaled by the infected individuals are not known , nor is it known how many spores were released from the facility ., However , human dose-response information has been inferred using atmospheric data on the day the release likely occurred , the likely locations of the infected individuals when they were exposed , and the epidemiology of the tabulated cases ., Meselson et al . 7 calculated that the attack rate at a ceramics factory 2 . 8 km downwind of the Sverdlovsk release was approximately 1–2% ( 18 out of about 1500 employees were infected , including 10 out of 450 employees working in a single unpartitioned building ) ., Wilkening 15 analyzed the Sverdlovsk case data and applied a series of theoretical dose-response models , finding that both the spatial ( distance from release ) and temporal ( incubation period , assumed to vary with dose ) distribution of cases are consistent with dose-response curves that predict a slow decrease in the probability of infection as the dose decreases ., They conclude that these data do not support a distinct exposure threshold below which no one is infected and above which everyone is infected ., In the absence of other human data , experimental studies involving non-human primates provide the best available data from which to gain insights into potentially appropriate dose-response relationships for humans ., We summarize three candidate data sets and dose-response models that have been applied to them ., Note that , while these studies generally use death as an endpoint and express their results in terms of lethal dose ( LD ) , we make the assumption that LD and ID are equivalent , i . e . , that infection with inhalational anthrax invariably leads to death in the absence of treatment ., Two of the following three studies do not make note of infected animals that survived ., The third study found evidence of infection in two surviving animals sacrificed at the termination of an experiment , but noted that “these animals were undoubtedly in the early stage of disease and presumably would have developed systemic disease and died , had the experiment not been terminated” 24 ., There is also evidence that humans with inhalational anthrax infection have a fatality rate approaching 100% in the absence of treatment ., Holty et al . , in reviewing 82 of the best-documented human inhalational anthrax cases 25 , found only one instance of an infected and untreated person ( an at-risk veterinarian thought to have some prior immunity ) who did not progress to the fulminant stage of disease ., They found only two cases ( 3% ) of humans surviving the fulminant stage of disease under any circumstance , and both of those cases received treatment ., Glassman 26 reports on data from unpublished work performed by Jemski in which 1 , 236 cynomolgus monkeys ( Macaca fascicularis ) were exposed to aerosols of B . anthracis ., While the raw data are not published , the paper reports that a log-probit model was fit to the data , resulting in a dose that is lethal to 50% of animals exposed ( LD50 ) of 4 , 130 spores ( 95% confidence interval 1 , 980 to 8 , 630 ) and a probit slope of 0 . 669 probits per base-ten log dose ( 95% confidence interval 0 . 520 to 0 . 818 ) ., Under our definition of the log-probit model ( see Materials and Methods ) , the best fit parameters are ID50\u200a=\u200a4 , 130 and m\u200a=\u200a0 . 291 ( Table 1 , model J ) ., Extrapolation using these values results in ID10 of 50 spores and ID1 of 1 spore ., Without raw data , it cannot be determined whether any of the monkeys in the Jemski experiments were exposed to low doses and , if so , whether any of those doses proved fatal ., Furthermore , without the full data set it is not possible to evaluate whether alternative dose-response models would have fit the data better than the log-probit model , which has been outperformed by other models in fitting other data sets 18 ., Two studies 11 , 15 applied a log-probit model based on the Jemski data to analyses of human exposure scenarios , although they applied ID50\u200a=\u200a8 , 600 ( the upper limit of the 95% confidence interval reported by Glassman ) ., Two studies contain raw data from a substantial number of monkeys exposed to a range of dose amounts ., Druett et al . 27 exposed rhesus monkeys ( Macaca mulatta ) to aerosols of B . anthracis spores resulting in a range of inhaled doses estimated between about 35 , 000 to 200 , 000 spores ., We summarize the data from these experiments in Table S1 ., The authors also fit a log-probit model to their data ( Table 1 , model D1 ) resulting in optimal parameters equivalent to ID50\u200a=\u200a53 , 000 spores ( 95% confidence interval 30 , 000 to 52 , 000 ) and m\u200a=\u200a1 . 39 ., Haas 18 reported a fit of the exponential model ( model D2 ) to this data set and also stated that the best fit log-probit and beta Poisson models did not provide a statistically significantly improved fit compared to the exponential model ., The second study containing raw data , Brachman et al . 24 , exposed cynomolgus monkeys to B . anthracis-contaminated air from a goat hair mill ., The data consist of estimated dosage and the number of deaths from anthrax , sacrifice , or other cause on each day across three model runs and are shown graphically in 24 ., We visually estimated the daily exposure data from their figures and manually adjusted those estimates until they were consistent with the cumulative dose numbers reported in the source text ., Our estimates of these numerical data are shown in Tables S2 , S3 , S4 ., The authors did not fit a dose response model to their data , but two more recent studies have done so ., Haas 18 used an averaging technique 28 to fit a time-independent exponential model ( Table 1 , model B1 ) to the data , and Mayer et al . fit a time-dependent exponential model and an extended exponential model ( Table 1 , models B2 and B3 ) ., The published literature also includes quantitative human inhalational anthrax dose-response estimates based on the opinion or judgment of experts ., For example , biodefense experts from the US Army Institute of Infectious Diseases ( USAMRIID , Fort Detrick , MD ) state the infective dose ( presumably ID50 ) of inhalational anthrax for humans is 8 , 000–50 , 000 spores 29 , 30 ., An expert elicitation of seven anthrax subject matter experts 31 indicated an ID10 of 1 , 000–2 , 000 spores , an ID50 of 8 , 000–10 , 000 spores , and an ID90 of 50 , 000–100 , 000 spores ., Webb and Blaser 16 extended those expert-derived estimates to age-specific values for the ID10 and ID50 , but without providing quantitative evidence or reasoning used to derive these estimates ., Several dose response models have been proposed and applied based entirely or in part on the values from these expert elicitations ( Table 1 , models E1–E5 ) ., We evaluate the previously published models against the criteria listed in Materials and Methods in Table, 1 . Versions of six of the models in Table 1 ( J and E1–E5 ) have been applied in recent mathematical modeling or simulation studies of human exposure to anthrax 9–11 , 15 , 16 , 32 ., Models J , D1 , D2 and B1–B3 were fit to one of three non-human primate dose-response data sets and , therefore , satisfy criterion 1 ( although model J is based on a data set by Jemski for which the raw data are not published , which limits transparency ) ., Models E1–E5 do not have a clear basis in quantitative dose-response data , but are instead based entirely or partly on assumptions , recommendations , or expert opinions for which the reasoning has not been made clear in published accounts ., All models except for three of the log-probit models with steeper slopes ( E1 , E2 , and D1 ) produce dose-response curves with shapes that either were shown to be consistent with the Sverdlovsk data in Wilkening 15 or produce similar estimates to the models tested in that study and , therefore , satisfy criterion, 2 . The models taking the exponential form ( E5 , D2 , and B1–B3 ) are based on simple assumptions about the fate of individual spores inhaled in the lung , satisfying criterion 3 , while the other models make use of statistical distributions with no clear basis in assumed mechanisms of infection ., Model E5 produces incubation period estimates as an extension of the assumptions that form the basis of the model and , therefore , satisfy criterion 4 ., Models B2 and B3 produce estimates for the time course of infection but not for the incubation period ., I . e . , they specify time to infection take-off ( initial germination of inhaled spores ) and to death , but not to onset of symptoms ., The other previously existing models do not contain time components for disease progression among those infected ., Although an incubation period distribution can be added to any dose-response model exogenously ( as was done by Wilkening 15 to a version of model J and model E2 ) , our preference under criterion 4 is for models in which the incubation period estimates are derived ab initio in conjunction with a dose-response model ., Of the five models with a quantitative basis in expert opinion , model E5 has the most ( three ) of the desired characteristics of an anthrax dose-response model ., However , while some of the time-based parameters of this model have been estimated from non-human primate data and human data from Sverdlovsk 32 , the full dose-response model is incomplete without assuming a fixed point on the dose-response curve ( e . g . , the ID50 ) which does not have a firm basis in those data ., Non-human primate data sets can be used to fill that need ., Model J based on the Jemski data does not satisfy criteria 3 and 4 , and the raw data are not available to attempt further modeling with improved characteristics ., Therefore , we focus on models fit the Druett et al . and Brachman et al . data sets in the following sections ., We checked the results for the optimal parameters of the log-probit model D1 and the exponential model D2 when fit to the Druett et al . data ., Our best fit parameters for the log-probit model confirm the results of model D1 ., For the exponential model , our best fit parameter is r\u200a=\u200a1 . 43×10−5 ( 95% confidence interval 0 . 92×10−5 to 2 . 19×10−5 ) , which is twice the estimate of model D2 ., We have listed our novel result as model D3 , and we explain the source of the difference from model D2 below ., We also fit the beta Poisson model to the data set , and the result produced a nearly identical curve to model D3 , so we did not list it in Table, 1 . The exponential model contains one fewer parameter than the beta Poisson model and is , therefore , more parsimonious , so the beta Poisson model need not be considered further , as it does not improve the fit to the data ., Models D1 and D3 have a statistical deviance ( defined in Materials and Methods ) of and 10 . 3 and 11 . 3 , respectively , which are less than the corresponding 95th percentile chi-squared statistics ( 14 . 1 and 15 . 5 ) with degrees of freedom equal to the number of dose points ( 9 ) minus the number of parameters in each model ( 2 and 1 ) ., Under this criterion , both models provide an adequate fit to the data 33 ., The deviance under model D1 is lower than under D3 , which suggests a better fit , but the difference is less than the difference in the chi-squared statistics , so that the exponential model would be chosen as the best combination of fit and parsimony 33 ., The ID estimates shown for models D1 and D3 in Table 1 illustrate the sensitivity of extrapolated estimates to model choice ., The ID50 estimates , which are within the range of the doses actually supplied to the animals , agree closely , whereas the estimates for doses farther below the lowest dose from the data set ( ≈35 , 000 spores ) differ substantially ., While the extrapolations from the exponential model are better supported according to the statistical criteria described above , even a small amount of additional data at lower doses could have shifted support to the estimates of the log-probit model ., Dose-response models fit to the Druett et al . data have not been applied to mathematical models or simulations of human anthrax exposure , to our knowledge ., While both the exponential and log-probit models provide adequate fits to the data and , therefore , satisfy our first criterion , the exponential model better satisfies our other criteria: it is derived from testable , mechanistic assumptions , while the log-probit model is not 18 , and it produces a less steep dose-response curve that is more consistent with the Sverdlovsk data 15 ., However , neither model can satisfy our criterion of providing a time-to-infection component without making additional unsupported assumptions , as the time of death was not reported in the Druett et al . data ., Therefore , we turn to the Brachman et al . data , which have the ability to support a model that satisfies all four of our criteria ., We fit a novel Exposure–Infection–Symptomatic illness–Death ( EISD ) model to the Brachman et al . data set 24 , resulting in Model B4 ( Table 1 ) ., The overall model , summarized here and described in detail in Materials and Methods , contains five parameters ., The exponential dose-response model parameter r , the probability of one spore germinating before being cleared , governs the probability that infection will eventually occur after exposure to a given dose ., Among those infected , the time from exposure to infection , defined as the time of the first successful spore germination leading to a sustained population of vegetative cells in the host , is governed by the parameters r and θ , the rate of clearance of spores from the lung ., The time from infection to the onset of symptomatic illness is represented by the fixed parameter T , and the time from the onset of symptoms to death is governed by the parameters a and b , which are shape and scale parameters of a gamma distribution ., Estimates for three of these five model parameters are available from independent data of B . anthracis infections in humans and in non-human primates ., Brookmeyer et al . 32 calculated the probability-per-time for clearance of spores from the lung , θ , to be 0 . 07 per day , based on data from examination of the lungs of non-human primates at varying times after inhalation 34 ., Data are also available for the time between the onset of symptoms and death in humans ., Holty et al . 25 assembled data from 82 human inhalational anthrax cases , occurring between 1900 and 2001 , that met their inclusion criteria concerning sufficient documentation of anthrax infection , symptoms , and treatment ., Their data set includes , for 75 of the cases , the time from the onset of symptoms to death , if it occurred , and/or to appropriate antibiotic therapy , if it occurred , which may have prevented or delayed death ., We used a maximum likelihood procedure , designed to account for time censoring ( see Materials and Methods ) , to fit a gamma distribution for the time between onset of symptoms and death to their compiled data set ., We determined the shape parameter a\u200a=\u200a5 . 43 and scale parameter b\u200a=\u200a0 . 864 , which results in an average time of 4 . 7 days , with a standard deviation of 2 . 0 days ., By fixing those values of the three parameters θ ( rate of clearance of spores from the lung ) , a ( shape parameter ) , and b ( scale parameter ) , we estimated the values for the remaining parameters r ( probability of one spore germinating before being cleared ) and T ( delay between initial spore germination and onset of symptoms ) from the Brachman data ., The best fit model estimates r\u200a=\u200a6 . 4×10−5 ( 95% confidence interval of 4 . 0×10−5–9 . 5×10−5 ) and T\u200a=\u200a2 . 3 days ( 0–5 . 4 ) ., The optimal deviance of 129 is less than the corresponding 95th percentile chi-squared statistic ( 170 ) with 142 degrees of freedom ( the number of daily dose points minus the number of optimized parameters in the model ) , suggesting that the model provides an adequate fit to the data ., The optimal value of r leads to an ID50 of 11 , 000 spores ( 7 , 200–17 , 000 ) , ID10 of 1 , 700 spores ( 1 , 100–2 , 600 ) and ID1 of 160 spores ( 100–250 ) ., The optimal value of T , when combined with the dose-dependent delay from exposure to infection , produces dose-dependent incubation periods ( exposure to symptom onset ) ., For an ID50 dose , the median incubation period is estimated to be 9 . 9 days ( 7 . 7–13 . 1 ) ., For ID10 , the estimate is 11 . 8 days ( 9 . 5–15 . 0 ) and for a low dose of ID1 , the estimate is 12 . 1 days ( 9 . 9–15 . 3 ) ., Our best fit model to the Brachman data satisfies all four criteria that we propose for a defensible anthrax dose-response model that is useful for quantitative risk assessment ., All parameter values are transparently derived from human and non-human primate data , the model is derived from biological assumptions about the establishment of infection and progression of disease , the model provides estimates for dose-dependent infection probability and distribution of incubation period , and the shape of the dose-response curve is consistent with what was observed in the Sverdlovsk data ., We compare the results from this model to others in the literature in the following sections ., We compare the uncertainty range of the dose-response curve ( probability of infection at any time after exposure to a given dose ) produced by model B4 to the curves from selected models shown in Table 1 , focusing on low doses ( Figure 1 ) ., Models E3 , E4 , and E5 from Table 1 are in agreement with model B4 , as those curves fall entirely within the shaded region representing the 95% range ., Models E1 and E2 are in agreement for doses above 200 and 400 spores , respectively , but they produce a significantly lower probability of infection for lower doses ., Model J produces a significantly higher infection probability at doses less than 5 , 000 spores ., Models D1 and D3 produce lower infection probabilities at all doses ., Our optimal estimate for the exponential model parameter r fit to the Druett data set ( model D3 ) is two times higher than the value calculated by Haas 18 ( model D2 ) for the same model fit to the same data set ., The lower infectivity produced by model D2 results from an estimated respiration rate of 2 . 4 L/min rather than the value of 1 . 2 L/min reported and used for calculations in the original paper 27 ., See Table S1 for our calculation of the doses inhaled by each group of non-human primates in the original study ., For the Brachman data , our model B4 produces an estimated range for the exponential model parameter r , and thus for infectivity at a given dose , that is somewhat higher than the values calculated by Haas 18 ( model B1 ) and Mayer et al . 35 ( model B2 ) for exponential model fits to the same data set ( Figure 2 ) ., There were several assumptions made by the three studies that contributed to the differing infectivity results among these three models ., In theory , the averaging technique used by model B1 should have produced the same value as the other two models for the exponential parameter r 28 ., An important reason why the model B1 result is lower is that its calculation resulted from an incorrect assumption of a higher total cumulative dose for runs three and four of the Brachman experiments than was reported in the original paper ( see Table S5 ) ., We recalculated r using the technique of model B1 with the correct dose values and found r\u200a=\u200a3 . 8×10−5 , which is very close to the result of model B2 ., Also , model B1 included animals that died of non-anthrax causes during the Brachman experiments in the group of survivors; if those cases had been excluded entirely , their estimate for r would have increased slightly ., The main reason why our novel result for the parameter r ( model B4 ) is less than both model B2 and corrected model B1 , is that models B1 and B2 both assumed that all animals sacrificed and not found to be infected at the end of each run would not have become infected had the experiment continued ., Our modeling process allows for the possibility that animals dying of other causes or sacrificed could have become infected with B . anthracis at later dates had they lived ., Model B4 estimates that there was approximately a 7% , 4% , and 4% chance of infection after the day the animals were sacrificed in Brachman runs 3 , 4 , and 5 , respectively , assuming no further exposure ., If those probabilities are accurate , then there likely would have been a few more deaths from anthrax across the three runs had the animals lived longer ., Our model B4 also differs from model B2 in its assumptions and results for the time course from exposure to death in anthrax cases ., In their procedure for model B2 , Mayer et al . 35 independently assumed that the delay between infection take-off and death was 1 , 2 , 3 , or 4 days with equal probability ( an assumption not based clearly on data ) ., They then optimized their equivalent to our parameter θ to account for the remaining portion ( exposure to infection take-off ) of the overall delay between exposure and death , finding an optimal value of θ\u200a=\u200a0 . 11 ., We chose different assumptions that rely more directly on quantitative data , fixing θ\u200a=\u200a0 . 07 based on data of spore clearance rates in non-human primates and expressing the symptoms onset to death delay with a gamma distribution fit to rigorously reviewed human anthrax case data , leaving the infection to symptoms onset delay T to be optimized ( resulting in T\u200a=\u200a2 . 3 days ) ., Model B2 does not provide estimates of the incubation period that can be compared to our estimates from model B4 , because model B2 does not specify the time of symptoms onset in its formulation ., However , both models do provide estimates for the time from exposure to death ( the endpoint of the Brachman experiments ) ., We find that our model B4 produces significantly longer estimates than model B2 for this time interval ., For example , after a single ID10 exposure , our model B4 estimates a median time from exposure to death , among those infected and untreated , to be 16 . 6 days ( 95% confidence range , 14 . 4 to 19 . 8 days ) , while model B2 estimates 8 . 6 days ., It is unclear why these time progression estimates differed so widely , given that the two models were fit to the same data set ., We evaluated model B2 against our optimization criterion ( the minimized deviance Y , defined in Materials and Methods ) and found that it provided a poorer fit to the Brachman data by our measure ( Y\u200a=\u200a158 compared to Y\u200a=\u200a129 for our model B4 ) ., Our model B4 also outperforms model B2 in describing the distribution of human exposure-to-death time estimates from the Sverdlovsk release reported by Abramova et al . 36 ( Figure 3 ) ., To further explore the implications of our assumptions in constructing model B4 compared to model B2 , we tested the sensitivity of our results to our choice of θ ( Figure S1 ) ., We reran our optimization procedure fixing θ\u200a=\u200a0 . 11 , which results in optimized values of r\u200a=\u200a5 . 6×10−5 and T\u200a=\u200a3 . 4 days ., The new r value is a small decrease in the infectivity estimate compared to our model B4 result , causing an increase in the ID50 estimate from 11 , 000 to 12 , 000 spores ., The new value of θ caused the optimized time from initial germination to symptom onset to increase by about 1 . 1 days; however , the new value of θ causes the median time from exposure to initial germination to decrease by about 3 . 6 days at low doses ., Therefore , applying θ\u200a=\u200a0 . 11 instead of 0 . 07 would have decreased our median incubation time and time-to-death estimates at low doses by about 2 . 5 days , not enough to fully account for the 8-day difference described above between models B4 and B2 ., A final difference between model B2 and B4 is that the model B2 parameters were only fit to runs 3 and 4 of the Brachman experiments , whereas we made use of runs 3 , 4 , and 5 ( see Tables S2 , S3 , S4 ) in producing model B4 ., We found that deleting the data from Brachman run 5 had a negligible effect on our infectivity results ( the optimal r value was unchanged to two significant digits ) , so the additional data we incorporated did not contribute to the differing infectivity results of the two models ., Next , we compare the incubation period distribution produced by our model B4 to three other estimates of the incubation period for human inhalational anthrax found in the literature 12 , 32 , 37 ( Figure 4 ) ., Our model is unique in that , while the shape of the dose-response curve being consistent with the Sverdlovsk data was a criterion for model choice , we did not actually use incubation period data or time-to-death data from Sverdlovsk to determine parameter values ., Therefore , we also check our models estimates against the Sverdlovsk data ( Figure 4 ) as a validation for the utility of applying model B4 to a human outbreak ., The Institute of Medicine ( IOM ) performed a detailed review 12 of data from analyses of Sverdlovsk patients: Abramova et al . 36 reported on 41 autopsy-confirmed cases , among which 30 cases had known dates of symptom onset; Meselson et al . 7 compiled data from 77 cases , 60 with known symptoms timing , but no additional confirmed cases beyond those that were reported by Abramova et al . ; Brookmeyer et al . 38 analyzed 70 cases with known symptoms onset dates , but again , no additional cases beyond the Abramova data that were confirmed by autopsy or microbiological testing ., The IOM committee reviewing these data wrote “in its analysis of previous anthrax incidents , the committee required either microbiologic or histopathologic confirmation of infection with B . anthracis when determining the minimum incubation period of patients with inhalational anthrax” 36 ., We chose to follow the lead of this committee and used only the autopsy-confirmed Abramova et al . data in Figure 4 to test the performance of our model ., These data , when choosing April 2 , 1979 as the assumed date of release and exposure ( an assumption supported by compelling evidence 12 ) , consist of 30 estimated incubation periods ranging from 5 to 40 days , with median 13 days and mean of 16 . 0 days ., Of the data from the other two studies excluded from this set , the IOM cast doubt in particular on unconfirmed reports of shorter incubation periods , as low as 2 days , which are not well supported 12 ., We compared the incubation period distribution provided by our model B4 under the assumption of exposure to the ID1 ( consistent with an approximate 1% attack rate observed at given locations downwind of the Sverdlovsk release 7 ) to the distribution of the estimated incubation periods of confirmed Sverdlovsk cases ( Figure 4 ) ., Our model appears to provide a good match to these data , as most points fall within our 95% confidence range , despite the fact that these data were not used in fitting parameter values for our model ., We find that model B4s consistency with these Sverdlovsk incubation time data is robust to assuming that infected cases were exposed to a much higher dose ( ID50 ) and to the alternate a | Introduction, Results, Discussion, Materials and Methods | Anthrax poses a community health risk due to accidental or intentional aerosol release ., Reliable quantitative dose-response analyses are required to estimate the magnitude and timeline of potential consequences and the effect of public health intervention strategies under specific scenarios ., Analyses of available data from exposures and infections of humans and non-human primates are often contradictory ., We review existing quantitative inhalational anthrax dose-response models in light of criteria we propose for a model to be useful and defensible ., To satisfy these criteria , we extend an existing mechanistic competing-risks model to create a novel Exposure–Infection–Symptomatic illness–Death ( EISD ) model and use experimental non-human primate data and human epidemiological data to optimize parameter values ., The best fit to these data leads to estimates of a dose leading to infection in 50% of susceptible humans ( ID50 ) of 11 , 000 spores ( 95% confidence interval 7 , 200–17 , 000 ) , ID10 of 1 , 700 ( 1 , 100–2 , 600 ) , and ID1 of 160 ( 100–250 ) ., These estimates suggest that use of a threshold to human infection of 600 spores ( as suggested in the literature ) underestimates the infectivity of low doses , while an existing estimate of a 1% infection rate for a single spore overestimates low dose infectivity ., We estimate the median time from exposure to onset of symptoms ( incubation period ) among untreated cases to be 9 . 9 days ( 7 . 7–13 . 1 ) for exposure to ID50 , 11 . 8 days ( 9 . 5–15 . 0 ) for ID10 , and 12 . 1 days ( 9 . 9–15 . 3 ) for ID1 ., Our model is the first to provide incubation period estimates that are independently consistent with data from the largest known human outbreak ., This model refines previous estimates of the distribution of early onset cases after a release and provides support for the recommended 60-day course of prophylactic antibiotic treatment for individuals exposed to low doses . | Anthrax poses a potential community health risk due to accidental or intentional aerosol release ., We address the need for a transparent and defensible quantitative dose-response model for inhalational anthrax that is useful for risk assessors in estimating the magnitude and timeline of potential public health consequences should a release occur ., Our synthesis of relevant data and previous modeling efforts identifies areas of improvement among many commonly cited dose-response models and estimates ., To address those deficiencies , we provide a new model that is based on clear , transparent assumptions and published data from human and non-human primate exposures ., Our resulting estimates provide important insight into the infectivity to humans of low inhaled doses of anthrax spores and the timeline of infections after an exposure event ., These insights are critical to assessment of the impacts of delays in responding to a large scale aerosol release , as well as the recommended course of antibiotic administration to those potentially exposed . | medicine, bacterial diseases, infectious diseases, anthrax, population modeling, infectious disease modeling, biology, computational biology | null |
journal.pcbi.1006892 | 2,019 | How Dendrites Affect Online Recognition Memory | To function well in a complex world , our brains must somehow stream our everyday experiences into memory as they occur in real time ., An “online” memory of this kind , once termed a “Palimpsest” 1 , must be capable of forming durable memory traces from a single brief exposure to each incoming pattern , while preserving previously stored memories as long and faithfully as possible ( Fig 1 ) ., This combined need for rapid imprinting and large capacity requires that the memory system carefully manage both its learning and forgetting processes , but we currently know little about how these processes are implemented and coordinated in the brain ., A number of quantitative models have been proposed for palimpsest-style online memories , and have addressed a variety of different issues , including: how memory capacity scales with network size , how metaplastic learning rules can increase memory capacity , and the tradeoff between initial trace strength and memory lifetimes 1–8 ., A few studies with a more empirical focus have addressed the biological mechanisms underlying recency vs . familiarity memory 9; the coordination of online learning with long-term memory processes; and the details of memory-related neuronal response properties during online learning tasks 10–12 ., Nearly all previous models of online learning have assumed that the neurons involved in memory storage are classical point neurons” , that is , simple integrative units lacking any representation of a cell’s dendritic tree ., This simplification is notable , given the now substantial evidence from both modeling and experimental studies that dendritic trees are powerful , functionally compartmentalized information processors that can augment the computing capabilities of individual neurons in numerous ways 7 , 13–59 ., Beyond their contributions to the computing functions of neurons , it is also increasingly apparent that dendrites help to organize and spatially compartmentalize synaptic plasticity processes 7 , 40 , 60–86 ., Thus , given that dendrites can act as both signaling and learning units within a neuron , it is important to understand how having dendrites could affect the brain’s online learning and memory processes ., In this paper , we focus on the role that dendrites may play in familiarity-based recognition , a function most closely associated with the perirhinal cortex 87 , 88 ., Here , we introduce a mathematical model that allows us to calculate online storage capacity from the underlying parameter values of a previously proposed dendrite-based memory circuit 7 ., The model includes biophysical parameters ( dendritic learning and firing thresholds , network recognition threshold ) , wiring-related parameters ( number of axons , number of dendrites , number of synapses per dendrite ) , and input pattern statistics ( pattern density , noise level ) ( see Table 1 ) ., As an example of the model’s use , we study the interactions between memory capacity , dendrite size , and pattern statistics , and cross-check the results using full network simulations ., We found that dendrites containing a few hundred synapses ( as opposed to a few tens or a few thousand ) maximize storage capacity , providing the first normative theory that accounts for the actual sizes of dendrites found in online memory areas of the brain ., The network structure and plasticity rules have been previously described in 7 , but are repeated here for clarity ., A population of neurons with a total of M separately thresholded dendrites receives inputs from NA input axons ( Fig 2b ) ., Each dendrite receives K synaptic contacts randomly sampled from the NA axons , for a total number of synapses NS=M∙K ., The connectivity matrix is assumed to be fixed ., Input patterns are binary-valued vectors x = {x1 , … , xNA} for which component xi is 1 if the ith axon is “firing” and 0 otherwise ., We quantify density/sparsity of the patterns by the fraction of axons fA firing in each pattern; the value of fA ranged from 0 . 008 to 0 . 18 in this study , as we found empirically in previous work that sparse patterns maximize capacity in this type of memory 7 ., To model a biologically realistic form of input variability , we assumed that each active axon ( xi=1 ) produces a burst of spikes , where the number of spikes in the burst is drawn from a binomial distribution with mean μburst=Nburst·Pburst=4 spikes/burst ., Pburst ranged from 1 ( no noise ) to 0 . 4 ( high noise ) , with Nburst varying inversely ., Inactive axons ( xi=0 ) were assumed to produce no spikes ., We denote the noisy spike count version of an input component x~i~xi∙Binom ( Nburst , Pburst ) ., Synapses are characterized by both a weight wij , where the subscript indicates a connection between axon i and dendrite j , and an additional scalar parameter αij , representing the synapse’s “age” ., The weight of each synapse is binary-valued , and can change between weak ( w = 0 ) and strong ( w = 1 ) states when the dendrite containing the synapse undergoes a learning event; the conditions that trigger a learning event are discussed below ., The age variable at each synapse tracks the number of learning events that have occurred in the parent dendrite since the synapse last participated in learning ., Two different measures of a dendrite’s activation level determine how the dendrite responds to an input , and whether it undergoes a learning event ., The “presynaptic” activation measure is based on the activity levels of the set of axons Dj that make contact with the jth dendrite, apre ( j ) =∑iϵDjx˜i ., In words , apre ( j ) is the total number of presynaptic spikes arriving at all the synapses impinging on the jth dendrite , regardless of their postsynaptic weights , and is thus a measure of the maximum response the dendrite could muster to that input pattern assuming all of the activated synapses were strong ( w=1 ) ., The more conventional “postsynaptic” activation level takes account of the synaptic weights in the usual way:, apost ( j ) =∑iϵDjwij·x˜i ., When the postsynaptic activation level exceeds the “firing” threshold θF , the dendrite is said to fire , that is , generates a response rj = 1 ., The responses of all dendrites within a neuron sum linearly to produce the neuron’s response ( Fig 2b ) , and the responses of all neurons in the network sum linearly to produce the overall network response r ., The overall response of the network can therefore be written directly as a sum over all the M dendritic responses:, r=∑jϵ1 , Mrj, so that the network can be viewed as a single “super neuron” with M dendrites ., Finally , an input pattern is classified as “familiar” if r≥θR , and “novel” if r<θR , where θR is the recognition threshold ( Fig 2b ) ., The goal of learning is to ensure that learned patterns going back as far as possible in time produce suprathreshold network responses ( r≥θR ) , while randomly drawn patterns do not ., Learning of any given pattern occurs in only the small fraction of dendrites that cross both the presynaptic and postsynaptic learning thresholds ( apre ( j ) >θLpre and apost ( j ) >θLpost ) ., When this occurs , a “learning event” is triggered in the dendrite , and all active synapses belonging to that dendrite “learn” , as follows ., If an active synapse is currently in the weak state , it is “potentiated” ( i . e . both strengthened and “juvenated”: wij→1 , αij→0 ) , or if it is already in the strong state , then it remains strong but is juvenated ( wij=1 , αij→0 ) ., All strong synapses in the dendrite that are not active during the learning event remain strong but grow older ( wij=1 , αij→αij+1 ) ., Thus αijcounts the number of learning events that have occurred in the dendrite since the synapse last learned , and thus represents the age of the most recent information that that synapse is involved in storing ., Note that a synapse’s age variable counts learning events within its parent dendrite only , and any given dendrite learns only rarely , so the counter need have only a small number of distinct values , on the order of ~12 under the simulation conditions explored in this paper ., To maintain a constant fraction of strong synapses ( we used fs=0 . 5 ) , and thereby to prevent saturation of the memory , in each dendrite undergoing learning , a number of strong synapses are depressed ( wij→0 ) equal to the number of weak synapses potentiated during that learning event ., A key feature of the learning rule is that the synapses targeted for depression are those that learned least recently ( i . e . having the largest values of αij ) , so that the information erased during depression is the “oldest” stored information ., This “age-ordered depression” strategy substantially increases online storage capacity 5 , especially in a 2-layer dendrite-based memory where the very sparse use of synapses during pattern storage gives each strong synapse , and the information it represents , the opportunity to grow old 7 ., One of the key quantities involved in calculating storage capacity is L , the length of the age queue within a dendrite ( see Fig 3 ) ., An approximate expression for L is given here; the derivation can be found in the Methods ., L is a measure of the time a pattern feature persists in a dendrite , and given that age queues progress at roughly equal rates in all the dendrites involved in storing a pattern , it also effectively measures a pattern’s lifetime in memory–counted in units of dendritic learning events ., L can be understood intuitively through an oversimplified example: If 10 synapses are strengthened on a dendrite during a learning event , and there are 120 strong synapses on the dendrite , then L would be ~12 ., That is , after ~12 learning events have elapsed since a pattern was first stored , the 10 synapses involved in storing the pattern are now the oldest on the dendrite and must be depressed , and the memory is lost ., The actual expression for L is more complex as it takes into account the fact that strong synapses do not inexorably progress to the ends of their age queues–they can be rejuvenated one or more times during the course of their lifetimes , in which case the same strong synapse participates in the representation of more than one pattern ., To convert from L to a number of training patterns , we must multiply L by the approximate number of patterns per dendritic learning event , or “learning interval” 1PL , where PL is the probability that an arbitrary dendrite learns a particular pattern ., This gives an expression for capacity:, C≈LPL=1PLlog ( 1−fS ) log ( 1−θLpreK⋅μburst ) −1, ( 2 ), Although PL is conceptually simple , its expression is complicated since it depends on pattern density , noise level , two learning thresholds , dendrite size , and fS , and so it is omitted here for clarity ( see in the Methods section for the full expression and some discussion ) ., The expression for C measures how long patterns persist in memory , but a different calculation is needed in order to predict the memory’s recognition performance , that is , the false positive and false negative error rates ϵ+ and ϵ- that we can expect to obtain during a pattern’s lifetime ., These error rates depend on the separation of the distributions of responses to trained vs . untrained patterns ( Fig 1 ) ., These two distributions can be computed from the network parameters to determine whether the allowable error rate tolerances θ+ and θ- will be met during the lifetime calculated in Eq 2 ( see Methods ) ., How can the expression for online storage capacity ( Eq 2 ) be exploited ?, Given that one of the unique features of our model is that dendrites are the learning units , we used the model to determine how capacity varies with dendrite size , which in turn allows us to determine the optimal dendrite size ., In particular , we asked: for a fixed total number of synapses in the memory network ( NS=M∙K ) , if the goal is to maximize online storage capacity , is it better to have many short dendrites ( i . e . large M , small K ) , a few long dendrites ( small M , large K ) , or something in between ?, Furthermore , how does the optimal dendrite size vary with properties of the input patterns , such as pattern density and input noise level ?, To address these questions , we fixed network parameters Nsand fs and then for varying combinations of the pattern-related parameters ( fA , Nburst , Pburst ) , we computed C as a function of dendrite size K , using values of the learning , firing , and recognition thresholds ( θLpost , θLpre , θF , θR ) optimized for each value of K through a semi-automated grid search ., The “optimal” dendrite size under a particular set of input conditions was the value of K that maximized capacity , subject to the constraint that immediately after training , responses to trained patterns were strong enough , and responses to random patterns were weak enough , that both the false positive ( ϵ+ ) and false negative ( ϵ- ) error rates fell below specified tolerances ( we used 1% for both ) ., Note that though K appears explicitly only once in Eq 2 , as a result of the capacity optimization process , all of the thresholds , and consequently θLpre and PL in Eq 2 depend implicitly on K . The net effect of these dependencies is analyzed in detail in the sections below on penalties for long and short dendrites ., Capacity is plotted in Fig 4a as a function of K for pattern density values ranging from 0 . 8% to 18% ., In the case with fA=1 . 5% , capacity peaked at ~30 , 000 patterns when dendrites each contained 256 synapses , and declined substantially for both short ( K<100 ) and long ( K>1000 ) dendrites ., As the pattern density increased ( to 18% ) or decreased ( to 0 . 8% ) , peak capacity varied nearly 5-fold , favoring sparser patterns , but over the more than 20-fold range of pattern densities tested , peak capacity always occurred for dendrites ranging from 100–500 synapses ( grey shaded area ) ., Focusing on the high-capacity ( sparse ) end of the range with fA<3% , peak capacity was confined to the narrower range of 200–500 ( i . e . “a few hundred” ) synapses ., We also observed that sparser patterns led to a preference for longer dendrites , an effect we unpack below using full network simulations ., It is important to clarify that the higher recognition capacity seen for sparser patterns does not result from the fact that sparser patterns contain less information , thereby reducing storage costs per pattern ( see S1 Text ) ., We also note that in the more realistic conditions modeled in the full network simulations ( see below and Fig 5 ) , peak capacity saturates at slightly higher pattern activation densities ( around 1 . 5% ) than is predicted by the analytical model , and the optimal pattern density may be higher still under conditions of increased background noise ( S1 Fig shows strong susceptibility to background noise even at 3% pattern density ) ., To test the effect of pattern noise on capacity , we varied the input noise level by choosing combinations of Nburst and Pburst whose product was always μburst=4 spikes , but that yielded narrow or broad spike count distributions for each active pattern component ( Fig 4b , see histogram insets ) ., In this way , we varied the degree to which a trained pattern resembled itself upon repeated presentations ., The variation in event counts arising from the above scheme could be viewed as representing either variation in the number of action potentials arriving at the presynaptic terminal from trial to trial , or variation in the number of synaptic release events caused by a given number of action potentials , or a combination of both effects ., As expected , higher noise levels reduced peak capacity ( Fig 4b ) , except in the long dendrite range ( K>1000 ) where central limit effects rendered dendrites insensitive to this type of noise ., In keeping with this effect , optimal dendrite size increased slightly as the noise level increased , but again , peak capacity was consistently seen for dendrites in the “few hundred” synapse range ., Even higher levels of noise were not considered because a simple , biologically available saturation strategy that maps multiple release events into a relatively constant post-synaptic response can largely mitigate the effects of this type of noise ., ( We did not include a multi-input saturation mechanism in our model to avoid the added complexity ) ., To verify that the preference for dendrites in the few hundred synapse range was not an artifact of “small” network size , we generated capacity curves from Eq 2 for networks scaled up 256-fold from a base size of N = 5 . 12 million synapses to ~1 . 3 billion synapses ., The results are shown on a log plot in Fig 4c ., As shown in Fig 4d , the scaling power for dendrite sizes K = 64 , 256 , and 1024 were , respectively , 0 . 98 , 0 . 97 , and 0 . 97 , confirming earlier observations that storage capacity in an optimized dendrite-based memory grows essentially linearly with network size 7 ., All the while , the preference for dendrites containing a few hundred synapses remained essentially invariant ., To cross-check the results of the analytical model , we simulated a full memory network , and measured capacity empirically as a function of K . Unlike the analytical case , in which capacity was assumed to be proportional to the calculated length of dendritic age queues , in the network simulations we performed explicit old-new recognition memory tests , and optimized system parameters to achieve false positive and false negative error rates of 1% ., In the interests of greater biological realism , we replaced the hard dendritic firing threshold and binary input-output function with a continuous sigmoidal input-output function given by 11+e-sx-θF , and optimized over the slope parameter s along with the 4 threshold parameters ., In addition , we relaxed the strict assumption of the analytical model that every input to the network was statistically independent of every other , and instead arranged for each input axon to form ρ synaptic contacts within the memory area , rather than just one ., This “redundancy” factor , ρ , set by default to 200 , introduced some degree of correlation in the input patterns , and lowered peak capacity somewhat , but had no effect on our main conclusions ., Fig 5a depicts one such simulation with 5 . 12 million synapses ., In the top panel , blue dots show responses to trained patterns , red dots show responses to randomly drawn ( untrained ) patterns that establish the baseline trace strength ( green dashed line ) above which stored pattern traces must rise to be recognized ., Consistent with the analytical model , responses to trained patterns remain essentially constant during an extended post-training period , in this example spanning ~10 , 000 patterns ., After the flat post-training phase , in contrast to the relatively abrupt fall in trace strength envisioned by the analytical model , a more gradual decline is seen , reflecting the variable times at which the synapses encoding each pattern reach the end of their age queues in different dendrites ., Note that the false negative error rate begins to climb during this trace decay period , as the lower fringe of the trained response distribution ( blue ) progressively merges with the untrained background distribution ( red ) ., In this simulation , capacity was reached at ~21 , 000 patterns , which by our specification is the point where both false positive and false negative error rates equaled 1% ., Mirroring the approach taken with the analytical model , multiple simulations were run with varying firing , learning , and recognition thresholds to find the combination of parameters that maximized capacity for each value of K , subject to the same error rate constraints as before ., As an additional check of the analytical model , we histogrammed synapse ages within a dendrite ( for many dendrites ) ( Fig 5b ) , and found that they conformed to a geometric distribution as predicted ( red line shows a fitted exponential decay ) , up to the “cliff” at the end of the age queue ( blue dashed line ) ., Capacity was measured for dendrite sizes between 32 and 4 , 096 synapses , and the results are shown in Fig 5c and 5d , which are the analogues of Fig 4a and 4b , respectively ., When compared to the curves produced by the analytical model , the capacity curves produced by full network simulations had similarly placed capacity peaks and similar qualitative dependence on pattern density and noise levels ., In one minor difference , we noted that under the more realistic conditions modeled in the full network simulations , peak capacity saturated at slightly higher pattern activation densities ( around 1 . 5% ) than was predicted by the analytical model ( Fig 4a ) ., To determine whether the predictions regarding optimal dendrite size would survive under even more challenging “real world” operating conditions , we added increasing amounts of background noise ( spurious spikes added to nominally inactive pattern components ) , on top of the pre-existing burst noise and pattern correlations ., As in the case of burst noise , the background noise level varied between 2 extremes: zero noise , which maximized capacity , and a “high noise” level that reduced storage capacity by roughly a factor of 2 compared to the no-noise case ., As in the case of burst noise , we did not consider very high noise levels on the grounds that the deleterious effects of background noise can be compensated by a relatively simple mechanism , for which there is evidence: pre-synaptic terminals with low release probability for “singleton” spikes , along with paired pulse facilitation 89 , would allow the effects of sporadic background spikes to be suppressed while maintaining strong responses to signal-carrying bursts ., Even at background noise levels capable of causing a significant reduction in peak capacity , the effect of background noise on optimal dendrite size was negligible ( S1 Fig ) ., Only at very high levels of background noise , where capacity was reduced more than twofold , did optimal dendrite size change significantly , moving outside of the of the “few hundred” synapses per dendrite range ( S1 Fig ) ., Next we examined the effect of increasing correlations in the input patterns ., Given that a single axon can in fact form many thousands of synaptic contacts , corresponding to a much higher redundancy factor than we used in our base simulation , we ran simulations using redundancy factors ρ=5 , 000 and ρ=10 , 000 ( Fig 5f ) , which meant that groups of 5 , 000 or 10 , 000 synapses scattered across the memory were activated identically ., Given previous reports that input correlations can be very deleterious to capacity 10 , we speculated that these drastic reductions in the effective dimensionality of the input patterns would severely challenge a memory architecture that was designed to perform optimally with random inputs , or at least significantly alter its behavior ., As shown in Fig 5f , however , even in the high-redundancy case ( with a 10 , 000-fold reduction in input space dimensionality ) , peak capacity dropped by only a factor of ~2 compared to the case with ρ=200 , with little to no change in optimal dendrite size ., We next took advantage of the full network simulations to probe the mechanisms that lead to the capacity costs associated with both short and long dendrites ., Fig 5e shows two important quantities: the average number of dendrites ( μLD ) and synapses ( μLS ) used to store a single pattern in the simulations from Fig 5c ., The significance of these quantities is discussed below as we work through the distinct capacity penalties for long and short dendrites ., As shown in Fig 5e , as dendrites grow longer , dendrite usage per stored pattern drops from a value around 10 ( at peak capacity ) to a “floor” of roughly ~7 dendrites at the long-dendrite end of the range , whereas synapse usage climbs steeply from a baseline of around 150 synapses ., To understand the source of the lower bound of ~7 on the average number of dendrites used to store each pattern , it is useful to consider the situation that holds when , in the interests of resource efficiency , we attempt to store each pattern with the minimum possible trace strength: one dendrite ., One dendrite firing in response to a familiar pattern is in principle sufficient for recognition , if it is reliable ( i . e . occurs > 99% of the time ) , and if the network’s response to untrained patterns is reliably zero ( i . e . > 99% of the time ) ., In a large network , given that each dendrite participates in learning with equal ( small ) probability , the distribution of the number of dendrites that undergoes a learning event is approximately Poisson with mean μLD=PL·M ., Given that a Poisson distribution is characterized fully by its mean , setting μLD=1 by adjusting the learning thresholds , which control PL , means that one dendrite will undergo a learning event for each presented pattern–on average–which is the goal ., However , with a mean of 1 , the probability that zero dendrites learn is surprisingly high: ~37% ( Fig 6a , top plot ) ., Thus , in aiming to use a single dendrite to encode a pattern on average , more than a third of all patterns presented to the network would produce no memory trace at all , leading to a false negative error rate far above the 1% acceptable threshold ., To avoid this pitfall , it is critical to reduce the probability to below 1% that zero dendrites learn , which according to the Poisson distribution requires a mean μLD=5 dendrites ., This requires a remarkable 5-fold increase in PL relative to the theoretical minimum , with a corresponding 5x increase in synapse resource consumption ( Fig 6a , middle plot ) ., Worse , given increased variability in the number of learning dendrites as well as increased readout failures due to input noise and correlations , storage capacity turns out to be maximized when an even higher value of PLis used , achieved by further loosening the learning thresholds , which for our combination of system parameters leads to the empirically obtained optimal value of μLD=~7 dendrites at the long-dendrite end of the range ., Given this floor of ~7 dendrites , it becomes clear why synapse usage increases as dendrites grow longer: the number of synapses used in a dendrite that undergoes a learning event is roughly proportional to the dendrite length K , since the number of synapses that learn is roughly proportional to the number of synapses activated in the dendrite , which is proportional to dendrite size ., Tied to this increase in synapse usage per pattern , as the total number of dendrites M in the system decreases ( because each one contains a larger fraction of the synapses ) , the frequency with which each dendrite must participate in learning increases , which speeds the per-pattern rate at which synapses move along their age queues ., Thus , from a capacity standpoint , it is ideal to choose system parameters such that the minimum encoding bound of 7 dendrites is actually used ( or whatever minimum number of dendrites is needed , given the settings of the error rate thresholds and noise level ) , but having met this lower bound , dendrites should be kept as short as possible ., The reasons capacity declines as dendrites grow shorter are complex , and are discussed only briefly here ( see the S1 Text and S3 and S4 Figs for more details ) ., We first consider why dendrite usage increases for short dendrites , rather than remaining at the minimal encoding bound ., Short dendrites are intrinsically more susceptible to variability in crossing their learning and firing thresholds , since fewer active synapses are involved ., As dendrites become very short , this requires the network to increase dendrite usage far above the nominal lower bound of μLD=5 ., For example , under sparse activation ( fA=1% ) , medium noise conditions ( Pburst=47 , Nburst=7 ) with dendrites containing ~200 synapses , when the system is optimized for capacity , μLD≈15 ( blue solid curve in Fig 5e ) , substantially more than the number of dendrites used under maximum capacity conditions ., While this increase in dendrite usage is more than offset by the reduced dendrite size , which tends to reduce synapse usage , the total number of synapses altered during learning in fact remains approximately constant , implying that a larger fraction of synapses is modified within each short dendrite that engages in learning ., This higher synapse burn rate in short dendrites leads to shorter age queues , and in the end lowers capacity ., Why are dendrites of “medium” size optimal for storage capacity in the context of an online familiarity-based recognition memory ?, The simplest explanation is that short dendrites suffer from one set of disadvantages , and long dendrites suffer from another , leaving the optimal dendrite size somewhere in the middle ., Short dendrites have relatively noisier post-synaptic response distributions because fewer synapses contribute to the response ., As a result , a larger fraction of the synapses on a short dendrite must be modified during learning to ensure that the dendrites response to previously trained patterns remains comfortably at the upper tail of the untrained pattern response distribution ., Increasing the fraction of synapses used within a dendrite during each learning event shortens the dendrites age queue , which comes at a capacity cost ., This effect leads to a preference for longer dendrites ., But long dendrites also have their disadvantages ., An online recognition memory should aim to store the weakest possible trace of each learned pattern , which in our framework corresponds to learning in a small number of dendrites near the minimum encoding bound ( corresponding to ~7 dendrites under the conditions used in our study; see Fig 5e ) ., This means that the longer the dendrites become , the more synaptic resources are consumed by each dendrite that learns , since the number of synapses used per dendrite during a learning event is roughly proportional to dendrite size ., Clearly from this perspective , its best to keep dendrites as short as possible ., The compromise between the need to keep dendrites long enough to avoid noise and age queue problems , and short enough to avoid excessive synapse use per learning dendrite , puts the optimal size around a few hundred synapses for biologically reasonable values of pattern activation density and noise ., Of course , our assumptions regarding biologically reasonable pattern activation densities and noise levels are informed guesses rather than certain knowledge , and are not likely to be universal constants across brain areas , species and operating conditions ., It is therefore possible that the natural dendrite sizes found in medial temporal lobe memory areas are determined in part by factors other than capacity optimization according to Eq 2 ., For example , developmental constraints , energy constraints , space constraints , and combinations thereof , may have been responsible for pushing the actual dendrite size in one direction or another , away from the optimal length as determined by capacity considerations alone ., Nonetheless , it is useful to capture basic relationships between biophysical parameters , wiring parameters , input pattern statistics , and capacity , as a starting point for a more complete online memory model ., That mid-sized dendrites optimize capacity can be understood from another perspective ., Eq 2 shows capacity is given by the ratio of L , the length of a dendrites age queue , to PL , the probability that a dendrite learns ., PL , in the denominator , grows larger as dendrites grow in size because the same average number of dendrites is always used to learn , but when dendrites are long , there are fewer of them to choose from ., L , in the numerator , grows smaller as dendrites shrink in size because of the higher value of fpot needed to compensate for noise effects ., Balancing these two effects , capacity is maximized for dendrites of intermediate size , for which L is not too small , and PL is not | Introduction, Results, Discussion, Methods | In order to record the stream of autobiographical information that defines our unique personal history , our brains must form durable memories from single brief exposures to the patterned stimuli that impinge on them continuously throughout life ., However , little is known about the computational strategies or neural mechanisms that underlie the brain\s ability to perform this type of online learning ., Based on increasing evidence that dendrites act as both signaling and learning units in the brain , we developed an analytical model that relates online recognition memory capacity to roughly a dozen dendritic , network , pattern , and task-related parameters ., We used the model to determine what dendrite size maximizes storage capacity under varying assumptions about pattern density and noise level ., We show that over a several-fold range of both of these parameters , and over multiple orders-of-magnitude of memory size , capacity is maximized when dendrites contain a few hundred synapses—roughly the natural number found in memory-related areas of the brain ., Thus , in comparison to entire neurons , dendrites increase storage capacity by providing a larger number of better-sized learning units ., Our model provides the first normative theory that explains how dendrites increase the brain’s capacity for online learning; predicts which combinations of parameter settings we should expect to find in the brain under normal operating conditions; leads to novel interpretations of an array of existing experimental results; and provides a tool for understanding which changes associated with neurological disorders , aging , or stress are most likely to produce memory deficits—knowledge that could eventually help in the design of improved clinical treatments for memory loss . | Humans can effortlessly recognize a pattern as familiar even after a single presentation and a long delay , and our capacity to do so even with complex stimuli such as images has been called almost limitless ., How is the information needed to support familiarity judgements stored so rapidly and held so reliably for such a long time ?, Most theoretical work aimed at understanding the brains one-shot learning mechanisms has been based on drastically simplified neuron models which omit any representation of the most visually prominent features of neurons—their extensive dendritic arbors ., Given recent evidence that individual dendritic branches generate local spikes , and function as separately thresholded learning/responding units inside neurons , we set out to capture mathematically how the numerous parameters needed to describe a dendrite-based neural learning system interact to determine the memorys storage capacity ., Using the model , we show that having dendrite-sized learning units provides a large capacity boost compared to a memory based on simplified ( dendriteless ) neurons , attesting to the importance of dendrites for optimal memory function ., Our mathematical model may also prove useful in future efforts to understand how disruptions to dendritic structure and function lead to reduced memory capacity in aging and disease . | learning, medicine and health sciences, action potentials, neural networks, engineering and technology, nervous system, signal processing, membrane potential, social sciences, electrophysiology, neuroscience, learning and memory, cognitive psychology, cognition, memory, nerve fibers, neuronal dendrites, computer and information sciences, animal cells, axons, cellular neuroscience, psychology, anatomy, synapses, cell biology, physiology, neurons, information theory, biology and life sciences, cellular types, background signal noise, cognitive science, neurophysiology | null |
journal.pgen.1004790 | 2,014 | The Red Queen Model of Recombination Hotspots Evolution in the Light of Archaic and Modern Human Genomes | Meiotic recombination is a highly regulated process , initiated by the programmed formation of double-strand breaks ( DSBs ) ., These DSBs are subsequently repaired , using homologous chromosomes as a template , thus leading to crossover ( CO ) or non-crossover ( NCO ) recombination events ., In mammals , as in many other eukaryotes , the formation of at least one CO on each chromosome is required for the proper disjunction of chromosomes during meiosis ( for review see 1 ) ., Hence , the recombination machinery must be tightly controlled to promote a sufficient number of COs on each chromosome , while ensuring that all DSBs can be efficiently repaired to produce viable gametes ., Recombination events are not randomly distributed across the genome , but cluster in hotspots , typically 1 to 2 kb long 2–8 ., About 33 , 000 recombination hotspots have been identified in the human genome , which account for 60% of COs and 6% of the sequence 2 ., Many independent observations have clearly demonstrated that in human and mouse , the location of hotspots is primarily determined by the zinc finger protein PRDM9 , through its sequence-specific DNA-binding domain 9–12 ., PRDM9 contains a SET domain , which catalyzes histone H3 Lys4 trimethylation ( H3K4me3 ) at hotspot loci 4 , 12–14 ., PRDM9 is highly polymorphic , specifically in its DNA binding domain , and the location of recombination hotspots differs among individuals carrying different alleles 9 , 11 , 12 , 15 ., At the population scale , the set of recombination hotspots that are the most frequently used can be inferred from patterns of linkage disequilibrium 3 or of genetic admixture 16 ., These analyses revealed that more than 90% of recombination hotspots are shared between European and African populations 16 ., This strong overlap is due to the fact that the same major allele of PRDM9 ( allele A ) is present at high frequency both in European and African populations 11 ., Interestingly , this A allele presents affinity for the 13-bp motif CCTCCCTNNCCAC , which was initially identified on the basis of its enrichment within human recombination hotspots 17 ( we will hereafter refer to this sequence motif as HM – for human hotspot motif ) ., It has been shown that the location of recombination hotspots is not conserved between human and chimpanzee 18–20 ., This rapid shift is presumed to be due to the fact that the major PRDM9 alleles present in each species have different DNA binding specificities 10 , 19 ., There is clear evidence that PRDM9 has evolved under strong positive selection , in primates as well as in many other animal lineages , specifically at those sites involved in DNA sequence recognition 21 , 22 ., This indicates that PRDM9 has been under selective pressure to switch to new targets 21 , 22 ., However , the reasons for this selective pressure remain mysterious ., One interesting hypothesis , proposed by Myers and colleagues 10 , is that the turnover of PRDM9 alleles might be a consequence of the self-destruction of recombination hotspots by the process of biased gene conversion ( BGC ) 23–25 ., Indeed , the repair of DSBs is expected to lead to the conversion of recombination-prone alleles by hotspot-disrupting alleles 23–25 ( we will hereafter refer to this form of BGC as dBGC , for DSB-driven BGC ) ., In agreement with the dBGC model , it was shown that the HM motif was subject to accelerated evolution in the human lineage 10 ., The authors suggested that the progressive degradation of recombination hotspots through dBGC might lead to a loss of fitness ., Indeed , there is evidence that lower CO rates are associated with lower fertility , possibly due to improper chromosome disjunction 26 ., Hence , the loss of PRDM9 target motifs might favor the increase in frequency of new PRDM9 alleles , targeting different motifs 10 ., Simulation studies have shown that this model , termed the red queen theory of recombination hotspots , might explain the rapid turnover of recombination hotspots 27 ., It is however not established whether this model is quantitatively realistic ., Notably it has been argued that the number of human recombination hotspots ( ∼30 , 000 ) largely exceeds the number of COs per meiosis ( ∼60 ) and hence is unlikely to be limiting 22 ., The comparison of human and chimpanzee PRDM9 genes revealed multiple non-synonymous changes driven by positive selection in each lineage 22 ., If the red queen model is correct , this implies that this turnover process ( loss of PRDM9 targets by dBGC leading to a selective pressure that favored new PRDM9 allele ) occurred several times since the divergence between human and chimpanzee ., Thus , one strong prediction of the red queen model is that the life expectancy of PRDM9 target motifs should be much shorter than the human/chimpanzee divergence time ., In other words , the key issue is to determine whether , during the lifespan of a given PRDM9 allele , the loss of its target motifs by dBGC is fast enough to have a significant impact on genome-wide recombination patterns ., To address this issue , we first determined when the HM motif started to be the target of PRDM9 allele A in the human lineage ., For this , we analyzed the genome sequence of a Denisovan individual 28 , an archaic human that diverged from the modern humans about 400 , 000–800 , 000 years ago 29 ., We then used polymorphism data to quantify the strength of dBGC on HM motifs in extant human populations ., This combined analysis of polymorphism and divergence , made possible thanks to Denisovan genomic data , demonstrates that the life expectancy of human recombination hotspots is very short , and brings support for the red queen theory of recombination hotspots ., The major human allele of PRDM9 ( allele A , present at a frequency of 84% in European populations and 50% in African populations 11 ) recognizes a specific sequence motif , whose core consensus is CCTCCCTNNCCAC 9 , 10 ., This motif promotes recombination specifically in humans , not in chimpanzee , and is particularly active in the context of THE1 transposable elements 10 , 19 ., As predicted by the self-destructive dBGC drive model , it was previously shown that this motif has accumulated an excess of substitutions specifically in the human lineage , after its divergence from chimpanzee , and that the HM loss rate was particularly strong within THE1 elements 10 ., Based on the dBGC model 25 , the authors proposed that HM had been active for a period of time corresponding to the last 20% to 40% of the time since the human-chimpanzee split 10 ., This estimate was however based on poorly known parameters , and was therefore provided as a conservative upper bound 10 ., To obtain a more direct dating of the onset of the HM motif activity , we used the Denisovan genome so as to determine when the HM motifs started to be subject to dBGC during the evolution of modern and archaic humans ( Figure 1 ) ., We analyzed the evolution of HM motifs both within and outside human recombination hotspots ., For this , we used recombination maps inferred by HapMap from patterns of linkage disequilibrium in human populations 2 ., These maps reflect the average crossover rates across human populations over many generations ., We will hereafter refer to these data as human historical recombination rates ., Given that the list of human historical hotspots is currently available only for autosomes , we excluded sex chromosomes from our analyses ., We first identified HM motifs ( N\u200a=\u200a5 , 704 ) in the reconstructed autosomal sequences of the human-chimpanzee ancestor ( HC ) , and then counted base replacement changes along the four branches of the phylogeny ( hereafter termed modern human , Denisovan , Hominini and chimpanzee branches , Figure 1 ) , by comparing sequences of reference genomes to the ancestral one ( see methods ) ., It should be noted that the detected base changes include both fixed and polymorphic mutations ., To quantify the excess of base changes ( if any ) on HM motifs along each branch of the phylogeny , we used as a reference the rate of base change within a control motif ( CM: CTTCCCTNNCCAC , N\u200a=\u200a5 , 483 ) , which differs from HM by the second position and does not show any effect on the recombination pattern 30 ( Figure 2 ) ., We counted base changes only at informative sites ( i . e . we ignored the two N positions ) and excluded the second position , which differs between HM and CM motifs ., Thus , we only examined sites that are a priori expected to have the same rate of mutation ( and possibly sequencing errors ) in HM and CM motifs ., We considered a motif to be lost as soon as it was subject to one mutation in one informative site ., To minimize errors in the inference of motif losses , it is necessary to avoid regions with low sequencing quality or erroneous alignment ., Thus , we created three levels of filters ( F1 , F2 and F3 ) successively applied to our data so as to keep three subsets of motifs ., A motif is discarded from a subset if at least one informative site does not pass the filter ., Filter F1 retains all aligned sites common to human , chimpanzee and Denisovan , while filter F2 favors a more accurate HC ancestral sequence reconstruction ., Finally , the most stringent filter F3 accounts for sequence errors specific to ancient DNA in Denisovan ( see methods ) ., Unless explicitly mentioned , results presented below correspond to the F2 dataset , totalizing 4 , 440 HM and 4 , 393 CM motifs present in the human-chimpanzee ancestor ., In the modern human branch , we observed that the HM loss rate ( 1 . 8% ) is more than four times higher than the CM loss rate ( 0 . 4%; green branch in Figure 1 ) ., As expected , the HM loss rate is much higher within THE1 elements ( 6 . 7% vs . 1 . 7%; proportion test: p\u200a=\u200a8 . 2×10−5 ) ( Table 1 ) ., However , the excess of HM losses is not limited to THE1 elements: at non-THE1 loci , the HM loss rate is significantly higher than the CM loss rate ( 1 . 7% vs . 0 . 4% , p\u200a=\u200a2×10−8 ) ., Conversely , we observed no significant difference in HM and CM loss rates along the Chimpanzee branch ( in grey in Figure 1 ) , as expected given that the HM motif is not a target of PRDM9 alleles in chimpanzees 10 , 19 ., This negative control confirms that there is no intrinsic difference in mutation rate between the two motifs , and hence that the CM motif is a good reference to detect accelerated evolution of the HM motif ., These observations are consistent with the self-destructive dBGC drive model ., However , they could also be explained by a possible mutagenic effect of recombination ., To distinguish between these two possibilities , we analyzed the derived allele frequency ( DAF ) spectra of mutations in HM and CM motifs: under the hypothesis that the increased HM loss rate is simply due to a higher mutation rate ( and not a fixation bias , like BGC ) , the two DAF spectra are expected to be identical ., We included in these analyses all modern-human mutations detected as polymorphic by the 1000 genomes project 31 , as well as fixed ones ., We observed that the DAF spectrum of HM mutations is shifted towards higher frequencies compared to CM mutations ( Figure 3 ) , with an average mean DAF almost three times higher ( 13% vs . 5%; Wilcoxon test p\u200a=\u200a1 . 9×10−6 ) ., Overall 3 . 7% of HM mutations detected in the modern human branch are fixed , compared to 0 . 2% for CM mutations ( Proportion test p\u200a=\u200a1 . 6×10−4 ) ., This demonstrates that the accumulation of HM losses in the human branch is a consequence of a fixation bias , as predicted by the dBGC model 32 ., In chimpanzee , HM is not a target of PRDM9 and hence is not expected to be subject to any fixation bias ., Consistent with this prediction , the mean DAF of HM mutations in chimpanzee ( 0 . 28 ) is not higher than that of CM mutations ( 0 . 35 ) , and overall there is no significant difference in the DAF spectra of these two categories of mutations ( Figure S1 ) ., We note however that , given the small sample size ( 66 and 67 polymorphic mutations in HM and CM motifs , respectively ) , the power to detect fixations biases is lower in chimpanzees than in humans ., In the Hominini branch , ancestral to Denisovans and modern humans ( in blue in Figure 1 ) , the HM loss rate appears slightly higher than the CM loss rate , but this difference is not statistically significant ., In the Denisovan branch ( in red in Figure 1 ) , the HM loss rate ( 1% ) is two times higher than the CM loss rate ( 0 . 5% ) ., This excess is weaker than that observed in the modern human branch , but it is still significant ( p\u200a=\u200a0 . 025 ) ., Additionally , the rate of homozygosity of these mutations ( computed using the diploid sequence of the Denisovan individual ) is higher for HM than for CM ( 0 . 80 vs . 0 . 69 ) ., This trend is consistent with the hypothesis that in Denisovans , as in modern humans , HM mutations segregated on average at higher frequency than CM mutations ., The fact that the signature of dBGC on HM motifs is weaker in Denisovan compared to human might be explained by slightly different sequence affinities of their PRDM9 alleles , or by a lower population frequency of HM-targeting PRDM9 alleles in Denisovans ., This weaker signature of dBGC might also be due to the fact that the effective population size was smaller in Denisovans compared to modern humans 33 , which is expected to enhance the effects of random genetic drift , and hence to decrease the strength of dBGC 32 ., Given that ancient DNA is prone to sequencing errors , we repeated our analyses with more stringent criteria to keep only data with the highest sequence quality ( filter F3 ) ., In that F3 subset , we found the same two-fold excess of HM losses compared to CM losses in the Denisovan branch ( Table S1 ) ., Overall , the three filters ( F1 , F2 or F3 ) lead to the same conclusion: there is a strong signal of dBGC on HM motifs in the terminal branches ( stronger in humans than in Denisovans ) , and a weak signal of dBGC in the Hominini branch ( Table S1 , S2 and Figure 1 ) ., In each of these three branches , the observed excess of HM losses relative to CM losses is the strongest with the most stringent F3 filter ( Table S1 , S2 ) ., However , given that the power of statistical tests decreases as the sample size decreases , the slight excess of HM losses in the Hominini branch is detected as statistically significant only in the F1 dataset ( Table S2 ) , and the signal of dBGC in the Denisovan branch becomes non-significant in the F3 dataset ( Table S1 ) ., All these observations indicate that HM has been subject to dBGC both in Denisovans and modern humans lineages , and suggest that HM started to be a target of PRDM9 shortly before the Denisovan/modern human split ., Given that the CM motif is not recombigenic ( Figure 2 ) , the shift in DAF spectra observed between CM and HM mutations ( Figure 3 ) , can entirely be attributed to dBGC acting on HM motifs ., To estimate the intensity of dBGC against HM motifs , we fitted a population genetic model to the DAF spectra of CM and HM mutations , considering CM mutations as neutral references ( see methods ) ., Since dBGC behaves like selection on semi-dominant mutations 32 , we used the model of Eyre-Walker et al . 34 to quantify it ., Under the simplifying assumption that all HM informative sites are subject to the same dBGC strength , the population scaled dBGC coefficient ( G\u200a=\u200a4Neg ) estimated on all HM motifs is 8 . 55 ( 95% confidence interval =\u200a2 . 76–2655 ) ., This result is robust to the number of categories used to describe DAF spectra ( Table S3 ) ., It should be noticed that large values of G are difficult to estimate accurately because above a given threshold ( G>20 ) , all values of this parameter are expected to give very similar DAF spectra ( Figure S2 ) ., This explains why the upper bound of the confidence interval of this estimate of G is very high ., The strength of dBGC at a given locus is proportional to the absolute difference in recombination rate between the original ( hot ) allele and the mutant ( colder ) allele 25 ., This difference can be large only if the recombination rate at this locus is high ., Hence HM motifs that are located at lowly recombining loci are not expected to undergo dBGC ., It is important to note that the recombination rate at HM motifs is highly variable across the genome: 8% of HM motifs concentrate 60% of all crossover events located in the vicinity of HM motifs ( ±2 kb ) ( Figure 4A ) ., It is therefore expected that the intensity of dBGC should be stronger for HM motifs located in a genomic context prone to recombination ., To test that prediction , we re-estimated G from DAF spectra , in three subsets of equal sample size , binned according to the local historical recombination rate ( measured on a 2 kb window centered on motif position ) ., As expected , G increases with increasing historical recombination rates , from G\u200a=\u200a0 . 96 in the first tercile to G\u200a=\u200a14 . 64 in the third tercile ( Table S4 ) ., To get a better picture of the distribution of G across all HM motifs , we fitted a simple model where the dBGC coefficient at a given locus is directly proportional to the local crossover rate at this locus ( Text S1 ) ., The distribution of G inferred by the model ( given the observed distribution of recombination rates around HM motifs ) , indicates a median value of G\u200a=\u200a57 . 5 for HM motifs located within historical hotspots ( Figure 4 ) ., The 8% most highly recombining motifs are predicted to be subject to very strong dBGC ( on average , G\u200a=\u200a174 , CI: 29–291 ) ., The dBGC model predicts that the small subset of HM motifs located in a highly recombining context should accumulate substitutions extremely rapidly ., In agreement with that prediction , we observed that , along the modern human branch , the loss rate is almost 3 times higher for HM motifs located within historical hotspots compared to other HM motifs ( 3 . 5% vs . 1 . 2%; p\u200a=\u200a4 . 6×10−7 ) ., Overall , 55% of the HM motifs detected as being mutated along the modern human branch are located within historical recombination hotspots ( compared to 28% for motifs that have remained intact ) ( Table 2 ) ., Thus , on average , the historical recombination rate at HM motifs mutated in the modern human branch is more than two times higher than that at intact HM motifs ( 11 . 2 cM/Mb vs . 4 . 9 cM/Mb; Figure 5 ) ., Notably , we observed the same pattern with present-day recombination rates , inferred from pedigree-based genetic maps 35 ( Figure S3 ) ., Moreover this pattern is observed even for the subset of HM mutations that are fixed in human populations ( Figure S4 ) ., These observations show that mutations of HM motifs that were fixed in modern humans are generally located in loci that still have a high recombination activity in present-day populations ., Hence , although mutations of HM motifs diminish the local recombination rate , they generally do not directly convert a hotspot into a coldspot ., Interestingly , HM motifs that are located outside of historical recombination hotspots also show a signature of dBGC ., This signature is weaker than for HM located within hotspots , but still clearly significant: there is a 3-fold excess of HM losses compared to CM losses in the modern human branch ( Table 2 ) , and HM mutations segregate at higher frequencies than CM mutations ( 10% vs . 4%; p\u200a=\u200a0 . 0014 ) ., The analysis of present-day recombination rates confirmed the absence of recombination hotspots at these mutated HM sites ( Figure S5 ) ., This suggests that these HM losses occurred in ancient recombination hotspots that are not active anymore ., Overall , we detected 78 HM losses along the modern human branch , whereas only 19 would have been expected if the loss rate were the same as that of CM motifs ., Among these 59 extra losses that can be attributed to dBGC , 23 occurred at loci that are not detected as recombination hotspots ( Table 2 ) ., Thus , among all loci that used to be recombination hotspots in the human lineage and that have lost the HM motif by dBGC , 39% are no longer active ., With only one single individual sequenced , it is not possible to establish recombination maps in Denisovans ., However , different analyses can be performed to test whether recombination hotspots identified in modern human populations correspond to hotspots in Denisovans ., A first approach to detect past recombination activity consists in analyzing substitution patterns , so as to infer the equilibrium GC-content ( denoted GC* ) along different branches of the phylogeny ( see methods ) ., Many lines of evidence indicate that in primates , recombination is driving the evolution of GC-content via the process of GC-biased gene conversion ( gBGC ) , which results from a bias in the repair of AT:GC mismatches in heteroduplex DNA during meiotic recombination 36 , 37 ., Notably , it has been shown that GC* strongly correlates with present or past recombination rates 38–40 ., We therefore measured GC* separately for each branch of the phylogeny at loci corresponding to the 32 , 981 human historical recombination hotspots 2 ., As expected , we observed a strong peak of GC* centered on the middle of historical recombination hotspots , in the modern human branch ( Figure 6D ) ., In agreement with previous results 19 , this peak is absent in the chimpanzee branch ( Figure 6A ) , consistent with the fact that human and chimpanzee recombination hotspots do not overlap ., Interestingly , we observed only a very limited bump of GC* in the Hominini branch ( Figure 6B ) ., This indicates that , up to a recent time , shortly before the Denisovan/modern human split , loci corresponding to human historical recombination hotspots were not subject to gBGC ., Surprisingly , we observed no peak of GC* in the Denisovan branch ( Figure 6C ) ., This result was unexpected: given our observations indicating that the HM motif started to be a target of PRDM9 before the split between modern humans and Denisovans , we presumed , a priori , that the two populations should share the same recombination hotspots ., We first hypothesized that the absence of peak of GC* could be due to the fact that , owing to the relatively low effective population size in Denisovan , gBGC was too weak to leave any detectable signature ., To test this hypothesis , we investigated whether we could detect the hallmarks of gBGC in Denisovan , by analyzing correlations between GC* ( inferred along different branches of the phylogeny ) and recombination rates , measured in 1 Mb-windows ., At this genomic scale , recombination rates are well conserved between human and chimpanzee 19 and hence are expected to be also conserved in Denisovan ., As predicted by the gBGC model , and in agreement with previous results 38–40 , we observed a significant correlation between human historical recombination rates and GC* along the modern human branch ( R2\u200a=\u200a13%; p<10−74 ) ., This correlation is as strong for GC* computed in the Denisovan branch ( R2\u200a=\u200a14%; p<10−80; Figure S6 ) , which indicates that , genome-wide , the signature of gBGC is as visible in Denisovan as it is in human ., Thus , the absence of peak of GC* in the Denisovan branch at human recombination hotspots loci cannot be attributed to a possibly weaker gBGC effect in Denisovan ., Instead , it indicates that recombination hotspots were not shared between humans and Denisovans ., To further test this conclusion , we used an independent approach ., The self-destructive drive model predicts that HM motifs located in recombination hotspots should be subject to stronger dBGC than other HM motifs ., Thus , if the location of recombination hotspots was conserved , then HM motifs located in loci corresponding to human recombination hotspots should show an enhanced signature of dBGC not only in human ( as shown previously ) , but also in the Denisovan branch ., As already mentioned , we observed an excess of HM losses compared to CM losses in Denisovan ( Figure 1 ) , which indicates that there is a detectable signature of dBGC on HM in Denisovan ., However , the HM loss rate is not different between HM loci that correspond to human historical hotspots and other HM loci ( respectively 0 . 8% and 1 . 1% , p\u200a=\u200a0 . 579; Table 2 ) ., Thus , we see no evidence for stronger dBGC in Denisovan at the location of human historical hotspots ., Finally , it has been shown that HM motifs located within THE1 transposable elements are particularly prone to recombination in humans 10 , 17 ., As expected , we observed a markedly elevated HM loss rate within THE1 elements in human ( 6 . 7% ) ., In contrast , we did not detect any mutation in the Denisovan branch among HM motifs located in THE1 ( loss rate =\u200a0%; proportion test: p\u200a=\u200a0 . 0067; Table 1 ) ., This suggests that contrarily to human , HM motifs located within THE1 elements were not associated to elevated recombination rates in Denisovan ., All these observations concur to the conclusion that fine-scale recombination rates were not conserved between Denisovans and humans ., This therefore suggests that the A allele of PRDM9 was either absent or present at very low frequency in Denisovans ., The major PRDM9 allele in Denisovans was probably similar to the A allele ( since it also had affinity for the HM motif ) , but it targeted recombination hotspots to a different subset of HM motifs ., The lifetime of HM motifs can be predicted using standard population genetic approximation 25 ., Under the simplifying assumption that motif mutations are immediately either lost or fixed in the population , the probability that a hotspot motif accumulates at least one disrupting substitution after T generations , can be approximated by:where µ is the mutation rate ( 1 . 2×10−8 mutations/bp/generation in humans 41 ) , k the length of the motif ( here k\u200a=\u200a11 for the HM motif ) and G the population-scaled dBGC coefficient ( G\u200a=\u200a4Neg ) 25 ., Given the distribution of G estimated previously , this model predicts that after 100 , 000 generations ( i . e . about 3 MYR 29 ) , overall , 18% of HM motifs should be lost ( CI: 5%–23% ) ., But importantly , for the subset of most highly-recombining HM motifs ( top 8% of HM motifs , which concentrate 60% of HM-associated recombination events ) , the predicted loss rate is extremely high ( 87%; CI: 32%–96% ) ., Thus the model indicates that if the dBGC drive against HM remains as strong as it is in extant human populations , then the subset of highly-recombining HM motifs should be rapidly lost ., We observed that in many cases , the loss of HM does not totally abolish the hotspot activity ., This is most probably due to the fact that the affinity of PRDM9 depends not only on the HM motif , but also on interactions with other sites in its vicinity ., However , our observations indicate that losses of a HM motif by dBGC in the human branch were associated with hotspot extinction in 39% of cases ., Given that the human branch is relatively short ( 14 , 000–28 , 000 generations ) , this suggests that within the next 100 , 000 generations , the loss of HM motifs should be accompanied by the loss of recombination hotspots activity ., PRDM9 is the major determinant of the location of recombination hotspots in humans and mice 9–12 , 15 , 16 , 35 , 42 ., At the population scale , the chromosomal distribution of recombination events is therefore expected to depend on the allelic composition at the PRDM9 locus ., The location of human historical recombination hotspots reflects the DNA binding specificity of the A allele 9 ., This allele is present at high frequency both in African and European populations , and as expected , most historical recombination hotspots are shared between these populations 16 ., This implies that the majority of human historical recombination hotspots are older than 50 , 000 years ., To determine more precisely the age of historical hotspots ( i . e . to determine when the A allele started to reach substantial frequency within populations ) , we searched for signatures of recombination hotspot activity by analyzing patterns of sequence evolution across different branches of the phylogeny , before and after the divergence between modern humans and Denisovans ., We used the fact that when a locus is recombining at a high rate ( at the population scale ) , it then becomes subject to two forms of BGC: BGC in favor of mutations disrupting PRDM9 target motifs ( dBGC ) , and BGC in favor of GC-alleles ( gBGC ) ., Along the modern human branch , we observed clear signatures of dBGC against HM motifs: these motifs accumulated an excess of mutations , which tend to segregate at higher allelic frequencies ., Moreover , we showed that the strength of this fixation bias in favor of HM-disrupting mutations increases with increasing local recombination rate ., All these observations are perfectly consistent with the fact that HM is targeted by the major allele of PRDM9 in human populations ( allele A ) ., Interestingly , we also observed an excess of HM losses along the Denisovan branch , which suggests that HM started to be a target of PRDM9 before the population split between Denisovans and modern humans ., However , several independent lines of evidence indicate that recombination hotspots were not shared between Denisovans and modern humans: in Denisovan , at loci corresponding to human recombination hotspots , we observed no signature of gBGC and no evidence of stronger dBGC against HM motifs ., Moreover , in Denisovan , contrarily to human , HM motifs located within THE1 elements are not subject to accelerated loss ., The fact that fine-scale recombination rates were not conserved between humans and Denisovans might a priori seem in contradiction with the observation that the same motif ( HM ) was subject to accelerated loss in both lineages ., However , the affinity of PRDM9 to its targets is not determined by this 13-bp motif alone , but also depends on interactions with surrounding sites 17 , 43 ., For example , in human , the HM motif is much more prone to recombination when located within the context of THE1 17 ., Overall , among the 6 , 671 HM motifs found in human autosomes , only 1 , 358 ( 20% ) overlap with one of the 32 , 987 recombination hotspots identified by HapMap 2 ., Thus , only a subset of HM motifs in the human genome are in a context for which the A allele of PRDM9 presents a high affinity ., It is therefore possible that the major PRDM9 allele ( s ) present in Denisovan populations had affinity to HM , but within a different context ., In summary , the fact that Denisovans and humans had different hotspots but similar target motifs suggests that they had slightly different PRDM9 alleles , with distinct context specificity ., This conclusion is compatible with two scenario:, i ) the A allele of PRDM9 was already present at a substantial frequency in the ancestral population ( before the population split between Denisovans and modern humans ) , but was lost ( or present at very low frequency ) in the Denisovan lineage or, ii ) the A allele increased in frequency specifically in the modern human branch ., Schwartz and colleagues 44 recently reported the partial genotyping of PRDM9 alleles present in the genomes of two archaic humans ( the Denisovan genome analyzed here and that of an Altai Neandertal ) ., In the Denisovan individual , they found evidence for the presence of an allele with a Zn-finger array composition different from that of the A allele , but compatible with rare PRDM9 alleles found in African populations 44 ., By analyzing the copy number of Zn-finger repeat units , we further show that in fact the genotype of this Denisovan individual does not correspond to any known human PRDM9 allele ( Text S2 , Table S5 , Figure S7 , Datase | Introduction, Results, Discussion, Methods | Recombination is an essential process in eukaryotes , which increases diversity by disrupting genetic linkage between loci and ensures the proper segregation of chromosomes during meiosis ., In the human genome , recombination events are clustered in hotspots , whose location is determined by the PRDM9 protein ., There is evidence that the location of hotspots evolves rapidly , as a consequence of changes in PRDM9 DNA-binding domain ., However , the reasons for these changes and the rate at which they occur are not known ., In this study , we investigated the evolution of human hotspot loci and of PRDM9 target motifs , both in modern and archaic human lineages ( Denisovan ) to quantify the dynamic of hotspot turnover during the recent period of human evolution ., We show that present-day human hotspots are young: they have been active only during the last 10% of the time since the divergence from chimpanzee , starting to be operating shortly before the split between Denisovans and modern humans ., Surprisingly , however , our analyses indicate that Denisovan recombination hotspots did not overlap with modern human ones , despite sharing similar PRDM9 target motifs ., We further show that high-affinity PRDM9 target motifs are subject to a strong self-destructive drive , known as biased gene conversion ( BGC ) , which should lead to the loss of the majority of them in the next 3 MYR ., This depletion of PRDM9 genomic targets is expected to decrease fitness , and thereby to favor new PRDM9 alleles binding different motifs ., Our refined estimates of the age and life expectancy of human hotspots provide empirical evidence in support of the Red Queen hypothesis of recombination hotspots evolution . | In eukaryotic genomes , recombination plays a central role by ensuring the proper segregation of chromosomes during meiosis and increasing genetic diversity at the population scale ., Recombination events are not uniformly distributed along chromosomes , but cluster in narrow regions called hotspots ., The absence of overlap between human and chimpanzee hotspots indicates that the location of these hotspots evolves rapidly ., However , the reasons for this rapid dynamic are still unknown ., To gain insight into the processes driving the evolution of recombination hotspots we analyzed the recent history of human hotspots , using the genome of a closely related archaic hominid , Denisovan ., We searched for genomic signatures of past recombination activity and compared them to present-day patterns of recombination in humans ., Our results show that human hotspots are younger than previously thought and that they are not conserved in Denisovans ., Moreover , we confirm that hotspots are subject to a self-destruction process , due to biased gene conversion ., We quantified this process , and showed that its intensity is strong enough to cause the fast turnover of human hotspots . | biochemistry, genomics, genetic polymorphism, genome evolution, human genomics, genetics, biology and life sciences, dna, population genetics, dna recombination, evolutionary biology, molecular evolution, computational biology, homologous recombination | null |
journal.pbio.2006288 | 2,019 | Recombination rate variation shapes barriers to introgression across butterfly genomes | The genealogical relationships among closely related species can be complex , varying across the genome and among individuals ., This phylogenetic heterogeneity can be caused both by incomplete lineage sorting ( ILS ) in ancestral populations and by introgressive hybridisation , causing some parts of the genome to have genealogies that are discordant with the species branching pattern or ‘species tree’ ., Genome-scale studies have revealed that particular genomic regions such as sex chromosomes and chromosomal inversions can have distinct phylogenetic histories 1–3 , possibly reflecting systematic differences in the extent of introgression across the genome ., Indeed , the establishment of barriers to introgression in certain parts of the genome is a key part of the speciation process 4–8 ., The heterogeneous landscape of species relationships can therefore carry information about the ‘barrier loci’ that contribute to the origin and maintenance of species ., Barrier loci can be associated with extrinsic ( imposed by the environment ) or intrinsic ( affecting viability of fertility ) selective pressures , and can act at both prezygotic or postzygotic levels 6 , 9 ., The barriers between closely related subspecies or ecotypes that interbreed frequently are often restricted to just a few loci that contribute to local adaptation , resulting in narrow ‘islands’ of genetic differentiation between populations 1 , 10–12 ., As speciation proceeds , we expect an accumulation of barrier loci , leading to reduced gene flow and more widespread genetic differentiation across the genome 8 , 13 ., Recently , it has become evident that patterns of genomic differentiation between more strongly isolated species are often complex and reflect not only barriers to introgression but also within-species processes that cause variation in effective population size ( Ne ) across the genome , including localised selective sweeps and background selection 14–16 ., Relative measures of genetic differentiation , such as the fixation index ( FST ) , which are sensitive to variation in Ne , therefore provide a poor proxy for the strength of a local barrier 14 , 15 , 17 ., However , it is possible to largely avoid the confounding effects of positive and background selection by using methods that intrinsically account for heterogeneity in Ne and directly estimate the ‘effective migration rate’ or the level of admixture and how it varies across the genome , either using summary statistics 18 or through model-based inference 19 ., Provided that there has been sufficient introgression between the species , regions of the genome in which admixture is reduced can be inferred to have experienced selection against foreign genetic variation ., If species barriers are highly polygenic ( made up of many loci ) 20 and each locus has only a weak effect on fitness , their individual localised effects on levels of admixture might be difficult to detect , analogous to the difficulties in studying polygenic adaptation more generally 21–23 ., Whereas it may not be possible to identify all barrier loci in such a situation , we can test hypotheses about the architecture of barriers by studying genome-wide patterns of admixture ., In particular , barriers made up of many loci of small effect are expected to be more porous to introgression where recombination rates are higher ., Foreign chromosomes that enter a population through hybridisation and backcrossing will be more rapidly broken down over subsequent generations in regions with higher recombination rates ., This will tend to separate clusters of foreign deleterious alleles , reducing selection against them , and also break their linkage with neutral ( or mutually beneficial ) foreign alleles at other loci , allowing these to avoid removal by selection 19 , 24–27 ., A correlation between the recombination rate and the inferred rate of effective migration has been observed between subspecies of house mice 28 , subspecies of Mimulus monkeyflowers 19 , 29 , in hybrid populations of swordtail fishes 30 , and even between humans and Neanderthals 30 , 31 , suggesting that loci experiencing selection against introgression among close relatives can be widespread in the genome ., Therefore , a combination of extensive hybridisation and polygenic barrier loci could theoretically produce predictable large-scale heterogeneity in phylogenetic relationships across the genome , with regions of higher recombination rate showing greater discordance with the species tree ., Discordance caused simply by ILS in ancestral populations is also expected to be elevated in regions of higher recombination rate , due to their typically larger Ne 32 ., However , the effects of introgression should be distinguishable in that particular discordant topologies—those that group hybridising species pairs—should be overrepresented ., We explored species relationships and barriers to introgression among species of Heliconius butterflies ., Many Heliconius species are divided into geographically distinct ‘races’ with distinct warning patterns , which signal their distastefulness to local predators ., Selection favouring locally recognised warning patterns maintains narrow islands of divergence at a few wing-patterning loci between otherwise genetically similar races 1 , 10 , 33 ., However , there are also more strongly differentiated pairs of sympatric species that hybridise rarely and have strong postzygotic barriers , leading to higher genome-wide genetic differentiation 1 ., We studied three such species—H ., melpomene ( ‘mel’ ) , H . cydno ( ‘cyd’ ) , and H . timareta ( ‘tim’ ) —which form at least two independent zones of sympatry separated by the Andes mountains ., Whereas mel is found throughout much of South and Central America , cyd is largely restricted to the west of the Andes and the inter-Andean valleys , where it overlaps with the western populations of mel , and tim occurs only on the eastern slopes of the Andes , where it co-occurs with the eastern populations of mel ., In addition to strong assortative mating based on chemical cues , along with visual cues in the case of cyd and mel 34–39 , both species pairs show ecological differences as well as partial hybrid sterility 36 , 37 , 40–44 ( and see 36 for a review ) ., Nevertheless , previous studies have revealed surprisingly pervasive admixture between these species in sympatry , most likely explained by a low rate of ongoing hybridisation over an extended period of time 1 , 45 , 46 ., There is also considerable heterogeneity in the relationships among these populations across the genome 1 ., Adaptive introgression in Heliconius is well documented ., Mimicry between sympatric races of mel and tim has been facilitated by exchange of multiple wing-patterning alleles 47 , 48 , and at least one case of introgression between mel and cyd has allowed the latter to mimic other unpalatable species 49 ., However , the extent to which introgression among these species might be selected against remains unclear ., Using 92 whole-genome sequences , we asked whether the heterogeneous relationships observed among these species reflect the influence of polygenic barriers to introgression that vary in their strength across the genome ., Then , taking advantage of high-resolution linkage maps for these species 50 , we show that admixture is correlated with recombination rate , consistent with polygenic species barriers leading to widespread selection against introgression ., This selection also explains broader variation in admixture at the chromosomal scale ., Overall , our results highlight the pervasive role of natural selection in shaping the ancestry of hybridising species ., We analysed whole-genome sequence data from 92 butterflies representing nine populations from the three focal species , mel ( 5 populations or ‘races’ , 10 individuals each ) , cyd ( two races , 10 individuals each ) , and tim ( two races , 10 individuals each ) , along with two individuals from an outgroup species H . numata ( ‘num’ ) ( S1 Table ) ., Our sampling included four regions of sympatry: two on the west of the Andes where cyd co-occurs with western races of mel ( hereafter mel-W ) and two on the eastern slopes of the Andes where tim co-occurs with eastern races of mel ( hereafter mel-E ) , as well as an allopatric population from French Guiana ( hereafter mel-G ) ( Fig 1A ) ., Principal components analysis ( PCA ) and a phylogenetic network based on whole-genome single-nucleotide polymorphism ( SNP ) data show clear distinctions between the three species , as well as between mel-W , mel-E , and mel-G ( Fig 1B ) ., By contrast , pairs of races of the same species from the same broad geographic area ( i . e . , west of the Andes , east of the Andes , or French Guiana ) are not clearly distinct in the PCA , indicating nearly panmictic populations in each species in each area , despite variation at a few wing-patterning loci , as shown previously 51 , 52 ., The tight clustering and lack of intermediate individuals in the PCA indicates that none of the sampled individuals result from recent hybridisation , consistent with observations that hybridisation is very rare on a per-individual basis ., However , large reticulations in the network are consistent with extensive introgression , which has probably occurred gradually through rare hybridisation events spread across millions of generations 1 ., These results therefore highlight the contrast between the strong barriers that exist between species—even in sympatry—and the continuity that exists within species , with the Andes mountains and wide Amazon basin presenting the only major sources of discontinuity among sampled populations of the same species 51 , 52 ., We explored species relationships across the genome using Twisst 56 , which quantifies the frequency ( or ‘weighting’ ) of alternative topological relationships among all sampled individuals in narrow windows of 50 SNPs each ., Consistent with previous results , topology weighting indicates that large-scale introgression has shaped the relationships among these species ., Examples of local genealogies and their corresponding topology weightings are shown in S1 Fig . All 15 possible rooted topologies that describe the relationship between cyd , tim , mel-W , and mel-E ( rooted with num as the outgroup ) are represented at considerable levels across the genome ( Fig 2 ) ., Moreover , fewer than 0 . 5% of windows have completely sorted genealogies ( i . e . , all groups cluster according to a single topology , resulting in a weighting of 1 , see S1 Fig for an example ) ( Fig 2C ) ., Coalescent simulations using an appropriate split time ( approximately 1 . 5 million years ago 45 , 46 , 57 ) and population size ( 2 million 52 ) show that , in the absence of introgression , we would expect far less phylogenetic discordance and more complete lineage sorting ( more windows with a weighting of 1 for a single topology ) than seen here , unless the population sizes were much larger ( S2 Fig ) ., However , the addition of moderate gene flow between sympatric pairs produces levels of discordance and lineage sorting similar to our empirical results ( S2 Fig ) ., The two most common topologies across the genome are T3 and T6 , which differ entirely in the relationships among the ingroup taxa ( Fig 2A ) ., T3 matches the expected species branching order , in which cyd and tim are sister species and mel-W groups with mel-E ( cyd , tim , mel-W , mel-E ) 51 , 58 ., We refer to this as the ‘species topology’ ., T6 , by contrast , groups populations by geography: cyd with mel-W , and tim with mel-E ( cyd , mel-W , tim , mel-E ) ., We refer to this as the ‘geography topology’ ., We therefore hypothesise that the history of these species can be modelled as a branching process following the species topology , with considerable introgressive hybridisation beginning at some point after the species diverged that increases the rate of coalescence between sympatric populations from distinct species , as in the geography topology ., Although the geography topology has a slightly higher average weighting , the species topology occurs far more frequently with a weighting of 1 than any other topology ( Fig 2C ) ., In other words , the only pairs of populations that consistently show complete monophyletic clustering are mel-W with mel-E and cyd with tim ( for an example , see S1 Fig ) ., Our simulations agree that , although extensive introgression can also produce high rates of monophyly , this level of monophyly between allopatric populations is only expected if they are sister taxa ( S2 Fig ) , thus supporting T3 as the true species branching pattern ., The species topology also agrees with ecological trends such as host plant use , as well as both larval and adult morphology , which support a sister relationship of cyd and tim 59 , 60 ., Topology weightings for the Z chromosome are dramatically different from the genome-wide averages ( Fig 2B ) ., There is much less discordance , and the species topology has by far the highest weighting ., Moreover , the geography topology and others consistent with introgression have comparatively low weightings ., The contrasting abundance of discordant topologies on the autosomes could be explained not only by much higher levels of introgression affecting the autosomes , as shown previously 1 , but also by their larger Ne , and resulting slower rate of lineage sorting compared to the Z . However , as shown by our simulations ( S2 Fig ) , only extensive introgression can explain the strong skews in the topology weightings on the autosomes—with the geography topology ( T6 ) being the most abundant—whereas others such as T9 , which groups allopatric nonsister taxa , is among the least abundant genome wide ., Similarly , the third and fourth most highly weighted topologies across autosomes ( T14 and T5 ) both group one pair of sympatric taxa ( mel-E with tim and mel-W with cyd , respectively ) and otherwise match the species topology ( Fig 2B ) ., Taken together , these patterns are all consistent with extensive postspeciation introgression between sympatric pairs , with a strong reduction in introgression on the Z chromosome ., Our sampling design allows us to make inferences about biases in the direction of gene flow ., T14—which has tim nested within the mel clade , suggesting introgression predominantly from mel-E into tim ( Fig 2B ) —is the third most abundant topology genome wide and has a higher weighting than T11 , which implies gene flow in the opposite direction ., A bias towards introgression into tim would be expected given the much smaller range and lower Ne of tim , which also has the lowest nucleotide diversity of all the taxa studied 1 ., Hybrids that backcross into tim will therefore provide a larger relative contribution to the gene pool than those that backcross into mel ., Likewise , T5 is more abundant than T4 , suggesting that most introgression in the west of the Andes has been from cyd into mel-W ( Fig 2B ) ., This direction was also inferred to be the most likely in a previous study using coalescent modelling 46 and is consistent with the fact that F1 hybrids show mate preference for mel in experiments with Panama populations 34 ., Our simulations , in which the direction of introgression was biased in both cases , also show similar imbalances in the weightings for the same topologies ., In addition to the difference between autosomes and the Z chromosome , topology weightings vary considerably among and within the autosomes ., To highlight this heterogeneity , we first focus on the two most abundant patterns of relatedness: the species and geography topologies ( Fig 3A ) ., The species topology has the highest weighting in narrow peaks on some of the autosomes , whereas elsewhere the geography topology has the higher weighting ., In other words , throughout large parts of the genome , samples of mel-W and mel-E tend to be more closely related to their respective sympatric counterparts , cyd and tim , than to one another ., However , the species topology tends to occur in sharp peaks , which frequently have a weighting approaching 1 , as discussed above ( S3 Fig; note that these narrow peaks are not visible in Fig 3A due to smoothing ) ., There are strong trends in the abundance of the species and geography topologies across the 20 autosomes ., All species in this clade have 10 short and 10 long autosomes ., The latter formed through 10 independent fusions in the ancestor of Heliconius , which had 30 autosomes 61 , 62 ., The species topology is less abundant on the 10 short autosomes compared to the 10 long , fused autosomes , and there is a fairly linear increase in its weighting with chromosome length ( Fig 3B ) ., By contrast , the geography topology shows decreasing abundance with chromosome length and tends to be far more abundant on the short chromosomes ., There is also a fairly consistent within-chromosome trend , with higher weightings for the geography topology and lower weightings for the species topology towards the outer third of the chromosomes compared to the chromosome centres ( Fig 3C ) ., This reverses in the outer 5% of chromosomes , where the species topology is again strongly supported ., The above trends might be partly explained if ILS in ancestral populations was more common on short chromosomes and away from the centres and very ends of chromosomes , leading to increased discordance in these regions ., Indeed , we previously found a negative correlation between chromosome length and Ne in mel 52 ., However , the trends shown by other topologies suggest that these patterns also reflect variation in the extent of introgression between genomic regions ., The 5 topologies that group allopatric pairs and therefore likely reflect discordance due to ancestral ILS alone ( i . e . , T7 , T8 , T9 , T12 , and T13 ) show only weak relationships with chromosome length , and no clear pattern within chromosomes ( S4 Fig ) ., Furthermore , T14 , which is consistent with introgression between mel-E and tim , is less abundant than the species topology ( T3 ) on long chromosomes and at chromosome centres but is more abundant on short chromosomes and in chromosome peripheries ( S4 Fig ) ., Such a switch in rank is not expected if the short chromosomes and peripheries simply experience more ancestral ILS , but it is consistent with differences in the extent of introgression between short and long chromosomes and between centres and peripheries ., T5 , which is consistent with introgression between cyd and mel-W , does not show any clear relationship with chromosome length or relative chromosome position ( S4 Fig ) ., This implies that there may be less consistent variation in the extent of introgression between cyd and mel-W in different regions ., However , T4 , which reflects introgression between the same pair but in the opposite direction ( from mel-W into cyd ) does show weak patterns similar to the geography topology ., Overall , topology weighting reveals quantitative variation in species relationships both within and among chromosomes that are consistent with heterogeneity in the level of introgression across the genome ., However , topology weighting does not explicitly distinguish between introgression and shared ancestral variation ., We therefore set out to explicitly test the hypotheses that ( 1 ) there is heterogeneity in the level of admixture across the genome and ( 2 ) that this heterogeneity can be explained by variation in the strength of selection against introgression ., We used the summary statistic fd 18 to quantify admixture separately between cyd and mel-W and between tim and mel-E ., This approach also measures an excess of genealogical clustering of sympatric nonsister taxa ., However , fd provides a normalised measure that is approximately proportional to the effective migration rate 18 ., Building on previous work , we first investigated the degree to which fd might be influenced by variation in Ne across the genome ., Ne tends to be reduced in regions of reduced recombination rate due to linked selection ., By means of simulations , we find that , across a large range of realistic population sizes , fd is a reliable estimator of admixture ., Furthermore , fd outperforms the commonly used divergence statistics FST and dXY , which are both highly sensitive to Ne ( S5 Fig ) ., When population sizes are very large , fd tends to underestimate the true level of admixture ., This is caused by a loss of information when population sizes are large relative to the split times: the lack of lineage sorting means that there is insufficient information available to accurately quantify admixture ., The population sizes for which this is relevant are at the upper end of estimates for these species 52 ., Moreover , this error would cause a conservative bias in our results , as we expect reduced admixture in low-recombination regions , where Ne is expected to be the smallest ., Most important for our subsequent analysis , high background selection in regions of low recombination—which is known to influence measures such as FST—is not likely to strongly bias our estimates using fd ., We therefore conclude that fd provides a suitable , albeit conservative , measure to test the hypothesis that species barriers are enhanced in regions of reduced recombination rate ., Computation of fd requires the use of a ‘control’ population that is ideally allopatric and unaffected by introgression ., To confirm the robustness of our results , we computed fd with several different sets of populations , varying the control population , as well as splitting or joining each of cyd , tim , mel-W , and mel-E into their two constituent subpopulations ( S6 Fig ) ., Patterns of admixture estimated by fd show considerable heterogeneity across the genome ( Figs 4A and S7 and S8 ) ., As expected , admixture is minimal across the Z chromosome in both pairs , indicating a strong barrier to introgression ., There is also heterogeneity in admixture proportion across the autosomes ., This is most striking between tim and mel-E , where some regions exhibit deep troughs , implying strong , localised species barriers ., Some of this heterogeneity likely reflects individual barrier loci of large effect ., Indeed , the known wing-patterning loci provide a useful example ., The pattern differences between cyd and mel-W are determined by regulatory modules around 3 major genes: wnt-A ( Chromosome 10 ) , cortex ( Chromosome 15 ) , and optix ( Chromosome 18 ) 49 , 63–67 ., These probably act as strong barriers to introgression between cyd and mel-W , due to increased predation against hybrids with intermediate wing patterns 43 ., By contrast , the shared wing patterns of tim and mel-E are thought to result from adaptive introgression of wing-patterning alleles ., As expected , there is a strong reduction in admixture between cyd and mel-W in the vicinity of all 3 genes ( S7 Fig ) , while there are peaks of admixture between the comimetic tim and mel-E populations in the corresponding regions ( S8 Fig ) ., We hypothesised that many loci across the genome contribute to the species barriers , which leads to the expectation that the level of admixture will be correlated with the recombination rate 19 ., We quantified variation in recombination rate across the genome using high-resolution linkage maps ( based on 963 offspring in total ) 50 as well as using LDHelmet 68 , which estimates the population recombination rate ( ρ ) based on linkage disequilibrium ( LD ) in the genomic data from natural populations ., On a broad scale , the map-based estimates are highly concordant with the population-based estimates , and the latter are also strongly conserved across the different species ( Figs 4B and S9 ) ., There is considerable variation in recombination rate across the genome , allowing us to investigate whether admixture proportions are correlated with recombination rate ., There is a strong positive and nonlinear relationship between admixture proportion and recombination rate in both species pairs ( Figs 4C and S11 ) ., Correlations between fd and both the ρ and crossover recombination rate are highly significant , even after thinning windows to reduce effects of serial correlation ( S2 Table ) ., Strong reductions in admixture , implying barriers to introgression , are concentrated in genomic regions in which recombination rates are below 2 cM/Mb ., However , there is also more variability in admixture proportions in these low-recombination regions , with some showing high estimated levels of admixture ( Fig 4C ) ., This might imply that some regions do not harbour loci that contribute to the species barrier , although the variance in admixture proportions may also be increased in low-recombination regions due to increased genetic drift resulting from enhanced linked selection ., The correlations persist when considering only regions of intermediate recombination rate ( 2–8 cM/Mb ) ( S2 Table ) ., The relationship between admixture and recombination rate is stronger in the tim and mel-E pair , implying that the more heterogeneous pattern of admixture across the genome between this pair is more consistent with a model in which barrier loci are widespread and recombination rate modulates the strength of the barrier to introgression ., Admixture proportions are less well predicted by the map-based estimates of crossover recombination rate compared to the inferred ρ ( Fig 4C and S2 Table ) ., This probably partly reflects inaccuracy in fine-scale recombination rate estimated from the linkage maps ., However , it may also be that ρ ( = 4Ner ) provides a more meaningful predictor for the admixture proportion , as it is a composite of the per-generation recombination rate ( r ) and local Ne ., Due to linked selection , parts of the genome with a low recombination rate and a high density of selected sites are expected to have locally reduced Ne and therefore reduced ρ ., Indeed , ρ is strongly negatively correlated with the proportion of coding sequence per window ( referred to as ‘gene density’ hereafter ) ( Spearman’s rho = −0 . 681; p < 0 . 001; S10 Fig ) ., However , there is also a weaker but significant negative relationship between gene density and the crossover recombination rate ( Spearman’s rho = −0 . 248; p < 0 . 0001; S10 Fig ) ., This implies that linked selection in regions of low recombination rate may be further enhanced by a higher density of selected loci ., As the conditions that enhance linked selection are the same as those expected to strengthen barriers to introgression ( i . e . , a high ratio of selected loci relative to the recombination rate , also called the ‘selection density’ 19 ) , it is to be expected that ρ would provide a better predictor of barrier strength and therefore admixture proportion ., As expected , there is a negative relationship between admixture proportion and gene density ( S12 Fig ) ., However , the fact that regions with a high gene density also tend to have lower recombination rates makes it difficult to determine whether such regions harbour a higher physical density of barrier loci , but this seems likely given the arguments above ., The correlation between admixture and recombination rate remains clear when individual chromosomes are considered separately , with 18 and 20 of the 21 chromosomes showing a significant correlation for mel-W/cyd and mel-E/tim , respectively ( S13 Fig ) ., This further highlights the genome-wide nature of this trend ., The above trends are also robust to using different allopatric ‘control’ populations when estimating admixture proportions ( S11 and S12 Figs ) , with the exception that using very closely related control populations leads to very low estimated rates of admixture , for which the relationships with recombination rate and gene density are not clear due to the large number of windows at which the estimated admixture proportion is 0 ., See S11 and S12 Figs for details ., Average chromosomal admixture proportions are negatively correlated with chromosome length ( Fig 5A and 5B ) ., This is expected given the extremely strong negative correlation between physical chromosome length ( in base pairs ) and average recombination rate ( Spearman’s rho = −0 . 98; p = 4 . 6 × 10−6 ) ( Fig 5C ) ., By contrast , there is no clear relationship between chromosome length and gene density ( Spearman’s rho = 0 . 058; p = 0 . 8 ) ( Fig 5D ) ., The broadly enhanced barrier to introgression on long chromosomes is therefore more consistent with an effect of increased linkage , rather than an increased density of barrier loci ., As in the trends above , the correlation with chromosome length is stronger for admixture between tim and mel-E ( Spearman’s rho = −0 . 66; p = 7 ×10−4 ) than for admixture between cyd and mel-W ( Spearman’s rho = −0 . 51; p = 0 . 012 ) ., Between tim and mel-E , the shortest chromosomes experience about 50% more admixture than the longest chromosomes , with the exception of Chromosome 2 , which has strongly reduced admixture compared to other short chromosomes with similarly high recombination rates ., This might reflect a higher density of barrier loci on this chromosome , which seems possible because it also has the highest gene density of all chromosomes ( Fig 5D ) ., Given that the 10 longer chromosomes all arose through fusions in the ancestor of the genus 62 , we also investigated whether there was some additional effect of the fusions themselves in causing reduced admixture , apart from the obvious indirect effect of lower average recombination rates on longer chromosomes ., There is no consistent trend of reduced admixture in the vicinity of the fusion points themselves ( S7 and S8 Figs ) ., We tested for a general difference between fused and unfused chromosomes independent of recombination rate by binning 100 kb windows by local recombination rate and then comparing bins of equivalent recombination rates between fused and unfused chromosomes ( S13 Fig ) ., In both species pairs , most bins showed significantly reduced admixture on the fused chromosomes ., This suggests that their reduced recombination rate alone may be insufficient to explain the extent of reduction in admixture , implying a higher density of barrier loci on fused chromosomes , despite no significant difference in gene density ( Fig 5D ) ., However , the same pattern would be expected if admixture proportions in 100 kb windows are influenced by recombination rates in surrounding regions , so it is difficult to distinguish the effect of lower global recombination rates on longer chromosomes from a difference in the density of barrier loci ., As indicated by topology weighting , there is an effect of position along the chromosome on the proportion of admixture between mel-E and tim , where admixture increases on average towards the distal region of the chromosome but decreases again at chromosome ends ( Fig 5F ) ., This is not seen in the proportion of admixture between cyd and mel-W ( Fig 5E ) , which appears to show a weak decline moving away from the chromosome centre and a sharper decline at the end ., Unlike in many other taxa 27 , there is no consistent decrease in recombination rate towards chromosome centres ., By contrast , both the crossover recombination rate and ρ show a sharp decrease at the chromosome ends ( Fig 5G ) ., Gene density is roughly uniform across chromosomes on average ( Fig 5H ) ., Theory predicts that , given a uniform recombination rate and distribution of selected loci , species barriers should weaken towards chromosome ends due to a decrease in the number of linked deleterious alleles on one side of the focal locus ( i . e . , an ‘edge effect’ ) , leading to increased admixture towards chromosome ends 25 ., Therefore , the different trends seen in the two species pairs might imply a different balance between this edge effect , which should weaken species barriers , and reduced recombination , which should strengthen them ., Introgression effectively acts to rewrite the evolutionary hist | Introduction, Results, Discussion, Materials and methods | Hybridisation and introgression can dramatically alter the relationships among groups of species , leading to phylogenetic discordance across the genome and between populations ., Introgression can also erode species differences over time , but selection against introgression at certain loci acts to maintain postmating species barriers ., Theory predicts that species barriers made up of many loci throughout the genome should lead to a broad correlation between introgression and recombination rate , which determines the extent to which selection on deleterious foreign alleles will affect neutral alleles at physically linked loci ., Here , we describe the variation in genealogical relationships across the genome among three species of Heliconius butterflies: H . melpomene ( mel ) , H . cydno ( cyd ) , and H . timareta ( tim ) , using whole genomes of 92 individuals , and ask whether this variation can be explained by heterogeneous barriers to introgression ., We find that species relationships vary predictably at the chromosomal scale ., By quantifying recombination rate and admixture proportions , we then show that rates of introgression are predicted by variation in recombination rate ., This implies that species barriers are highly polygenic , with selection acting against introgressed alleles across most of the genome ., In addition , long chromosomes , which have lower recombination rates , produce stronger barriers on average than short chromosomes ., Finally , we find a consistent difference between two species pairs on either side of the Andes , which suggests differences in the architecture of the species barriers ., Our findings illustrate how the combined effects of hybridisation , recombination , and natural selection , acting at multitudes of loci over long periods , can dramatically sculpt the phylogenetic relationships among species . | Many species occasionally hybridise and share genetic material with related species ., Interspecific gene flow may be counteracted by natural selection at particular ‘barrier loci’ ., As a result , a pair of species can end up sharing more genetic variation in some parts of their genome than in others , and the tree of relationships in a group of species can differ from one part of the genome to another ., We studied relationships and barriers among three species of Heliconius butterflies using whole-genome sequences from nine populations ., We find that species relationships vary dramatically and predictably across the genome because the species barriers are more porous in genomic regions with higher recombination rates ., This occurs because recombination determines how broadly the surrounding genome is affected by a barrier locus ., The genome-wide pattern suggests that barrier loci are widespread across the genome ., One consequence is that smaller chromosomes , which have higher recombination rates , tend to have weaker species barriers than longer chromosomes ., The relationships among populations on small chromosomes therefore tend to be predicted by geography , rather than by which species they belong to ., Our work shows how hybridisation , recombination , and selection interact to reshape species’ relationships . | taxonomy, z chromosomes, population genetics, phylogenetics, data management, phylogenetic analysis, population biology, computer and information sciences, sex chromosomes, chromosome biology, autosomes, evolutionary systematics, genetic loci, cell biology, heredity, genetics, biology and life sciences, gene flow, introgression, evolutionary biology, evolutionary processes, chromosomes | null |
journal.pcbi.1006551 | 2,018 | Modeling driver cells in developing neuronal networks | Coordinated neuronal activity is critical for a proper development and later supports sensory processing , learning and cognition in the mature brain ., Coordinated activity represents also an important biomarker of pathological brain states such as epilepsy 1 ., It is therefore essential to understand the circuit mechanisms by which neuronal activity becomes coordinated at a population level ., A series of experimental results indicates that non-random features are clearly expressed in cortical networks 2–4 , in particular neuronal sub-networks , termed cliques 5 , have been shown to play a fundamental role for the network activity and coding both in experiments 6–10 as well as in models 11–14 ., The identification of these small highly active assemblies in the hippocampus 6 and in the cortex 7–9 poses the question if these small neuronal groups or even single neurons can indeed control the neural activity at a mesoscopic level ., Interestingly , it has been shown that the stimulation of single neurons can affect population activity in vitro as well as in vivo 15–25 ., The direct impact of single neurons on network and behavioral outputs demonstrates the importance of the specific structural and functional organization of the underlying circuitry ., Neurons having such a network impact were recently termed operational hubs 26 or driver cells 24 ., It is thus critical to understand how specific network structures can empower single driver cells to govern network dynamics ., This issue has been addressed experimentally in some cases ., More specifically , in the developing CA3 region of the hippocampus , single GABAergic hub neurons with an early birthdate were shown to coordinate neuronal activity ., These cells have a high functional connectivity degree , reflecting mainly the fact that they are activated at the onset of Giant Depolarizing Potentials ( GDPs ) , as well a high effective connectivity degree 17 ., This therefore represents a simple case where the circuit mechanism , promoting a cell to the role of hub , is due to their exceptional number of anatomical links ., But the picture can be quite different in other brain regions , as recently demonstrated in the developing Entorhinal Cortex ( EC ) 24 , where the driver cell population comprises both cells with a high functional out-degree , as well as low functionally connected ( LC ) cells ., In order to understand the circuit mechanisms by which even a LC cell can influence population bursts we have upgraded and modified a network model based on excitatory leaky integrate-and-fire ( LIF ) neurons 12 , previously developed to reproduce the functional properties of hub neurons in the developing hippocampal CA3 area 17 ., In such a model the population bursts ( PBs ) , corresponding to GDPs in neonatal hippocampus 27 , were controlled by the sequential and coordinated activation of few functional hubs ., Notably , the perturbation of one of these neurons strongly impacted the collective dynamics and brought even to the complete arrest of the bursting activity , similarly to what experimentally found for the developing hippocampus in 17 ., The model described in this paper contains two main differences with respect to the hippocampal model 12 ., Firstly , it comprises both inhibitory and excitatory neurons , to account for the fact that , even though GABA acts as an excitatory neurotransmitter at early postnatal stages , some more developed neurons have already made the switch to an inhibitory transmission at the end of the first postnatal week in mice ( P8 ) , where most experimental data was obtained 28–30 ., In that respect , the maturation profile of the entorhinal-hippocampal circuit was recently analyzed and it was shown that stellate cells in layer 2 of the medial EC were the first to mature , followed , in chronological order , by pyramidal cells in MEC layer 2 and CA3 31 ., This indicates a higher probability to find inhibitory GABAergic synapses in EC with respect to CA3 and it justifies our choice to include some inhibitory neuron in the model ., Secondly , the developmental profile of the network is regulated only by the correlation between neuronal excitability and connectivity , while in 12 a further correlation was present ., The anti-correlation between intrinsic excitability and the synaptic connectivity reproduces to some extent the homeostatic regulation of the intrinsic excitability described during neural development 32 ., This model nicely mimicked the experimental observations in the EC similarly displaying the presence of driver cells with both low and high functional connectivity ., The paper is organized as follows ., Firstly , we will report experimental evidences of how the stimulation of single LC drivers can impact network synchonization in the developing EC ., Secondly , we will show that our simple model can nicely reproduce such results ., This to validate its applicability in order to understand the synchronization dynamics of the EC at the early stage of the development ., Then , we will present a full characterization of the numerical model leading to a complete understanding of the mechanism underlying the PB generation and the impact of driver LC cells on population dynamics ., We will conclude with a discussion of the obtained results and of their possible relevance in the context of neuroscience ., The main experimental observation at the rationale of this work is the existence of driver cells ( or Operational Hubs 26 ) in the mice EC during developmental stage 24 ., Driver cells have been identified using calcium imaging experiments and they were characterized by the capability to impact network synchronization ( namely , GDPs′ occurence ) when externally activated/stimulated through intra-cellular current injection ., Two classes of driver cells were identified:, ( i ) those with high directed functional connectivity out-degree , early activated and playing a critical role in the network synchronizations ( driver hub cells ) and, ( ii ) those recruited only in the later stages of the synchronization build up , which therefore are low functionally connected ( driver LC cells ) ., The experimental setup used to identify , target and probe the single-handedly impact of neurons on spontaneous EC synchronization is schematized in Fig 1 ( a . E ) and S1 Fig . In brief , the functional connectivity of the cells has been measured during the spontaneous activity session , which preceded the single neurons’ stimulation session , both lasting two minutes ., A directed functional connection from neuron A to B was established whenever the firing activity of A significantly preceded the one of neuron B ( more details can be found in Methods ) ., The functional out-degree D i O of a neuron i corresponded to the percentage of imaged neurons which were reliably activated after its firing ., Neurons in the 90% percentile of the connectivity distribution were classified as hub neurons early activated in the network synchronization ., The protocol used for probing the impact of single neurons on the network dynamics was organized in three phases , each of two minutes duration: ( 1 ) a pre-stimulation resting period; ( 2 ) a stimulation period , during which a series of supra-threshold current pulses at a specific frequency νS have been injected into the cell; ( 3 ) a final recovery period , where the cell is no more stimulated ., The frequencies νS employed to stimulate the single neurons have been selected to be of the order of magnitude of the average GDP frequency with the aim of revealing cell-network interaction ., A cell was identified as a driver whenever the distributions of the Inter-GDP-Intervals ( IGIs ) were significantly different in the stimulation period with respect to the pre-stimulation and recovery periods ( see Methods for more details ) ., A further indicator that we use to visualize the effect of the stimulation on the GDP deliver is the shift of the IGI phase Φ measured with respect to the pre-stimulation phase ( for the definition see Methods Eq ( 1 ) and in 33 ) ., At a population level the stimulation may have an inhibitory ( excitatory ) effect corresponding to a slow-down ( acceleration ) of the GDP frequency associated with an increase ( decrease ) of the measured IGI corresponding to a positive ( negative ) phase shift ., Two examples of driver LC cells , with D0 ≃ 7 − 8% , are reported in Fig 1 in the panels ( b-d . E ) and ( e-g . E ) ., In the first case , upon stimulation the network dynamics accelerated , as testified by the decrease of the average IGI ( Fig 1 ( b . E ) ) and by the negative instantaneous phase shift of GDPs ( Fig 1d . E ) ., In the second case , the stimulation led to a pronounced slow down of the average network activity ( as shown in Fig 1 ( e . E ) ) together with an increase of the instantaneous phase with respect to control conditions ( Fig 1 ( g . E ) ) ., In both cases the removal of the stimulation led to a recovery of the dynamics similar to the control ones ., A further extreme case of a silent cell , i . e . not spontaneously active and therefore with a zero ( out-degree ) functional connectivity , is shown in Fig 2 ., This cell , when stimulated with different stimulation frequencies νS , revealed opposite effects on the network behaviour ., At lower stimulation frequency ( νS = 0 . 33 Hz ) the cell activity induced an acceleration of the population dynamics ( see Fig 2, ( a ) –2 ( c . E ) ) , while at higher stimulation frequency ( νS = 1 Hz ) of the same neuron we observed a slowing down of the network dynamics ( see Fig 2, ( d ) –2 ( f . E ) ) ., By considering a much larger pool of driver cells we have verified that the value of νS does not induce a systematic trend towards a slow down or an acceleration of the GDPs ( for more details see S2 Text and S3 Fig ) ., We also tested the possibility that single neuron stimulation could modify other features of network synchronization besides the GDP frequency , in particular we focused on the width of the GDPs , i . e . on their time duration as defined in S1 Text ., However , as shown in S2 Fig we could not find any statistical differences between stimulation epoch and pre/post stimulation periods ( see S1 Text for details ) ., This doesn’t mean that other more subtle features of the GDPs could not be impacted by manipulating single neurons , but probably these modifications could be hardly discernible in view of the limited time resolution associated to calcium imaging ., In order to mimic the impact of single neurons on the collective dynamics of a neural circuit , we considered a directed random network made of N LIF neurons 34 , 35 composed of excitatory and inhibitory cells and with synapses regulated by short-term synaptic depression and facilitation , analogously to the model introduced by Tsodyks-Uziel-Markram ( TUM ) 36 ., In particular , synaptic depression was present in all the connections , while facilitation only in the connections targeting inhibitory neurons ., This in agreement with recent experimental investigations of Layer II of the medial enthorinal cortex reporting evidences of short-term synaptic depression among excitatory neurons ( namely , stellate and pyramidal cells ) as well as among fast-spiking interneurons and pyramidal cells and of short-term facilitation in the connections from stellate cells towards low-threshold-spiking interneurons 37 ( see Methods for more details ) ., As shown in 36 , 38 , 39 , these networks exhibit a dynamical behavior characterized by an alternance of short periods of quasi-synchronous firing ( PBs ) and long time intervals of asynchronous firing , thus resembling cortical GDPs’ occurrence in early stage networks ., Similarly to the modeling reported in 12 , we considered neuronal intrinsic excitabilities negatively correlated with the total connectivity ( in-degree plus out-degree ) ( for more details see Definition of the Model in Methods and S4 Fig ) ., The introduction of these correlations was performed to mimic developing networks , where both mature and young neurons are present at the same time associated to a variability of the structural connectivities and of the intrinsic excitabilities ., Experimental evidences point out that younger cells have a more pronounced excitability , most likely due to the fact that their GABAergic inputs are still excitatory 40–42 , while mature cells exhibit a higher number of synaptic inputs and they do receive inhibitory or shunting GABAergic inputs 17 , 43 ., The presence of inhibition and facilitation are the major differences from the model developed in 12 to simulate the dynamics of hippocampal circuits in the early stage of development , justified by the possible presence of mature GABAergic cells in the network ., Using this network model , we studied the effect of single neuron current injection Istim on network dynamics , thus altering the average firing frequency of the neuron during the stimulation time , similarly to what done in the experiments ., In the numerical investigations , at variance with the experiments , the stimulation delivered to the neurons is an unique supra-threshold step of duration of 48 seconds ., In Fig 1 two representative driver LC cells are reported for comparison with the experiments ., The first cell ( panels ( b-d . S ) of Fig 1 ) was a silent neuron in control conditions ( therefore with DO = 0 ) , that once stimulated could enhance of ≃ 30% the PB emission , thus leading to a decrease of the instantaneous phase Φ with respect to control condition ., Panels ( e-g . S ) refer to a second neuron characterized by a low functional output connectivity , namely DO = 3% , whose stimulation led to a depression in the PB frequency ( as shown in panels ( e . S ) and ( f . S ) ) joined to an increase of the instantaneous phase of the network events with respect to control conditions ( as shown in panel ( g . S ) ) ., These results compare quite well with the experimental findings reported in the same figure ., Furthermore , analogously to what found in the experiment , Fig 2, ( a ) –2 ( f . S ) shows a silent neuron in control condition that once stimulated could lead to both enhancement or depression of the population activity depending on the level of injected current during stimulation ., A full characterization of the network model concerning the impact on the network dynamics of each single neuron stimulation in relation to neuronal type , current injected and functional connectivity is detailed below ., In order to explore the full dynamical range associated to the impact of single neuron stimulation on the network dynamics , we examined the response of the model network to two types of single neuron perturbations , i . e . single neuron deletion ( SND ) and single neuron stimulation ( SNS ) by employing the protocols introduced in 12 ., In particular , the SND experiment consisted in recording the activity of the network in a fixed time interval Δt = 84 s when the considered neuron was removed from the network itself ., While , the stimulation of the single neuron ( SNS ) was performed with a step of DC current of amplitude Istim for a time window Δt = 84 s ., The recording of the activity in control condition was lasting 84 s as well , in order to compare directly the number of observed PBs during control and perturbation period ., In particular , we tested the response of the network to a broad range of stimulation amplitudes varying from 14 . 5 mV ( slightly below the firing threshold for an isolated neuron Vth = 15 mV , see Methods ) to 18 . 0 mV with a step of 0 . 015 mV , inducing in the stimulated neuron a maximal firing frequency of ≃70 Hz ., Typically the stimulated neuron fired with a frequency much higher than the frequency of neurons under control conditions ( i . e . in absence of any perturbation ) ., As an example , for a stimulation current Istim = 15 . 90 mV the targeted neuron fired at a frequency ν ≃ 32 − 36 Hz well above the average ( ≃3 Hz ) and the maximal ( 22 Hz ) frequency of all neurons in control conditions ., The SND represented an extreme version of the SNS , where the neuronal removal corresponded to the injection of an hyperpolarizing current inhibiting the neurons from firing spontaneously or in response to any synaptic input ., In both SNS and SND experiments the impact of single neuron perturbation on the collective dynamics was measured by the variation of the PB frequency relative to control conditions ., In general , we have classified a neuron as a driver cell whenever upon stimulation it is able to modify the PB frequency of at least 50% with respect to control conditions ., In the specific , in analogy to what done in 24 , for SNS experiments we considered both enhancement and decrease in the PB activity ., On the other hand , SND allowed us to directly identify the driver neurons which are fundamental for the PB build-up ., Therefore in this case we limited to consider those cells whose SND led to a population decrease of at least 50% ., The choice of this threshold is based on the analysis of the distribution of the number of PBs delivered during SNS and SND experiments compared to the PB variability observed in control conditions ., This was preliminarily measured by considering the average and the standard deviation of the number of population bursts obtained by considering for each network realization 100 different initial conditions over a time window of 84 s ., As we were interested in strong impact on the network dynamics , we decided to consider as significant the variations of the population activity which were well beyond the statistical fluctuations in the population bursting measured by three standard deviations ., The choice of 50% fulfilled this condition in all the analyzed networks ., Fig 3, ( a ) and 3, ( b ) report a comparison of the impact of SND and SNS ( with representative injected current of 15 . 90 mV ) on the PB activity ., The removal of any of the four neurons labeled as ih1 , eh1 , eh2 , eh3 was able to arrest completely the bursting dynamics within the considered time window , while in other two cases ( for neurons ih2 and eh4 ) the activity was reduced of 60% with respect to the one in control conditions ., For clarity , the used labels i/e stand for inhibitory/excitatory and h for hub , as we will show later this is related to the functional role played by these cells ., For all the other neurons , the SND manipulation induced a non relevant modification in the number of emitted PBs , within the variability of the bursting activity in control conditions ( Fig 3, ( a ) ) ., The SNS confirmed that the neurons ih1 , eh1 , eh2 , eh3 were capable to arrest the collective dynamics ., Neurons eh4 and ih2 poorly impacted PB dynamics for the reported injected current , although for different values of Istim they were able to strongly influence the network dynamics ( as shown in the subsection Tuning of PBs frequency upon hubs’ and driver LC cells’ stimulations ) ., At variance from what found in a purely excitatory network 12 , the SNS revealed also the presence of other 18 driver cells not identified by the SND capable to impact the occurence of PBs in the network ( Fig 3, ( b ) ) ., For an equivalent random network , without any imposed correlation , SNS or SND affected the dynamics in a neglibile way producing a maximal variation of the bursting activity of 25-30% with respect to the control conditions ( see S5, ( a ) and S5, ( b ) Fig ) ., To summarize , the presence of correlations among the neuronal intrinsic excitabilities and the corresponding structural connectivities was crucial to render the network sensible to single neuron manipulation ., Differently from purely excitatory networks where SNS and SND experiments gave similar results , the inclusion of inhibitory neurons in the network promoted a larger portion of neurons to the role of drivers , and their properties will be investigated in the following ., The role played by the neurons in the simulated network was elucidated by performing a directed functional connectivity ( FC ) analysis ., In the case of the spiking network model , in order to focus on the dynamics underlying the PB build-up , the FC analysis was based on the first spike fired by each neuron in correspondence of the PBs ., An equivalent information was provided in the analysis of the EC by considering the calcium signal onset to calculate the directed functional connectivity ., The six neurons playing a key role in the generation of the PBs ( eh1−4 , ih1−2 ) were characterized by high values of functional out-degree , namely with an average functional degree DO = 68% ± 8% , ranking them among the 16 neurons with the highest functional degree ., Given the high functional out-degree and their fundamental role in the generation of the PBs ( as shown by the SND in Fig 3, ( a ) ) , we identified these neurons as driver hub cells ., The high value of DO reflected their early activation in the PB , thus preceding the activation of the majority of the other neurons ., Next , we examined the structural degree of the neurons , specifically we considered the total structural degree KT , which is the sum of the in-degree and out-degree of the considered cell ., As shown in Fig 3 ( f ) , we observed an anti-correlation among DO and KT where neurons with high functional connectivity are typically less structurally connected than LC neurons ., This was particularly true for the six driver hubs , previously examined , since they were characterized by an average KT = 15 ± 3 , well below the average structural connectivity of the neurons in the network ( ≃ 20 ) ., Concerning the excitability , the six driver hubs despite being in proximity of the firing threshold ( slightly above or below ) as shown in S6, ( a ) Fig , they were among the 25% fastest spiking neurons in control condition , ( as shown in Fig 3, ( c ) ) ., In particular , the three neurons eh1 , eh2 , ih2 were supra-threshold , while neurons eh3 , eh4 , ih1 were slightly below the threshold ., When embedded in the network their firing activity was modified , in particular three couples of neurons with similar firing rates can be identified , namely ( eh1 , ih1 ) , ( ih2 , eh2 ) and ( eh3 , eh4 ) , as reported in Table 1 ., The direct structural connections present among these couples ( see also Fig 3 ( g ) ) could explain the observed firing entrainments , as discussed in details in the next subsection ., When compared to the other hub neurons , the much lower activity of ( eh3 , eh4 ) , corresponding to twice the average frequency of the PBs in control condition , was related to the fact that these two neurons fired only in correspondence of the ignition of collective events like PBs and aborted bursts ( ABs ) , the latter being associated to an enhancement of the network activity but well below the threshold we fixed to detect PBs ., This will become evident from the discussion reported in the subsection Synaptic resources and population bursts ., As already mentioned , besides the six driver hubs , the SNS experiments revealed the existence of a different set of 18 drivers , whose activation also impacted the population dynamics , although they had no influence when removed from the network and therefore they were not relevant for the PBs build up ., These neurons represented in Fig 3 with squares were characterized by a low FC , namely D0 = 13% ± 15% ., Therefore , we have termed them driver LC cells representing the ones which reproduced the behaviour of the driver LC cells identified in the EC ( see Figs 1 and 2 and reference 24 ) ., In the following we will refer to them as el… or il1 according to the fact that they are excitatory or inhibitory neurons , respectively ( note that only one LC driver was inhibitory ) ., As shown in Fig 3, ( c ) , LC drivers were not particularly active ( with firing frequencies below 1 Hz in control conditions ) and in some cases they were even silent ., Notably , under current stimulation they could in several cases arrest PBs or strongly reduce/increase the activity with respect to control conditions as shown in Fig 3, ( b ) for a specific level of current injection and also as discussed in detail in the following sections ., Compared to the driver hubs , driver LC cells had a lower degree of excitability ( essentially they were all sub-threshold , see S6, ( a ) Fig ) , which resulted in a later recruitment in the synchronization build up , and as a consequence in a lower functional out-degree ., Therefore , driver LC cells were not necessary for the generation of the PBs , playing the role of followers in the spontaneous network synchronizations ., As shown in Fig 3 ( f ) , driver LC neurons were charaterized by a higher structural connectivity degree KT with respect to driver hubs , namely KT = 23 ± 3 , and the most part of them were structurally targeting the driver hubs either directly ( i . e . path length one ) or via a LC driver ( i . e . path length two , centered on a LC driver ) ., In Fig 3 ( f ) , the two groups of drivers , hubs and LC cells , can be easily identified as two disjoint groups in the plane ( KT , D0 ) ., These results indicated that driver hubs are not structural hubs , while the low functional connectivity neurons are promoted to their role of drivers due to their structural connections ., This latter aspect will be exhaustively addressed in subsection Tuning of PBs frequency upon hubs’ and LC cells’s stimulation ., In order to deepen the temporal relationship among neural firings leading to a PB , we examined the spikes emitted in a time window of 70 ms preceding the peak of synchronous activation ( see Methods for details ) ., The cross correlations between the timing of the first spike emitted by each driver hub neuron during the PB build up are shown in S7 Fig ( Upper Sequence of Panels ) ., The cross correlation analysis demonstrated that the sequence of activation of the neurons was eh1 → ih1 → ih2 → eh2 → eh3 → eh4 ., The labeling previously assigned to these neurons reflected such an order ., A common characteristic of these cells was that they had a really low functional in-degree DI as reported in Table 1 indicating that they were among the first to fire during the PB build-up ., In particular , eh1 had a functional in-degree DI zero , revealing that it was indeed the firing of this neuron to initialize all the bursts and therefore it could be considered as the leader of the clique ., A detailed inspection of the firing times , going beyond the first spike event , revealed the existence of more than one firing sequence leading to the collective neuronal activation: i . e . the existence of different routes to PBs ., This is at variance with what found in 12 for a purely excitatory network , where only one route was present and all the PBs were preceded by the same ordered sequential activation of the most critical neurons ., In particular , the neuron eh1 fired twice before the PBs ( see Fig 3, ( e ) ) , usually in-between the firing of eh2 and that of the pair ( eh3 , eh4 ) , and this represents the main route , occurring for ≃ 85% of the PBs ., Along the second route ( present only for the ≃ 7% of the PBs ) , eh1 was firing the second time at the end of the sequence ., The neuron eh1 fired essentially by following its natural period T 1 = τ m ln ( I e h 1 b - v r ) / ( I e h 1 b - V t h ) = 52 ., 15 ms , and its second occurrence in the firing sequence depended on the delay among the firing of the other neurons ., As a matter of fact we verified that the elimination of the second spike emitted by eh1 from the network dynamics didn’t prevent , and didn’t delay , the onset of the PB and had only a marginal effect on the firing of a very limited number of neurons in the PB ., Therefore we can conclude that it is not essential to the PB build up ., The two routes leading to the PB build-up are shown in Fig 3, ( e ) ., To observe a PB the six driver hubs should fire not only in an ordered sequence , as shown in Fig 3, ( e ) , but also with defined time delays , their average values with the associated standard deviations are reported in S1 Table for the two principal routes ., These results clearly indicate that the six driver hubs are arranged in a functional clique whose activation was crucial for the PB build-up ., In the period between the occurence of two PBs , the driver hubs in the clique could be active , but in that case they did not show the precise sequential activation associated to the main and secondary route , see the out-of-burst results reported in the Lower Sequence of Panels in S7 Fig . A remarkable exception is represented by the case of the ABs , in that case PBs are not triggered despite the presence of the right temporal activation of all the hubs in the clique , due to the lack of synaptic resources ( as discussed in details in subsection Synaptic resources for population bursts ) ., Out of PBs and ABs , we registered clear time-lagged correlations only for those neuronal pairs sharing direct structural connections ( shown in Fig 3 ( g ) ) : namely , eh1 → ih1 , ih2 → eh2 and eh2 → ( eh3 , eh4 ) ., The firing delays of these neuronal pairs were not particularly altered also out of burst with respect to those measured during the burst build-up and reported in S1 Table ., As shown in Fig 3 ( g ) , the eh3 neuron represented the cornerstone of the clique , receiving the inhibitory input coming from the structural pair ( eh1 , ih1 ) and the excitatory one from the pair ( ih2 , eh2 ) , with the activity of the neurons within each pair perfectly frequency locked ., More specifically , eh1 entrained the activity of ih1 ( below threshold in isolation ) so that both neurons before a PB fired with a period quite similar to the natural period of eh1 ., The other pair ( ih2 , eh2 ) was controlled by the inhibitory action of ih2 that slowed down the activity of eh2 , whose natural period was 60 . 6 ms , while before a PB ih2 and eh2 both fired with a slower period , namely 72 ± 2 ms . As it will be explained in details in the next two subsections , the two requirements to be fulfilled for the emergence of PBs are the availability of sufficient synaptic resources at neurons eh3 and eh4 and the coordinated activation of eh1 ( and ih1 ) with the pair ( ih2 , eh2 ) , in the absence of any synaptic connection between the two pairs ., Next we analyzed the relation between the evolution of synaptic resources in the hub driver cells and the onset of the PB ., The availability of synaptic resources was measured by the effective efferent synaptic strength XOUT as defined in Eq ( 8 ) ., In particular , we considered the available resources only for the hub neurons eh3 and eh4 which were the last neurons of the clique to fire before the PB ignition ., We have examined only these two hub neurons , because whenever eh3 and eh4 fired , a burst or an AB was always delivered ., Neurons eh3 , eh4 were receiving high frequency excitatory inputs from eh2 ( although the natural firing of eh2 was slowed down by the incoming inhibition of ih2 ) and high frequency inhibitory inputs from ih1 ( entrained by the eh1 , the neuron with highest firing frequency in the network ) ., This competitive synaptic inputs resulted in a rare activation of eh3 compared to the higher frequency of excitatory inputs arriving from eh2 ., The period of occurrence of the ABs was comparable to the average interval between PBs ( namely , TPB = 1 . 4±1 . 0 s ) and ABs were preceded by the sequential activations of the six critical neurons of the clique in the correct order and with the required delays to ignite a PB ., The number of observed ABs was 66% of the PBs , thus explaining why the average firing period of eh3 and eh4 was T e h 3 = 0 ., 8s≃ T P B / ( 1 + 0 . 66 ) , since their firing always triggered a PB or an AB ., To understand why in the case of ABs the sequential activation of the neurons of the clique did not lead to a PB ignition , we examined the value of synaptic resources for regular and aborted bursts , as shown in Fig 4, ( a ) ., From the figure it is clear that X e h 3 O U T and X e | Introduction, Results, Discussion | Spontaneous emergence of synchronized population activity is a characteristic feature of developing brain circuits ., Recent experiments in the developing neo-cortex showed the existence of driver cells able to impact the synchronization dynamics when single-handedly stimulated ., We have developed a spiking network model capable to reproduce the experimental results , thus identifying two classes of driver cells: functional hubs and low functionally connected ( LC ) neurons ., The functional hubs arranged in a clique orchestrated the synchronization build-up , while the LC drivers were lately or not at all recruited in the synchronization process ., Notwithstanding , they were able to alter the network state when stimulated by modifying the temporal activation of the functional clique or even its composition ., LC drivers can lead either to higher population synchrony or even to the arrest of population dynamics , upon stimulation ., Noticeably , some LC driver can display both effects depending on the received stimulus ., We show that in the model the presence of inhibitory neurons together with the assumption that younger cells are more excitable and less connected is crucial for the emergence of LC drivers ., These results provide a further understanding of the structural-functional mechanisms underlying synchronized firings in developing circuits possibly related to the coordinated activity of cell assemblies in the adult brain . | There is timely interest on the impact of peculiar neurons ( driver cells ) and of small neuronal sub-networks ( cliques ) on operational brain dynamics ., We first provide experimental data concerning the effect of stimulated driver cells on the bursting activity observable in the developing entorhinal cortex ., Secondly , we develop a network model able to fully reproduce the experimental observations ., Analogously to the experiments two types of driver cells can be identified: functional hubs and low functionally connected ( LC ) cells ., We explain the role of hub neurons , arranged in a clique , for the orchestration of the bursting activity in control conditions ., Furthermore , we report a new mechanism which can explain why and how LC drivers emerge in the structural-functional organization of the entorhinal cortex . | action potentials, medicine and health sciences, neural networks, nervous system, population dynamics, membrane potential, electrophysiology, neuroscience, network analysis, computational neuroscience, population biology, mood disorders, computer and information sciences, depression, animal cells, mental health and psychiatry, cellular neuroscience, cell biology, anatomy, synapses, single neuron function, neurons, physiology, biology and life sciences, cellular types, computational biology, neurophysiology | null |
journal.pbio.0050311 | 2,007 | Resolving the Fast Kinetics of Cooperative Binding: Ca2+ Buffering by Calretinin | In all eukaryotic cells , Ca2+ signals play a crucial role in the regulation of many cellular processes , including gene expression , cytoskeleton dynamics , cell cycle , cell death , neurotransmission , and signal transduction ., To achieve its role as messenger , the intracellular Ca2+ concentration ( Ca2+ ) is very tightly regulated in time , space , and magnitude ., The spatiotemporal characteristics of short-lived and often highly localized changes in intracellular Ca2+ result from a complex interplay between Ca2+ influx/extrusion systems , mobile/stationary Ca2+-binding proteins ( CaBPs ) , and intracellular sequestering mechanisms ., Understanding the kinetics of cellular Ca2+ transients and its influence on Ca2+-regulated processes requires a precise knowledge of the Ca2+ sensitivities and binding properties of all the components involved , including the binding dynamics to buffering and signaling CaBPs ., However , uncertainties in current models studying intracellular Ca2+ signaling arise mostly from the lack of accurate data on the binding properties of specific molecules involved in Ca2+ handling , considerably limiting the value of such modeling 1 ., An important step towards the goal of precisely describing intracellular Ca2+ transients was the study by Nagerl et al . 2 , in which the relevant parameters ( affinities and on- and off-rates of Ca2+ binding ) for the CaBP calbindin D-28k ( CB ) were determined in vitro by flash photolysis of caged Ca2+ ., Cooperative binding of Ca2+ , known to play a significant role in multisite CaBPs such as calmodulin 3 and calretinin ( CR ) 4 , has never been directly determined in rapid kinetic experiments , but only inferred from steady-state conditions using Hill 5 and Adair-Klotz models 6 , 7 ., Cooperativity first evidenced by oxygen binding to hemoglobin 8 is considered one of the most imperative functional properties of molecular interactions in biological systems , even considered to be the great secret of life , second only to the structure of DNA 9 ., Cooperativity is the ability to influence ligand binding at a site of a macromolecule by previous ligand binding to another site of the same macromolecule ., Many proteins show increased ( positive cooperativity ) or decreased ( negative cooperativity ) affinity for a ligand after binding of a first ligand ., Over the last 100 years , hemoglobin has been a paradigm for cooperative ligand binding and allostery ., Oxygen binding to hemoglobin resulted in four commonly used descriptions for cooperativity ( for review see 10 ) : the Hill 11 , the Adair-Klotz 6 , 7 , the Monod-Wyman-Changeux ( MWC ) 12 , and the Koshlan-Némethy-Filmer ( KNF ) 13 models ., Yet all these models describe cooperativity only when the binding reactions are at equilibrium ., Since temporal aspects of most ligand binding processes are essential for correct physiological functioning , it is imperative to consider the kinetics of cooperative binding ., To date , studies determining the kinetics of cooperative binding to biologically active molecules have been carried out using the MWC model , in which cooperativity is established by assuming two allosteric isoforms with different binding properties ., These studies were limited to special cases where transition rates between allosteric isoforms are much slower than the binding rates 14 , 15 or where binding and unbinding rates could be measured independently 16 ., For all other cases , the mathematical description becomes too complex for simple interpretations 10 , 17 ., The Hill equation is perhaps the oldest and most widely used description for the relative amount of binding by a cooperative molecule , and the cooperative binding is described with two constants: the dissociation constant ( Kd ) reports on the concentration of ligand at which the cooperative molecule is half occupied and the Hill number ( nH ) describes the steepness of the binding curve at the value of Kd , denoting a simple quantification of cooperativity ., Although not representing a mathematically correct description of cooperative binding at equilibrium—a fact that is stated in the original work 11—the Hill equation has proven to be extremely useful , as it describes occupancy as a function of ligand concentration with merely two constants that are easy to interpret intuitively ., With this in mind , we wanted to resolve the kinetics of Ca2+ binding to CR and to find an intuitively “accessible” quantitative description of the binding kinetics ., CR belongs to the superfamily of EF-hand Ca2+-binding proteins ., This superfamily is named after the common Ca2+ binding structure—the EF-hand—first described as the C-terminal E-helix–loop-F-helix Ca2+ binding site in parvalbumin 18 ., Most members have an even number of EF-hand domains organized in pairs 19 , representing a structurally conserved architectural unit ., CR has six EF-hand domains 4 , 20–24 , which can be subdivided into two independent domains: one with the cooperative pair of binding sites I and II , and another with binding sites III–VI 23 ., Sites III–VI can be further subdivided into one cooperative pair , sites III and IV and sites V and VI 4 ., Of the latter pair , only site V binds Ca2+ , whereas site VI is “inactive” 4 ., Thus , CR has two pairs of cooperative binding sites ( I–II and III–IV ) and one independent binding site ( V ) for Ca2+ ., We tried several approaches to describe the kinetics of these five binding sites based on the published models , and discovered a new and simplified kinetic model that quantitatively resolves the kinetics of cooperative binding ., This new model also revealed unexpected and highly specific nonlinear properties of cellular Ca2+ regulation by CR ., We determined the kinetics of Ca2+ binding to CR by Ca2+ uncaging using DM-nitrophen ( DMn ) and measuring the changes in Ca2+ with the fluorescent Ca2+ indicator dye Oregon Green BAPTA-5N ( OGB-5N ) as previously described 2 , 25 ., Changes in the OGB-5N fluorescence were observed immediately after photolysis of DMn in solutions containing various concentrations of CR ( Figure 1A ) ., A rapid rise in Ca2+ ensued from a resting concentration of approximately 2 . 4 μM to 11 μM ., At this initial Ca2+ , approximately 99 . 5% of the DMn is in the Ca2+-bound form , ensuring that ( 1 ) virtually every uncaged DMn molecule will release Ca2+ , and ( 2 ) the amount of free DMn capable of rebinding uncaged Ca2+ ions 25 , 26 is considerably limited ., This is evidenced by the negligible drop in Ca2+ in the absence of CR , leading to an almost step-wise increase in Ca2+ ( Figure 1A ) ., The presence of CR ( 31 and 62 μM ) resulted in a CR-dependent drop in Ca2+ ., Figure 1B depicts a simplified reaction scheme of the experiment ., To determine the association and dissociation rates of Ca2+ binding to CR , all these different reactions were incorporated into a mathematical model ( see below ) ., The aim was to find a mathematical description for the Ca2+-binding properties of CR that best fits the experimental fluorescence traces generated under various conditions ., As a starting point to determine CRs kinetics of Ca2+ binding , we relied on steady-state Ca2+-binding properties determined previously ., With the same human recombinant CR , selected Hummel-Dryer experiments yielded a Hill coefficient of 1 . 3 for the four binding sites , with a Kd of 1 . 5 μM 4 ., However , by flow dialysis , the steady-state binding of Ca2+ to human CR could be described with the following macroscopic constants ( K1 through K5 ) 2 . 2 × 105 M−1 , 3 . 2 × 105 M−1 , 4 . 7 × 105 M−1 , 8 . 0 × 105 M−1 , and 2 . 0 × 104 M−1 4 ., The resulting binding curve derived from these values could be accurately fitted with two Hill equations; one equation described four cooperative binding sites with a Kd of 2 . 5 μM and a Hill coefficient of 2 . 4 , and the other one described a single independent site with a Kd of 53 μM ., In agreement with this data , equilibrium dialysis experiments with chick CR revealed a Hill coefficient of 1 . 9 22 ., Even Hill coefficient values of up to 3 . 7 for Ca2+-induced tryptophan ( Trp ) fluorescence changes in rat CR have been reported 21 ., However , these conformational changes measured by Trp fluorescence do probably not linearly relate to Ca2+ binding ., Thus , the absolute values should be interpreted with caution , but nonetheless , cooperativity of Ca2+ binding also occurs in rat CR ., Based on these Hill coefficients of 1 . 3 , 1 . 9 , and 2 . 4 , we conjectured that CR has four binding sites for Ca2+ with positive cooperativity , with a Hill coefficient of approximately 2 and one independent binding site for Ca2+ ., In accordance with these and other earlier findings on the structure and physiology of CR ( see Introduction ) , we modeled the protein as possessing two pairs of cooperative binding sites ( BIBII ) and ( BIIIBIV ) , and one independent binding site BV ., Because the properties of the two cooperative pairs in CR were considered indistinguishable in the steady-state study 4 , thus indicating that the cooperative binding sites are fairly similar , we assumed the properties of both cooperative pairs to be identical:, Such an assumption is most useful for reducing the number of variables , thus increasing the reliability of fitting procedures by constraining the model ., The most straightforward approach to determine the kinetics of a system would consist of fitting the Ca2+ decay with a set of exponential functions ., However , rebinding of Ca2+ to free DMn affects the decay kinetics ., In addition , the changing properties of the binding sites that underlie cooperativity are expected to cause a shift in the kinetic properties of binding during the decay phase ., As a consequence , the relative contribution of multiple decay time constants is continuously shifting ., Although fitting the Ca2+ decay with exponentials might result in time constants for a given trace , this does not allow accurate deduction of the Ca2+-binding kinetics of CR ., A mathematical model simultaneously describing all processes taking place in the recording chamber , including a “total” description of the cooperative and noncooperative binding of Ca2+ to CR is expected to yield more reliable information on the kinetic properties of CR ., To model the cooperativity , we started out by including an allosteric influence between the binding sites of the pairs ., This was achieved by setting two states ( R and T ) for a particular binding site , with each having its own set of rate constants ., A binding site is in the “tensed” state ( T ) , with a low affinity for Ca2+ , when no Ca2+ is bound to the other site in the pair , whereas a binding site is in the “relaxed” state ( R ) , with a high affinity for Ca2+ , when the other site already has a Ca2+ ion bound ., We assumed that binding of Ca2+ to one site always leads to a rapid transition T→R in the other site and that an unbinding of Ca2+ from one site always leads to a rapid transition R→T in the other site ., This allowed us to incorporate the transition rates between states R and T in the binding and unbinding rate constants , further simplifying the model ( Figure 2A ) ., CR was thus modeled as if consisting of two independent proteins described by the following binding reactions:, This cooperative part of the model can be easily related ( see Discussion ) to the Adair model 6 , which provides the most general description of equilibria in terms of stochiometric binding ., For the independent site of CR , we used a standard equilibrium equation:, where kon ( R or T ) and koff ( R or T ) are the association and dissociation rate constants for the individual cooperative binding sites depending on their Ca2+-binding status , and kon ( V ) and koff ( V ) are the rate constant for the independent site ., The total concentrations of the different “virtual” parts are:, Despite the simplifying assumptions concerning the cooperative sites , the model that allows a fitting routine to proceed is fairly complex , because it has a considerable number of degrees of freedom ., Thus , a procedure was developed that significantly constrains the fit to minimize the variance of the fit results ., Simultaneously fitting combined sets of uncaging data ( Figure 3; see Materials and Methods ) obtained under different experimental conditions sufficiently constrained the model to yield consistent results ., We performed a number of individual uncaging experiments generated at one of seven initial conditions A–G ( Table in Figure 3 ) ., These conditions varied in the initial free Ca2+ ( hence , total Ca2+ ) , total CR , total DMn , as well as on the lot number of OGB-5N , with each lot having slightly different properties ( see Materials and Methods ) ., Under each condition ( A–G ) , we performed 12 to 25 uncaging experiments , each one with a different flash energy of the UV laser leading to different amounts of uncaged Ca2+ ., In total , 123 traces were obtained , covering a wide rage of uncaging energies and , subsequently , a wide range of increases in Ca2+ ( see gray areas in Figure 3; for all 123 individual traces , see Figure S1 ) ., We set out to find a satisfactory model that would be able to fit all curves obtained under conditions A–G and all tested uncaging intensities ., The obtained results from the modeling should be able to describe all experimental curves with a unique set of parameters describing the kinetic properties of CR ., To confine the fits , 38 sets of 14 pseudo-randomly picked traces consisting of two traces from each initial condition A–G were generated ., The 38 sets were chosen randomly , with the precondition that every trace of a specific starting condition A–G was represented equally ., To create 38 sets , each individual measurement was picked at least three and at most eight times ., On average , each trace was picked 4 . 3 times ( for details , see Figure S2 ) ., Each set of traces was fitted with the model , and the fitted parameters describing the properties of CR were constrained to be identical for all individual traces within one set ., The only variable parameter between traces was the amount of uncaging that was fitted individually for each trace ., An example of a dataset of 14 traces ( two sets of data points • for each condition A–G ) and the fitted traces ( red or blue lines , see Figures S1–S3 for additional details ) are shown in Figure 3 ., Fit results for this dataset are depicted in Figures 4 and 5 ( yellow symbols ) together with the results of the fits on the other 37 sets ., The new model was programmed as to fit the Kd ( V ) and kon ( V ) for the independent site ., However , to aid the choice of starting values , the cooperative part of the model was set up such that kon ( R ) , kon ( T ) , the apparent Kd ( Kd ( app ) ) for the pairs , and the Hill number ( nH ) could be fitted ., This was achieved by adding a calculation step that determined Kd ( R ) and Kd ( T ) from the latter two parameters ( see Protocol S1 ) :, Previously determined steady-state parameters ( apparent ) Kds and nH of CR 4 , 20–22 served as starting points in the modeling and helped to further constrain the model ., Various combinations of kon starting values between 105 and 108 M−1s−1 were tested , but this did not significantly influence the outcome of the fit , indicative of the “robustness” of the modeling procedure ., Occasionally a particular set out of the 38 yielded an atypical fit with values significantly deviating from the general population of results ., If this was the case , we followed up with two approaches ., First , we tested whether any of the other 37 sets of fluorescence traces could also be fitted with these deviating values , which in almost all cases yielded unsatisfactory fits ., Second , we tested whether the deviating set of traces could also be fitted with the more homogeneous values of the general population of sets by choosing starting parameters closer to these values ., In this case , the deviating set could always be fitted with values comparable to the homogeneous constants ., It should be noted that the critical parameters ( Kd ( V ) , kon ( V ) , kon ( R ) , kon ( T ) , Kd ( app ) , and nH ) were never constrained or fixed to a certain value ., The atypical fit results were probably caused by local minima in the error function of the fit routine ., Such local minima are expected , based on the fairly large number of degrees of freedom where the parameters are not completely independent ., Deviations in one parameter can be partially compensated by “shifting” other parameters ., Initially , we used a model that did not include cooperativity and found that most of the 38 individual sets could be fit reasonably well ., For a given individual dataset , the quality of the fits were similar between a model with a Hill coefficient of either 1 or 1 . 9 ., But when comparing the fit results of all 38 sets , most of the fitted parameters showed strong deviations ( up to five orders of magnitude , depending on how the model was exactly defined ) when using nH = 1 . The high variability of the binding parameters found when assuming nH = 1 is shown in Figure S3 ., This indicated that there is no unique solution to describe CRs Ca2+-binding properties without cooperativity , in line with previous steady-state findings of nH values between 1 . 3 and 2 . 4 4 , 22 ., Thus , only when including cooperativity and starting the modeling procedure with previously determined steady-state parameters for CR 4 , 20–22 did we find a congruent set of values for the fitted parameters for all 38 sets of 14 traces ., An accurate model describing CRs Ca2+-binding dynamics should be able to fit all the experimental traces obtained under any condition ., This should be the case at lower resting Ca2+ , when Ca2+-free binding sites of any affinity in any state are abundant ( Figure 3A–3D and 3G ) , but also at higher Ca2+ , when mostly the lower affinity independent site V is available ( Figure 3E and 3F ) ., Furthermore , the model is also able to closely describe the Ca2+ signals after a relatively large uncaging , when Ca2+ is so high that the buffering by CR is relatively small ( see upper trace in Figure 3F ) ., The goodness of fit of the fit procedure can be appreciated by the averaged error for the 38 fits ( Figure 3H , black bars ) , which shows systematic deviations ( if any ) ., Errors were found to be extremely small for the first 20 ms after the flash , and at time points greater than 20 ms , the averaged fits show a small systematic undershoot of the experimental data , yet never exceeding 1 . 5% of the actual amplitude ( Figure 3H , black bars ) ., The larger errors towards the end of the traces are likely due to the small amplitudes of the signals at these time points , which increase the relative error when there is a constant absolute deviation ., But even the largest deviations of the 38 fits ( the striped bars showing the largest deviation in either direction ) never exceeded 5% of the measured amplitude ., The average absolute error ( not allowing for positive and negative errors to cancel each other ) was maximally 2 . 1% and again only found towards the end of the traces ., Thus , the fit procedure , applying our model , allowed accurate quantifying of the kinetic properties of CR ., All 38 results for the fitted values were plotted as a log-normal cumulative probability distribution because they have a log-normal distribution ( Figures 4 and 5 ) , except for the nH value , which was normally distributed ( Figure 5E ) ., These results were then fitted with a normal distribution to determine average and standard deviation ., The results of these fits are shown in Table 1 . To summarize , we conclude that CR can be described with one independent binding site with a Kd of 36 μM , a kon of 7 . 3×106 M−1s−1 , and a koff of 240 s−1 together with two identical cooperative pairs of binding sites with an initial ( T state ) Kd of 28 μM with a kon of 1 . 8×106 M−1s−1 and a koff of 53 s−1 that will dramatically change to a ( R state ) Kd of 68 nM with a kon of 3 . 1 × 108 M−1s−1 and a koff of 20 s−1 once the cooperative partner site has already bound Ca2+ ., The Kd ( app ) for the cooperative sites is 1 . 4 μM with an nH of 1 . 9 ., To compare the results obtained with our new “simple” model , we subjected our experimental data to a previously described cooperative model , the MWC model ., There , the protein as a whole can switch between two states: one state in which all the binding sites are in the T state , and another one in which all the binding sites are in the R state ( Figure 2B ) ., The equilibrium between the T and R states is described with the equilibrium constant L , which is also dependent on the number of Ca2+ ions bound ( see Figure 2B ) ., It can be shown 12 , 27 that:, where k+ and k− are rates of the transition between the two states R and T . The indices ( 0 and i in Equations 8 and 9 , respectively ) indicate the number of Ca2+ ions bound ., K is often indiscriminately used for both association and dissociation equilibrium constants; here , we denote K as association equilibrium constants , whereas we use Kd for dissociation equilibrium constants ., Furthermore , the pair of cooperative binding sites can transition from R to T and back , independently of the number of sites that are occupied , thus all transitions defined by L0 , L1 , and L2 are possible ., However , for steady-state purposes , one transition ( L0 ) suffices , as described in the original MWC model 12 ., Although steady-state properties are independent of the transitions allowed , the kinetic properties will highly depend on the number of allowed transitions ., We chose to allow all possible transitions because it was used in earlier kinetic fits with this model 14 , 16 ., We also attempted to fit the data with a MWC model in which only the L0 transition was possible; however , the resulting fits showed large deviations ( >20% ) and generated traces with significant deviations from the experimental data ., From Equation 9 , we can derive the information that while KT < < KR and L0 > > 1 , the equilibrium between the T and R states is shifted towards the lower affinity T state when no or little Ca2+ is bound , whereas it shifts towards the higher affinity R state when plenty of Ca2+ is bound 27 , which causes the cooperative effect ., The identical 38 sets of 14 experimental traces as used above were fitted with the MWC model ., Here also , both cooperative sets were considered to be identical and the cooperative part of the model was set up such that kon ( R ) , kon ( T ) , the apparent Kd ( Kd ( app ) ) for the pairs , and the Hill number ( nH ) , could be fitted ., This was achieved by adding a calculation step that determined Kd ( R ) and Kd ( T ) from the latter two parameters ( see Protocol S1 ) :, As discussed above , L is dependent on the number of Ca2+ ions bound to CR ., To establish this dependence , we changed the forward and backward rate constants between the R and T states equally:, We started with the same values for the ( apparent ) Kds and nH as above with various combinations of kon starting values between 105 and 108 M−1s−1 ., As with the first model , occasionally a set of traces was fitted with values that deviated significantly from the general population , but again we found that only with one general set of constants , all 38 sets could be accurately fitted; the details are reported in Table 2 . For comparison , we depicted the fit and the fit errors obtained with the MWC model , using the identical set of traces used for the new model ( compare Figure 3A–3G , blue vs . red traces; errors for MWC fit are not shown , but are comparable to the fits with the other model ) , and observed that the MWC model can fit the data with a similar accuracy as our new model ., The average results from all traces were obtained as described for the new model ( Figures 4 and 5 ) ., For the independent site ( Figure 4A and 4B ) , results based on the MWC model ( blue symbols ) were essentially identical to the ones found with the new model ( red symbols , compare also Tables 1 and 2 for the independent site V ) ., Also , the properties of the apparent Kd and the nH of the cooperative sites were similar between the two models ( Figure 5A , compare green and pink circles , and Figure 5E , compare blue and red symbols ) ., This is a first indication that both models quantitatively describe the same process ., Obviously , the detailed descriptions for the cooperative sites , applying either the new or the MWC model , deviate from one another , based on the differently modeled processes as described in Figure 2 . Both models accurately fitted the data in a quantitative manner; based on the quality of the fits , they were indistinguishable , and technically can both be used to quantify the kinetic properties of Ca2+ binding to CR ., In particular , results for the independent site ( V ) are virtually identical ( Figure 4 ) ., For the cooperative sites , both models describe binding sites that have a similar steady-state/affinity profile ( Figure 5A; apparent Kd and Figure 5E; nH ) ., Our uncaging experiments were performed over a fairly narrow range of resting Ca2+ ( 2 . 0–5 . 3 μM ) , dictated by the constraints that most ( >99% ) of the DMn should be in the Ca2+-bound form to obtain valuable data ., At lower resting Ca2+ , the unbound DMn , present at higher concentrations than the free CR , would rapidly rebind most of the released Ca2+ 25 , making CRs relative contribution to the Ca2+ decay small and difficult to distinguish ., Because buffering kinetics depend on both the on-rates and the concentration of free buffer , the overall Ca2+ binding speed to DMn would be much higher than that to CR , thus masking Ca2+ binding to CR ., With much less Ca2+-bound DMn , the changes in Ca2+ would be quite small and difficult to detect ., Such technical constraints do not allow performing the experiments over the whole “physiological” range of Ca2+ , e . g . , from 10 nM to 100 μM ., Thus , we could not exclude that the two models describe systems with different kinetic properties outside the boundaries of our experiments ., This possibility was tested by examining the behavior of each model as a filtering system for Ca2+ signals regulated by resting levels of Ca2+ ., The filtering properties of both models were determined over a range of conditions covering the whole physiological range that CR is expected to encounter ., CR ( 500 μM ) was subjected to a wide frequency range ( 0 . 3 Hz to 10 kHz ) of small ( 1 nM ) sinusoidal perturbations of the Ca2+ at a wide range of starting Ca2+ ( 1 nM to 100 μM ) ., The resulting Ca2+ “waves” were close to sinusoidal ., We used their amplitude as output of the filter to determine the transfer function of CR ( Figure 6 ) ., The attenuation of the sine wave is plotted as a function of frequency and resting Ca2+ ., Both models “filter” Ca2+ signals in a very similar way; the signals are less attenuated as the frequency gets higher ( CR acts as a high-pass filter ) or when CR becomes fully occupied as the starting Ca2+ gets higher ., Remarkably , both models show a similar strong increase in attenuation at the lower frequencies , when the Ca2+ concentration gets close to the apparent Kd value of the cooperative sites , i . e . , approximately 1 . 5 μM ., The similarity of the transfer function of CR using either model indicates that they quantify the kinetics of Ca2+ binding by CR in a similar way ( but via different mechanisms ) over the whole physiological range of conditions ., Both our new model , which is closely linked to the Adair-Klotz model 6 , 7 , and the MWC model can be used equally well to quantify the Ca2+-binding kinetics of CR ., However , we consider the new model to facilitate the “intuitive” understanding of how the kinetic properties of the cooperative sites relate to the binding kinetics at the level of the whole protein or at the macroscopic level ., The Adair-Klotz model is the most general description of equilibria in terms of stochiometric binding ., It describes the steady-state equilibrium using the constants ( K1 , K2… . Kn ) for the successive binding ( or macroscopic ) steps , but not as the affinity constants of the individual ( or microscopic ) binding sites:, where, for which in equilibrium , the fractional occupation ( ν ) of a protein P is described by the Adair-Klotz equation:, Usually , rate constants are not denoted in the macroscopic equilibrium equation ( Equation 15 ) ; instead , only the K values are denoted , which is sufficient for steady-state descriptions ., The equilibrium constants for the new model ( KT and KR ) and the MWC model ( KT , KR , and L0 ) can often rather easily be translated into macroscopic K values 10 ( also see Protocol S1 ) ., Therefore , it is fairly simple to relate any of the steady-state constants of cooperative models ( new , MWC , and KNF ) to the more generally used Adair-Klotz equation ., In addition to the calculation of the steady-state equilibrium ( Equation 17 ) , the macroscopic Adair-Klotz model compiles the binding of multiple binding sites into an intuitively easy-to-understand sequential binding model ., It would even be more insightful if one could also obtain the rate constants for each macroscopic step ., Unfortunately , the macroscopic rate constants are generally extremely hard to define when cooperative mechanisms are involved ., For example , the macroscopic kon ( 1 ) for the MCW model depends on the relative amounts of totally unoccupied molecules in the R and T states ( see Figure 2B ) ., At steady state , this equilibrium is fairly straightforward , as this is simply defined by L0 ., However , when the balance is disturbed by a sudden change in Ca2+ concentration , it will disturb the equilibrium between unoccupied molecules in the R and T states ., This equilibrium will settle over some time according to k+ and k− ., During this time , the relative amount of binding sites in states R and T is dynamically changing , making the macroscopic kon ( 1 ) itself dynamic ., With most cooperative models , the macroscopic rate constants will be dynamic because they are dependent on most perturbations ., This makes the rate constants very difficult to interpret ( and to calculate ) ., However , for the new model of cooperativity , the macroscopic rate constants are easily defined and are truly constant ., For instance , for two cooperative sites as described in this paper:, and, Through these simple relationships and according to our data , CR can be quantitatively described as a mixture of two “virtual” CaBPs ., The cooperative part can be described as:, and the independent part as:, where As described above , at a starting Ca2+ around the apparent Kd for the cooperative binding sites , CR will more effectively buffer perturbations at lower frequencies ( Figure 6 ) ., Thus , the Ca2+-buffering kinetics of CR clearly depends on the starting Ca2+ that determines the distribution between states T and R of the cooperative binding sites ., While more cooperative sites get occupied , more Ca2+ binding will take place through the second faster binding step as described in Equation 20 ., To better understand the cooperative nature of Ca2+ binding by CR , we simulated with the new model a 1 μM step in Ca2+ from a resting Ca2+ of 10 nM in the presence of 100 μM CR ., In comparison , we also simulated the widely used synthetic Ca2+ buffers BAPTA ( Kd = 160 nM , kon = 2 × 108 M−1s−1 , Maxchelator software version 10/02 , see Materials and Methods ) and EGTA ( Kd = 70 nM , kon = 1 × 107 M−1s−1 2 ) ., Under these conditions , the Ca2+ decay kinetics mediated by 100 μM CR could be faithfully reproduced by either 7 . 8 μM BAPTA or 153 μM EGTA ( Figure 7A ) ., Now , to compare the binding kinetics of CR to the two synthetic chelators without cooperative binding , we kept all parameters constant , i . e . , 1 μM steps in | Introduction, Results, Discussion, Materials and Methods | Cooperativity is one of the most important properties of molecular interactions in biological systems ., It is the ability to influence ligand binding at one site of a macromolecule by previous ligand binding at another site of the same molecule ., As a consequence , the affinity of the macromolecule for the ligand is either decreased ( negative cooperativity ) or increased ( positive cooperativity ) ., Over the last 100 years , O2 binding to hemoglobin has served as the paradigm for cooperative ligand binding and allosteric modulation , and four practical models were developed to quantitatively describe the mechanism: the Hill , the Adair-Klotz , the Monod-Wyman-Changeux , and the Koshland-Némethy-Filmer models ., The predictions of these models apply under static conditions when the binding reactions are at equilibrium ., However , in a physiological setting , e . g . , inside a cell , the timing and dynamics of the binding events are essential ., Hence , it is necessary to determine the dynamic properties of cooperative binding to fully understand the physiological implications of cooperativity ., To date , the Monod-Wyman-Changeux model was applied to determine the kinetics of cooperative binding to biologically active molecules ., In this model , cooperativity is established by postulating two allosteric isoforms with different binding properties ., However , these studies were limited to special cases , where transition rates between allosteric isoforms are much slower than the binding rates or where binding and unbinding rates could be measured independently ., For all other cases , the complex mathematical description precludes straightforward interpretations ., Here , we report on calculating for the first time the fast dynamics of a cooperative binding process , the binding of Ca2+ to calretinin ., Calretinin is a Ca2+-binding protein with four cooperative binding sites and one independent binding site ., The Ca2+ binding to calretinin was assessed by measuring the decay of free Ca2+ using a fast fluorescent Ca2+ indicator following rapid ( <50-μs rise time ) Ca2+ concentration jumps induced by uncaging Ca2+ from DM-nitrophen ., To unravel the kinetics of cooperative binding , we devised several approaches based on known cooperative binding models , resulting in a novel and relatively simple model ., This model revealed unexpected and highly specific nonlinear properties of cellular Ca2+ regulation by calretinin ., The association rate of Ca2+ with calretinin speeds up as the free Ca2+ concentration increases from cytoplasmic resting conditions ( ∼100 nM ) to approximately 1 μM ., As a consequence , the Ca2+ buffering speed of calretinin highly depends on the prevailing Ca2+ concentration prior to a perturbation ., In addition to providing a novel mode of action of cellular Ca2+ buffering , our model extends the analysis of cooperativity beyond the static steady-state condition , providing a powerful tool for the investigation of the dynamics and functional significance of cooperative binding in general . | The binding of a ligand to a protein is one of the most important steps in determining the function of these two interactive biological partners ., In many cases , successive binding steps occur at multiple sites such that binding at one site influences ligand binding at other sites ., This concept is called cooperative binding , and constitutes one of the most fundamental properties of biological interactions ., The functional consequences of cooperativity can be accurately resolved when reactions are at equilibrium , but mathematical complexity has prevented insights into the dynamics of the interactions ., We studied the protein calretinin , which binds Ca2+ in a cooperative manner and plays an important role in shaping Ca2+ signals in various cells ., We used two models , a widely tested one and a novel , mathematically simplified one , to resolve the dynamics of a cooperative binding process ., The cooperative nature of Ca2+ binding to calretinin results in accelerated binding as calretinin binds more Ca2+ ., This behavior constitutes an important new insight into the regulation of intracellular Ca2+ that cannot be matched by noncooperative artificial Ca2+ buffers ., Our simple mathematical model can be used as a tool in determining the kinetics of other biologically important molecular interactions . | biochemistry, biophysics, neuroscience | A novel and relatively simple mathematical model for the kinetics of cooperative binding reveals, the tuning of Ca2+-buffering kinetics due to cooperative binding in calretinin. |
journal.pgen.1000927 | 2,010 | Epigenetic Regulation of a Murine Retrotransposon by a Dual Histone Modification Mark | Our present view on transcriptional regulation has substantially advanced in recent decades ., The classical model , that the presence of promoter sequences and the availability of transcription factors determine the expression status of corresponding genes , has been extended to a model , in which the accessibility of the DNA is central to transcriptional control ., In eukaryotes , DNA is packed and compacted into chromatin with the nucleosome consisting of DNA and histone proteins as the basic unit ., The degree of compaction – either into inaccessible heterochromatin or open euchromatin – has major implications for the transcriptional potential of associated DNA ., A way to regulate chromatin accessibility is the posttranslational chemical modification of histone proteins ., It can alter chromatin structure and switch genes from a transcriptional repressed to an active state and vice versa 1 ., The best-studied histone modification is the acetylation of N-terminal lysine residues , which correlates with open chromatin and active gene transcription ., In contrast , histone deacetylation is linked to gene silencing and histone deacetylases ( HDACs ) are considered as transcriptional co-repressors 2 ., Only recently , a more general role of histone deacetylation during transcriptional regulation is emerging 3 ., The use of chemical HDAC inhibitors and knock out/down technologies resulted in the identification of an increasing number of target genes controlled by histone acetylation in different cellular contexts 4–6 ., Based on sequence homology , mammalian HDACs have been divided into four classes , of which class I enzymes seem to fulfil more basic cellular functions whereas class II enzymes are thought to play specialised cell-type specific roles ., HDACs are often found to interact with proteins important for stable gene silencing via DNA methylation and constitution of heterochromatin , but also with transcriptional regulators mediating only transient repression ., Crosstalk between histone acetylation and other epigenetic marks is an important feature of HDAC function 7 ., Hence , HDACs are central components of multiple silencing complexes containing additional enzymatic activities such as DNA and histone methylation ., Recent annotation of multiple complete genomes has revealed that a large fraction of eukaryotic genomes consists of repetitive sequences , mainly derived from transposable elements ( TEs ) 8 , 9 ., Most of those sequences are remnants of once active TEs now incapable of transposition because of their host-mediated inactivation followed by subsequent functional erosion and the accumulation of mutations and deletions ., However , some elements remain intact and constitute a constant threat to the integrity of the host genome ., Potentially functional elements can act as insertion-mutagens via targeting protein coding genes or causing chromosome breakage ., Even transpositionally inactive TEs have the potential to trigger illegitimate recombination and genome rearrangement , or to influence neighbouring genes by causing alternative splicing , premature termination , and modulating gene expression patterns 10 , 11 ., Recent estimations suggest that TEs provoke around 0 . 1% ( in humans ) or 10–12% ( in mice ) of all spontaneous germ line mutations , and an increasing number of reports document the contribution of somatic transposon activity to pathological situations 11–13 ., Consequently , organisms were challenged to evolve multiple lines of defence to restrict TE activity , targeting crucial steps in their life cycle 11 , 14 ., One of the most efficient host-directed strategies is to block TE transcription ., This is achieved applying epigenetic mechanisms such as DNA methylation , modification of chromatin structure , and the action of small RNAs 13 , 15–19 ., Different subclasses of small RNAs ( siRNAs , rasiRNAs and piRNAs ) seem to guide and target the silencing machinery , while chromatin modifiers and DNA methyltransferases ( DNMTs ) accomplish the transformation into a heterochromatic , transcriptionally silent state 20 ., Numerous proteins , involved in the establishment , maintenance and read-out of DNA methylation and chromatin modification patterns , are crucial for TE silencing in germline and somatic cells 13 ., However , the exact contribution of DNA methylation , chromatin remodelling , modification of histones and their interplay has not been completely resolved ., Analysis in several model organisms revealed that the different silencing-mechanisms are of varying importance depending on the host organism ., Some ( e . g . yeast and Drosophila ) lack an efficient DNA methylation machinery and therefore depend on alternative silencing pathways , while others ( e . g . mammals and plants ) widely use DNA methylation in concert with other mechanisms to silence TEs 18 , 21 ., Despite the tight collaboration of multiple silencing layers , numerous TEs maintain their capacity to escape host-mediated silencing under certain conditions ., Different types of internal and external stimuli are able to trigger TE activity in a wide range of organisms ., Many of those stimuli can be subsumed as cellular stress and the correlation between TE activity and stress has been in the focus of interest for several decades 22 , 23 ., Recent examples of mammalian TEs , showing increased transcriptional activity upon external stimuli , include members of all major TE orders: SINEs 24 , 25 , LINEs 26 , and LTR elements 27 ., To better understand the role of histone acetylation in transcriptional regulation of TEs in mammalian somatic cells , we monitored the expression of selected , potentially active TEs in mouse fibroblasts with impaired histone deacetylase capacity ., In our present study we found that chemical inhibition of class I and II HDACs , but not genetic inactivation of HDAC1 alone , resulted in local hyperacetylation and expression of a defined subset of VL30 LTR retrotransposable elements ., Furthermore , we show that not only histone acetylation , but histone phosphorylation in concert with acetylation ( phosphoacetylation ) was triggering transcriptional induction of VL30 elements ., We suggest a model , in which external stress signals cause the activation of MAP kinase pathways , ultimately leading to the phosphoacetylation of histones and efficient transcription of VL30 elements ., To confirm that our experimental setup indeed perturbed epigenetic states in mouse fibroblasts , we performed the following experiments: firstly , we monitored the expression levels of proteins targeted by chemical or genetic inhibition ., Protein expression levels of the class I deacetylases HDAC1 , HDAC2 , HDAC3 and the maintenance DNA methyltransferase DNMT1 were surveyed via Western blot ., As shown in Figure 1A , loss of HDAC1 in immortalized fibroblasts led to upregulation of HDAC2 , its closest homologue ., This compensatory mechanism may buffer some effects caused by the loss of HDAC1 , but still total cellular HDAC activity was diminished by about 30% in HDAC1−/− compared to HDAC1+/+ cell lines ( data not shown ) , demonstrating an important enzymatic function of HDAC1 in fibroblasts ., In our experimental setup TSA treatment for 24h did not influence expression of HDAC1 , HDAC2 or HDAC3 , whereas DNMT1 levels were slightly decreased ., A reduction of DNMT1 levels after histone deacetylase inhibitor treatment has been previously reported 28 , 29 ., Aza-dC treatment resulted in slightly reduced amounts of DNMT1 , consistent with earlier reports showing that Aza-dC can trap DNMT1 by covalently linking the enzyme to DNA 30 or lead to proteasomal degradation of the DNMT1 protein 31 ., Protein expression of HDAC1 , HDAC2 and HDAC3 remained unaffected ., Next , we evaluated effects on the biological targets of HDACs and DNMT1 , i . e . the acetylation states of histones and CpG methylation levels of DNA ., To examine CpG methylation levels at defined targets we employed the MS-SNuPE technique 32 ., We quantified DNA methylation levels of three CpG sites within the regulatory LTR region of an IAP retrotransposons , highly methylated in somatic cells 33 ., Whereas neither loss of HDAC1 , nor chemical inhibition of HDACs with TSA for 24h affected levels of DNA methylation , Aza-dC treatment for 24h severely reduced DNA methylation of the IAP LTR region to 50% ( Figure 1B ) ., Furthermore , we analysed histone acetylation levels in wildtype fibroblasts , either untreated ( HDAC1+/+ ) or treated with TSA and Aza-dC and HDAC1-deficient cells ( HDAC1−/− ) ., To that end we extracted histones and performed Western blot analyses using antibodies recognising acetylated histones ., Previous reports had shown that in murine embryonic stem cells genetic inactivation of Hdac1 was accompanied by hyperacetylation of histones H3 and H4 , detectable with antibodies recognising acetylated histone H3 ( H3ac ) and acetylated histone H4 proteins ( H4ac ) 34 ., In our fibroblast system , however , we failed to detect global hyperacetylation of histones in HDAC1−/− cells as compared to HDAC1+/+ cells using such H3ac or H4ac antibodies ( Figure 1C ) ; only when we probed the blots with antibodies specifically recognising the acetylation state of individual lysine residues , we detected weak hyperacetylation occurring at defined residues ( data not shown ) ., In contrast to this limited site-specific response , inhibition of all class I and II HDACs with TSA led to a robust and global hyperacetylation detectable with H3ac and H4ac antibodies ., Finally , interfering with DNA methylation by Aza-dC treatment moderately enhanced global H3 and H4 acetylation levels ( Figure 1C ) ., After defining global effects of HDAC and DNMT inhibition in our cell system , we moved on to specifically address the role of histone acetylation in the regulation of TEs ., We analysed the acetylation state of chromatin associated with defined TEs in wildtype fibroblasts without treatment ( HDAC1+/+ ) , or treated either with TSA or Aza-dC , and HDAC1-deficient fibroblasts ( HDAC1−/− ) ., To monitor their acetylation status , we performed chromatin-immunoprecipitation ( ChIP ) experiments with H3ac and H4ac antibodies ., Furthermore , we included antibodies recognising the C-terminal part of histone H3 to survey nucleosome occupancy and unspecific rabbit IgGs to control the specificity of immunoprecipitation reactions ., Quantitative Real-Time PCR analyses allowed the quantification of H3 and H4 acetylation levels associated with a representative selection of ten murine TEs , including a DNA transposon ( Tigger ) , non-LTR retrotransposons ( LINEs , SINEs ) , and LTR retrotransposons ( VL30 , IAP , ETn , MERVL , MT , ORR1 ) ( Table 1 ) ., We chose LINEs and SINEs , because they constitute the majority of murine TEs and populate the genome in more than 660 , 000 and 1 . 5 million copies , respectively ., LTR elements are less abundant , but harbour the most active TEs within the mouse lineage; MERVL , MT and ORR1 elements are expressed during early embryogenesis 35 and VL30 , IAP and ETn elements have been responsible for a number of germ line transposition events 12 ., The transposon Tigger was included as representative of the class of DNA elements , which is functionally extinct in vertebrates ., As shown in Figure 2A ( upper panel ) , loss of HDAC1 did not influence histone H3 acetylation at any transposon type analysed as compared to wildtype cells ., In contrast , TSA treatment was sufficient to specifically enhance H3 acetylation at SINE B1 and B2 elements and VL30 LTR elements , the latter ones near the promoter as well as throughout the body of the element ., Histone H4 acetylation of TE-associated chromatin was analysed similarly in a subsequent series of experiments ., Again , loss of HDAC1 did not alter histone acetylation at tested transposons , but TSA-induced inhibition of all class I and II HDACs led to increased H4 acetylation at all repeat families investigated ( Figure 2A; lower panel ) ., Thus , TSA treatment exhibited differential effects on histone H3 and histone H4 acetylation of TEs ., Whereas H3 histones were found hyperacetylated at specific elements only , H4 hyperacetylation occurred as a global response ., Interestingly , the elements hyperacetylated at H3 upon TSA treatment , SINEs and VL30 , also showed highest enrichment of H4 acetylation compared to other TEs ., Next , we analysed histone acetylation levels upon loss of DNA methylation ., Incubation of cells with Aza-dC increased H3 acetylation at SINE B1 and VL30 elements ( Figure 2B , upper panel ) , although to a lesser extent than TSA ., Furthermore , IAP element promoter sequences exhibited slightly increased H3 acetylation levels in response to Aza-dC ., Finally , Aza-dC treatment resulted in elevated H4 acetylation levels of SINEs , VL30 and IAP elements ( Figure 2B; lower panel ) ., Since hyperacetylation of histones in general correlates with transcriptional activation , we tested whether increased histone acetylation levels were mirrored by changes in transcriptional activity of TEs ., Therefore we monitored mRNA levels of the ten mouse transposons , listed in Table 1 , in untreated wildtype fibroblasts ( HDAC1+/+ ) , TSA or Aza-dC treated wildtype cells , and HDAC1-deficient fibroblasts ( HDAC1−/− ) by real-time PCR analyses ., We also analysed the expression of another LTR element , MuLV , which comprises endogenous as well as exogenous family members 36 ( Figure S1 ) ., As shown in Figure 3 , upper panel and Figure S1 , loss of HDAC1 did not influence TE expression ., In contrast , general HDAC inhibition via TSA strongly increased VL30 element expression in this fibroblast cell line ( Figure 3 , upper panel ) , as well as in other commonly used fibroblast cell lines such as Swiss 3T3 and NIH/3T3 ( data not shown ) ., Enhanced activity of VL30 elements correlated with hyperacetylation of VL30 chromatin upon TSA treatment ., This suggests that modulation of histone acetylation regulates VL30 element activity ., Expression of SINE B1 , the second element showing H3 and H4 hyperacetylation after TSA treatment , was modestly increased ., Besides IAPs showing a moderate two-fold increase , all other TE candidates investigated did not exhibit major changes in transcriptional activity ., This expression pattern mirrors the unchanged local acetylation state of histone H3 , but not the one of histone H4 acetylation , which was elevated upon TSA treatment ., Therefore , H3 acetylation of the transposons analysed correlates with transcriptional activity , whereas H4 hyperacetylation is not sufficient to trigger transcriptional activation ., The role of DNA methylation in transposon silencing is well established and it has been shown earlier that loss of DNA methylation in somatic cells correlates with VL30 , IAP and MuLV expression 37–39 ., Indeed , Aza-dC treatment in fibroblasts caused highly elevated VL30 , IAP and MuLV mRNA levels , but had no effect on other TEs tested ( Figure 3 , lower panel and Figure S1 ) ., However , it was surprising that the majority of murine elements tested could not be reactivated with Aza-dC ., This finding suggests that those transposons are not permissive for transcriptional activation in our fibroblast system , either because they lack regulatory features necessary for efficient transcriptional activation or additional mechanisms contribute to their silencing 40 ., ChIP analyses of TE-associated chromatin showed that two elements that were highly expressed upon Aza-dC treatment ( VL30 and IAP elements ) also gained hyperacetylation after DNMT inhibition ( Figure 3 , lower panel ) ., SINE B1 and SINE B2 elements , both showing increased acetylation upon Aza-dC treatment , were not or only modestly activated ., In summary , we could establish a correlation between the transcriptional activity of TEs and the acetylation state of associated chromatin ., In particular histone H3 acetylation correlated well with increased expression of certain TEs ., Furthermore , we identified VL30 elements as the single candidate transposon efficiently regulated via acetylation of associated chromatin ., Retrotransposition is a complex process depending on multiple steps , each of which can be targeted by regulatory mechanisms 11 ., Alterations in RNA stability , translation efficiency , nuclear shuttling , editing of template RNA , or integration efficiency can have severe impact on the rate of successful transposition ., Treatment of cells with the HDAC inhibitor TSA has strong effects on the chromatin state ., Since chromatin constitutes the substrate for the last step of transposition , i . e . integration , we tested whether histone modification patterns might influence transposition efficiency , independent of TE transcription ., Genetic screens in yeast have uncovered numerous host factors either restricting or enhancing transposon activity 41–45 ., Strikingly , a substantial fraction of the proteins identified has a role in chromatin biology; for instance , SIN3 and RPD3 , both major components of yeast HDAC complexes , were found to restrict Ty1 activity and modulate target site selection ., Interestingly , SIN3 and RPD3 do not act on the transcriptional level , as they do not alter Ty1 mRNA levels 45 ., To test whether changed histone acetylation levels induced by TSA treatment might have similar effects in mammalian cells , we determined transposition levels of four mammalian retroelements in untreated and TSA treated cells ., To this end we took advantage of an experimental system described earlier 46–48 , in which episomally delivered retrotransposons are marked with a neo-transgene , which gets activated only after its successful retrotransposition ( Figure S2A ) ., Following this approach we quantified the transposition efficiency of murine IAP and MusD LTR retrotransposons , and murine and human L1 elements in untreated and TSA treated cells ., In order to avoid potential interference with endogenously transcribed IAP , MusD and L1 elements , we performed the experiments in the human U2OS osteosarcoma cell line ., In addition , we monitored transcript levels of episomally delivered retroelements via real-time PCR ( Figure S2B ) ., As presented in Figure S2C , TSA treatment led to a 2- to 4-fold increase in transposition rates , a moderate increase compared to findings reported elsewhere , where mutation of TE restricting factors in yeast led to an increase between 5 and several hundred times 45 ., In our system , enhanced transposition was accompanied by a 2- to 3-fold increase in TE mRNA abundance ., These data suggest that TSA treatment led to enhanced expression of marked elements , which in turn resulted in increased transposition rates ., Therefore , under our experimental conditions , inhibition of deacetylases by TSA was not sufficient to influence TE transposition rates independent of TE expression , suggesting transcription as the rate-limiting factor for transposition ., In contrast to single copy genes , TEs exist in multiple copies dispersed throughout the genome ., Individual elements within a TE family are different in respect of genomic environment , influencing chromatin status and transcriptional potential ., Furthermore , transposons are under high selective pressure and mutations mediated by host and/or transposon-intrinsic factors lead to the generation of pools of elements with individual sequence specificities within given TE families ., Therefore , a family of elements is a collection of similar , but heterogeneous individuals , differing in external and internal features ., Upon external signals increased transcriptional activity could originate from the upregulation of all TEs within a family , of a subgroup of elements exhibiting specific features , or even a single copy with unique intrinsic properties ., In order to distinguish between these different scenarios , we aimed to determine the chromosomal origin of VL30 and IAP derived transcripts detected in untreated , TSA and Aza-dC treated wildtype fibroblasts ., RNA from three independent biological samples each ( untreated , TSA , Aza-dC ) was collected , reversely transcribed , and subjected to PCR amplification ., To amplify IAP derived transcripts , primers annealing in the highly conserved gag region were used ( Figure S3A ) ., For VL30 transcript amplification primers within the 3′ UTR and 3′LTR R region were chosen ( Figure 4A ) ., To exclude a bias due to individual PCR reactions , six independent PCR reactions for each condition were pooled , PCR products were cloned and sequenced ., As control , the same procedure was also performed with genomic DNA as template ., This approach allowed us to amplify chromosomal copies of both candidate elements and to estimate the diversity of IAP and VL30 elements in our cell system ., Each sequenced clone corresponding to an individual transcribed ( or genomic ) TE was then analysed as follows: to identify genomic elements serving as potential sources of transcription , sequences were blasted against the annotated murine genome ( http://www . ncbi . nlm . nih . gov/blast/Blast . cgi ) ., Genomic sequences exhibiting at least 98% identity to the cloned sequences were considered to show sufficient homology to serve as potential sources of transcription and were analysed further ., If more than one genomic sequence showed identity >98% , the sequence with the highest E-value was chosen ., That way ∼90% of the cloned sequences were assigned to distinct genomic elements ( IAP: 115 of 127 clones; VL30: 111 of 122 clones ) ., Individual characteristics , such as chromosomal localisation and length were examined for each element as summarised in Table S1 ., Next , we defined whether the transcribed IAP and VL30 elements were derived from multiple genomic sources , a small number of elements , or even unique elements ., Therefore , we analysed the diversity of the identified genomic TEs , constituting bona fide sources of transposon transcription ., Following this approach , each IAP transcript could be assigned to a different genomic element , indicating that IAP transcripts originated from multiple loci throughout the genome ( see chromosomal origins and individual features of elements listed in Table S1 ) ., In untreated cells , VL30 transcripts were of heterogeneous origin , with a broad level of sequence diversity similar to the chromosomal VL30 elements that were amplified from genomic DNA ., In contrast , upon TSA and Aza-dC treatment , VL30 transcripts could be assigned to only four and two genomic loci , pointing towards a limited number of VL30 elements serving as source of expression ., TSA treatment led to the accumulation of transcripts related to the genomic loci NT_039207 4840bp on chromosome 2 ( 36 of 41 sequences ) , NT_039341 4902bp on chromosome 6 ( 3 of 41 ) , NT_039462 5167bp on chromosome 8 ( 1 of 41 ) , and NT_039649 4919bp on chromosome 17 ( 1 of 41 ) ., Aza-dC treatment resulted in transcripts similar to NT_039207 4840bp on chromosome 2 ( 13 of 26 sequences ) and NT_039341 4902bp on chromosome 6 ( 13 of 26 ) ( Table S1 ) , Alignment of VL30 sequences directly obtained after cloning and sequencing further showed that transcripts assigned to the genomic locus on chromosome 2 ( NT_039207 4840bp ) were almost identical , indicating a single locus as transcriptional source ., In contrast , transcripts assigned to the locus on chromosome 6 ( NT_39341 4902bp ) were slightly divergent , demonstrating that they originate from a group of related VL30 elements , of which only one copy is annotated in the published mouse genome ( data not shown ) ., The long terminal repeat regions of LTR elements are enriched with regulatory sequences important for transcriptional regulation ., Therefore , we determined the structure of the LTR regions associated with expressed IAP and VL30 elements in more detail ., First , we analysed the LTRs associated with transcribed IAP elements , which revealed a marked length polymorphism between different LTRs: the shortest LTR , identified with an IAP element expressed in untreated fibroblasts , comprised 322bp , whereas the longest LTR , found in an IAP element expressed after Aza-dC treatment , was 456bp long ., Alignment of IAP LTR sequences disclosed that this difference was due to the presence or absence of nucleotides in a highly variable CT-dinucleotide-rich stretch within the R region of the LTR ( Table S1 and Figure S3B ) ., The region encoded a varying number of tandem repeat sequences containing a 13bp core sequence ( TCTCTCTTGCTTC ) , previously described as polymorphic marker for different IAP subtypes 49 ., Interestingly , the average length of this stretch was significantly longer in LTRs associated with IAPs expressed upon Aza-dC treatment ( Figure S3C ) , indicating that upon loss of DNA methylation an increased number of tandem repeats might be favourable for expression ., As shown in earlier studies the structure of the VL30 LTR region is crucial for stimulus mediated activation of VL30 expression ( reviewed in 27 ) ., Based on their highly divergent U3 regions , VL30 LTRs have been classified into four distinct subtypes: I , II , III and IV 50 ., To allocate the VL30 elements isolated in our transcript analysis to the four subtypes , we isolated the 5′LTR sequences from the genomic sequences , aligned them together with VL30 LTR reference sequences ( NVL3: subclass I , BVL1: subclass III , NVL1/2: subclass IV ) and generated a phylogenetic tree ., As depicted in Figure 4B , all elements including the references were located in three distinct clades corresponding to subclass I , III and IV elements ., Elements expressed in untreated cells mainly comprehend subtype III containing LTRs , which also associate with most genomic elements ( Figure 4B and Table S1 ) ., As described above , transcripts arising upon TSA and Aza-dC treatment , mainly originate from two types of loci: one VL30 insertion on chromosome 2 ( NT_039207 4840bp ) and several other loci showing high similarity to another element on chromosome 6 ( NT_039341 4902bp ) ., Interestingly , whereas the VL30 element on chromosome 6 is flanked by LTRs containing subtype I U3 regions at both ends , the element located on chromosome 2 exhibits an unusual structural feature: it is associated with two different LTRs , a 5′ LTR containing a U3 region belonging to subtype I , but a 3′ LTR containing a U3 region of subtype IV ( Figure 4A , lower part ) ., Transposon transcription depends on the 5′ LTR , therefore we conclude that expression of this VL30 element is controlled by LTR sequences comprising a subtype I U3 region ., 40 out of 41 transcripts isolated from TSA treated fibroblasts were associated with a LTR containing a subtype I U3 region , suggesting that this LTR design might be favourable for VL30 expression upon TSA treatment ., However , 36 of those 40 were derived from a single genomic source ., Therefore we wanted to clarify if increased VL30 transcription upon TSA treatment is exclusively caused by activation of this particular locus mirroring a locus specific phenomenon or if it reflects a general feature of subtype I elements ., We surveyed the chromatin signature at the 5′LTRs of five genomic VL30 elements via ChIP using flanking primers: the highly responsive hybrid element on chromosome 2 ( NT_039207 4840bp ) , two other subtype I elements ( NT_039649 4919bp , NT_039341 4902bp ) , one subtype III ( NT039589 5225bp ) and one subtype IV element ( MW_001030889 5248bp ) ( Figure 4C ) ., Interestingly , we detected increased levels of H3K4me3 ( a mark for active promoters ) upon TSA treatment at all five VL30 elements , with the highest levels at the hybrid element , followed by the two other subclass I elements ( Figure 4C ) ., The subclass III and IV element showed the lowest response ., A similar pattern could be observed for H3ac and H4ac ., In addition we monitored the repressive histone mark H3K9me2 , which was low in all elements and remained unchanged upon TSA treatment ., Another repressive mark , H3K27me3 showed variable levels with no clear correlation with VL30 subtypes ., Together these data suggest an increased responsiveness to activation of the hybrid element , followed by other subclass I elements and finally subclass III and IV elements ., Based on the finding that VL30 elements respond to histone acetylation , we examined the potential activation mechanism for VL30 elements in detail ., In the recent past VL30 elements have been studied extensively and others showed that VL30 elements expression is highly sensitive to a plethora of internal and external stimuli 27 , 39 , 51 ., Interestingly , many of those stimuli can be described as internal or external stress in the broadest sense , and a considerable number among them is able to trigger signalling cascades such as the ERK or p38 pathways ., Signalling via these pathways ultimately leads to phosphorylation and activation of several transcription factors but also to phosphorylation of histone H3 ., Phosphorylation of serine 10 at histone H3 ( H3S10p ) during interphase is a rare event that correlates with the activation of specific target genes 52 , 53 ., Importantly , the presence of the H3S10ph mark at promoter-associated chromatin coincides with acetylation of neighbouring lysines K9 and K14 leading to a dual histone modification termed phosphoacetylation 54–56 ( Figure 5A ) ., Histone H3 phosphoacetylation correlates with the activation of the so called “immediate early” genes c-fos and jun-B , as well as late inducible genes such as HDAC1 57 ., The finding that the HDAC inhibitor TSA was sufficient to highly induce VL30 expression in logarithmically growing fibroblasts , where ERK pathways are activated through the action of growth factors , prompted us to ask whether both stimuli , histone acetylation and histone phosphorylation , were required for full VL30 activation ., To answer this question , we performed all following experiments with serum-arrested Swiss 3T3 fibroblasts ., Serum withdrawal arrests cells in G0 and eliminates mitotic H3 phosphorylation , a modification that targets the majority of histone H3 molecules during mitosis and is associated with chromatin condensation ., Furthermore , the absence of the serum-dependent activity of different kinase pathways allows to selectively activate specific signalling cascades , thereby triggering H3 phosphorylation ., To induce the ERK pathway in fibroblasts , cells were arrested in G0 for 72h and then treated with foetal calf serum ( FCS ) ., Alternatively , anisomycin ( Aniso ) was used to activate the p38 stress pathway ., Additionally cells were treated with TSA to induce hyperacetylation ., To analyse simultaneous phosphorylation and acetylation of histone H3 we used an antibody that recognises histone H3 only in the presence of the dual H3S10phK14ac mark ( Figure 5A , Figure S4 ) ., Figure 5B shows the effect of different compounds on histone H4 acetylation and histone H3 phosphoacetylation in G0 arrested Swiss 3T3 fibroblasts ., As expected , TSA led to the accumulation of hyperacetylated histones as detected by the H4ac antibody; H3 phosphoacetylation was low in untreated and TSA treated G0 cells and slightly induced upon Aniso or FCS treatment ., Combined treatment with Aniso/TSA or FCS/TSA led to enhanced phosphoacetylation of histone H3 ., It was shown previously that , even upon Aniso/TSA stimulation , the fraction of phosphoacetylated histone H3 is only minute 52 , 58 and that the mark is localised to a restricted number of genomic loci in Aniso/TSA treated fibroblasts ( Simboeck E . , unpublished results ) ., Hence we tested , if phosphoacetylated histones were associated with VL30 elements ., To this end we isolated chromatin from G0 arrested fibroblasts that were untreated or treated with TSA , Aniso , or Aniso/TSA ., We performed ChIP experiments using the H3S10ph/H3K14ac antibody or an unspecific control antibody and surveyed phosphoacetylation levels of the VL30 5′LTR regulatory region via semi- | Introduction, Results, Discussion, Materials and Methods | Large fractions of eukaryotic genomes contain repetitive sequences of which the vast majority is derived from transposable elements ( TEs ) ., In order to inactivate those potentially harmful elements , host organisms silence TEs via methylation of transposon DNA and packaging into chromatin associated with repressive histone marks ., The contribution of individual histone modifications in this process is not completely resolved ., Therefore , we aimed to define the role of reversible histone acetylation , a modification commonly associated with transcriptional activity , in transcriptional regulation of murine TEs ., We surveyed histone acetylation patterns and expression levels of ten different murine TEs in mouse fibroblasts with altered histone acetylation levels , which was achieved via chemical HDAC inhibition with trichostatin A ( TSA ) , or genetic inactivation of the major deacetylase HDAC1 ., We found that one LTR retrotransposon family encompassing virus-like 30S elements ( VL30 ) showed significant histone H3 hyperacetylation and strong transcriptional activation in response to TSA treatment ., Analysis of VL30 transcripts revealed that increased VL30 transcription is due to enhanced expression of a limited number of genomic elements , with one locus being particularly responsive to HDAC inhibition ., Importantly , transcriptional induction of VL30 was entirely dependent on the activation of MAP kinase pathways , resulting in serine 10 phosphorylation at histone H3 ., Stimulation of MAP kinase cascades together with HDAC inhibition led to simultaneous phosphorylation and acetylation ( phosphoacetylation ) of histone H3 at the VL30 regulatory region ., The presence of the phosphoacetylation mark at VL30 LTRs was linked with full transcriptional activation of the mobile element ., Our data indicate that the activity of different TEs is controlled by distinct chromatin modifications ., We show that activation of a specific mobile element is linked to a dual epigenetic mark and propose a model whereby phosphoacetylation of histone H3 is crucial for full transcriptional activation of VL30 elements . | The majority of genomic sequences in higher eukaryotes do not contain protein coding genes ., Large fractions are covered by repetitive sequences , many of which are derived from transposable elements ( TEs ) ., These selfish genes , only containing sequences necessary for self-propagation , can multiply and change their location within the genome , threatening host genome integrity and provoking mutational bursts ., Therefore host organisms have evolved a diverse repertoire of defence mechanisms to counteract and silence these genomic parasites ., One way is to package DNA sequences containing TEs into transcriptionally inert heterochromatin , which is partly achieved via chemical modification of the packaging proteins associated with DNA , the histones ., To better understand the contribution of histone acetylation in the activation of TEs , we treated mouse fibroblasts with a specific histone deacetylase inhibitor ., By monitoring the expression of ten different types of murine mobile elements , we identified a defined subset of VL30 transposons specifically reactivated upon increased histone acetylation ., Importantly , phosphorylation of histone H3 , a modification that is triggered by stress , is required for acetylation-dependent activation of VL30 elements ., We present a model where concomitant histone phosphorylation and acetylation cooperate in the transcriptional induction of VL30 elements . | genetics and genomics/epigenetics, molecular biology/histone modification | null |
journal.ppat.1006471 | 2,017 | Cross-modulation of pathogen-specific pathways enhances malnutrition during enteric co-infection with Giardia lamblia and enteroaggregative Escherichia coli | Childhood malnutrition and its resultant host developmental , metabolic , and immunologic sequelae continue to affect 156 million children less than five years of age worldwide 1 ., Impaired child growth attainment is epidemiologically associated with, 1 ) alterations in resident intestinal microbiota ( dysbiosis ) 1 , 2;, 2 ) increased susceptibility to multiple concurrent and recurrent enteric pathogens 3 , 4;, 3 ) intestinal dysfunction together with markers of increased intestinal myeloid 5 and T-cell activation ( termed Environmental Enteropathy ( EE ) ) 3 , 6; and, 4 ) perturbations in gut microbial-host co-metabolism 7 ., Emerging data from the Malnutrition and Enteric Diseases ( Mal-ED ) multisite international study has revealed that cumulative pathogen exposures confer a high associated risk for poor growth 8 ., These exposures are diverse , with prokaryotic pathogens enteroaggregative Escherichia coli and the oftentimes persistent protozoan Giardia lamblia among the most commonly detected 9 , 10 ., Elucidating mechanistic pathways by which these diverse microbial triggers interact to potentiate the malnourished condition could improve restorative interventions for malnourished children ., Indeed , data from randomized control therapeutic trials expose knowledge gaps in our biological understanding of microbial drivers of malnutrition 11–15 ., Inconsistent or only partial benefits are achieved from interventions that target a single component of this pathogenesis such as nutrient supplementation alone 11 or in combination with broad-spectrum antibacterials 12 , 13 , anti-parasitic drugs 14 , or anti-inflammatory agents 15 ., Furthermore , field studies of endemic pediatric Giardia have associated Giardia with decreased risk of severe diarrhea and inflammatory biomarkers of EE , yet increased risk for growth impairment , suggesting that for some pathogens , novel pathways may contribute to impaired child development 9 ., Using murine models of malnutrition that result in diet-dependent changes in the microbiota 16 we have published that challenge with the human EAEC isolate ( strain 042 ) 17 or G . lamblia ( assemblage B , strain H3 , cysts ) 18 is sufficient to impair growth and disrupt mucosal architecture , but with unique intestinal pathologies ., We have not previously investigated whether and how these individual pathogens interact with the resident microbiota or with one another , in part due to the use of antimicrobial mediated microbiota depletion to support human pathogen colonization in mice 17–21 ., Also , although xenotransplantation of feces from discordant healthy and malnourished children into gnotobiotic murine recipients demonstrate the functional ability of human-derived microbial communities to selectively recapitulate phenotypes of undernutrition , dysbiosis , 1 , 2 , 22 , and disrupted metabolism 1 , 23 , these studies are only beginning to examine the influence of enteropathogenic bacteria accompanying the dysbiosis 22 , 23 and have yet to uncouple these effects from direct or residual influences of intestinal eukaryotes also present in donor feces 1 ., Thus , while existing murine models provide insight into individual nutritional and microbial triggers that influence gut function , metabolism , and host growth , none to date have intentionally examined the integrated effects of dysbiosis with sequential and diverse pathogen exposures common in malnourished children ., To address how gut microbial adaptations to undernutrition combine with cumulative enteropathogen burdens to influence host growth , mucosal immune responses , and metabolism , we developed a new integrated model of protein-malnutrition induced microbial disruption and multi-pathogen enteropathy ., G . lamblia and EAEC , pathogens commonly detected in malnourished children , were selected as pathogens of interest ., In addition to identifying new pathogen-specific pathways that contribute to malnutrition , we demonstrate co-modulation of mucosal immune and metabolic responses that converge to worsen host growth ., Furthermore , gut microbial-mediated proteolysis was amplified in the increasingly wasted host along with exhaustion of co-metabolic adaptations in energy regulatory and compensatory metabolic pathways ., Murine intestinal microbiota can differentially prevent prolonged Giardia lamblia colonization , even in T and B-cell deficient hosts ., Thus , we and other investigators have used continuous antibiotics ( ampicillin , vancomycin , neomycin ) ( Abx ) in drinking water to enhance G . lamblia infection 18–21 ., Using this Abx cocktail we previously published that G . lamblia ( Assemblage B , strain H3 cysts ) challenge results in detectable shedding at 104−105/gram feces by qPCR of the 18S small ribosomal subunit through the first 5–7 days post-challenge ( early infection ) ., But unlike clearance following G . lamblia ( Assemblage A , strain WB trophozoites ) challenge , G . lamblia H3 shedding increases by ~ 2 logs after day 9 and remains consistent through 4–6 weeks together with small intestinal trophozoite colonization ( persistent infection ) 18 ., To test the hypothesis that the disrupted intestinal 16S community during protein malnutrition 16 would functionally impair microbiota-mediated colonization resistance 24 , we eliminated Abx from the model ., Weaned mice were fed either a protein deficient ( 2% protein ) diet ( PD ) or an isocaloric but protein sufficient ( 20% protein ) control diet ( CD ) for 15 days prior to challenge with 106 G . lamblia H3 ( Assemblage B ) cysts ( Fig 1A ) ., We previously published that this duration of acclimation on diet is sufficient to establish discrepant 16S rRNA genetic profiles 16 ., Small intestinal tissues harvested at 5 days ( early ) and 28 days ( persistent ) post-infection ( dpi ) demonstrated higher abundance of Giardia on day 5 in mice fed PD and only mice fed PD remained infected through 28 days by qPCR ( Fig 1B ) ., Histopathology confirmed the presence of mucosal-associated Giardia trophozoites in H3 cyst-challenged mice fed PD ( Fig 1C ) ., In separate experiments , we confirmed that parasites persisted in mice fed CD and challenged with 106 G . lamblia H3 cysts if concurrently treated with Abx ., Giardia was detected in the duodenum of abx-treated mice fed CD at 35 dpi , and regardless of diet , Giardia was detected in stools through 42 dpi ( S1 Fig ) ., Finally , consistent with the greater infectious potential of the partially stomach-acid resistant parasite cyst stage compared with the excysted trophozoite stage , we confirmed that regardless of Abx , only H3 cysts and not axenized H3 trophozoites were sufficient to achieve consistent Giardia colonization by both light microscopy and qPCR ( S1 Fig ) in this model ., Similar to what we observed previously in abx-treated mice 18 , protein deficiency combined with Giardia to impair host growth ( P<0 . 05 PD-Giardia vs uninfected CD-fed control in Fig 1D ) ., Giardia infection in mice fed a PD diet also had greater duodenal bacterial abundance , measured by both universal 16S rRNA qPCR/gram tissue and V3-V4 specific amplicon product , than infected mice fed CD ( P<0 . 05 ) ( Fig 1E and 1F ) ., Consistent with several features of microbial alterations in both malnourished children 25 and protein-deficient diet fed mice 26 , the duodenal 16S rRNA composition in mice fed PD demonstrated an increased Firmicutes:Bacteroidetes ratio ( S1 Fig ) that was mainly driven by an increase in the abundance of Clostridiales ( Fig 1G ) ., Giardia tended to enhance this skew towards Firmicutes together with a reduction in Bacteroidetes from 12%—7% ( Fig 1H ) ., Thus , rather than excluding Giardia , the small intestinal microbiota in mice fed the protein deficient diet permitted persistent Giardia infection whereas abx were necessary for prolonged parasite detection in mice fed the CD diet ., To examine whether PD in this model had interfered with protective mucosal responses against Giardia we performed flow cytometry on upper small intestinal lamina propria in the same mice ., Regardless of infection , mice fed PD demonstrated a skew toward increased myeloid cells ( CD11b+ ) with reciprocal reductions in T-cells ( CD3ε+ ) among CD45+ cells analyzed compared with mice fed the CD-diet ( Fig 1I ) ., In addition , mice fed PD demonstrated a reduction in B-cell frequency ( B220+ ) at 5 dpi compared with infected mice fed CD ( P<0 . 05 ) ( Fig 1I ) ., Concurrently , we analyzed mucosal production of key cytokines that promote Giardia clearance , such as IL-6 and IL-17A 27 , 28 , compared with those that resembled the profile of prolonged infections in children ( increased IL-4 ) 28 ( Fig 1J ) ., Corresponding to persistent infection , mice fed PD demonstrated a trend toward decreased IL-6 , IL-17A , and IL12p40 , with a significant increase in IL-4 ( P<0 . 05 for IL-4 ) , irrespective of Giardia infection ., Having established that Giardia incorporates into a disrupted intestinal microbiota in mice fed the PD diet , we next investigated the role of resident microbiota as determinants of growth impairment during Giardia infection in PD diet fed mice ( Fig 2A ) ., Continuous exposure to the antibacterials that have no anti-giardial activity ( Abx ) prevented Giardia-induced growth impairment ( Fig 2B ) even despite an early increase in Giardia fecal shedding ( Fig 2C ) and a similar intestinal Giardia burden through 14 dpi ( Fig 2D ) ., The Abx exposure resulted in a fecal dominance of Lactococcus ( >98% ) regardless of infection ( Fig 1S ) ., In non-Abx treated mice , there were no significant differences in fecal 16S rRNA composition between infected and non-infected mice ( Fig 2E ) , although Enterobacteriaceae tended to be over-represented in Giardia infected mice ( Fig 2F ) ., Targeted qPCR to determine absolute abundance of predominant taxa ( Firmicutes and Bacteroidetes ) as well as Enterobacteriaceae identified reductions in both Firmicutes and Bacteroidetes to below the limit of detection in the duodenum and 3–4 log decreases in the feces during Abx treatment regardless of infection ( Fig 2G ) ., In non-Abx treated Giardia-infected mice there was a 1 . 5 log increase in Firmicutes ( P<0 . 05 ) in the duodenum ( Fig 2G ) ., In addition , consistent with findings that increased numbers of E . coli in small intestinal aspirates recovered from patients with giardiasis correlate with greater symptom severity 29 , increased fecal Enterobacteriaceae abundance at 15 dpi in non-Abx treated mice was a better predictor of poor growth in individual Giardia-infected mice than Giardia burden in either stool or duodenum ( Fig 2H ) ., To test whether alterations in intestinal microbiota and mucosal immune responses during persistent Giardia infection would enhance or diminish growth impairment during enteropathogen co-infection , we next developed a sequential co-infection model using one of the most common pathogen isolated in malnourished children , EAEC 10 ., For these experiments we used EAEC042 that elicits acute myeloid cell inflammation during other nutrient deficient states 17 ., First , we established that challenge with 109 EAEC042 in mice fed the PD diet but not mice fed CD diet led to rapid weight loss ( 7% body weight compared with uninfected PD-fed controls ( P<0 . 001 3 dpi ) ( Fig 3A ) and mucosal inflammation that persisted through 14 days post-EAEC challenge ( Fig 3B ) ., Next , we acclimated mice on either the PD or CD diet for 15 days prior to Giardia exposure and then sequentially challenged with EAEC042 during the persistent phase ( 14 dpi ) of Giardia infection ( Fig 3C ) ., The two pathogens combined to enhance weight loss in mice fed the PD diet ( ~100% greater loss of initial weight , P<0 . 05 in co-infected mice compared with uninfected PD-fed controls ) ( Fig 3D ) ., In mice fed CD and co-infected with both pathogens , no weight loss was observed ( Fig 3D ) ., Giardia did not influence EAEC042 stool shedding which was 2 logs greater in mice fed the PD diet as determined by qPCR of the EAEC-specific aap target ( Fig 3E ) ., Giardia , however , did alter inflammatory markers of environmental enteropathy when present alone and during EAEC co-infection in protein deficient fed mice ., Myeloperoxidase ( MPO ) a product of activated neutrophils , was variably detected in the mice fed protein deficient ., Fecal MPO tended to be elevated in response to either pathogen alone , but paradoxically decreased to levels similar to uninfected controls in co-infected mice ( Fig 3F ) ., Calprotectin ( Cp ) , another marker of myeloid cell activation , was elevated only in EAEC042 mono-infected animals , but was decreased in any Giardia infected group ( Fig 3F ) ., Lipocalin-2 ( LCN ) , a marker of either neutrophil or epithelial cell activation was elevated only in Giardia-infected mice regardless of EAEC co-infection ., Immune responses in the mucosal compartment were also altered in persistent Giardia infected mice later challenged with EAEC ., EAEC infection led to significant increases in myeloid lineage ( CD11b+ cells ) in the ileum at 17 dpi Giardia challenge and 3 dpi EAEC challenge ( Fig 3G ) ., The increased proportion of lymphocytes ( both CD3ε+ and B220+ cells ) at 28 dpi Giardia challenge and 14 dpi EAEC challenge ( Fig 3H ) was similar in either EAEC mono-infected or co-infected mice ., In contrast , total LPLs , particularly lymphocytes ( CD3ε+ and B220+ cells ) were decreased during persistent ( 17 dpi ) Giardia infection compared with uninfected controls ( S2 Fig ) ., Using a broad-based luminex 32-plex panel we performed an unbiased analysis of cytokine and chemokine responses on all protein deficient diet fed mice at 17 dpi Giardia challenge and 3 dpi EAEC challenge ., We detected 28 of 32 targets in at least 2 mice in each group that are shown as fold change relative to uninfected controls ( Fig 3I ) ., Both pathogens modulated the cytokine/chemokine response alone and during co-infection ., In all conditions , IL1α , a pro-inflammatory alarmin that is released by enterocytes during intestinal injury 30 , was elevated in all groups and reached significance in Giardia infected mice ( ~30-fold ) and robustly increased in co-infection ( ~80-fold ) ., IL-9 was significantly elevated in Giardia mono-infected and by ~20-fold in co-infected mice , together with a tendency towards greater IL-4 and IL-13 ., Each group demonstrated a decrease in IFNγ , that was significant in EAEC mono-infected mice ., CCL5 was elevated ( ~2 fold ) in EAEC infected and co-infected mice ., Consistent with the early expansion of myeloid cells in EAEC infected mice , CXCL8 ( IL-8/KC ) was also elevated in EAEC and co-infected mice ( ~1 . 6 fold ) ., CCL11 ( eotaxin ) was uniquely elevated ( ~6-fold ) only in co-infected mice ., This change corresponded to a trend toward increased eosinophils ( CD45+SiglecF+ ) among myeloid cells in co-infected compared with EAEC mono-infected mice ( 52% vs 39% , ns ) at later timepoints ( Fig 3H ) ., Changes in select cytokines and chemokines in Giardia mono-infected mice from early ( 5 dpi ) to persistent ( 17 dpi ) timepoints compared with uninfected controls revealed alterations in mucosal immune responses during persistent Giardia infection ., Giardia lead to progressive increases in IL-1α ( P<0 . 05 ) and IL-2 ( P<0 . 05 ) as well as IL-4 and IL-13 , but IFNγ progressively decreased ( P<0 . 05 ) in persistently infected mice ( S2 Fig ) ., To determine whether either pathogen alone or the pathogens in combination altered gut microbial host metabolism , we performed 16S rRNA sequencing in feces simultaneously with urinary metabolic profiling ( metabonomics ) using 1H nuclear magnetic resonance ( NMR ) spectroscopy in mice fed the PD diet ( Fig 4 ) ., In this experiment , weaned mice were highly susceptible to weight loss following 106 G . lamblia H3 cyst challenge , that was further potentiated following EAEC co-infection six days later ( Fig 4A ) ., Focusing first on 16S rRNA sequencing , in Giardia mono-infected mice , phyla-level changes at day 7 and day 13 showed consistent relative increases in Firmicutes and reductions in Verrucomicrobia ( Akkermansia mucinophila ) in Giardia mono-infected mice ( Fig 4B ) ., Anaerobes such as Clostridiales members ( day 7 and day 13 after G . lamblia challenge ) and Turicibacter ( day 7 after G . lamblia challenge ) as well as Enterococcus sp ., ( day 13 after Giardia challenge ) accounted for the Firmicutes expansion in Giardia mono-infected mice ( S3 Fig ) ., Either Giardia or EAEC mono-infected mice had a reduction in Bifidobacterium pseudolongum ., In EAEC mono-infected or co-infected mice Enterobacteriaceae were increased relative to uninfected controls or Giardia mono-infection ( S3 Fig ) ., Phyla-level 16S rRNA composition in co-infected mice otherwise more closely resembled EAEC mono-infection ., Orthogonal projection to latent structures-discriminant analysis ( OPLS-DA ) coefficient plots identified a range of urinary metabolic perturbations induced by Giardia infection on both day 7 and day 13 ( Fig 4C ) ( Q2Y = 0 . 40; P = 0 . 02 vs uninfected PBS controls ) many of which were also elevated in co-infected compared with EAEC mono-infected mice ( Q2Y = 0 . 82; p = 0 . 001 ) ( Fig 4D ) ., We observed no significant difference in the OPLS-DA metabolic profiles of Giardia infected mice between day 7 or 13 days post-challenge ( Q2Y = 0 . 33 R2X = 0 . 24 , P = 0 . 12 ) ., Significantly altered metabolites are summarized in a heat map in Fig 4E along with their correlation to class membership ., Focusing first on metabolites unique to Giardia infection ( Fig 4E ) , consistent with Giardia trophozoite reliance upon on host-derived lipids for membrane synthesis and optimal growth ( ie . lecithin , gylcocholic and taurocholic bile ) 31 , Giardia-infected mice demonstrated increased excretion of bile acid constituents , phosphatidylcholine ( PC ) coupled with choline breakdown metabolites methylamine ( MA ) and dimethylamine ( DMA ) and the taurine metabolite isethionate ., These indicators of bile acid deconjugation and lipid breakdown were present on both day 7 and day 13 post-Giardia challenge ( Fig 4E ) ., Increases in MA and DMA occurred independent of a concurrent increase in the microbial-dependent precursor trimethylamine ( TMA ) or its hepatic oxidized metabolite TMAO , a biochemical pattern resembling that observed with Kwashiorkor-type malnutrition 1 , and were thus suggestive of increased choline availability in the small intestine rather than downstream gut microbial-dependent choline breakdown ., Alanine , a by-product of Giardia glucose fermentation , was elevated at day 13 post-Giardia challenge in mono-infected mice , while pipecolic acid , one of the most abundant amino acid byproducts of Giardia metabolism in vitro 32 was identified in Giardia-infected mice at both timepoints , regardless of co-infection ., Giardia also enhanced gut microbial-host co-metabolites of aromatic amino acids including tyrosine ( 4-cresol glucuronide ( 4-CG ) and 4-cresyl sulfate ( 4-CS ) and 4-hydroxyphenylacetyl ( 4-HPA ) sulfate ) , tryptophan ( 3-indoxyl sulfate ( 3-IS ) and indole-3-acetylglycine ( IAG ) ) , and phenylalanine ( phenylacetylglycine ( PAG ) ) ., Increases in urinary β-oxidation metabolites , accumulation of the early tricarboxylic acid cycle intermediate cis-aconitate , and changes in muscle metabolites creatine and creatinine indicated altered host energy utilization in Giardia-infected mice ., In addition , methylated nicotinamide derivatives capable of regulating energy expenditure ( N-methylnicotinamide ( NMND ) and nicotinamide-N-oxide ( NAO ) ) 23 were increased in Giardia infected mice ., Consistent with the finding that increased urinary NMND predicts catch-up growth in undernourished children 7 , persistently Giardia-infected mice fed the protein deficient diet developed ‘overshoot’ growth gains compared with uninfected age and diet-matched controls upon re-nourishment ( switched from the PD to the CD diet on 42 dpi ) ( S3 Fig ) ., The Giardia-induced changes in gut microbial host co-metabolites of proteolysis either persisted ( 4-HPA sulfate , IAG ) or overlapped ( PAG , 4-CG , 4-CS ) with those seen during EAEC infection alone , and these metabolites were even further magnified in co-infected mice ( Fig 4E ) ., However , EAEC-mediated increases in TMA and TMAO ( Fig 4E ) , indicative of microbial-dependent choline breakdown , were reversed in Giardia co-infected mice ( Fig 4E ) , and resembled the metabolic perturbation in choline metabolism of Giardia infection alone ( Fig 4E ) ., Similarly , elevated taurine excretion in Giardia- infected mice persisted through co-infection ( Fig 4E ) ., Whereas either infection increased lipid oxidation evident in increased β-oxidation breakdown products ( hexanoylglycine , butyrylglycine , and isovalarylglycine ) along with the β-oxidation pathway precursor acetyl-carnitine , metabolism during co-infection shifted away from β-oxidation as indicated by a decrease in acetyl-carnitine and downstream β-oxidation metabolites ., Concurrently , co-infection led to an inversion of creatine:creatinine ratios , suggesting altered muscle metabolism compared with either infection alone ( Fig 4E ) ., Finally , host energy expenditure adaptations via the nicotinamide pathway ( NMND and NAO ) during Giardia infection alone ( Fig 4E ) were extinguished following EAEC co-infection ( Fig 4E ) ., Multiple and diverse pathogen exposures are hypothesized to cause intestinal dysfunction , also termed Environmental Enteropathy ( EE ) , in malnourished children ., In the present study , we modeled co-infection with two of the most commonly isolated pathogens in malnourished children , Giardia lamblia and enteroaggregative Escherichia coli ( EAEC ) ., We used protein deficiency in weaned mice to investigate how microbial-specific pathways intersect to impair host growth , mucosal inflammation , and metabolism during malnutrition ., Our integrated nutritional , microbial , immunological , and metabolic observations add insight into how changes in resident microbiota combine with cumulative enteropathogen exposures to interfere with host growth and metabolic adaptations to protein malnutrition ., For the first time we demonstrate that a resident microbiota that is permissive to enteropathogen colonization also simultaneously promotes growth impairment during persistent Giardia infection ., Although Giardia was insufficient to induce intestinal inflammation characteristic of EE-like changes , despite evidence of mucosal injury ( IL1α ) , the parasite had a profound effect on gut microbial-host co-metabolism ., EAEC , on the other hand , promoted robust expansion of lamina propria cells coupled with secretion of myeloid ( CXCL8 ( IL-8 ) ) and lymphoid ( CCL5 ) chemokines ., Together , these pathogens synergistically increased signals of intestinal injury , IL1α , and CCL11 ., EAEC-dependent increases in myeloid cells were preserved in co-infected mice; however , persistent Giardia infection resulted in diminished myeloid cell specific activation markers ( Cp and to a lesser degree MPO ) consistent with parasite-mediated alterations in host immune pro-inflammatory responses ., Strikingly , these non-invasive co-pathogens resulted in an increasingly proteolytic microbiota that dominated the co-metabolic profile ( specifically leading to increased tryptophan , tyrosine , and phenylalanine co-metabolites ) , despite relatively restricted changes in the 16S rRNA composition ., Simultaneously , host metabolic adaptation to protein deficiency progressively declined , eventually resulting in the loss of host-mediated nicotinamide-pathway energy regulation , and disrupting lipid oxidation up-regulation and muscle metabolism in the increasingly malnourished host ., A working model of these specific pathogen-mediated microbial , immunologic , and metabolic alterations are shown in Fig 5 ., Our findings support that an ability to better compete for restricted resources in the intestinal environment is one mechanism whereby enteropathogens may more successfully infect malnourished hosts ., For example , the protein deficient diet contains 0 . 34% rather than the 3 . 4% arginine contained in the 20% protein sufficient control diet 16 ., In weaned mice , this protein deficient diet recapitulates several dysbiotic features described in malnourished children: altered maturation of the fecal intestinal microbiota 16 , 33 , increased susceptibility to Giardia and EAEC , and increased microbial-mediated tryptophan breakdown 7 , 16 ., In the present study we also observed an altered Firmicutes:Bacteroidetes ratio in the protein deficient diet-fed mice 25 , that was modestly increased at early timepoints after Giardia infection ., In contrast arginine-supplementation has been shown to increase the abundance of Bacteroidetes relative to Firmicutes in the small intestine 34 ., Since Giardia is a microaerophilic protozoan that can utilize either glucose or arginine for growth and replication , we speculate that a diet-dependent decrease in Bacteroidetes reduced bacterial competition for arginine ., A limitation of 1H-NMR profiling is an inability to directly detect arginine metabolites ( ornithine , citrulline ) , and thus we could not determine whether Giardia infection was sufficient to further magnify host arginine deficiency ., However , consistent with Giardia use of arginine in order to evade host immune defenses through arginine-deiminase ( ADI ) 35 , the continued decline in IFNγ in mice infected with Giardia could be a result of the actions of Giardia ADI to skew dendritic cell TLR-responses away from pro-inflammatory cytokines 36 ., Furthermore , a reduction in B-cells as seen in Giardia infected mice , is similar to other models of arginine deficiency 37 ., Also , unlike arginine-mediated increases in Bacteroidetes that can enhance TLR-dependent mucosal immune responses 38 similar to some specific Lactobacilli that facilitate Giardia clearance in mice 34 , the protein deficient diet led to decreases in pro-inflammatory cytokines associated with Giardia 27 , 28 , 39 , 40 or EAEC 41 clearance: namely IL-6 , IL-17A , and IFNγ ., Rather , reciprocal increases in IL-4 , a correlate of prolonged duration of Giardia shedding in children 18 , 42 were seen ., Interestingly , like the arrested maturation of the microbiota during protein deficient conditions 16 , this relative shift toward a predominately Th2-type cytokine mileu also resembles that of the neonatal period 43 ., This Th2-type cytokine shift can also be differentially induced via upregulation of thymic stromal lymphopoietin ( TSLP ) in response to Firmicutes-rich altered Schaedler flora 44 ., These collective findings suggest that the isolated protein deficiency in this model establishes a threshold nutrient deficiency that is sufficient to disrupt microbiota-mediated pathogen exclusion , and adds insights into why postnatal Giardia acquisition ( up to 6 months of age ) may be such a vulnerable period for longitudinal growth impairment 7 , 45 ., Despite variation among reports , many epidemiologic studies in malnourished children reveal that early and persistent Giardia associates with impaired growth attainment 9 , 45 despite inversely decreased stool markers of EE-like inflammation myeloperoxidase ( MPO ) 7 and the T-cell activation marker neopterin 9 , 46 ., Also , there is an apparent decreased risk for acute diarrhea and diminished markers of systemic inflammation in children infected with Giardia 47 that may be abolished following multi-nutrient supplementation 48 ., It was critical , therefore , to examine how Giardia interacted alone and during co-infection ., Previous findings have shown that bacteria cultivated from jejunal aspirates of patients with symptomatic giardiasis elicit more inflammation in germ free mice than axenized Giardia trophozoites 49 , and that Giardia increased bacterial mucosal translocation , even after parasite clearance in some animal models 50 , 51 ., This led us to hypothesize that bacteria may similarly influence growth outcomes during giardiasis ., Using continuous antibiotic exposures , we show for the first time that these interactions are crucial for host growth attainment ., However , unlike the same antibiotic cocktail that led to reduced CD8+T-cell activation and consequently , decreased host-mediated immunopathogenesis in another Giardia model 52 , we did not see significant inflammation in the mucosa of protein deficient diet fed infected mice ., Therefore , the primary driver of growth impairment in this model appears to be a Giardia-mediated disruption in microbial-host metabolism ., These data support that one mechanism of Giardia-mediated growth faltering is through an altered intestinal ecology 53 ., For example , pipecolic acid , a byproduct of lysine degradation that is significantly increased in Giardia spent media 32 , was uniquely detected only in Giardia-infected or co-infected mice , and could represent a pathway whereby intestinal parasites limit luminal availability of essential amino acids in the undernourished host 16 ., Our findings of increased phosphatidylcholine ( PC ) , choline , and taurine/isothionate in Giardia infected mice , may also indicate disrupted lipid metabolism through the parasite’s consumption of bile salts ( independent of known expression of bile-salt hydrolases ) and acquisition/turnover of exogenous lipids in the small intestine via phospholipid-transporting transmembrane proteins ( such as flipases ) 32 , as well as choline kinases and phosphatidylcholine synthases 32 ., These perturbations in bile acid and/or lipid homoestasis could have implications for growth in malnourished children 31 , 54 ., Finally , urinary alanine , a unique byproduct of Giardia glucose fermentation under low-oxygen tension 55 , was elevated together with relative increases in fecal Clostridiales , suggesting an increased anaerobic environment in the Giardia infected mice on a protein deficient diet ., Other microbial-dependent urinary metabolites that are altered during Giardia infection are not known to be direct products of Giardia metabolism: such as MA , DMA , and TMA as well as metabolites of aromatic amino acid breakdown ( ie ., PAG ( phenylalanine ) , 4-CS/4-CG ( tyrosine ) , and 3-IS/3-IAG ( tryptophan ) ) ., Also , since EAEC alone also fueled microbial-dependent proteolysis of aromatic amino acids with the exception of 4-HPA sulfate , a breakdown product of tyramine that may be another unique metabolite of Giardia 32 , we suspect these markers of amino acid catabolism indicate products of bacterial metabolism ., Since decreases in the dietary constituents sucrose and tartrate in either Giardia or EAEC infected mice suggested reduced exogenous protein intake , this metabolic shift could have resulted from microbial degradation of host derived proteins , potentially released from injured or sloughted epithelial cells or leakage across disrupted tight-junctions ., Furthermore , since these same proteolytic metabolites have been identified in undernourished children 7 , and were present together with an uncoupling of TCA intermediates in Kwashiorkor-associated dysbiosis 1 our findings raise the need to elucidate the role of Giardia and EAEC in metabolic-based studies in human infections ., Also , follow-up integrated proteomic and metabolomics analyses across various Giardia and EAEC strains could greatly expand the presently limited systems biology databases of these and other enteropathogens 16 , 56 ., Our data raise important considerations for host mucosal immune consequences of multi-enteropathogen infections in malnourished children ., The lack of intestinal inflammation seen in these protein deficient diet fed mice during persistent Giardia infection is reminiscent of the majority of intestinal biopsies in children with Giardia infection 9 ., This is in contrast to the persistent inflammation seen in symptomatic chronic giardiasis in adult humans 39 , 57 as we previousl | Introduction, Results, Discussion, Methods | Diverse enteropathogen exposures associate with childhood malnutrition ., To elucidate mechanistic pathways whereby enteric microbes interact during malnutrition , we used protein deficiency in mice to develop a new model of co-enteropathogen enteropathy ., Focusing on common enteropathogens in malnourished children , Giardia lamblia and enteroaggregative Escherichia coli ( EAEC ) , we provide new insights into intersecting pathogen-specific mechanisms that enhance malnutrition ., We show for the first time that during protein malnutrition , the intestinal microbiota permits persistent Giardia colonization and simultaneously contributes to growth impairment ., Despite signals of intestinal injury , such as IL1α , Giardia-infected mice lack pro-inflammatory intestinal responses , similar to endemic pediatric Giardia infections ., Rather , Giardia perturbs microbial host co-metabolites of proteolysis during growth impairment , whereas host nicotinamide utilization adaptations that correspond with growth recovery increase ., EAEC promotes intestinal inflammation and markers of myeloid cell activation ., During co-infection , intestinal inflammatory signaling and cellular recruitment responses to EAEC are preserved together with a Giardia-mediated diminishment in myeloid cell activation ., Conversely , EAEC extinguishes markers of host energy expenditure regulatory responses to Giardia , as host metabolic adaptations appear exhausted ., Integrating immunologic and metabolic profiles during co-pathogen infection and malnutrition , we develop a working mechanistic model of how cumulative diet-induced and pathogen-triggered microbial perturbations result in an increasingly wasted host . | Malnourished children are exposed to multiple sequential , and oftentimes , persistent enteropathogens ., Intestinal microbial disruption and inflammation are known to contribute to the pathogenesis of malnutrition , but how co-pathogens interact with each other , with the resident microbiota , or with the host to alter these pathways is unknown ., Using a new model of enteric co-infection with Giardia lamblia and enteroaggregative Escherichia coli in mice fed a protein deficient diet , we identify host growth and intestinal immune responses that are differentially mediated by pathogen-microbe interactions , including parasite-mediated changes in intestinal microbial host co-metabolism , and altered immune responses during co-infection ., Our data model how early life cumulative enteropathogen exposures progressively disrupt intestinal immunity and host metabolism during crucial developmental periods ., Furthermore , studies in this co-infection model reveal new insights into environmental and microbial determinants of pathogenicity for presently common , but poorly understood enteropathogens like Giardia lamblia , that may not conform to existing paradigms of microbial pathogenesis based on single pathogen-designed models . | medicine and health sciences, protein metabolism, pathology and laboratory medicine, giardia, diet, parasitic protozoans, protozoans, nutrition, malnutrition, digestive system, giardia lamblia, infectious diseases, pathogenesis, gastrointestinal tract, biochemistry, anatomy, host-pathogen interactions, co-infections, biology and life sciences, metabolism, organisms | null |
journal.pntd.0003295 | 2,014 | Pathophysiologic and Transcriptomic Analyses of Viscerotropic Yellow Fever in a Rhesus Macaque Model | Yellow fever virus ( YFV ) is a member of the flavivirus genus and is endemic or intermittently epidemic in 45 countries ( 32 in Africa and 13 in South America ) 1 , 2 ., YFV causes ∼200 , 000 cases and 30 , 000 deaths annually 3 ., There are two main life cycles for YFV: in the urban cycle , YFV is transmitted between humans primarily through the bite of infected Aedes aegypti mosquitoes; in the jungle cycle , YFV is transmitted between nonhuman primates via Hemagogus mosquitoes in South America and Aedes africanus in Africa 4 ., The clinical symptoms of yellow fever ( YF ) can be quite broad , ranging from mild disease to severe manifestations including liver and kidney failure and hemorrhage 3 , 5 ., YF is characterized by three stages ., During the incubation period , which lasts 3–4 days , virus can be detected in the blood and patients may experience fever , myalgia , and nausea ., This is usually followed by remission with abatement of symptoms for 24–48 hours ., In some patients , this is followed by the return of symptoms at a more severe level and the onset of jaundice ., Deepening jaundice , rising pulse , hypotension , and hypothermia , appear before death , which occurs in 20 to 50% of cases 6 ., The current live attenuated YFV vaccines are effective , but are not without complications 7 , 8 ., YF vaccine-associated neurotropic disease ( YEL-AND ) and YF vaccine-associated viscerotropic disease ( YEL-AVD ) are rare but represent serious adverse events 7 , 9–13 ., YF vaccines are contraindicated in infants <9 months of age , people with primary immunodeficiencies , malignant neoplasms , organ transplant , AIDS or other clinical manifestations of HIV , and thymus disorders ( thymoma , myasthenia gravis , or thymic ablation ) 14 , 15 ., Therefore the development of a safer vaccine is highly desirable 16 ., In addition , as with all pathogens , increased travel increases the risk of outbreaks in areas with high-density vectors and an unvaccinated population ., These concerns are further compounded by the lack of approved antivirals for YF ., In order to develop new vaccine and therapeutic strategies , we need a better understanding of the pathophysiology of YF ., Small animal models of YFV infection such as Golden hamsters 17–19 or mice that are genetically deficient in IFNαβγ receptor expression ( AG129 mice ) 20 have been developed ., Although these small animal models offer several advantages , they also have caveats ., For example , the hamster model requires the use of a hamster-adapted strain of YFV ( YFV-Jimenez ) , and unfortunately many immunological reagents are not readily available for this species ., Infection of the AG129 mouse model with the vaccine strain of YFV-17D results in lethal viral encephalitis but not viscerotropic disease 20 ., More importantly , the host immune response to YFV infection cannot be adequately studied in an immune deficient host such as AG129 mice ., In contrast , non-human primates ( NHP ) provide a very robust model for studying YFV since these animals represent a natural reservoir during the jungle cycle of transmission 15 and the clinical manifestations following lethal YFV challenge of rhesus macaques closely mimic severe forms of human viscerotropic disease 21 ., Indeed , large YF outbreaks in NHP populations have been reported in areas where human epizootics have occurred ., For instance , between October 2008 and June 2009 , over 2000 howler monkeys succumbed to YF in Brazil during the same time as 21 confirmed human cases 22 ., In this study , we characterized viral dissemination; changes in immune cell frequencies both in peripheral blood and lymphoid tissues; as well as changes in cytokine and liver enzyme levels in NHP infected with YFV-DakH1279 ., Data presented herein show that a profound loss of peripheral lymphocytes in the blood precedes characteristic liver pathology and provides an early indicator of fatal YF in this model ., In addition , we examined the transcriptome in peripheral blood mononuclear cells ( PBMC ) collected 3 days after infection with YFV-DakH1279 compared with the attenuated vaccine strain , YFV-17D ., This analysis revealed that striking changes in gene expression are evident at this early time point and provide glimpses into the molecular basis of YFV virulence ., YFV-DakH1279 ( originally isolated from a YF patient in Senegal in 1965 ) was obtained from the World Reference Center for Emerging Viruses and Arboviruses after approval from Dr . Robert Tesh ( University of Texas Medical Branch , Galveston , TX ) ., The initial inoculum was passaged once in a young rhesus macaque ( ∼103 TCID50 ) and the animal developed viscerotropic disease and required humane euthanasia at 5 days post infection ., Serum from the YFV-DakH1279 infected macaque collected at necropsy was then passaged once on C6/36 cells grown in EMEM supplemented with 10% FBS and antibiotics at 28°C , 6% CO2 to prepare a low-passage virus stock for in vivo pathogenesis studies at a titer of 9 . 4×105 infectious units/mL ., Since YFV-DakH1279 does not form plaques , cytopathic effect ( CPE ) , or measurable focus forming units , we used a flow cytometry-based tissue culture limiting dilution assay ( TC-LDA ) to determine the infectious virus titer ., The TC-LDA functionally similar to a tissue culture infectious dose-50 ( TCID50 ) was performed by incubating serial dilutions of virus in replicate wells of C6/36 mosquito cells and stained intracellularly with a YFV-specific monoclonal antibody , 3A8 . B6 as previously described 23 ., RNA isolation was performed using the ZR Viral RNA kit per the manufacturers instructions ( Zymo Research ) ., Briefly , 200 µL of serum was transferred to a tube containing 30 µL of ZR Viral RNA Buffer ., This mixture was bound to a Zymo-Spin IC Column by centrifugation at 16 , 000× g for 2 minutes ., The flow-through was discarded , and the column was washed twice with 300 µL of RNA Wash Buffer ., Residual wash buffer was removed by centrifugation , and the purified RNA was eluted with 12 µL of RNase-free water ., Purified RNA was reverse transcribed using the High Capacity cDNA Reverse Transcription Kit ( Applied Biosystems ) following the manufacturers instructions for 20 µL reactions ., YFV genome copy numbers were then measured by quantitative PCR ( qPCR ) using the following forward ( 5′CAC GGA TGT GAC AGA CTG AAG A 3′ ) and reverse ( 5′CCA GGC CGA ACC TGT CAT 3′ ) primers and probe ( 5′ 6-FAM- CGACTGTGTGGTCCGGCCCATC 3′BHQ ) ., Standard curve was established using the following amplicon as template ( CGA CTG TGT GGT CCG GCC CAT CCA CGG ATG TGA CAG ACT GAA GAG GAT GGC GGT GAG TGG AGA CGA CTG TGT GGT CCG GCC CAT CGA TGA CAG GTT CGG CCT GG ) ., In these experiments , cDNA was subjected to 10 min@95°C followed by 40 cycles of 15 sec@95°C/1 min@60°C ., Experiments were carried out using a StepOnePlus Real-Time PCR system ( Applied Biosystems ) ., Viral load in a subset of serum samples were also measured using the TC-LDA method described above 23 ., All Rhesus macaques were handled in strict accordance with the recommendations described in the Guide for the Care and Use of Laboratory Animals of the National Institute of Health , the Office of Animal Welfare and the United States Department of Agriculture ., All animal work was approved by the Oregon National Primate Research Center ( ONPRC ) Institutional Animal Care and Use Committee ( PHS/OLAW Animal Welfare Assurance # A3304-01 ) ., The ONPRC is fully accredited by the Assessment and Accreditation of Laboratory Animal Care-International ., Animals were housed in adjoining individual primate cages allowing social interactions , under controlled conditions of humidity , temperature and light ( 12-hour light/12-hour dark cycles ) ., Food ( commercial monkey chow supplemented by treats and fruit twice daily ) and water were available ad libitum ., Environmental enrichment consisted of commercial toys ., All procedures were carried out under Ketamine anesthesia by trained personnel under the supervision of veterinary staff and all efforts were made to minimize animal suffering ., After infection , trained personnel monitored animals 4 times a day ., Monkeys were humanely euthanized by the veterinary staff at ONPRC in accordance with endpoint policies ., Euthanasia was conducted under anesthesia with ketamine followed by overdose with sodium pentobarbital ., This method is consistent with the recommendation of the American Veterinary Medical Association ., Twenty female rhesus macaques ( Macaca mulatta ) 8–16 years of age were used in these studies ., Animals were assigned to Animal Biosafety Level-3 ( ABSL-3 ) housing in successive cohorts ranging from 2 to 4 animals and infected subcutaneously with YFV-DakH1279 at doses ranging from 25 to 5×104 infectious units ( n\u200a=\u200a2–4/dose ) ., Blood samples were collected on days 0 , 3 , 4 , 5 , 6 , 7 , 10 , and 14 post-infection ., Complete blood counts and liver enzymes were determined every time a blood sample was collected ., Animals were euthanized if 4 out the 6 criteria listed below were reached:, 1 ) >80% decrease in number of circulating lymphocytes;, 2 ) ALT levels >1000 U ( normal <100 U ) ;, 3 ) bile acid ( BA ) levels >100 U ( normal <10 ) ;, 4 ) total bilirubin ( TBIL ) >1 . 5 mg/dl ( normal <0 . 5 mg/dl ) ;, 5 ) weight loss >30%; and, 6 ) viral loads >107 genomes/ml serum ., We used the cohort infected with the 5×104 TCID50 ( i . e . , the first cohort ) to develop the humane endpoints listed above ., In this first challenge experiment , one of the animals euthanized presented only with high viral loads and lymphopenia , while blood chemistry profiles were within normal ranges ., Following necropsy , the histopathology analysis showed minimal organ damage ( Table 1 ) , which suggested that this animal might have survived the challenge ., Based on those observations , we made the decision to require humane euthanasia when 4 out of the 6 criteria were met ., In subsequent experiments , two animals that did not meet the humane euthanasia endpoints were necropsied 7 and 10 dpi ., The truncation of the study time course in the case of these two animals was due to Institutional Animal Care and Use Committee policy that requires animals be housed no longer than 24 hours without a companion animal in the room ., This would have required additional animals to be assigned to ABSL-3 and the euthanasia of these additional animals ., For humane reasons it was decided therefore , to necropsy the experimental animals before the full 14 days ., At the time of necropsy , blood , liver , kidney , spleen , bone marrow , and axillary , inguinal and mesenteric lymph nodes were harvested from all animals ., Three additional animals were infected with 1 standard dose ( 0 . 5 ml ) of YFV-17D ( YF-Vax , Sanofi Pasteur , formulated to contain no less than 4 . 74 log10 PFU/0 . 5 ml ) subcutaneously ., Blood samples were collected prior to and 3 days post-infection for gene expression analysis ., Total white blood cell count , lymphocyte , platelet , red blood cell counts , hemoglobin , and hematocrit values , were determined from EDTA blood with the HemaVet 950FS+ laser-based hematology analyzer ( Drew Scientific , Waterbury , CT ) ., Serum was analyzed for alkaline phosphatase ( ALP ) , alanine aminotransferase ( ALT ) , gamma glutamyltransferase ( GGT ) , bile acid ( BA ) , total bilirubin ( TBIL ) , albumin ( ALB ) , and blood urea nitrogen ( BUN ) using a VetScan VS2 ( Abaxis veterinary diagnostics , Union City , CA ) ., PBMC were surface stained with antibodies against CD8β ( Beckman Coulter , Brea , CA ) , CD4 ( eBioscience , San Diego , CA ) , CD20 ( Beckman Coulter , Brea , CA ) , HLA-DR ( eBioscience ) , and CD14 ( Biolegend , San Diego , CA ) ., Samples were fixed with 4% paraformaldehyde for 4 hrs before removal from the BSL-3 ., The samples were acquired using the LSRII instrument ( Beckton Dickenson , San Jose , CA ) and the data were analyzed using FlowJo software ( TreeStar , Ashland , OR ) ., Aliquots of plasma samples ( stored at −80°C ) were thawed and heat inactivated for 60 min at 55°C for removal from the BSL-3 ., Heat inactivated serum samples must be tested for residual live virus before removal from the BSL-3 ., Samples were then analyzed with Milliplex Non-Human Primate Magnetic Bead Panel containing the following analytes: TNFα , IL-6 , IL-12/23p40 , IL-8 , MCP-1 , IL1Ra , soluble CD40L , IL-15 , IFNγ , IL-4 and IL-17 as per manufacturers instructions ( Millipore Corporation , Billerica , MA ) ., Heat inactivation decreased the detection of the cytokines in this kit as follows: TNFα by 42% , IL-6 by 53% , IL-12/23p40 by 56% , IL-8 by 20% , MCP-1 by 33% , IL1Ra by 96% , IL-15 by 49% , IFNγ by 73% , and IL-17 by 19% ., We were unable to determine the impact of heat inactivation on the levels of CD40L and IL-4 since the levels of these analytes were below detection in the test samples we subjected to this treatment ., Tissues were collected and placed in neutral-buffered formalin for paraffin embedding ., Sections were cut at 5 µm , deparaffinized and stained with hematoxylin and eosin , or blocked with 5% normal goat serum and 5% bovine serum albumin for immunostaining with primary antibodies specific for YFV antigen ( mouse anti-YF clone 3A8 . B6; 1 . 5 µg/µL , a generous gift from Dr . Ian Amanna ) , B cells ( mouse anti-human CD20 , Dako; 1∶475 ) , or T cells ( rabbit anti-human CD3 , Dako; 1∶200 ) ., Secondary antibodies used were: biotinylated goat-anti-mouse IgG and biotinylated goat-anti-rabbit IgG ( Vector; 1∶300 ) ., DAB chromagen with hematoxylin counterstain ( Vector ) was used to visualize CD20+ B cells and CD3+ T cells ., VIP substrate with methyl green counterstain ( Vector ) was used to visualize YFV antigen ., The sections were then analyzed and images captured using an Axioplan microscope ( Carl Zeiss ) with a Spot Insight camera ( Diagnostic Insturments Inc . ) Microarray assays were performed in the OHSU Gene Profiling Shared Resource ., One million PBMC were resuspended in Trizol ( Invitrogen ) and RNA was extracted using RNeasy Micro Kit ( Qiagen ) according to the manufacturers protocol ., Total RNA was treated with RNase-free DNase ( Qiagen ) followed by purification and concentration with the RNA Clean & Concentrator-5 kit ( Zymo Research ) ., Following clean-up , 25 ng of total RNA from each sample were amplified and biotin-labeled using the Ovation RNA Amplification System V2 , Ovation WB Reagent , and Encore Biotin Module ( NuGEN Technologies ) as per manufacturer recommendations ., Labeled hybridization targets were mixed with hybridization solution containing hybridization controls ( Affymetrix ) according to NuGEN Technologies protocol and hybridized with the GeneChip Rhesus Macaque Genome Array ( Affymetrix ) ., This array contains 52 , 024 probe sets interrogating over 47 , 000 M . mulatta transcripts ., Arrays were scanned using the GeneChip Scanner 3000 7G and image quality was determined immediately following each scan ., Image processing was performed with Affymetrix GeneChip Command Console v3 . 1 . 1 and probe set summarization and CHP file generation were performed using Affymetrix Expression Console v1 . 1 software ., All microarray data analysis steps were performed in the statistical environment R , using Bioconductor packages ( R Development Core Team , 2008 ) ., The probe set-to-gene mappings for the Rhesus chip were downloaded from the Affymetrix site ., All ambiguous probe sets on this chip were treated in the gene enumeration steps of this study in the following manner: controls and probe sets matching no or several loci in the Macaca mulatta genome were ignored in the downstream analysis steps ., In addition , redundant probe sets that represent the same locus several times were counted only once ., The normalization of the raw data CEL files was performed with the Robust Multi-array Average ( RMA ) algorithm using the default settings of the corresponding R function 24 ., The quality of the Affymetrix Gene Chips was assessed with analysis routines provided by the affyPLM library 25 ., For each probe , log2 fold change ( log2FC ) expression was calculated as the difference of log2 expression at 3 days post-infection ( dpi ) relative to 0 dpi ., Analysis of differentially expressed genes ( DEGs ) was performed with the LIMMA package using the normalized expression values 26 ., The Benjamini and Hochberg method was selected to adjust p-values for multiple testing and control false discovery rates ( FDRs ) 27 ., As confidence threshold for identifying DEGs we chose an adjusted p-value of <\u200a=\u200a0 . 05 and absolute log2FC superior to 1 ., Functional analysis of statistically significant gene expression changes was performed using Ingenuity Pathways Knowledge Base ( IPA; Ingenuity Systems ) and Gene Ontology ( GO ) 28 ., In addition , we also used previously published microarray data from resting and activated human immune cells ( GSE22886; IRIS database ) to define genes specific to each immune cell type as previously described 29 ., Genes specific to innate immune cells were further defined as the union of genes significantly up-regulated in resting or activated dendritic cells , natural killer cells , monocytes or neutrophils ., Genes specific to adaptive immune cells were defined as the union of genes significantly up-regulated in naïve or activated T or B cells ., For all gene set enrichment analyses , a right-tailed Fishers exact test was used to calculate a p-value determining the probability that each biological function assigned to that data set was due to chance alone ., An enrichment score ( ES ) , defined as −log10 ( p-value ) as calculated using a right tailed Fishers exact test , was calculated ., In addition , we used the IPA regulation z-score algorithm which identifies biological functions that are expected to be activated or inhibited in infected animals vs . controls , and which is designed to reduce the chance that random data will generate significant predictions ., Z-scores ≥2 , indicate that the function is significantly increased and z-scores ≤−2 , indicate that the function is significantly decreased ., Raw microarray data have been deposited in NCBIs Gene Expression Omnibus and are accessible through GEO series accession number GSE51972 ., Infection of rhesus macaques with 25 TCID50 to 5×104 TCID50 of YFV-DakH1279 resulted in a fulminating disease that typically lasted 4–7 days ( Figure 1 ) ., Higher doses of YFV-DakH1279 resulted in slightly higher and earlier viremia than lower doses of virus ( Figure 1A ) ., Peak viremia occurred between days 3 and 7 post-infection and in lethal cases reached 109 to 1013 YFV genome equivalents/mL as measured by qRT-PCR shortly before the animals required humane euthanasia ., We have previously showed a 1∶1 relationship between the levels of virus measured by qRT-PCR and the levels of infectious virus ( 23 , R2\u200a=\u200a0 . 89 , p<0 . 0001 ) , indicating that the YFV genome equivalents shown here are representative of the levels of infectious virus in circulation ., Overall , we observed 75% lethality at 25 and 100 TCID50 ( 3/4 animals in each group ) by 5–7 days post-infection ( dpi ) ; 84% lethality at 103 TCID50 ( 5/6 animals ) by 4–6 dpi; 50% mortality at 104 infectious units ( 1/2 animals ) ; and 75% lethality at a dose of 5×104 TCID50 by 4–5 dpi ( 3/4 animals ) ( Figure 1B ) ., All animals that controlled viral replication to below 106 genome equivalents/mL during the first week of infection survived ., Interestingly , two animals that survived at least 14 days after infection had received the lower challenge doses of virus ( 25 or 100 infectious units/animal ) but presented with two successive rounds of viremia that occurred at 3–5 dpi and then again at 10–14 dpi ., In line with previous observations 21 , animals that required humane euthanasia exhibited signs of significant liver injury ., Unlike healthy liver from uninfected animals ( Figure 2A ) , the livers of the animals that required euthanasia were discolored and contained hemorrhagic foci ( Figure 2B ) ., Histological examination revealed wide spread hepatocyte degeneration and necrosis , vacuolation and fatty changes ( increased prevalence of lipid droplets ) ( Figure 2D , Table 1 ) that were not present in healthy liver ( Figure 2C ) ., We also detected councilman bodies , the hallmark of YF disease in the liver and extensive hemorrhage throughout the livers ( Figure 2D ) ., On rare occasions , we observed eosinophilic intranuclear inclusions ( Torres bodies ) ., YFV antigen was detected by immunohistochemistry in all liver sections obtained at necropsy from animals that required humane euthanasia ( Figure 2F and 2H ) ., In contrast , liver sections obtained from animals that survived YFV infection showed no evidence of YFV antigen ( Figure 2E and 2G ) ., In addition to these histological changes , we found sharp increases in serum levels of alanine aminotransferase ( ALT ) , bile acids ( BA ) , total bilirubin ( TBIL ) and alkaline phosphatasase ( ALP ) within 6–8 hours before the animals were euthanized ( Figure 3 ) ., Levels of ALT reached 2000–9000 U/L in the most severe cases ( normal <100 U/L ) ( Figure 3A ) ; BA levels reached ∼100 umol/L ( normal <10 umol/L ) ( Figure 3B ) ; TBIL reached 1–1 . 5 mg/dl ( normal <0 . 5 mg/dl ) ( Figure 3D ) ., Changes in ALP on the other hand were less pronounced ( Figure 3C ) with one animal reaching 580 U/L ( normal <200 U/L ) ., In line with these observations , ALT , TBIL and BA showed significant curvilinear correlation with viral load ( p<0 . 0001 ) , whereas ALP levels showed no correlation ( p\u200a=\u200a0 . 3 , Figure 3 ) ., Animals that survived infection exhibited little or no change in plasma levels of these key liver enzymes ( Figure 3 ) ., Our analysis also revealed , that in contrast to kidney from uninfected animals ( Figure 4A ) , evidence of kidney injury as indicated by renal tubular degeneration and epithelial tubular necrosis ( Figure 4B , Table 1 ) ., We also detected granular bilirubin casts in dilated distal convoluted tubules and proteinaceous casts in kidneys from animals that required humane euthanasia ( Figure 4B–D ) ., Interestingly , YFV antigen was not detected in kidney tissue sections ( Figure 4E–F ) , indicating that this is not a major site of active viral replication ., Kidney dysfunction at the higher challenge doses was also indicated by a rise in blood urea nitrogen ( BUN ) , averaging 1 . 5 and 1 . 9 -fold increase from baseline at TCID50 104 and 5×104 ) ., There was a significant correlation between challenge dose and fold changes in BUN ( R2\u200a=\u200a0 . 97 , p\u200a=\u200a0 . 002 ) ., There was also a curvilinear correlation between BUN values and viral load ( p\u200a=\u200a0 . 007 ) with changes in BUN observed only approximately 6–8 hours before euthanasia ( Figure 4G , F ) ., We monitored changes in hematological parameters ( Figure S1 ) and circulating white blood cells ( WBC ) throughout infection ( Figure 5 , Figure S2 ) ., Hematocrits , hemoglobin levels and platelet counts were stable until a few hours before the animals required humane euthanasia when small but significant decreases in platelet counts ( p<0 . 01 ) and hematocrit ( p<0 . 01 ) levels and a trend towards reduced hemoglobin levels ( p\u200a=\u200a0 . 07 ) were detected ( Figure S1 ) ., We observed a modest decrease in total WBC counts between days 4–6 post-infection in animals with a lethal infection that required humane euthanasia ( Figure 5A ) ., This decrease was most evident in animals with the highest viral loads ., In contrast , we found a severe loss of circulating lymphocytes , which declined by 71%±29 . 5 in animals that required euthanasia compared to a 23%±15 . 4 decline in animals that survived challenge ( Figure 5B ) ., Neutrophils declined at 3 dpi in most animals but increased slightly in others , resulting in an irregular pattern following YFV infection ( Figure 5C ) ., Indeed , we detected a significant negative correlation between viral load and extent of lymphocyte loss ( R2\u200a=\u200a0 . 46 , p<0 . 0001; Figure 5D ) whereas no correlation between neutrophils and viral load was noted ( Figure 5E ) ., We further characterized the loss in lymphocyte subsets by measuring changes in both frequency and absolute numbers of CD4+ and CD8+ T cells , and CD20+ B cells ( Figure S2 ) ., Frequencies of peripheral CD20+ B cells , CD4+ T cells and CD8+ T cells rapidly decreased , reaching nadir levels by about day 4 post-infection in the animals that ultimately required euthanasia ( Figure S2 A–C ) ., In the four animals that survived , frequencies of CD20+ B cells , CD4+ T cells and CD8+ T cells also declined but to a lesser extent and in two of the animals recovered to pre-infection levels days 10–14 post-infection ( Figure S2 A–C ) ., Numbers of circulating CD14+ monocytes in peripheral blood also decreased within 24 hours before the animals were euthanized ( Figure S2 ) ., In vitro studies were performed to determine if YFV-17D or YFV-DakH1279 replicate in rhesus PBMC but no reproducible viral replication of either strain of virus in primary PBMC was found , indicating that it is unlikely that lymphocyte loss was due to direct viral infection ., To further investigate the virus-induced lymphopenia , we examined lymphoid tissue collected from animals that survived and those with a lethal infection requiring humane euthanasia ( Figure 6 ) ., Histological analysis showed that , in contrast normal cellular turnover observed in germinal center ( GC ) in spleen and lymph nodes in surviving YFV-DakH1279 infected animals ( Figure 6A , Table 1 ) , significant GC necrosis as indicated by increased apoptotic bodies and numerous tangible body macrophages was observed in animals that required euthanasia ( Figure 6B; Table 1 ) ., Severity of GC necrosis correlated with infectious dose and was primarily observed in animals infected with 5×104 and 104 ( Table 1 ) ., Moreover , several of the spleens examined were congested with evidence of hemorrhage ., As described for the kidneys , we did not detect viral antigen in lymphoid tissue despite the GC necrosis ( Figure 6C , 6D ) ., We also examined distribution of CD20+ B cells and CD3+ T cells by immuno-histochemistry ( IHC ) in the spleen ( Figure 6E–H ) ., This analysis showed decreased B cell staining in the germinal center in animals that required euthanasia ( Figure 6F ) ., We analyzed changes in serum cytokine levels associated with YFV-DakH1279 infection ., Analysis of cytokines was affected by the heat inactivation step required to remove samples from the BSL-3 ( see methods and materials ) ., Levels of IL-4 , IL-5 , IL-8 , IL-12/23p40 , IL-17 , G-CSF , GM-CSF , sCD40 , and RANTES were either unchanged in post-infection samples or below levels of detection ., In contrast , increased levels of IL-6 , IL-15 , MCP-1 and IFNγ were detected especially shortly before euthanasia ., Levels of each cytokine/chemokine showed a significant correlation with viral load ( p<0 . 001 , with an R2 ranging from 0 . 25 for IFNγ to 0 . 68 for MCP-1 ) ( Figure 7 ) ., To further explore the molecular basis of YFV pathogenesis , we performed gene expression profiling in PBMC isolated from three animals 0 dpi and 3 dpi with 103 TCID50 of YFV-DakH1279 which required humane euthanasia ( Figure 8 ) ., As a comparison , we included PBMC collected from three animals infected with one standard dose of the YFV-17D vaccine ( 6×104 infectious unit ) on 0 dpi and 3 dpi ., PBMCs isolated from animals infected with 103 YFV-DakH1279 were used because this dose elicits profound viscerotropic disease and the severe lymphopenia in animals infected with 5×104 TCID50 made it difficult to obtain sufficient high quality RNA for microarray analysis from enough animals within this group ., Day 3 was chosen because it preceded the severe lymphopenia observed in wild type YFV-DakH1279 infection ( lymphocyte fold change 0 dpi and 3 dpi: 1 . 15 , 0 . 85 and 0 . 94 respectively ) ., No significant changes in lymphocyte numbers were observed following YFV-17D infection either ( lymphocyte fold change 0 dpi and 3 dpi: 0 . 99 , 1 . 14 and 1 . 18 respectively ) ., Statistical analysis of gene profiles at 3 dpi compared to baseline levels ( 0 dpi ) revealed that YFV-DakH1279 infection induced a more pronounced transcriptional response than YFV-17D infection ., Specifically , 765 differentially expressed genes ( DEGs ) were detected following infection with YFV-DakH1279 ( 337 were downregulated and 428 genes were upregulated ) ., In contrast , only 46 differentially expressed genes were identified following infection with YFV-17D ( 6 downregulated and 40 upregulated ) ., Only 3 genes were shared between the two lists of DEGs: KLRC1 , CPA3 and RSAD2 ., All three genes , which are involved in the innate immune response to viral infection , were upregulated following YFV-17D or YFV- DakH1279 infection ., We further characterized each transcriptional signature by performing functional enrichment 30 ., This analysis revealed that DEGs after YFV-DakH1279 and YFV-17D infection belonged to different biological processes ( Figure 8 , 9 ) ., The only shared process between YFV-DakH1279 and YFV-17D was that of “immune response” ( Figure 8A , 9A ) ., However , while YFV-17D infection only induced up-regulation of immune response genes ( Figure 8A ) , approximately two-thirds of the DEGs associated with immune response were down-regulated after YFV-DakH1279 infection ( Figure 9A ) ., Genes specific to innate immune cells were enriched in both signatures ( Enrichment Score ( ES ) =\u200a4 for YFV-17D signature and 3 . 1 for YFV-DakH1279 ) ., Specifically , the 8 most highly upregulated genes following YFV-17D infection were associated with innate immune response to infection: LOC699418 ( eosinophil lysophospholipase-like ) , RTD1A ( theta defensin 1 α precursor ) , MNPA1 ( α defensin 1 α ) , CRISP-3 ( cysteine-rich secretory protein 3 ) , IFIT3 ( interferon-induced protein with tetratricopeptide repeats 3 ) , RSAD2 ( radical S-adenosyl methionine domain containing 2-IFN induced gene ) and CPA3 ( mast cell secreted carboxypeptidase A3 ) ., In contrast , 43 genes associated with innate immune cells were down-regulated following YFV-DakH1279 infection and only three genes were upregulated ( KLRC1 , CPA3 and RSAD2 ) ., However , it should be noted that 35 genes associated with inflammatory responses , notably STAT-1 ( important for signaling through type I , II or III interferons ) , IL-5 , and CD40L were upregulated following YFV-DakH1279 infection ( Figure 9A ) ., With regards to the adaptive immune response , PLS3 , also known as T-plastin was upregulated following YFV-17D infection ., Expression of this gene is essential for germinal center formation and development of T-dependent antibody responses in mice 31 , 32 ., In contrast , numerous genes specific to B and/or T cells were dysregulated after YFV infection ( ES\u200a=\u200a2 . 1 ) ., For instance , TNFSF11 , which is hypothesized to augment the ability of dendritic cells ( DC ) to stimulate naïve T cell proliferation , was downregulated 33 ., Similarly , SerpinB2 , believed to play a role in sculpting the adaptive immune response 34 is also downregulated and BTLA , a negative regulator of T and B cell responses is upregulated 35 ., Functional categories enriched only after YFV-17D infection included ubiquitination and ISGylation , cytoskeleton and cell adhesion , and epigenetic regulation ( Figure 8 B–D ) ., Interestingly , among the 765 DEGs after YFV-DakH1279 infection , 115 had metal ion binding activity and over 80% of those genes are specifically involved in zinc ion binding ( Figure 9B ) ., An additional 30 genes were involved in cell growth or apoptosis regulation ( Figure 9C ) and 226 were related to transcription ( GEO series accession number GSE51972 ) ., Finally , ingenuity pathway analysis revealed that the biological functions predicted to be the most activated after YFV-DakH1279 infe | Introduction, Methods and Materials, Results, Discussion | Infection with yellow fever virus ( YFV ) , an explosively replicating flavivirus , results in viral hemorrhagic disease characterized by cardiovascular shock and multi-organ failure ., Unvaccinated populations experience 20 to 50% fatality ., Few studies have examined the pathophysiological changes that occur in humans during YFV infection due to the sporadic nature and remote locations of outbreaks ., Rhesus macaques are highly susceptible to YFV infection , providing a robust animal model to investigate host-pathogen interactions ., In this study , we characterized disease progression as well as alterations in immune system homeostasis , cytokine production and gene expression in rhesus macaques infected with the virulent YFV strain DakH1279 ( YFV-DakH1279 ) ., Following infection , YFV-DakH1279 replicated to high titers resulting in viscerotropic disease with ∼72% mortality ., Data presented in this manuscript demonstrate for the first time that lethal YFV infection results in profound lymphopenia that precedes the hallmark changes in liver enzymes and that although tissue damage was noted in liver , kidneys , and lymphoid tissues , viral antigen was only detected in the liver ., These observations suggest that additional tissue damage could be due to indirect effects of viral replication ., Indeed , circulating levels of several cytokines peaked shortly before euthanasia ., Our study also includes the first description of YFV-DakH1279-induced changes in gene expression within peripheral blood mononuclear cells 3 days post-infection prior to any clinical signs ., These data show that infection with wild type YFV-DakH1279 or live-attenuated vaccine strain YFV-17D , resulted in 765 and 46 differentially expressed genes ( DEGs ) , respectively ., DEGs detected after YFV-17D infection were mostly associated with innate immunity , whereas YFV-DakH1279 infection resulted in dysregulation of genes associated with the development of immune response , ion metabolism , and apoptosis ., Therefore , WT-YFV infection is associated with significant changes in gene expression that are detectable before the onset of clinical symptoms and may influence disease progression and outcome of infection . | Yellow fever virus causes ∼200 , 000 infections and 30 , 000 deaths annually in Africa and South America ., Although this is an important human pathogen , the basis of yellow fever disease severity remains poorly understood ., Rhesus macaques are susceptible to yellow fever and develop similar symptoms as severe as those observed in humans ., In this study , we characterized disease progression in this model and observed a profound loss of lymphocytes that preceded the appearance of serum markers of virus-induced liver pathology ., This change might provide an early indicator of fatal yellow fever ., In addition , we also identified significant changes in gene expression in white blood cells that occur before any measurable disease symptoms and these genetic signatures may provide future targets for antiviral therapeutics and better diagnostics . | vaccines, yellow fever, immunopathology, infectious diseases, medicine and health sciences, clinical immunology, attenuated vaccines, biology and life sciences, immunology, viral diseases, vaccination and immunization, arboviral infections, vector-borne diseases, immune response | null |
journal.pcbi.1004299 | 2,015 | Adaptive Synaptogenesis Constructs Neural Codes That Benefit Discrimination | Adaptive synaptogenesis 1–4 is designed to allocate neural resources in a thrifty manner or in a manner to regulate function ., The three resources of concern are number of synapses , number of neurons , and firing-rate of the neurons ., Inspired by the Bienenstock-Cooper-Munro ( BCM ) algorithm 5 and its forcing of a neuron to a predefined activity level , adaptive synaptogenesis achieves a similar goal that not only guarantees the average activity of a postsynaptic neuron but does so in a way that rations synapses ., Previously , adaptive synaptogenesis was used as a mechanism to produce compressive coding with small information losses 6–10 ., It also successfully models developmental studies of ocular dominance 11–12 ., Both results are achieved by postsynaptic neurons discovering implicit correlational structures within the input data space ., Through the random acquisition and forced shedding of synapses , associated inputs find their way to the same neuron , and uncorrelated or anti-correlated inputs are ignored ., As thus conceived , adaptive synaptogenesis consists of ( 1 ) a random Bernoulli process that selects a new excitatory connection between nearby axon i and postsynaptic neuron j; ( 2 ) once formed , associative synaptic modification controls the strength of each existing synapse , and this control includes the possibility of potentiation , depression , or no change of a synaptic weight; but with enough long-term depression , ( 3 ) shedding of a synapse occurs when the weight is appropriately weak ( near zero for a sufficiently long time ) ( Fig 1 ) ., Critically , the possibility of forming a new synapse on neuron j is determined by j’s long-term average firing-rate ., Instead of compressive coding , the context for studying adaptive synaptogenesis here is self-taught discrimination ., The motivating idea is that if one studies a particular field and its subject matter over a long enough period of time ( perhaps the oft quoted ten thousand hours 13 ) and if one studies over a wide enough variety of representative examples , the allocation of neurons in the cerebral cortex is enhanced for this particular concentrated field of study ., After a detailed description of the neural algorithm and the input data structures , we establish a mathematical theory that quantifies relationships ( e . g . synaptic weights ) required for stability ( or lack thereof ) of the neurons formed by this algorithm ., Computational simulations follow these theoretical developments ., The simulations demonstrate the development of stable neuron configurations without turning-off the algorithm ., Moreover , these simulations also reveal the effect of input statistics—frequency of input patterns and the input correlational structure—on neuron allocation ., As shown , the form of adaptive synaptogenesis used here produces neuron allocations that are appropriately biased by the statistics of the input environment ( more experience produces more neurons devoted to the experience ) ., Also revealed is an important effect of the input world’s statistical structure that can help or hinder this proportional neuron allocation ., Here we study an adaptively constructed , feedforward network of McCulloch-Pitts neurons ., The inputs are vectors with binary elements , xi ( k ) ϵ{0 , 1} , and the outputs are vectors with binary elements , zj ( k ) ., For the jth neuron , postsynaptic excitation is linear , yj ( t ) = Σixi ( t ) ·cij ( t ) ·wij ( t ) with connection indicator cij ( k ) ϵ{0 , 1} , with all weights wij positive , and output zj ( k ) : = {1 if yj ( k ) ≥θ , and 0 otherwise} ., Threshold θ is 3 . 0 for dataset A and 0 . 8 for datasets B . The “sensory” input dimensions are 80 ( dataset A ) or 390 ( datasets B ) as described below ., The number of postsynaptic neurons simulated is 2000 per dataset ., Because there is no interaction between the outputs of these neurons ( i . e . there is no feedback or lateral inhibition ) and because there is no avidity rule 7 , each neuron develops its connections independent of all other neurons ., Each neuron is initialized with one connection from a randomly chosen input line with weight 0 . 2 ., There are three distinct aspects of synaptic modification: synaptogenesis , associative synaptic modification , and synaptic shedding ., A connection from input neuron i to output neuron j is indicated as cij ( t ) ϵ{0 , 1} ., The strength ( weight ) of this connection is wij ( t ) ϵ ( 0 . 01 , 1 ) ., Inputs are binary , i . e . , xi ( t ) ϵ{0 , 1} ., Excitation of a neuron on a timestep is linear , yj ( t ) = Σicij ( t ) ·wij ( t ) ·xi ( t ) ., There is no inhibition ., A neuron , whose excitation reaches threshold , fires by the rule zj ( t ) = {1 if yj ( t ) ≥θ , and 0 otherwise} ., A weight is updated according to 15; in particular , Δwij ( t ) = ε·cij ( t ) · ( xi ( t ) – Exi – wij ( t ) ) ·yj ( t ) ., On each timestep , a neuron’s moving-average of firing-rate , z¯j , is updated as z¯j ( t ) =z¯j ( t−1 ) ⋅ ( α ) +zj ( t−1 ) ⋅ ( 1−α ) ., Synaptogenesis is controlled by z¯j and a random variable: uij ϵ{0 , 1} where prob ( uij = 1 ) = γ ., A connection/weight is shed whenever it falls below 0 . 01; that is ,, if ( wij ( τ ) <0 . 01 ) , then\xa0cij ( τ+1 ) =0\xa0&\xa0wij ( τ+1 ) =∅ ., There are two sets of parameterizations ., There is one parameterization for dataset A and one parameterization for datasets B1 , B2 , and B3 ., However , many parameterizations were examined for each dataset , and in fact there are ranges of parameter settings for which the generic results presented below are valid ., In this case ‘valid’ means a parameter set that produces stable connections and postsynaptic neurons that exceed the desired minimum firing-rate ., For the results presented here , the parameterizations produce a relatively large number of synapses per neuron compared to other valid settings ., The parameterizations are listed in Table 1 . Neuron parameterizations that change between dataset A and datasets B1 , B2 , and B3 are threshold to fire ( 3 . 0 versus 0 . 8 ) and minimum desired firing-rate ( 9% versus 10% ) ., As an explicit part of the model , there are three time-scales:, i ) the shortest is the neuron update ( fire or not , given an input ) ;, ii ) the next in duration is synaptic modification of existing synapses , which occurs every timestep; and, iii ) the slowest time-scale , which occurs after each training block , synaptogensis and shedding ., In one cycle , all input vectors are presented to the network ., A training block of inputs equals 10 cycles ( e . g . if there are 50 input states , a block occurs after the network is presented with 500 inputs ) ., The length of the simulations are shown in Table 2 . For each postsynaptic neuron the input blocks are repeated until no synapses are gained or lost for 200 blocks ., At this time , a neuron’s synapses are assumed stable ., At this point , as shown in the results , the synaptic weights have achieved their predicted values ., S1 Code contains the Matlab program used for simulations ., There are 2000 neurons simulated per dataset; this large number serves the purpose of producing accurate statistics ., However , because the synaptic modification algorithms used here yield feed-forward networks with excellent data compression and little information loss , certain analyses only make sense when the number of neurons are much fewer than 2000; in particular we limit the number of randomly sampled neurons to 1 through 50 out of the 2000 ., Neuron construction is driven by repeated presentation of patterns from a predetermined dataset ., Four different input environments are studied ( Table 3 ) ., The first dataset ( dataset A , see Fig 2A and S1 Dataset ) has 80 input lines ., There are five orthogonal prototypes that define the corresponding categories: the five prototypes correspond to a higher probability of firing within one of five distinct sets of input lines 1–16 , 17–32 , 33–48 , 49–64 , or 65–80 ., The five exemplars are generated from these prototypes , and presented with relative frequency 0 . 1 , 0 . 15 , 0 . 2 , 0 . 25 , and 0 . 3 for a total of 100 input patterns ., In the case of dataset A , each prototype is randomly perturbed such that the total number of active input-lines for each generated pattern remains constant ., Specifically , two randomly selected active input-lines of the prototype are inactivated , and two randomly selected inactive input-lines of the prototype are activated ., Fig 2A illustrates the binary input vectors of dataset A . Note the small amount of overlap between the input patterns only occurs due to noise ., Dataset B1 is much more complex in terms of relationships between input vectors ., This set consists of nine categories each with its own prototype ., Each of the nine individual categories has the same relative frequency ( 11 . 1% ) ., The nine categories can be partitioned evenly into three super-categories: I , II , and III ., Fig 2B visualizes the input set ., These inputs are coded as 390-dimensional binary vectors ., Between super-categories , the input patterns are orthogonal ., Within a super-category , the prototypes and the patterns they generate overlap; the degree of overlap varies as a function of the super-category ., Within the three super-categories overlap increases from 5 to 10 , and then 15 input lines for super-categories I , II , and III , respectively ., Each super-category has some input lines that are activated by only a single category; other input lines of this super-category are shared by two of the three categories; and the remaining input lines of the super-category are shared by all three categories ., Super-category I has the least overlap: category A has 45 potentially active input lines that belong only to category A , 5 that belong to A and B , 5 that belong to A and C , and 5 that belong to A , B , and C for a total of 60 potentially active input lines per category ., Super-category II has more overlap between its three categories ( D , E , and F ) ., There are 30 potentially active input lines that belong only to category D , 10 that belong to D and E , 10 that belong to D and F , and 10 that belong to D , E and F for a total of 60 potentially active input lines ., Super-category III has the most overlap ., There are 15 potentially active input lines that belong only to category G , 15 that belong to G and H , 15 that belong to G and J , and 15 that belong to G , H , and J for a total of 60 potentially active input lines ., Thus , each category has a total of 60 potentially active input lines ., Exactly 20 of the 60 potentially active input lines from each category are pseudo-randomly chosen to be active for each pattern ., None of the super-category’s designated inactive input lines are turned into active input lines ., Datasets B1 , B2 , and B3 are carefully constructed to illustrate the effects of category overlap ( Fig 3 and S1 Dataset ) versus category probability on neuron allocation ., Datasets B2 and B3 are constructed in a similar way as dataset B1 , but with different relative frequencies for each category ., Dataset B2 has relative frequencies: 0 . 13 , 0 . 13 , 0 . 13 , 0 . 11 , 0 . 11 , 0 . 11 , 0 . 098 , 0 . 093 , and 0 . 093 ., Dataset B3 has relative frequencies: 0 . 18 , 0 . 17 , 0 . 15 , 0 . 12 , 0 . 11 , 0 . 087 , 0 . 063 , 0 . 058 , and 0 . 053 ., The most important idea of this section is that there is a mathematical derivation that characterizes the stable connectivities for a feedforward neuron whose connections are governed by adaptive synaptogenesis ., This theory’s convergence results provide a means for identifying stable configurations when simulations are performed ., Going in the other direction , and of secondary importance is establishing the relevance of the theory via simulations because the theory requires multiple infinities of sampling: thus , simulations must be used to establish the existence of parameterizations capable of achieving the predicted , stable connectivities ., The stable weight values on a neuron with a stable connectivity are a function of the subspace covariance matrix that arises from the set of input lines received by this neuron ., For example , one of our input environments is a random vector of 390 distinct input lines ( axons arising from distinct neurons which may or may not be correlated in activity ) ., Out of these 390 lines , a postsynaptic neuron may acquire a small fraction of this number , for example 20 input lines ., Such an acquired set defines a subspace of the original space , and just as there is a 390-by-390 covariance matrix associated with the full input space , there is a specific 20-by-20 covariance matrix associated with the subspace defined by this neurons input connections ., Then for this neuron ( call it j ) , we can also associate a dominant eigenvector and its associated eigenvalue arising from js covariance matrix ., A simple theorem states that the weights of these inputs are proportional to this subspace’s dominant eigenvector ., ( A pleasing result since this vector maximizes the information throughput compared to all other linear , n-by-1 input filters for a given y . ), Moreover , the theorem below tells us 1 ) the proportionality constant that scales this eigenvector to the stable weight values , 2 ) the average excitation of j , and 3 ) the variance of js excitation ., In what follows , we assume that ε is a small positive constant and assume synaptic modification has been going on for a long time with a fixed set of input lines ., Therefore the synaptic weights ,, i . e ., the column vector w ( t ) for neuron j , change very slowly and thus can be treated deterministically ., For a fixed set of input lines , each input-activation is random column vector X ( t ) ( with realizations x ( t ) ) with mean value EX ., Via the definition of excitation , y ( t ) = x ( t ) Tw ( t ) , the average excitation is EY = EXTw ., As noted above , the weights can be treated as a constant; thus in the limit , the mean excitation is EY = EXTw ( ∞ ) ., The variance of this excitation arises from the covariance matrix of the input to this neuron, j . That is , define j’s covariance matrix of its input space as Cov ( X ) : = EXXT − EXEXT , and then note that, w ( t ) TCov ( X ) w ( t ) =Ew ( t ) TXXTw ( t ) −Ew ( t ) TX\u2009EXTw ( t ) \u2009=EY2−EY2=Var ( Y ), ( 1 ), Finally , define λ1 to be the largest eigenvalue of this covariance matrix and e1 as its associated eigenvector of unit length ( the so-called dominant eigenvector ) ., Theorem ., Assuming a stable set of input weights is achieved via the synaptic modification equation Δwij = ε ( X ( t ) – EX – w ( t ) ) X ( t ) Tw ( t ) operating along with the shedding rule then ,, EY=λ1w ( ∞ ) :\u2009=limt→∞w ( t ) =ke1 , wherek=Var ( Y ) /EY ., ( 2 ), Proof ., By definition , stability implies limt→∞EΔw ( t ) =0 ., Then , taking this same expectation and limit on the other side of the synaptic modification equation yields, EΔw ( ∞ ) = 0 = ε ( ( EXXT − EXEXT ) w ( ∞ ) – w ( ∞ ) EY ) , or, ε ( Cov ( X ) w ( ∞ ) – w ( ∞ ) EY ) = 0 , which implies, Cov ( X ) w ( ∞ ) =w ( ∞ ) EY ., ( 3 ), Note that ( 3 ) is the eigen-equation , and the shedding rule guarantees all weights are positive while the synaptic modification equation guarantees wij is bounded from above because X–EX < 1 ., Therefore because y is bounded both below and above , convergence is implied ., With our old synaptic modification rule based on a correlation matrix of a non-negative input , the Perron-Frobenius ( PF ) theorem implies that the dominant eigenvector ( associated with λ1 ) is in the all-positive orthant ., Here however , without an all-positive covariance matrix , we must conjecture an extension to PF ( see below ) ., Thus , by this perturbation conjecture , the synaptic weights align with e1 ( proving 2 ) ., Now designate an unknown positive constant k and define w ( ∞ ) = ke1 ., Pre-multiplying ( 3 ) by e1T quickly yields ( 2 ) : e1TCov ( X ) w ( ∞ ) =e1Tw ( ∞ ) EY implying λ1e1Tke1=e1Tke1EY , producing the result EY = λ1 ., For ( 2 ) , pre-multiple both sides of ( 3 ) by w ( ∞ ) T , and note that by Eq ( 1 ) the left hand side is var ( Y ) while the right hand side yields k2EY ., Thus ,, k=Var ( Y ) /EY, If a neuron happens to acquire enough synapses , a valid central limit theorem ( with mean and variance of the excitation values ) would even tell us where threshold should be placed to produce the desired activity level ., That is , the right-hand tail , beginning at threshold , yields the fraction of times a neuron fires to a randomly sampled input ., This theorem assumes convergence of all algorithmic processes ., However there is an important exception to the convergence hypothesis ., Certain input configurations will never produce stable connectivities nor achieve their algorithmically guaranteed firing-rates ., Sensibly , such neurons might be killed-off; such neurons might lower their firing threshold; or from another perspective , such an input configuration will be very unlikely to exist ., For example , we must conjure an input environment in which a set of input patterns is orthogonal to all others ( thus very unlikely ) and the probability of a member of this set occurring is less than the receptivity cutoff ., For example , suppose synaptogenesis remains positive until a neuron fires 10% of the time ., Suppose a subset of patterns occurs 9% of the time and that this set of patterns is orthogonal to all the other patterns ., If a subspace of this set with its positively correlated input lines gains a controlling influence on a postsynaptic neuron , then any other input line not positively correlated but acquired through synaptogenesis will have its weight decreased by the synaptic modification equation and then it will eventually be shed ., Nevertheless , no matter how many positively correlated input lines are acquired , synaptogenesis continues never to halt ( because postsynaptic firing will converge to 9% , a value below the required 10% that halts synaptogenesis on such a neuron ) ., The theory of adaptive synaptogenesis was developed from observations of empirical neuroscience ( see 1 , 2 , and 16 for motivating studies ) , from the underlying assumption that in order for a neuron to be most useful , its afferent synapses must reflect the statistical structure of the input world , and from one more motivating idea ., We assume that there are desirable operating values for balancing costs versus information ( e . g . mean firing-rate or mean excitation vs . variance of excitation ) ., Then , as the outcome of the algorithm , adaptive synaptogenesis guarantees such desirable , predetermined values ., In this regard , BCM theory led the way , as it explicitly creates postsynaptic neurons with a particular average excitation 5 ., In this regard , BCM theory provided the inspiration for adaptive activity control over the long term ., More generally , the importance of producing an average activity level in a postsynaptic neuron became clearer with the demonstration 17 that neuron parameters ( such a axonal leak current ) imply a particular optimum firing-rate in order to maximize the bits per joule of an axon ., Given a neuron with such an optimized axon , the values of synaptic excitation must be important in terms of matching dendro-somatic-initial segment computation with the axon’s optimal firing-rate ., As well , its synapses should in some sense maximize incoming information 15 ., In any event , the BCM algorithm with initial full-connectivity conjoined with an appropriate shedding rule , may well produce identical results to what is found here ., Of course spike-timing rules will also work , again assuming full initial connectivity 18 ., Indeed , in its earliest version , the utility of adaptive synaptogenesis was understood in the context Barlow’s information-conserving compressive coding idea 19–20 , a clearly energy saving transformation with its reduction in both firing-rate and number of neurons while maintaining almost all of the information of the inputs ., The idea of using random connectivity to create network codes has always been part of our synaptogenesis algorithm; in fact , it is the baseline condition in 6–7 ., Independently , such ideas have been used to study efficient connectivity distributions 21 and abstract functions 22 ., As documented in our early work 7 , just random connectivity without shedding is still quite useful for compressive coding ., That is , these randomly formed networks produce large values of mutual information while decreasing statistical dependence ., However , as documented in the series of articles 6 , 8–9 , random connectivity with associative modification is inferior to using the algorithm that includes synaptic shedding of small weights ., Although we know of no first-principles theory for optimizing number of synapses , it is clear from synapse count data and the volume penalties incurred by synaptic structures 23 that only a minute fraction of an input space ( for example the lateral geniculate as the input to V1 ) can form synapses with any one postsynaptic neuron in the cerebral cortex ., In this light , it may be possible to tune adaptive synaptogenesis to achieve an appropriate range of synapses per neuron ., There are four differences between the adaptive synaptogenesis algorithm used previously and the current version: two of these ( A and B below ) are improvements that can be applied to the compressive algorithm , a third ( C ) is a specialization for neurons performing discrimination , and the fourth ( D ) is largely inconsequential in the context of the data structures used here ., There are three primary results here:, 1 ) extension of the adaptive synaptogenesis algorithm from data compression to discrimination;, 2 ) documentation of neuronal allocation as a function of a category’s relative frequency and of the statistical input structure; and, 3 ) when suitably formulated , adaptive synaptogenesis produces a stable connectivity in a stable input world . | Introduction, Methods, Results, Discussion | Intelligent organisms face a variety of tasks requiring the acquisition of expertise within a specific domain , including the ability to discriminate between a large number of similar patterns ., From an energy-efficiency perspective , effective discrimination requires a prudent allocation of neural resources with more frequent patterns and their variants being represented with greater precision ., In this work , we demonstrate a biologically plausible means of constructing a single-layer neural network that adaptively ( i . e . , without supervision ) meets this criterion ., Specifically , the adaptive algorithm includes synaptogenesis , synaptic shedding , and bi-directional synaptic weight modification to produce a network with outputs ( i . e . neural codes ) that represent input patterns proportional to the frequency of related patterns ., In addition to pattern frequency , the correlational structure of the input environment also affects allocation of neural resources ., The combined synaptic modification mechanisms provide an explanation of neuron allocation in the case of self-taught experts . | One neural correlate of being an expert is more brain volume—and presumably more neurons and more synapses—devoted to processing the input patterns falling within ones field of expertise ., As the number of neurons in the neocortex does not increase during the learning period that begins with novice abilities and ends with expert performance , neurons must be viewed as a scarce resource whose connections are adjusted to be more responsive to inputs within the field of expertise and less responsive to input patterns outside this field ., To accomplish this enhanced , but localized improvement of representational capacity , the usual theory of associative synaptic modification is extended to include synaptogenesis ( formation of new synapses ) and synaptic shedding ( rejection of synapses by a postsynaptic neuron ) in a manner compatible with the original , associative synaptic modification algorithm ., Using some mathematically simplifying assumptions , a theory is developed that predicts the algorithms eventual outcome on expert neuronal coding , and then without the simplifying assumptions , computational simulations confirm the theory’s predictions in long , but finite periods of simulation-time ( i . e . , finite-sampling leads to stable connections , and thus , stable expert encodings ) . | null | null |
journal.pcbi.1006097 | 2,018 | A machine learning based framework to identify and classify long terminal repeat retrotransposons | Transposable elements ( TEs ) are DNA sequences that can move and duplicate within a genome , autonomously or with the assistance of other elements ., The field of TE annotation includes various steps such as the identification and classification of TE families ., In this article , we focus on these activities since accurate identification and classification of TEs enable researches into their biology and can shed light on the evolutionary processes that shape genomes 1 ., TEs in eukaryotes can be classified according to whether reverse transcription is needed for their transposition ( Class I or retrotransposons ) or not ( Class II or DNA transposons ) ., A consensus for a universal TE classification has not been reached yet 3 , but this lack of consensus does not affect the focus of our study ., Here , we will follow the hierarchical system proposed by Wicker et al . 2 , which includes the levels of class , subclass , order , superfamily , family and subfamily ., Fig 1 presents an illustration of Wicker’s hierarchy considered in our study ., Class I is composed of five orders: LTR retrotransposons , DIRS-like elements , Penelope-like elements ( PLEs ) , long interspersed nuclear elements ( LINEs ) and short interspersed nuclear elements ( SINEs ) ., Similar in structure to retroviruses , LTR retrotransposons have long terminal repeats ( LTRs ) , two normally homologous non-coding DNA sequences that flank the internal coding region and that range in size from a few hundred base pairs to more than 5 kb ., Superfamilies within an order are distinguished by uniform and widespread large-scale features , such as the structure of protein or non-coding domains and the presence and size of the target site duplication ( TSD ) ., Families are defined by DNA sequence conservation and subfamilies on the basis of phylogenetic data ., Class II is divided into two subclasses , which are distinguished by the number of DNA strands that are cut during transposition ., Subclass 1 consists of TEs of the order TIR , which are characterized by terminal inverted repeats ( TIRs ) ., Subclass 2 groups the Helitron and Maverick orders ., Methods identifying TEs in a genome are homology-based , employ structural information or do not use prior information at all about the TEs to be identified 4–6 ., The latter methods , known as de novo repeat discovery methods , search for example for repeats in the genome ., A widely used method for TE identification is RepeatMasker 7 ., This tool screens a query sequence searching for repeats , taking into account their similarity with sequences from a reference library , using an optimal pairwise alignment algorithm ., Censor 8 works similarly as RepeatMasker but uses BLAST for the comparison ., Afterwards , both RepeatMasker and Censor remove overlaps and defragment detected repeats ., Loureiro et al . 9 show that machine learning can be used to improve the identification of TEs ., They assessed a set of ( non-machine learning based ) identification methods and learn a classifier that combines their predictions to determine whether a sequence is a TE or not ., Another classifier predicts the best method to determine the exact boundaries of a TE ., In their analysis , both RepeatMasker and Censor were the most accurate tools ., While Loureiro et al . demonstrate the benefit of using machine learning models to improve predictions , they do not use machine learning to obtain the predictions , which we address in this article ., A few methods have been proposed to classify TEs ., LtrDigest 10 evaluates a list of LTR retrotransposons generated by another tool called LTRharvest 11 , annotating these sequences w . r . t . the protein domains and other structural characteristics that were found in them ., LtrDigest can then be used for de novo ( unsupervised ) classification , i . e . , finding groups within the LTR retrotransposons without any predefined classification scheme ., To evaluate whether the resulting groups represent known LTR retrotransposon superfamilies , Steinbiss et al . 10 have matched representative sequences of the groups to a reference set of known transposon sequences using a fixed set of rules ., LtrSift 12 takes the LtrDigest output and clusters the candidate sequences ., It then tries to find patterns of shared cluster membership that might indicate multiple TE families , e . g . different Copia-like , Gypsy-like or Bel-Pao families ., It is a generic tool that uses sequence clusters to find family-specific patterns , based on the LtrDigest detected features ., These patterns are then used as evidence for family discrimination ., TEClass 13 classifies TE sequences into Class I and Class II TEs ., The Class I elements can further be classified into LTRs and non-LTRs , and the non-LTRs are classified into the SINE or LINE orders ., This classification is obtained by a hierarchy of binary classifiers based on machine learning support vector machines , using oligomer frequencies as features ., RepClass 14 consists of three independent classification modules: a module based on homology information , a module that searches for structural characteristics such as LTRs or TIRs , and a module that searches for target site duplication ., The three modules provide classifications at different levels of granularity , typically at the subclass or order level , sometimes at the superfamily level ., Finally , an integration module aims to compare , rank , and combine the results of the three modules providing a single tentative classification ., Pastec 15 also uses multiple features of TEs to classify TE sequences: structural features ( TE length , presence of LTRs or TIRs , presence of simple sequence repeats , etc . ) , sequence similarities to known TEs , and conserved functional domains found in HMM profile databases ., It provides classifications on the order level , including all orders from the classification hierarchy defined by Wicker et al . 2 , whereas TEClass and RepClass only consider a subset of the orders ., Importantly , none of the above classification systems is able to provide classifications for LTR retrotransposons at the superfamily level ., One exception is a recently introduced method called LtrClassifier 16 , which performs both annotation ( i . e . , identifying structural elements ) and classification ( but not identification ) for plant genomes , and returns predictions for the Copia and Gypsy superfamilies ., In this article we introduce TE-Learner , a framework for the identification of TEs of a particular order , and for the classification of these TEs on the superfamily level ., TE-Learner consists of three steps ., First , based on the characteristics of the order under consideration , it extracts from the genome a set of candidate sequences , which may include false positives ., Second , it automatically annotates these candidates with features ., Finally , the features are given as input to a machine learning model , which predicts whether a given candidate sequence is indeed a TE of the considered order , and if so , predicts its superfamily ., In particular , we present TE-LearnerLTR , an implementation of this framework for LTR retrotransposons , which include the superfamilies Copia , Gypsy and Bel-Pao 2 ., As features we consider the occurrence of conserved protein domains , which help TEs perform the transposition process ., The machine learning method we apply is random forests ., This last step is essential , since the model of the three superfamilies contains the same protein domains 2; for Gypsy and Bel-Pao some domains even occur in the same order ., As LTR retrotransposons have a high abundance in the genomes of Drosophila melanogaster 17 and Arabidopsis thaliana 18 , 19 , and as these genomes are well annotated , they present the ideal candidates for evaluating how well our proposed method can identify and classify the LTR retrotransposons without using any prior information about these genomes ., We present an extensive quantitative analysis on D . melanogaster and A . thaliana comparing the obtained results to three widely used methods ( each dealing with one of the two tasks considered ) and we show that TE-LearnerLTR outperforms the state-of-the-art methods w . r . t . predictive performance and runtime ., The novelty of our proposed method w . r . t . the available methods lies mainly in three aspects ., First , in contrast to the other methods , which focus on one task , here we consider the tasks of identifying and classifying TEs together ., Second , we propose a general framework for these tasks ., Even though the implementation we provide in this article focuses on LTR retrotransposons , our framework can be extended to other TE orders ., Third , in contrast to classification methods such as LtrDigest , LtrSift , RepClass , Pastec and LtrClassifier , our method is not based on a predefined set of rules ., Instead , we exploit the strength of machine learning to automatically derive rules from the available data , with no need of prior knowledge ., Our framework is the first step towards completely automatic identification and classification of TEs in superfamilies ., We address the following problem: given an unannotated genome , find subsequences in it corresponding to a particular order from the classification scheme 2 , and predict their superfamily ., We propose the following three-step framework , called TE-Learner: We now discuss TE-LearnerLTR , one particular implementation , for every step in detail , focusing on the LTR retrotransposon order ., In Step 1 we use a simple splitting strategy to obtain subsequences of the genome ., The features used in Step 2 are conserved protein domains known to occur in LTR retrotransposons , and the machine learning model used in Step 3 is a random forest ., Fig 2 shows a schematic representation of our framework based on this implementation ., Note the modularity of the framework: every step can be implemented independently of the other steps ., For instance , an alternative implementation could use an LTR pair detection tool in Step 1 , annotate the candidates with oligomer frequencies in Step 2 , and apply an artificial neural network in Step 3 ., Any machine learning classifier can be used , as long as it outputs a probability ., We evaluate the predictive performance of our framework on the genomes of D . melanogaster and A . thaliana ., We use version 6-15 of the annotated genome from Flybase ( http://flybase . org/ ) , as the official annotation for D . melanogaster , which was made publicly available in April 2017 ., We use the Flybase annotations “Transposable Elements” and “Repeat Regions” to constitute the golden standard in our experiments ., For A . thaliana , we used the Araport11 annotation , released in June 2016 , for genome TAIR10 , from The Arabidopsis Information Resource ( TAIR ) ( http://www . arabidopsis . org ) ., We will compare our results for the Copia , Gypsy and Bel-Pao superfamilies ( Bel-Pao only for D . melanogaster because there is no annotation for it in A . thaliana ) with those of three methods for TE identification or classification that can make predictions at the superfamily level: RepeatMasker , Censor and LtrDigest ., For each superfamily , we also compare the results to those of a baseline model ., We now discuss the specific parameter settings for each of the tools used in our framework , as well as for the methods we compare to ., Baseline: The baseline model starts from the TE candidates obtained in Step 2 of our framework and makes predictions solely based on the presence of one key protein domain ( as predicted by the RPS-Blast program ) : RNase_HI_RT_Ty1 , RNase_HI_RT_Ty3 , and RT_pep_A17 for Copia , Gypsy , and Bel-Pao , respectively ., As such , it evaluates the impact of the machine learning aspect ( step 3 ) in our framework ., RPS-Blast: We constructed the database used by RPS-Blast by taking for each domain of interest the set of sequences from the Conserved Domain Database ( CDD ) ( http://www . ncbi . nlm . nih . gov/Structure/cdd/cdd . shtml ) 28 , used to generate the original multiple sequence alignment , except that we excluded the sequences of the organisms used in our tests ( D . melanogaster and A . thaliana , respectively ) ., The reason to exclude the D . melanogaster or A . thaliana sequences is that we want to provide an evaluation as blindly as possible , without using any known information from the target organism ., The PSI-Blast ( Position-Specific Iterative Blast ) program was used to obtain the new PSSMs and Makeprofiledb application for creating the database for RPS-Blast ., FORF: The relational trees were built with default parameters , except for the minimum number of examples in a leaf , which was set to 5 ., No pruning was used ., The forests consist of 100 trees , with a feature sample size at each node equal to the square root of the number of possible features ., For the training of FORF we used sequences from Repbase ( http://www . girinst . org/server/RepBase/ ) , volume 17—issue 3 , one set for each superfamily of interest here ., In order to provide a fair evaluation , as before , we excluded from these sets the sequences of the target organism ., Note that each analysis was performed twice: each time leaving out one target organism ., The resulting sets were also used as the databases for RepeatMasker and Censor applications—as described further ., We ran the RPS-Blast program , with the PSSM database created in Step 2 of our framework , to search these sequences for regions related to the conserved domains of interest ( Table 1 ) , retrieving the same types of information obtained from the screening of the candidates ( Step 2 ) ., We observed that the longest predicted domain region in these sequences has a length smaller than 800 nucleotides , which indicates that the overlap size of 1 , 000 nucleotides we used in Step 1 of our framework , is sufficient ., We removed training sequences without domain hits and those that contained domain hits in both strands of their genomic sequence ., The resulting number of sequences is 3188 for Copia , 4718 for Gypsy , and 891 for Bel-Pao when leaving out D . melanogaster sequences ., Leaving out A . thaliana , the numbers become 3077 for Copia and 4728 for Gypsy ., These sequences constitute the positive training set ., For each superfamily , we also constructed a negative set , by sampling without replacement from the other superfamilies ., For Copia and Bel-Pao the negative set has an equal size as the positive set; however for Gypsy , given the size of its positive set , the negative set contains less sequences ( all Copia and Bel-Pao sequences ) , which still yields a balanced classification task ., RepeatMasker and Censor: These systems were run using their standard parameter settings ., For a fair evaluation , we used as reference library the same training sets as for FORF as described above ., As each of the training sets belongs to a particular superfamily , we can label hits with the corresponding superfamily ., Both applications were run on the complete genomes ., LtrDigest: This method was also run with its standard parameter settings on the complete genomes ., We only retained predictions with an assigned DNA strand and used the authors’ guidelines to assign a particular superfamily to each prediction as follows ., Every predicted sequence is annotated with protein domain hits ., If the sequence has a “Peptidase_A17” hit it is classified as BelPao; otherwise , if the sequence has a “Gypsy” hit , it is classified as Gypsy; otherwise , following 10 , if the sequence has an “INT” followed by an “RT” ( there may be other hits in between ) , it is classified as Copia and if an “RT” is followed by an “INT” , it is classified as Gypsy ., The remaining sequences are not classified ., We report the predictive performance of the different methods with precision-recall ( PR ) curves 29 ., The motivation for preferring PR curves over the more popular ROC curves is as follows ., Only a small fraction of the genome contains TE sequences of a specific superfamily , thus we are more interested in recognizing the positive cases , i . e . the candidate sequences that actually belong to the superfamily , than in correctly predicting the negatives ., Precision is the percentage of predictions that are correct and recall is the percentage of annotations that were predicted ., A PR curve plots the precision of a model as a function of its recall ., Assume the model predicts the probability that a new example is positive , and that we threshold this probability with a threshold t to obtain the predicted class ( positive or negative ) ., A given threshold corresponds to a single point in PR space , and by varying the threshold we obtain a PR curve: while decreasing t from 1 . 0 to 0 . 0 , an increasing number of examples is predicted positive , causing the recall to increase whereas precision may increase or decrease ( with normally a tendency to decrease ) ., A domain expert can choose the threshold corresponding to the point on the curve that looks most interesting ., To consider a prediction as a true positive , we do not require it to match the exact same boundaries of the corresponding annotation of the genome , as this would be an overly strict criterion ., Instead , we allow some tolerance by defining a true positive as a prediction which has a minimum overlap of 100 nucleotides with an annotation , or a prediction which overlaps a complete annotation and vice versa ., Our motivation for this evaluation is that a domain expert can inspect each prediction and determine the exact boundaries of the complete TE ., The random forests in TE-LearnerLTR only make predictions w . r . t . the superfamily for which they were built ., For example , one forest outputs the probability whether a sequence belongs to Copia or not ., However , one might be interested in having a model that can make predictions w . r . t . many superfamilies at the same time ., An advantage of such a model is that the user does not need to combine the results of individual models , avoiding conflicting predictions ., As our models output probabilities , one straightforward idea to obtain this more general model consists of selecting the superfamily with the highest probability ., To avoid that a superfamily with a very low probability is predicted , we include the category None ( i . e . , the sequence does not belong to any of the considered superfamilies ) , which is predicted when none of the probabilities exceeds a certain threshold ., In this setting , we construct a single average PR curve for all superfamilies together as follows ., When a sequence is predicted to have a certain superfamily , we consider it correct if the sequence indeed belongs to that superfamily ., The definition of precision and recall is then as before ., Thus , for precision , the denominator contains all candidate predictions , minus those predicted as None; for recall , the denominator contains all annotations ( for all considered superfamilies ) ., We compare our results to those of LtrDigest , which is also able to make predictions w . r . t . different superfamilies at the same time ., Before discussing each superfamily in detail , we first show for both genomes the number of predictions that were made and the average prediction length of each method and for each superfamily ( Table 3 ) ., Note that TE-LearnerLTR presents the same numbers of candidates and average length of candidates for the three superfamilies ., This happens because in our implementation Steps 1 and 2 output one common candidate set for the three superfamilies ., From the table it is clear that RepeatMaskerand Censor make a lot of predictions , which are on average much smaller than the predictions of TE-LearnerLTR ., LtrDigest on the other hand , makes much less predictions , which are considerably longer ., Figs 10 and 11 report the combined PR curves of TE-LearnerLTR and the point of LtrDigest ., For D . melanogaster , the point of LtrDigest obtains a slightly higher precision ( 0 . 80 ) than TE-LearnerLTR ( 0 . 77 ) at a recall of 0 . 15 ., For A . thaliana , LtrDigest and TE-LearnerLTR both obtain a precision of 1 , however , the latter obtains a higher recall ., Moreover , our combined model has the advantage of allowing the user to choose an appropriate threshold ., In this paper we have proposed a framework based on machine learning to identify and classify TEs in a genome ., We evaluated our approach on three Class I TE superfamilies in D . melanogaster , and two Class I TE superfamilies in A . thaliana , using a relational random forest model ., We found a better predictive performance ( w . r . t . F1 measure ) and runtime compared to three widely used methods for TE identification and classification ., In terms of F1 , the performance of RepeatMasker comes close to TE-LearnerLTR , however , it obtains a higher recall , because it is able to recover TEs that have no conserved protein domains ., The fact that we rely on these protein domains is a clear limitation of our method , yet , we are able to find TEs that other methods did not find ., This suggests that TE-LearnerLTR presents a viable alternative to the state-of-the-art methods , in case one prefers predictions with a very high precision , or as a complement to the other methods when one is interested in finding more TEs ., Furthermore , for our top predictions not confirmed by the official annotations , we validated their homology to known TEs of the corresponding superfamilies , showing that our method could be useful to detect missing annotations ., While our implementation has been focusing on LTR retrotransposons , it is possible to train it on other TE orders with superfamilies that have recognizable protein domains ., Alternatively , one could change the implementation of any of the steps of the framework: the machine learning model , the features used , and the candidate generation procedure ., For instance , to identify TEs from the TIR order ( a Class II order with Terminal Inverted Repeats ) , the first step could use software tools to identify a candidate set of sequences surrounded by TIRs ( such as 30 , 31 ) ., A possible direction for further work is to explore hierarchical classification methods in the machine learning step of the framework ., This would allow to exploit the underlying structure of the TE classification scheme ., Additionally , one could try to still boost the performance of the different steps of the framework , e . g . , by improving protein domain detection , or by including additional features ( including features not related to protein domains ) in the decision trees . | Introduction, Methods, Results, Discussion | Transposable elements ( TEs ) are repetitive nucleotide sequences that make up a large portion of eukaryotic genomes ., They can move and duplicate within a genome , increasing genome size and contributing to genetic diversity within and across species ., Accurate identification and classification of TEs present in a genome is an important step towards understanding their effects on genes and their role in genome evolution ., We introduce TE-Learner , a framework based on machine learning that automatically identifies TEs in a given genome and assigns a classification to them ., We present an implementation of our framework towards LTR retrotransposons , a particular type of TEs characterized by having long terminal repeats ( LTRs ) at their boundaries ., We evaluate the predictive performance of our framework on the well-annotated genomes of Drosophila melanogaster and Arabidopsis thaliana and we compare our results for three LTR retrotransposon superfamilies with the results of three widely used methods for TE identification or classification: RepeatMasker , Censor and LtrDigest ., In contrast to these methods , TE-Learner is the first to incorporate machine learning techniques , outperforming these methods in terms of predictive performance , while able to learn models and make predictions efficiently ., Moreover , we show that our method was able to identify TEs that none of the above method could find , and we investigated TE-Learner’s predictions which did not correspond to an official annotation ., It turns out that many of these predictions are in fact strongly homologous to a known TE . | Over the years , with the increase of the acquisition of biological data , the extraction of knowledge from this data is getting more important ., To understand how biology works is very important to increase the quality of the products and services which use biological data ., This directly influences companies and governments , which need to remain in the knowledge frontier of an increasing competitive economy ., Transposable Elements ( TEs ) are an example of very important biological data , and to understand their role in the genomes of organisms is very important for the development of products based on biological data ., As an example , we can cite the production biofuels such as the sugar-cane-based ones ., Many studies have revealed the presence of active TEs in this plant , which has gained economic importance in many countries ., To understand how TEs influence the plant should help researchers to develop more resistant varieties of sugar-cane , increasing the production ., Thus , the development of computational methods able to help biologists in the correct identification and classification of TEs is very important from both theoretical and practical perspectives . | invertebrates, retrotransposons, engineering and technology, brassica, animals, invertebrate genomics, animal models, decision analysis, drosophila melanogaster, model organisms, management engineering, artificial intelligence, experimental organism systems, genetic elements, plants, drosophila, arabidopsis thaliana, research and analysis methods, sequence analysis, computer and information sciences, decision trees, bioinformatics, proteins, biological databases, insects, animal genomics, arthropoda, biochemistry, machine learning, eukaryota, plant and algal models, sequence databases, database and informatics methods, protein domains, genetics, biology and life sciences, transposable elements, genomics, mobile genetic elements, organisms | null |
journal.ppat.1002487 | 2,012 | Two-Drug Antimicrobial Chemotherapy: A Mathematical Model and Experiments with Mycobacterium marinum | The concurrent use of multiple drugs , which is one of the mainstays of chemotherapy , is useful and in some cases necessary for the successful treatment of diseases such as tuberculosis ( TB ) , HIV/AIDS , malaria and various cancers ., Shortly after antimycobacterial agents became available for treating TB , it was recognized that single drug therapy almost invariably led to treatment failure due to the ascent of resistance , but that this could be mitigated by the use of multiple drugs with different modes of action 1–4 ., In its current form , standard tuberculosis treatment consists of a two-month combinatorial course of rifampin , isoniazid , pyrazinamide and ethambutol , followed by a four-month continuation phase of isoniazid and rifampin ., Despite the barrage of antibiotics and long term of combination therapy , Mycobacterium tuberculosis ( Mtb ) strains that are resistant to multiple drugs are an increasingly troubling component of the epidemiological landscape ., In 2009 , the World Health Organization estimated close to half a million cases of multidrug resistant ( MDR ) TB ( cases in which recovered strains were resistant to the most potent first-line antibiotics , rifampin and isoniazid ) 5 ., By mid-2010 , 58 countries had reported at least one case of extensively drug-resistant ( XDR ) TB ( MDR strains that are additionally resistant to any fluoroquinolone as well as at least one of the injectable drugs capreomycin , kanamycin and amikacin ) 5 ., The important issue is thus: how can the term of tuberculosis chemotherapy and the likelihood of treatment failure due to the evolution of resistance during the course of therapy be reduced ?, One approach to improving the efficacy of single drug therapy has been to design treatment regimes based on in vivo data of the changes in the concentration of the antibiotic , pharmacokinetics ( PK ) , and in vitro data on the relationship between the concentration of the drug and the rate of growth/death of the bacteria , pharmacodynamics ( PD ) 6–9 ., This PK/PD approach to the rational design of antibiotic treatment regimes has been employed for tuberculosis but almost exclusively for single antibiotics 10–19 ., To extend this approach to the multi-drug treatment regimes clearly needed to prevent acquired resistance , it is necessary to concurrently account for the PD of the different drugs , and most critically , how they interact 20–22 ., Drug interactions are generally classified as antagonistic , synergistic or additive ., In the case of bactericidal antibiotics , additive interactions are usually described in one of two ways , ‘Bliss Independence’ and ‘Loewe Additivity’ ., Bliss Independence asserts that each drug in a combination exerts its killing action independently of the other drugs 23 ., For example , if there are two drugs , A and B , and at particular concentrations they kill fa and fb ( 0<fa , fb<1 ) fractions of a bacterial population in an hour , at the end of the hour the viable cell density would be reduced to ( 1-fa ) ( 1-fb ) of its initial level ., For Loewe additivity , the fraction of surviving cells with both drugs would be 1-fa-fb , the constraint being that fa+fb<1 24 ., Antagonism and synergism can then be defined relative to one of these descriptions of additivity: drugs interact antagonistically if their combined cidal activity is less than would be predicted for an additive drug combination , and synergistically if the cidal activity is more ., Unfortunately , these definitions cannot be readily translated into the PD of two drugs as they do not account for how the rate or extent of killing would vary with the concentrations of the drug ., To address this , Greco and colleagues proposed a seminal Emax-based two-drug pharmacodynamic function which assumes that a single parameter can account for the interaction between both drugs 25 , 26 ., If the value of this parameter is zero , then the drugs are additive , with a negative value indicating antagonism and a positive value indicating synergy ., Although this and other Emax-based models have been used to characterize the nature of the interactions between different kinds of drugs , including antimicrobials 27–33 , there has been limited quantitative consideration of how two-drug PD models apply to the design and evaluation of antibiotic treatment regimes for bacteria , particularly those , like tuberculosis , where multiple drug therapy is essential 27 , 33 ., In this study , we explore the fit of Hill functions ( which subsume Emax models ) for the PD of the antimycobacterial antibiotics rifampin , amikacin , clarithromycin , streptomycin and moxifloxacin ., We then employ a Hill-function-based variant of the Greco model to explore the PD of the 10 possible pairs of these drugs ., As our experimental organism , we use Mycobacterium marinum ., In addition to being safer and more convenient to work with , M . marinum is a close genetic relative and shares numerous virulence determinants with Mtb ., It also recapitulates key immunopathological features of human tuberculosis infection in its natural poikilothermic hosts 34–36 ., To explore the potential clinical implications of these theoretical and in vitro PD studies , we use Monte Carlo simulations of antibiotic treatment and resistance that incorporate PD functions that best fit our data ., Of particular concern in this analysis are:, ( i ) the relative rates at which these different drug combinations clear the simulated infections ( their microbiological efficacy ), ( ii ) the likelihood of resistance to the two drugs evolving during the course of therapy ( their evolutionary efficacy ) , and, ( iii ) how that efficacy is affected by different forms of non-adherence to the treatment regime ., In Figure 1 we show the fit of the theoretical single-drug pharmacodynamic function ( Equation 1 ) to the PD data obtained from experiments with five antimycobacterial agents ., These data were generated by exposing M . marinum to the antibiotics at different concentrations and estimating net bacterial growth/death rates ( based on the increase or decrease in the density of viable bacteria ) over 72 hours ., The analyses of these time-kill data were restricted to 72 hours in order to ensure that bacteria were growing and/or being killed exponentially ., For single antibiotics , the Hill function provides a good fit for the relationship between the concentration of the drug and the growth/death rate of the bacteria ( Figure 1 , see R2 values ) ., This is also evident in Table 1 , where we list the estimates of the Hill function parameters for each of the drugs ., The maximum growth rates calculated from this function are very close to that estimated independently ( data not shown ) ., Moreover , the estimated zMICs ( MICs calculated from the Hill functions ) and MICs determined by the CLSI 37 recommended broth dilution method are , given the factor of two limitation of the latter , coincident ., The individual antibiotics exhibited different pharmacodynamic signatures reflected in the varying shapes of the PD function ( the parameter κ ) and the kill rate parameter ψmin , which ranged from −0 . 043 to −0 . 166 h−1 ., With the PD function parameter estimates for single antibiotics in hand , we proceeded to assess the validity of the two-drug pharmacodynamic function ( Equation 3 ) ., To accomplish this , we exposed M . marinum to combinations of antibiotics , each of which was at some multiple of its respective MIC , and estimated the growth/death rates of the bacteria over 72 hours ., Using the differential equation ( Equation 4 ) , the estimated single-drug Hill function parameters and different values of α , we compared the observed growth/death rates to those anticipated from the unique α model ., In Figure 2 we show the experimentally-observed changes in bacterial growth/death rates generated by different two-antibiotic combinations ( curves with markers ) together with those predicted from our model for different drug interaction parameters , the αs ( curves without markers ) ., Our estimates of these growth/death rates were limited to situations where the density of surviving cells exceeded 10 CFU per ml ., Both the experimental and theoretical analyses were conducted for all possible two-drug combinations of the antimycobacterial drugs used in the study ., For all the drug combinations , it is apparent that a single interaction parameter is insufficient to describe the dynamics over the entire range of concentrations assessed ., While the deviation of fit from this single α function varies among antibiotic pairs , in all cases , at lower drug concentrations the observed growth rate is greater than that anticipated from the model ., The fit with a single value of α does , however , get somewhat better at higher drug concentrations ., To get a better idea of the relationship between antibiotic concentration and α , we used Equation 5 to separately estimate this interaction parameter for different concentrations of the ten drug pairs ( Figure 3 ) ., For all antibiotic combinations , this interaction became relatively more synergistic with increasing drug concentration ., Interactions at sub-MIC concentrations were universally antagonistic , but could be mildly antagonistic , additive or synergistic at supra-MIC concentrations ( Figure 3 and Table S1 ) ., In addition , the rate of change in α from one concentration to the next was much greater at sub-MIC than at supra-MIC concentrations ., Interaction coefficients at the larger concentrations only changed to a limited extent and appeared to approach constancy , mirroring the results shown in Figure 2 . Although not providing a precise fit to these data , if we assume a two-phase interaction function , one for sub- and one for supra-MIC concentrations and use linear regressions to generate the α functions for each phase , a reasonable fit obtains ( Figure 3 and Table S1 ) ., For convenience , but also to make this approach to evaluating the pharmaco- and population dynamics of two-drug antibiotic treatment readily applicable , we restricted the above PD experiments to situations in which both antibiotics were at the same xMIC concentration ., In an effort to explore the robustness of the two-drug PD observed for these cases of symmetric drug concentrations , we performed time kill experiments for three asymmetric ( unequal xMIC concentrations ) situations:, ( i ) where both antibiotics are below their respective MICs ,, ( ii ) where one antibiotic is below its MIC and the other above and, ( iii ) where both are above their MICs ., When both antibiotics are below the MIC , there is antagonism similar to that observed for the symmetric case ., This can be seen in Figure S1 , where we present the observed growth rates and those anticipated for situations where there is no interaction between the drugs , α\u200a=\u200a0 ., As would have been anticipated from the symmetric combination results ( Figure 2 ) , at sub MICs the drugs together kill at a lower rate than expected were there no interactions between them i . e . they exhibit antagonism ., Moreover , the estimated αs for the combination of 0 . 1 and 0 . 5 xMIC concentrations of the antibiotics were generally less negative than those calculated for combinations of 0 . 1-0 . 1xMIC but more negative than those calculated for the 0 . 5-0 . 5 xMIC symmetric cases ( Table S2 ) ., Of particular concern in situations where one drug is below the MIC and the other above is that the substantial antagonism observed for below-MIC antibiotic concentrations would be manifest by sub-MIC drugs reducing the efficacy of supra-MIC antibiotics ., The results of our experiments indicate that this is not the case ( Figure S2 ) ., When combined with a sub-MIC concentration of a second drug , the rate of kill of the supra-MIC drug is no less than that when it is alone and in some cases greater ., To explore the effects of asymmetric concentrations for pairs of above-MIC antibiotics , we compared the observed death rate with that anticipated for no interaction between the antibiotics ., The results of these experiments suggest that there is either no interaction between the antibiotic pairs or there is the mild antagonism or synergy observed for the symmetric drug concentration experiments ( Figure S3 ) ., In sum , the results of these experiments with asymmetric drug concentrations are consistent with that anticipated from the symmetric concentration experiments depicted in Figure 3 . To evaluate how the pharmacodynamics estimated above would be manifest in a treatment regime , we use a simulation of the within-host population dynamics of bacteria in a two-drug therapy regime for tuberculosis ., In Figure 4 , we present a diagram of the model used for the analysis ( equations for the model can be found in Protocol S1 ) ., In designing this model and in choosing the dosing parameters , bacterial densities and PD parameters , we tried to mimic that which would be appropriate for mycobacterial chemotherapy ., The structure of our model is based on that suggested by D . Mitchison 38 ., It assumes two compartments , one in which the bacteria are actively proliferating and the other where they are dividing slowly and thereby responding differently to antibiotics 39 , 40 ., This compartment difference in antibiotic susceptibility is reflected in the pharmacodynamic Hill functions , such that the maximum and minimum rates of growth/death are proportional to the rate of replication in the two compartments ., The idea is that the slowly dividing subpopulation is relatively refractory to killing by the antibiotics , as would be the case for latent or persister cells in a tuberculosis infection ., We allow for four states of the bacteria , one that is susceptible to both drugs , S0 and L0 ( S and L for rapidly- and slowly-dividing populations respectively ) , S1 and L1 for those resistant to drug 1 , S2 and L2 for cells resistant to drug 2 , and S12 and L12 for cells that are resistant to both drugs ., These variables are both the densities ( cells/ml ) of bacteria in these states as well as their state designations ., By resistance we are assuming that these bacteria are totally refractory to the drugs , with MICs at least 100X that of the susceptible cells ., Resistance also engenders a 5% fitness cost which is manifest as a 5% lower maximal growth rate of bacteria in those states ., This assumed cost is in the range of what has been observed for M . marinum mutants resistant to the antibiotics considered in this study ( unpublished results ) ., We allow migration at rates fls ( from latent to susceptible ) and fsl ( from susceptible to latent ) cells per hour , representing either a physical or a physiological translocation between the compartments ., Resources for bacterial growth enter and are removed from the habitat ( host ) at a constant rate , w ml per hour ., The bacteria , however , are removed from the habitat at two rates , w for S0 , S1 , S2 and S12 , and wL for L0 , L1 , L2 , and L12 , where w> wL ., For the pharmacodynamic functions , we use the two-drug Hill functions with the biphasic model for the interaction coefficient described above ., For pharmacokinetics we assume that a fixed dose A1max and A2max of each drug is added every T hours ., In addition to washout at rate w , both drugs also decay at a rate d mg/L per hour ., In these simulations we assume that at the onset of treatment , the sensitive population is initially at a density of S0\u200a=\u200a5×107 in the main compartment 41 and L0\u200a=\u200a5×104 cells per ml in the refractory compartment ., As would be anticipated for hosts infected with numbers of bacteria that exceed the reciprocal of the mutation rates , we assume that there are minority populations of bacteria resistant to single antibiotics , S1 , S2 , L1 and L2 , with a relative frequency of 10−3 to the corresponding susceptible population 42 ., We also allow resistance to single drugs to evolve during the course of the simulations at rates proportional to the product of the number of individuals of each ancestral state and a mutation rate ., The actual generation of mutants occurs in a semi-stochastic manner , via a Monte Carlo routine ., At each time step ( Δt ) in the finite step size ( Euler ) simulation , the probability that a mutant would be generated is the product of the number of individuals of the genotype , Δt and the mutation rate µ ., When the random number is less than this product , a mutant is added to the noted population , e . g . when S1 is generated from S0 , a bacterium is added to the S1 state and one removed from the S0 state ., We use step sizes of Δt so that the probability of a mutant being added at a particular time interval is always less than 1 . For these simulations , µ takes values in the range of that estimated from fluctuation experiments for different antibiotics and M . marinum ( unpublished results ) ., There are no doubly resistant cells , S12 and L12 at the start of the simulations , but they can evolve by mutation from the single resistant states ., In Figure 5 , we follow the changes in density of the different bacterial populations in the main compartment ( 5a ) and in the refractory compartment ( 5b ) ., The PD parameter values used in this simulation are those in the range estimated in our experiments for the combination of rifampin ( A1 ) and amikacin ( A2 ) ., These antibiotics are inoculated every 24 hours at a concentration of 5X their respective MICs and decline in concentration due to flow and a decay rate , d\u200a=\u200a0 . 075 per hour ., With these parameters , the overall densities of the sensitive and single-resistant populations continue to decline during the course of the simulation ., In the main compartment this decline is punctuated by oscillations in density reflecting the waxing and waning of the antibiotic concentration , with net decline each hour ., The single resistant populations are cleared earlier than the sensitive for two reasons: their lower initial densities and their lower fitness relative to the sensitive bacteria ., This interpretation was confirmed by running simulations in which single resistant populations were at higher initial densities and had lower fitness costs ( data not shown ) ., Under these conditions , their resistance to single antibiotics does not make up for this fitness cost ., In the refractory compartment , the rate of change in cell density is lower and the oscillations are not manifest to the same extent as in the main compartment ., This occurs because the replication and washout rates are lower , as is the rate of kill by the antibiotics ., As a result of continuous migration of cells from and to the slower-growing population , the rate of decline in the density of cells in the main compartment is reduced whilst that in the refractory compartment increased relative to what would obtain were they the sole compartments or not connected ., Said another way , the existence of a refractory compartment prolongs the term of therapy ., To compare the relative efficacy of different combinations of antibiotics , we ran these simulations with the estimated PD parameter values obtained for the different combinations of drugs ., In addition to simulations with symmetric antibiotic concentrations for the two drugs , we also conducted these simulation experiments with asymmetric antibiotic concentrations ., The former were initiated with 5xMIC of both drugs and the latter with 5xMIC of one antibiotic and 2xMIC of the other ., As a result of flow and decay , the asymmetric drug concentration simulations include periods where both drugs are above the MIC , one above and one below , and both below ., The interaction coefficients used in these simulations are those estimated from the corresponding symmetric and asymmetric concentration experiments ., As our measure of the efficacy of treatment , we considered the time until the total density of bacteria was less than one ( time to clearance ) ., The results of these simulations are presented in Table 2 . While in some runs doubly resistant mutants emerged , ascended and thereby precluded clearance , these were not included in the Table 2 clearance data ., The frequencies of runs in which double resistance emerged are considered separately ., Although mutation is a stochastic process , there was effectively no between-run variation in the time before clearance ., For eight out of the ten combinations , clearance occurred in less than 1600 hours ., The rifampin + amikacin combination was the most effective , leading to clearance in 1080 hrs ., The combinations of clarithromycin + moxifloxacin and clarithromycin + streptomycin took substantially longer to clear the bacteria; compared to the rifampin + amikacin combination , the clarithromycin + moxifloxacin combination took some 4 times longer , with the clarithromycin + streptomycin combination taking approximately 11 times longer ., This is what would be anticipated from the relative pharmacodynamics of the different drug combinations ( Figure 2 ) ., As in the symmetric case , the majority of the antibiotic combinations in the asymmetric simulations cleared the infection over a relatively similar period , i . e . <2800 hours ., The reason that the average time to clearance is greater for the asymmetric concentrations is because there is a lower peak concentration for one of the two drugs , rather than equal peaks ., While clarithromycin + streptomycin and clarithromycin + moxifloxacin remained the least effective drugs , the most effective combination was streptomycin + moxifloxacin rather than rifampin + amikacin ., Compared to streptomycin + moxifloxacin , clarithromycin + moxifloxacin and clarithromycin + streptomycin took , respectively , approximately 2 . 5 and 6 times longer to clear the infection ., What is the relationship between the PD of the antibiotics and the likelihood of mutants resistant to both drugs emerging ?, To address this question , we separately performed 1000 simulation experiments using three sets of parameters reflecting the ‘extreme’ conditions of relative efficacy for the symmetric combinations: rifampin + amikacin , clarithromycin + moxifloxacin and clarithromycin + streptomycin ., The aggregate results from these simulation experiments are presented in column one of Table 3 . As can be seen , the two-drug resistant population emerged in only a few runs ., Although the relative number of runs in which resistance emerged for the different drug combinations is what would be anticipated from the clearance data in Table 2 , the differences were not statistically significant ( p∼0 . 525 ) ., With these parameters , the frequency of two-drug resistance emerging was low and was roughly the same for all three pairs of drugs ., In a number of epidemiological studies , non-adherence to the prescribed treatment regime has been associated with adverse therapeutic outcomes 43 , longer terms of treatment and acquired drug resistance 44 , 45 ., In practice , non-adherence takes a number of forms and depends on a variety of factors such as organization of treatment and care ( access to services , length , drug-type and other requirements for therapy , support services , etc ) individual interpretations of illness and wellness , drug side effects and the social context in which therapy is undertaken 46 ., How does non-adherence contribute to the amount of time required for microbiological cure and the likelihood of multi-drug resistance emerging within a host during the course of treatment ?, How sensitive are different drug combinations to the adverse outcomes of non-adherence ?, To address these questions , we considered three broadly-inclusive types of non-adherence that we call random , thermostat 39 , and drug holiday ( described below ) ., To explore the relationship between the PD of the drug combinations and the frequency of non-adherence with respect to the generation of the double resistant mutants , we conducted 1000 runs for each of the three aforementioned drug combinations and the different non-adherence scenarios ., The results of these simulations are presented in Table 3 . We model this scenario in the following manner: At each dosing period there is a probability P ( 0≤P≤1 ) that both drugs will be taken and a corresponding probability ( 1-P ) that neither will be taken ., To simulate this we use a Monte Carlo routine where if the random number , r≤P , the drugs are administered , but if r>P that dosing period is skipped ., In Figure 6, ( a ) , we illustrate this process for a single run where two-drug resistance emerges ., Non-adherence is reflected in a hiatus in the dosing and a rise in the density of all the bacterial populations ., There are periods , such as between 600 and 648 hours , where consecutive doses are missed ., This results in a substantial rise in the density of bacteria and thereby an increase in the likelihood of a doubly resistant mutant being generated ., With 10% random non-adherence ( P\u200a=\u200a0 . 9 ) , there was no significant difference among drug combinations in the probability of resistance arising ( p∼0 . 073 ) ( Table 3 , Column 2 ) ., With 20% random non-adherence ( P\u200a=\u200a0 . 8 ) there was a highly significant drug combination effect , p∼0 . 001 ( Table 3 , Column 3 ) ., The likelihood of multiple resistance arising with 20% non-adherence was negatively related to the microbiological efficacy of these different drug combinations ., The relationship between the probability of a doubly resistant population emerging for different levels of random non-adherence was also directly related to the microbiological efficacy of the drug combinations ., For the rifampin + amikacin combination , there was no significant difference among the 0 , 10% and 20% non-adherence regimes ( p∼0 . 435 ) ., For the other two pairs , there were significant p<0 . 001 relationships between the frequency of non-adherence and the likelihood of double resistance emerging ., We simulate this by incorporating a situation in which treatment ceases when the density of the rapidly growing population falls below 104 and doesnt commence again until the density exceeds 106 ., The situation we are mimicking is one in which patients cease taking their antibiotics when they are feeling better ( the bacterial densities are low enough not to be symptomatic ) and do not take the drugs again until the density is high enough to be symptomatic ., We illustrate this situation in Figure 6, ( b ) with a run in which two-drug resistance emerged ., In column 4 of Table 3 , we summarize the results of 1000 simulations of thermostat non-adherence for the three drug combinations ., With respect to our measure of microbiological efficacy , the thermostat non-adherence scenario seems paradoxical ., Two-drug resistance emerged far more frequently in the runs with the most microbiologically effective drug combination , indeed in all 1000 runs ., The reason for this is that the more effective drug combination reduced the density more rapidly than the less effective drug combinations ., As a result there were far more frequent periods where drugs were not taken and the single-resistant populations ascended to high-enough densities where two-drug resistant mutants were produced with a very high probability ., Under the parameter conditions of this simulation , the non-adherence threshold was never crossed in any of the 1000 simulations for either of the two less effective drugs ., We model this scenario in the following manner: Both drugs are taken for 4 consecutive dosing periods , at which time neither drug is taken for the subsequent 3 dosing periods ., This regime continues throughout the duration of simulated treatment ., We are mimicking a situation where holidays are imposed because the drugs may be costly , limited in their availability or induce debilitating side effects that are alleviated by terminating treatment for an interval ., In Figure 6, ( c ) we illustrate this situation for a run where two-drug resistance emerged ., As noted in the last column of Table 3 , the overall frequency of double resistance was on the order of 5% and similar for the two microbiologically less effective drug combinations ., For the most effective drug combination , relative to complete adherence , the drug holidays doubled the likelihood of two-drug resistance emerging ., With few exceptions , studies of the pharmacodynamics ( PD ) of antibiotics and bacteria have been restricted to single drugs 10–19 ., Some infections , particularly those that are long-term like tuberculosis , require multiple antibiotics for treatment to be effective ., It follows then , that for the rational design of treatment protocols for these infections , multidrug PD analyses are necessary ., Our results indicate that Hill functions provide an excellent fit for the single-drug PD for Mycobacteria marinum and each of the five antibiotics considered in this study , amikacin , clarithromycin , moxifloxacin , rifampin and streptomycin ., On the other hand , if , as is assumed in the classical model of Greco and colleagues 25 , 26 , the interactions between drugs is expressed as a single parameter with a constant value , two-drug Hill function models do not fit the PD observed for any of the 10 pairs of drugs considered ., In all cases , at lower antibiotic concentrations the interactions between the drugs is antagonistic; they are less effective together than anticipated from their action alone ., As the antibiotic concentrations increase , this drug-drug interaction becomes relatively more synergistic and approaches constancy ., To address this phenomenon , we allow for two phases of the drug-drug interaction , one for low ( sub-MIC ) and one for high ( supra-MIC ) concentrations with an antibiotic concentration-dependent function for the interaction term ., Albeit not as convenient as a unique parameter , these functions can be readily estimated from time-kill data ., Most importantly , the biphasic drug interaction Hill function models thus generated provide quantitatively accurate analogues of the PDs of all 10 pairs of antibiotics examined ., It has been hypothesized that there are subpopulations of bacteria within an infected TB host that exhibit differential growth rates and , by extension , variable susceptibility to antimycobacterial agents 38 , ., Here , we develop a simple mathematical model that accounts for this within-host bacterial heterogeneity by assuming that there are two ‘compartments’ , one that houses rapidly-growing and the other slowly-growing bacteria ., The model incorporates the possibility of non-adherence to therapy , which is considered to be one of the major contributory factors to TB treatment failure 43 , 45 , 51 ., Our computer simulations of tuberculosis chemotherapy employing the empirically estimated biphasic Hill functions suggest that there can be substantial differences among drug combinations in treatment efficacy , as measured by the time to clearance ., Of the ten antibiotic pairs we consider , rifampin + amikacin is the most effective and streptomycin + clarithromycin the least , with some eleven-fold difference in the time before clearance ., With the parameters used in our semi-stochastic model of treatment and assuming different probabilities for the occurrence of random non-adherence , either complete adherence or limited non-adherence to the therapeutic regime would not be manifest as a significant difference among drug combinations in the likelihood of the generation and ascent of two-drug resistant mutants ., However , with greater rates of non-adherence , the likelihood of two-drug resistance emerging becomes increasingly dependent on the drug combination employed ., The emergence of two-drug resistance due to random non-adherence is more likely for less microbiologically effective drug combinations than those that are more effective ., With externally imposed regular drug holidays , the likelihood of emergence of two-drug resistance is also inversely proportional to the microbiological efficacy of th | Introduction, Results, Discussion, Materials and Methods | Multi-drug therapy is the standard-of-care treatment for tuberculosis ., Despite this , virtually all studies of the pharmacodynamics ( PD ) of mycobacterial drugs employed for the design of treatment protocols are restricted to single agents ., In this report , mathematical models and in vitro experiments with Mycobacterium marinum and five antimycobacterial drugs are used to quantitatively evaluate the pharmaco- , population and evolutionary dynamics of two-drug antimicrobial chemotherapy regimes ., Time kill experiments with single and pairs of antibiotics are used to estimate the parameters and evaluate the fit of Hill-function-based PD models ., While Hill functions provide excellent fits for the PD of each single antibiotic studied , rifampin , amikacin , clarithromycin , streptomycin and moxifloxacin , two-drug Hill functions with a unique interaction parameter cannot account for the PD of any of the 10 pairs of these drugs ., If we assume two antibiotic-concentration dependent functions for the interaction parameter , one for sub-MIC and one for supra-MIC drug concentrations , the modified biphasic Hill function provides a reasonably good fit for the PD of all 10 pairs of antibiotics studied ., Monte Carlo simulations of antibiotic treatment based on the experimentally-determined PD functions are used to evaluate the potential microbiological efficacy ( rate of clearance ) and evolutionary consequences ( likelihood of generating multi-drug resistance ) of these different drug combinations as well as their sensitivity to different forms of non-adherence to therapy ., These two-drug treatment simulations predict varying outcomes for the different pairs of antibiotics with respect to the aforementioned measures of efficacy ., In summary , Hill functions with biphasic drug-drug interaction terms provide accurate analogs for the PD of pairs of antibiotics and M . marinum ., The models , experimental protocols and computer simulations used in this study can be applied to evaluate the potential microbiological and evolutionary efficacy of two-drug therapy for any bactericidal antibiotics and bacteria that can be cultured in vitro . | The goal of this investigation is the development and a priori evaluation of multi-drug treatment regimes that are effective in clearing long-term bacterial infections like tuberculosis , and also minimize the likelihood of multi-drug resistance arising during therapy ., To achieve this end , we use mathematical models and in vitro experiments with Mycobacterium marinum ( a close relative of M . tuberculosis ) and five different antimycobacterial agents to develop and validate realistic analogues of the pharmacodynamics of two-drug chemotherapy ., All ten drug pairs examined exhibited the same general biphasic drug-drug interaction properties: at low concentrations ( subMICs ) , the two drugs together were less effective than anticipated from their independent pharmacodynamics ( were antagonistic ) , but as their concentrations increased , the interactions between the antibiotics became relatively more synergistic ., Using computer simulations with these empirically estimated two-drug pharmacodynamic functions , we evaluated the relative efficacy of the different antibiotic combinations in terms of the anticipated rate of clearance of infections and the likelihood of resistance arising with and without non-adherence to a treatment regime ., The simulations predict different outcomes for each of the drug combinations ., The models and experimental methods used in this study can be applied to characterize any combinations of bactericidal antibiotics and evaluate their potential efficacy . | bacteriology, medicine, infectious diseases, theoretical biology, biology, evolutionary biology, microbiology, population biology, evolutionary processes | null |
journal.pgen.1007143 | 2,017 | Parallel evolution of the POQR prolyl oligo peptidase gene conferring plant quantitative disease resistance | Plant pathogens are major threats to biodiversity in natural ecosystems and to food security worldwide ., Plant disease resistance is mediated by an elaborate multilayered system of defense , sometimes including resistance ( R ) genes conferring complete resistance against a limited number of pathogen genotypes and encoded by nucleotide-binding site leucine-rich repeat ( NBS-LRR ) proteins 1 ., Substantial progress has been made in understanding the genetic and molecular bases of R-gene mediated resistance 2 , 3 ., However , for some important plant diseases , especially those caused by necrotrophic and broad host range pathogens , R genes of major effect are unknown ., The only available form of resistance to these diseases is Quantitative Disease Resistance ( QDR ) ., QDR is based on complex inheritance , involving numerous genes of small effect 4–6 ., QDR is frequent in crops and natural plant populations , and is of practical importance in agriculture because it is often more durable than R-mediated resistance 7 ., In addition , the identification of genes underlying QDR is expected to provide fundamental insights into the diversity of plant immune responses and prediction of evolutionary trajectories of natural populations ., To date , the molecular bases of QDR have been identified only in few cases and involve remarkably diverse molecular functions 6 , 8 , 9 ., Sclerotinia sclerotiorum is an Ascomycete fungus , causal agent of the white mold and stem rot diseases ., It is considered as a typical necrotrophic pathogen , using secreted proteins and metabolites to rapidly kill host cells and complete its infection cycle 10 , 11 ., S . sclerotiorum is also notorious for its remarkably broad host range , encompassing several hundred Eudicot species in about a hundred botanical families 12 , 13 ., S . sclerotiorum notably infects soybean , tomato and rapeseed on which it causes several hundred million dollar losses annually 14 , 15 ., Besides rapeseed , S . sclerotiorum is also a natural pathogen of other Brassicaceae such as Arabidopsis species 16 ., On A . thaliana natural populations , S . sclerotiorum causes symptoms ranging from high susceptibility to relative tolerance , corresponding to a typical QDR response 17 ., The role of a few A . thaliana genes in resistance to S . sclerotiorum is beginning to emerge ( for a review: 18 ) , notably through association genetics approaches 19 ., Thanks to its reduced linkage disequilibrium and extensive genotyping information , A . thaliana is an excellent model to deploy genome wide association ( GWA ) mapping to identify QDR genes 20 , 21 ., A better understanding of plant QDR genes function and molecular evolution is critical to increase the durability of disease resistance mechanisms used in the field , a major challenge for plant breeding and evolutionary biology ., S . sclerotiorum lineage gained the ability to infect Brassicacea and Solanaceaea plants , among others , after the divergence with S . trifoliorum lineage , about 8 . 2 million years ago 13 ., The divergence between ancestors of the Brassicaceae and Solanaceaea plant families is estimated to about 120 million years ago 22 ., While most NBS-LRR genes show dramatic lineage-specific expansions and contractions with diverse rates of sequence variation 23 , 24 , a few NBS-LRR gene clusters have likely been conserved over 100 million years in core eudicot genomes 25 ., The persistence of function and polymorphism after several million years of divergence has been documented in some orthologous R genes such as Rpm1 in A . thaliana 26 or members of the Mla family in Triticaea 27 ., This pattern is often indicative of balancing selection acting on resistant and susceptible haplotypes 28 ., Similar balancing selection has been identified for the RKS1 gene conferring quantitative disease resistance to the bacterial pathogen Xanthomonas campestris pv ., campestris ( Xcc ) in A . thaliana 21 ., However , our knowledge of the mechanisms underlying the evolution of QDR genes is very limited ., Furthermore , how genes associated with QDR against S . sclerotiorum evolved in the multiple lineages it infects remains largely unknown ., In this work , we reveal the parallel evolution , in distinct plant lineages , of sequence and expression polymorphisms associated with quantitative disease resistance against the fungal pathogen S . sclerotiorum ., Using genome wide association mapping in A . thaliana populations , we associated the prolyl-oligopeptidase ( POP ) gene POQR with quantitative disease resistance against S . sclerotiorum ., The phenotypic analysis of two null mutant lines confirmed that POQR confers partial resistance to S . sclerotiorum ., Next , we highlight the long term co-existence of POQR alleles in A . thaliana and associate high level of disease resistance with specific amino acid substitutions ., Furthermore , similar amino acid substitutions occurred independently in POQR in distinct plant lineages , following independent gene duplications ., Genome wide analysis of the POP family in A . thaliana and S . lycopersicum indicated that the emergence of POQR resulted from parallel, ( i ) gene duplications ,, ( ii ) amino acid substitutions and, ( iii ) gain of gene induction upon S . sclerotiorum challenge in these two species ., We report parallel divergence in gene expression upon S . sclerotiorum infection for POQR and multiple genes that duplicated both in A . thaliana and S . lycopersicum ., Our findings provide one of the very few functions known for prolyl-oligopeptidases in plants and reveal that the molecular evolution of quantitative resistance against generalist pathogens can follow the same trajectory several times independently in distinct lineages ., To identify genetic loci associated with QDR to S . sclerotiorum , we used Genome Wide Association ( GWA ) mapping in A . thaliana natural populations ., For this , we scored disease index six days after leaf inoculation on 84 A . thaliana European accessions ., We used the average disease severity index ( DSI ) from 6 to 16 plants per accession ( S1 File ) ., Observed phenotypes covered most of the range ( Fig 1A ) ., The most resistant accession was Ei2 ( 6915 ) with a DSI of 1 . 99 ± 0 . 39 , while the most susceptible accession was ALL15 ( 4 ) with a DSI of 5 . 92 ± 0 . 14 ., There was no obvious structure in the geographic distribution of this phenotype ( Fig 1B ) ., Next , we performed a genome wide association analysis using a mixed model algorithm on single nucleotide polymorphism data from the A . thaliana 250K chip ( S2 File , S1 Table ) ., This revealed two significant loci at the false-discovery rate ( FDR ) level of q < 1 . 0 . e-07 which corresponds to–log10 ( p-value ) of 5 . 25 ., The highest association value , -log10 ( p-value ) = 6 . 03 , corresponded to a missense SNP located on chromosome 1 ( position 7 061 677 ) within the predicted coding sequence of the gene At1g20380 ( Fig 1C and 1D ) ., At1g20380 encodes an uncharacterized prolyl-oligopeptidase hereby named POQR ( Prolyl-Oligopeptidase associated with Quantitative Resistance ) ., To directly test the role of POQR in quantitative disease resistance against S . sclerotiorum , we analyzed the phenotype of two Col-0 insertion mutant lines ., The poqr-1 line ( SALK_121407C ) and the poqr-2 line ( SALK_027815C ) had a T-DNA insertion in the second and tenth exon of the POQR gene respectively ( S1 Fig ) ., We detected truncated transcripts in healthy plants of both lines , encoding proteins truncated after amino acids 53 ( poqr-1 ) and 681 ( poqr-2 ) , instead of 732 in the wild type protein ., Quantitative RT-PCR analysis showed that POQR was induced ~5 . 2 times upon S . sclerotiorum infection in Col-0 wild type plants but this induction was abolished in poqr mutants ( S1 Fig ) ., The poqr-1 and poqr-2 mutant lines showed no obvious macroscopic developmental and growth defects ( S1 Fig ) ., We used a strain of S . sclerotiorum constitutively expressing the green fluorescent protein to determine the extent to which POQR affects the ability of S . sclerotiorum to colonize A . thaliana leaves ( Fig 2A and 2B ) ., The average area colonized by the fungus 24 hours after inoculation was ~61 mm2 in the Col-0 wild type , ~96 mm2 in the susceptible mutant rlp30-1 19 , ~87 mm2 and ~90 mm2 in the poqr-1 and poqr-2 mutants ( Student’s t test with Benjamini-Hochberg correction p-value = 0 . 04 and 0 . 035 respectively ) ., Therefore , POQR disruption results in a clear increase in leaf colonization by S . sclerotiorum ., In agreement , we measured an increase in disease lesion size on poqr mutant lines compared to wild type 36 hours after inoculation by S . sclerotiorum isolate 1980 ( S1 Fig ) ., In these assays , lesions on poqr1 were intermediate between Col-0 and poqr2 ., Consistent with previous report on natural accessions susceptibility to S . sclerotiorum 17 , Rubenzhnoe showed the smallest and Shadahra the largest lesions in our experiments ., To test whether poqr mutants are altered in a general biotic stress response pathway , we challenged A . thaliana wild type , control and poqr mutant plants with the bacterial pathogen Xanthomonas campestris pv ., campestris ( Fig 2C and 2D ) ., We scored symptoms following the DSI method of 21 at 10 days post inoculation ., Severely diseased plants ( DSI 3 or 4 ) were limited to 15% of Col-0 wild type plants at this time ., As expected , the Kashmir accession and rks1-1 mutant appeared clearly more susceptible than the wild type , with 73% and 50% of plants showing a DSI 3 or 4 respectively 21 ., Severely diseased plants ( DSI 3 or 4 ) represented 21% of poqr-1 plants , and 13% of poqr-2 plants , similar to the wild type ., To broaden our view of the plant pathogens to which POQR respond , we exploited publicly available A . thaliana gene expression data ( S2 Fig ) ., These data suggested that POQR is induced rather specifically during the response to leaf-infecting necrotrophic fungal pathogens such as Botrytis cinerea and Alternaria brassicicola ., These results confirm the identification of a new genetic component of A . thaliana QDR and demonstrate a positive and relatively specific role for POQR in QDR against S . sclerotiorum ., To get insights into the link between POQR natural diversity and its function in QDR , we first analyzed the distribution of DSI in A . thaliana accessions harboring either a cytosine or a thymine at the top associated SNP in our GWA mapping ( position 7061677 on chromosome I ) ( Fig 3A ) ., The median DSI was ~3 . 3 for accessions with a cytosine and ~4 . 8 for accessions with a thymine at this position ( Student’s t test p-value = 1 . 3e-05 ) ., Next , we analyzed the natural diversity of POQR protein sequence in 46 accessions from our European GWA population 29 ., We verified by PCR sequencing the N-terminal sequence of POQR in 8 A . thaliana accessions ( S3 File ) ., We constructed a maximum likelihood phylogenetic tree with POQR sequence from 46 European accessions plus the Col-0 accession , and used A . lyrata POQR sequence ( AL1G33310 ) to root the tree ( Fig 3B , S4 File ) ., This revealed three major well-supported clades , including 9 , 14 and 15 POQR sequences ., We mapped average DSI for each POQR isoform to highlight contrasted DSI values in each clade ., Clade A ( ‘Ancestral’ ) includes 9 accessions with median DSI of 4 . 6 , clade R ( ‘Resistant’ ) includes 14 accessions with median DSI of 3 . 2 and clade S ( ‘Susceptible’ ) includes 15 accessions with median DSI of 4 . 5 ., The divergence of clades R and S associates with significant differences in DSI ( Student’s t test p-value = 2 . 9e-03 ) and involved the substitution of POQR serine 5 into a proline ( S5P polymorphism ) , which corresponds to the SNP with the highest association value in our GWA analysis ( Fig 3C ) ., The S5P polymorphism showed a discontinuous phylogenetic distribution ., Accessions containing the S5P polymorphism are more resistant than sister lineages ., The S5P polymorphism is notably common to all accessions in clade R , which includes 8 of the 10 most resistant accessions included in this analysis ., To explore POQR diversity across plants , we searched for POQR homologs in the complete genome of 40 plant species and constructed a maximum likelihood phylogenetic tree using Volvox carteri POQR homolog as a root ( Fig 4A , S5 File ) ., We retrieved a total of 75 POQR homologs , with 11 species harboring a single POQR homolog , 24 species harboring two POQR homologs , 4 species with three POQR homologs and one species with four POQR homologs ( S5 File ) ., The most parsimonious tree defined six well delimited clades corresponding to, ( i ) Bryophyta ,, ( ii ) Brassicales ,, ( iii ) Solanales ,, ( iv ) Fabales ,, ( v ) other Eudicots and, ( vi ) Poales , suggesting that for genomes including multiple POQR paralogs , duplications are posterior to the divergence of the corresponding plant orders ., The Brassicales , Solanales , Fabales and Poales clades showed clear duplication patterns in all species analyzed ., This indicates parallel duplication of POQR ancestral gene early in the evolution of Brassicales , Solanales Fabales , and Poales , consistent with paleo-polyploidization by whole genome duplications in several angiosperm lineages between ~75 and 25 million years ago ( Mya ) 30 ., Polymorphism at position 5 of the POQR protein was associated with QDR against S . sclerotiorum in A . thaliana ., We highlighted the evolution of position 5 in the evolution of POQR across plants ( Fig 4A ) ., Among Bryophyta and Dicotyledones , 77% of POQR proteins showed a Serine at position five , while in Monocotyledones , all but one POQR proteins had an Alanine at position five ., With the exception of sequences from P . patens that had Glycines at position five , all 10 remaining POQR proteins had a Proline at position five ., The genome of Sphagnum fallax , A . thaliana , Salix purpurea and 6 Solanales species , including Solanum lycopersicum , all contain POQR homologs with a Serine at position five and others with a Proline at position five ( Fig 4B ) ., Similarly , the genome of Oryza sativa includes one POQR homolog with an Alanine at position five ( OsPOP1 ) and an ortholog with a Proline at position five ( OsPOP9 ) ., In A . thaliana Col-0 genome , At1g76140 is AtPOQR closest paralog ( 81 . 4% identity at the protein level ) and harbors a Serine at position five , instead of a Proline in AtPOQR ., In S . lycopersicum Heinz genome , Solyc08g022070 is the closest paralog of SlPOQR ( Solyc04g082120 , 78 . 4% identity at the protein level ) and harbors a Serine at position five , instead of a Proline in SlPOQR ., To study the evolutionary history of POQR sequence , we performed ancestral sequence reconstruction with FastML ( S6 File ) ., We compared POQR homologs from Brassicaceae , Solanaceae and Poales with their respective ancestral sequences prior gene duplication in these lineages ., This revealed the occurrence of S5P and F613Y substitutions in parallel in AtPOQR , OsPOP9 and in all seven Solanaceae species analyzed ( p-value of these substitutions occurring in three independent lineages is 7 . 5e-09 ) ., We noted two additional amino acid substitutions that occurred in parallel in AtPOQR and SlPOQR ( V119I and A533S ) that could have contributed to POQR sequence adaptation to function in QDR ., To support a role for POQR in QDR against S . sclerotiorum in tomato , we used a virus-induced gene silencing ( VIGS ) approach to silence POQR in tomato ( Fig 4C ) ., Eighteen days after delivery of the viral constructs , plants were inoculated with a GFP-expressing S . sclerotiorum strain ., Plants carrying the POQR-silencing construct showed an average ~57% reduction in POQR expression during S . sclerotiorum infection as compared to wild type plants and plants carrying the empty viral construct ( S3 Fig ) ., The area colonized by the fungus 24 hours after inoculation was measured using fluorescence imaging ., Consistent with results obtained in A . thaliana poqr mutant lines , we found that POQR silencing in tomato resulted in colonized areas ~1 . 42 times larger than in wild type plants and ~1 . 36 times larger than plants carrying the empty viral construct ., These results indicate parallel duplication and convergent sequence evolution of POQR homologs in multiple plant lineages ., The A . thaliana Prolyl-olypeptidase ( POP ) family includes 11 members harboring a peptidase S9 catalytic domain ( PF00326 ) ( S4 Fig , S7 File ) ., Two AtPOPs have been studied functionally: At5g20520 encodes WAV2 , a negative regulator of skewing root patterns 31 , and At4g14570 encodes an acyl-amino acid-releasing enzyme ( AARE ) mostly present in chloroplasts 32 ., Five AtPOPs , including POQR , also harbor a Prolyl-oligopeptidase N-terminal beta-propeller domain ( PF02897 ) which functions in limiting the access of the enzyme catalytic triad to small ( <30 aa ) unstructured peptides 33 , 34 ( S4 Fig ) ., In tomato , the POP family includes 13 members , four of which harbor the N-terminal beta-propeller domain ( S4 Fig ) ., In phylogenies including A . thaliana and S . lycopersicum POPs , At1g76140 and AtPOQR on the one hand , and Solyc08g022070 and Solyc04g082120 ( SlPOQR ) on the other hand , formed well resolved pairs , supporting the conclusion that they duplicated after the divergence of the two plant species ( Fig 5A , S7 File ) ., To further support convergent evolution of POQR homologs in A . thaliana and S . lycopersicum , we first analyzed global gene expression in response to S . sclerotiorum in A . thaliana and S . lycopersicum ( S8 File ) ., We detected 4 , 703 genes ( 16 . 4% ) significantly upregulated ( log fold change LFC>1 . 5 , adjusted p-value<0 . 01 ) and 5 , 813 genes ( 20 . 3% ) significantly down-regulated in A . thaliana ., In S . lycopersicum , 3 , 513 genes ( 10% ) were upregulated and 4 , 556 ( 12 . 9% ) were down-regulated ., We analyzed gene annotations from both species and found 50 gene ontology ( GO ) terms significantly enriched among up-regulated genes including defense-related mechanisms , detoxification processes and secondary metabolism ( S9 File ) ., We found 136 GO terms significantly enriched among down-regulated genes mainly related to photosynthesis ( S9 File ) ., We then focused on the expression of POP genes in response to S . sclerotiorum infection at the genome scale in A . thaliana and S . lycopersicum ( Fig 5B , S8 File ) ., In A . thaliana , four AtPOP genes were significantly ( adjusted p-value<0 . 01 ) down-regulated upon S . sclerotiorum inoculation , while three genes were significantly up-regulated ., In S . lycopersicum , six SlPOP genes were significantly down-regulated upon S . sclerotiorum inoculation , while SlPOQR was the only POP gene significantly up-regulated ., Among a total of 24 POP genes in the two species , AtPOQR and SlPOQR were the only ones induced more than 3 fold ( LFC>1 . 5 ) upon S . sclerotiorum inoculation ., Both AtPOQR and SlPOQR were significantly more induced upon S . sclerotiorum inoculation than their duplicate ( delta LFC>2 ) ., These results show that , after divergence of the two species , POQR ancestral gene duplicated in parallel in A . thaliana and S . lycopersicum , with one paralog ( POQR ) evolving responsiveness to S . sclerotiorum infection in each species ., To document divergence in duplicated genes expression upon S . sclerotiorum infection at the whole genome scale , we first analyzed the expression of 1843 A . thaliana duplicated gene pairs 35 ., For this , we calculated LFC of gene expression in S . sclerotiorum-infected plants compared to mock-treated plants , and the difference between LFC for two genes forming a duplicated pair ( delta LFC ) ( Fig 5C , S10 File ) ., Median delta LFC was 1 . 77 , corresponding to a ~3 . 4 fold change in gene expression responsiveness to S . sclerotiorum infection ., A total of 852 duplicated gene pairs ( 46% ) showed delta LFC≥2 , supporting the view that divergence of gene expression is relatively widespread between duplicated genes in A . thaliana 35 ., We next used Markov Cluster algorithm 36 on A . thaliana and S . lycopersicum predicted proteomes to identify 252 clusters of genes that duplicated in parallel in A . thaliana and S . lycopersicum ( a total of 504 A . thaliana genes and 504 S . lycopersicum genes ) ., This identified 42 gene clusters ( 19% ) with a delta LFC≥2 was observed after fungal infection in both A . thaliana and S . lycopersicum ( Fig 5C ) ., Among those , 24 clusters included paralogs induced at least 4 fold both in A . thaliana and S . lycopersicum ( Fig 5D , S11 File ) ., The corresponding genes are notably involved in cell redox homeostasis ( e . g . the uncoupling protein cluster UCP ) , in primary and secondary metabolism ( e . g . the adenylate kinase cluster ADK ) , transport ( e . g . the phosphate transporter 4 cluster PHT4 ) , signaling ( e . g . protein kinases ) , and response to abscisic acid ( e . g . the ARM repeat protein interacting with ABF2 cluster ARIA ) ., These genes are prime candidates for contributing to quantitative disease resistance mechanisms shared between plant species ., Plant QDR is a complex trait governed by multiple genes of minor effect , which renders the identification of the underlying molecular bases challenging 4 , 5 ., Whereas gene-for-gene resistance relies on receptor proteins that belong to the receptor-like kinase ( RLK ) or the Nod-like receptor ( NLR ) classes family , the limited number of QDR determinants known to date encodes a remarkably broad range of molecular functions , such as transporters 8 , 37 , 38 , kinase domain-containing proteins 9 , 21 or enzymes of the central metabolism 39 ., Gene variants conferring resistance to fungal pathogens in natural plant populations also remain poorly documented 40 ., We report the role of a prolyl oligopeptidase ( POP ) conferring QDR to S . sclerotiorum in A . thaliana ., Several classes of proteases are involved in plant defenses against fungal and bacterial pathogens 41–43 but little is known about POP functions in plants ., In animals , unlike other serine proteases , POP cleaves short peptides ( usually <30 amino acids ) after a proline residue 33 ., In human , this enzyme is associated with several neurological disorders , possibly through its action in the metabolism of neuropeptides or in the inositol phosphate signaling pathway 44 ., In Basidiomycota fungi , POPs are required for the maturation of amatoxins , toxic cyclic peptides derived from ~30 amino acid propeptides 45 , 46 ., Plants in the Caryophyllaceae family use a POP enzyme designated as peptide cyclase 1 ( PCY1 ) to synthesize cyclic peptides from propeptide precursors 47 ., Plant cyclic peptides are diverse and widespread , some exhibiting antifungal activity 48 ., POQR may therefore be involved in the maturation of plant antifungal peptides ., Alternatively , POQR may act directly on fungal secreted proteins promoting disease 10 to disable them ., Our inoculation assays with the bacterial pathogen Xanthomonas campestris pv ., campestris and a survey of publicly available A . thaliana gene expression data were in agreement with a possible activity of POQR on specific fungal processes ., Future work aiming at determining the spectrum of pathogens impacted by POQR-mediated resistance should provide valuable insights into POQR molecular function ., Our analyses associated the substitution of POQR serine in position 5 by a proline with enhanced QDR to S . sclerotiorum in A . thaliana ., Changes at POQR N-terminus induced by the S5P substitution may affect the ability of POQR to form dimers , a property of some Archean POPs 49 , the subcellular localization of POQR , or remotely affect the accessibility and function of POQR catalytic site ., Additional amino acid substitutions occurred in parallel in POQR homologs from multiple species , notably the F613Y substitution that arose at least three times independently in A . thaliana , rice and Solanaceae species ., This variant may create a novel tyrosine phosphorylation site important for POQR regulation , or as shown for the Y473F mutation in porcine POP 50 , modulate the range of conditions in which the POQR enzyme is active ., The analysis of POQR expression in two distantly-related plants species suggested that POQR induction was associated with resistance to S . sclerotiorum ., This observation is consistent with resistance increasing with the level of POQR accumulation ., Similarly , enhanced accumulation of a tomato chitinase was associated with resistance to Alternaria solani 51 , and expression of the atypical kinase gene RKS1 was correlated with QDR to X . campestris pv ., campestris 21 ., Besides , copy number variation and DNA methylation are polymorphisms altering the accumulation of gene products encoded by the Rhg1 locus , which confers QDR to soybean cyst nematodes 38 , 52 ., A detailed understanding of POQR molecular function will be required to establish causal relationships between polymorphism at the POQR locus and QDR to S . sclerotiorum ., Our current data points towards POQR transcriptional regulation and the modulation of POQR enzymatic activity as determinants of QDR ., Ancient whole genome duplication ( WGD ) events have had a major role in shaping the gene content of extant plant genomes , and contributed to the evolution of a range of new gene functions 53 , 54 ., Based on our analyses , a scenario for the convergent evolution of POQR and its recruitment in plant QDR can be inferred ( Fig 6 ) ., The Brassicales and Solanales lineages probably inherited a single POQR ancestral sequence at the time these two lineages diverged , about 120 million years ago ( Mya ) 22 , 55 ., Two WGD events were detected in the Brassicales ( At-α and At-β ) estimated respectively to ~40 and ~88 Mya 54 , 56 ., A WGD occurred in the Solanales ( Sl-T ) and is estimated at least ~64 Mya 55 , 57 ., The duplication of POQR ancestor in A . thaliana and S . lycopersicum lineages therefore occurred at least 40 million years ago ., The gene content of extant plant genomes is the result of frequent loss of duplicates created by WGDs 53 ., The maintenance rate of At-α duplicates has been estimated to ~14% in average , with only ~6 . 5% of defense-related gene duplicates being maintained 35 ., The maintenance of POQR duplicates over >40 million years in multiple plant species supports an important role for POQR in plants fitness ., This is in agreement with POQR locus explaining ~20% of phenotypic variation upon S . sclerotiorum infection in the A . thaliana population analyzed in this work ., After duplication , POQR ancestral genes underwent a number of parallel amino acid substitutions in the Brassicales and the Solanales , including the S5P and F613Y substitutions ( Fig 6 ) ., These substitutions are present in all POQR homologs from Solanales , and therefore likely occurred in the most recent common ancestor , at least ~30 million years ago ( estimated divergence time for Petunia genus , 58 ) ., By contrast , in the Brassicales , POQR S5P and F613Y substitutions were only found in a subset of A . thaliana accessions , placing their probable emergence after A . thaliana divergence ~6 million years ago 59 ., Remarkably , POQR alleles with amino acids S5 and F613 persisted in a significant number of A . thaliana accessions ., The co-selection of two gene variants within a population , such as observed for POQR , often results from complex evolutionary constrains referred to as balancing selection 28 ., Our results suggest that plants adapted to high fungal disease pressure by tuning POQR enzymatic activity and evolving gene induction upon fungal infection ., We observed similar molecular mechanisms for POQR evolution in A . thaliana at the infra-specific level and following duplication in the Solanales ancestor , between 30 and 64 million years ago ., The estimated emergence of POQR S5P and F613Y polymorphisms in Solanales predates the inclusion of these species in S . sclerotiorum host range 13 , suggesting that POQR evolution in this lineage was driven by the interaction with other fungal pathogens , or other environmental constraints ., Pigmentation in mammals is a well known example of phenotypic convergence at different levels , including convergence at the level of mutation ( mutational convergence ) 60 ., Indeed , the function of the Mc1r gene product was evolved similarly in beach mouse and woolly mammoth through the same R65C mutation 61 , 62 ., In insects , the ability to feed on plants producing cardenolide toxic compounds results from parallel amino acid substitutions in the ATPα1 subunit of a Na+ and K+ transporter , independent duplications and convergent expression polymorphism of the corresponding gene 63 ., Similarly our work associate parallel mutations , gene duplication and convergent expression polymorphism in POQR ., Epistasis is thought to reduce the rate of molecular convergence with species divergence 64 ., Nevertheless , mutational convergence at the intra- and inter-specific level has been documented recently for color vision in stickleback fishes , as an adaptation to blackwater 65 ., Our study provides another example of mutational convergence at the intra- and inter-specific level ., The co-existence of divergent alleles within plant populations has often been associated with fitness trade-offs , notably between growth and defense 66 ., Quantitative disease resistance has been proposed to rely in part on genes controlling plant growth in the absence of pathogens 4 ., Such a function seems unlikely for POQR , considering that A . thaliana poqr mutant lines did not show obvious developmental defects ., Instead , POQR may have specialized to function in QDR through sequence and expression polymorphisms while its duplicate copy maintained function in the control of plant growth ., Partial functional redundancy between POQR and its duplicate could explain the persistence of multiple variants in A . thaliana populations and across plant species 67 ., Such pleiotropic constraints have been proposed to facilitate convergent and parallel molecular evolution 64 ., In agreement , we found that 19% of genes duplicated both in A . thaliana and S . lycopersicum evolved responsiveness to S . sclerotiorum in a convergent manner ., Insights into the possible trade-offs mediated by these genes will be required to test for the interaction between pleiotropy and molecular convergence ., In summary , our work has shown that the evolution of genes mediating plant quantitative disease resistance displays some level of repeatability and predictability ., This finding has implication for our understanding of the evolution of quantitative traits in plants and for the design of innovative and sustainable crop breeding strategies ., Adaptive evolution of POQR involved a combination of gene copy number variation , sequence and expression polymorphism ., An important question for the future will be to assess the impact of convergent mutations on POQR function in several plant species and test for the role of putative fitness trade-offs in constraining the evolution of QDR genes ., Arabidopsis thaliana accessions and mutant plants were obtained from the European Arabidopsis Stock Centre ( NASC ) , ecotype ids for natural accessions used in this study are listed in S1 File ., All Arabidopsis mutant lines used in this study were in the Col-0 background , the Nottingham Arabidopsis Stock Centre accession number N1093 Col-0 was used as wild type ., Plants were grown in Jiffy pots under controlled conditions at 22°C , with a 9 hour light period and a light intensity of 120 μmol/m2/s 4 weeks prior to infection ., Homozygous T-DNA insertion mutant plants poqr-1 ( SALK_121407C ) and poqr-2 ( SALK_027815C ) were isolated by PCR screening with the primers FWD_poqr1 ( 5’-ATGTTGGTTGAGTTGACGGAG-3’ ) and REV_poqr1 ( 5’-TTGATCAGTCCCAAGGAAATG-3’ ) or FWD_poqr2 ( 5’-GATCTCTT | Introduction, Results, Discussion, Materials and methods | Plant pathogens with a broad host range are able to infect plant lineages that diverged over 100 million years ago ., They exert similar and recurring constraints on the evolution of unrelated plant populations ., Plants generally respond with quantitative disease resistance ( QDR ) , a form of immunity relying on complex genetic determinants ., In most cases , the molecular determinants of QDR and how they evolve is unknown ., Here we identify in Arabidopsis thaliana a gene mediating QDR against Sclerotinia sclerotiorum , agent of the white mold disease , and provide evidence of its convergent evolution in multiple plant species ., Using genome wide association mapping in A . thaliana , we associated the gene encoding the POQR prolyl-oligopeptidase with QDR against S . sclerotiorum ., Loss of this gene compromised QDR against S . sclerotiorum but not against a bacterial pathogen ., Natural diversity analysis associated POQR sequence with QDR ., Remarkably , the same amino acid changes occurred after independent duplications of POQR in ancestors of multiple plant species , including A . thaliana and tomato ., Genome-scale expression analyses revealed that parallel divergence in gene expression upon S . sclerotiorum infection is a frequent pattern in genes , such as POQR , that duplicated both in A . thaliana and tomato ., Our study identifies a previously uncharacterized gene mediating QDR against S . sclerotiorum ., It shows that some QDR determinants are conserved in distantly related plants and have emerged through the repeated use of similar genetic polymorphisms at different evolutionary time scales . | Plant disease resistance is mediated by molecular components depending on pathogens infection strategy ., Sclerotinia sclerotiorum is a devastating plant pathogenic fungus notorious for its ability to infect a wide variety of plant species by rapidly triggering cell death ., Plant exhibit a response designated as quantitative disease resistance ( QDR ) when challenged by S . sclerotiorum , the molecular bases of which are largely unknown ., Here we used genome wide association mapping a natural population of Arabidopsis thaliana host to identify the POQR prolyl oligopeptidase gene involved in QDR against S . sclerotiorum , and validate its function ., We associate specific amino-acid changes in POQR sequence to increased QDR in A . thaliana accessions ., Remarkably , the same changes occurred in multiple plant lineages ., We demonstrate the parallel evolution of POQR in A . thaliana ( Brassicaceae ) and tomato ( Solanaceae ) , showing that this gene evolved through duplication followed by similar sequence and expression divergence in these two distant plant lineages ., These findings expand on the diversity of molecular functions involved in QDR and suggest that the evolution of QDR against broad host range pathogens is repeatable to some extent . | biotechnology, taxonomy, brassica, phylogenetics, plant science, model organisms, data management, phylogenetic analysis, experimental organism systems, plant genomics, plant pathology, convergent evolution, plants, research and analysis methods, arabidopsis thaliana, computer and information sciences, gene expression, plant genetics, evolutionary systematics, evolutionary genetics, eukaryota, plant and algal models, genetics, biology and life sciences, genomics, evolutionary biology, plant biotechnology, evolutionary processes, organisms | null |
journal.pbio.1001937 | 2,014 | In Vitro Generation of Neuromesodermal Progenitors Reveals Distinct Roles for Wnt Signalling in the Specification of Spinal Cord and Paraxial Mesoderm Identity | The differentiation of embryonic stem cells ( ESCs ) to specific cell types offers insight into developmental mechanisms and has potential therapeutic applications ., For example the differentiation of neural progenitors ( NPCs ) from monolayers of ESCs seeded in serum free conditions is a model of neural induction and regional patterning 1 ., In the absence of additional signals , NPCs differentiated from ESCs adopt an anterior-dorsal neural ( telencephalon ) identity 1 , 2 ., The addition of Sonic Hedgehog ( Shh ) ventralises these neural progenitors , mimicking the in vivo role of Shh 3 , 4 ., Exposing NPCs to retinoic acid ( RA ) results in the repression of anterior identity and the induction of genes that typify hindbrain and anterior spinal cord ( cervical ) identity 5 ., This has been taken as support for the idea that newly generated NPCs are by default anterior and are then posteriorised by exposure to specific extrinsic signals 6 , 7 ., It is notable , however , that RA is actively excluded in the progenitors of the posterior spinal cord after gastrulation 8 and that commonly used ESC differentiation protocols do not efficiently generate neural cells of the more posterior spinal cord such as thoracic and lumbar spinal cord cells marked by posterior Hox gene expression , including Hoxc8–10 expression 9 ., The anterior and posterior nervous system has distinct origins 10–12 ., Anterior epiblast expresses Otx2 and contributes cells to the anterior nervous system 2 , 13 whereas spinal cord progenitors are located posteriorly 14–16 ., Clonal analysis indicates that the spinal cord shares a common lineage , at least in part , with the trunk paraxial mesoderm that forms the somites 15 ., The dual-fated neuromesodermal precursors ( NMPs ) of these tissues are located in the node-streak border ( NSB ) , caudal lateral epiblast ( CLE ) cell layer adjacent to the regressing node and the chordoneural hinge of the tail bud 13 , 14 , 17 , 18 ., Cells in these regions coexpress the neural marker Sox2 and nascent mesoderm marker Brachyury 8 , 19 , 20 ., Genetic lineage tracing experiments confirm that many spinal cord cells previously expressed Brachyury 21 indicating that as cells from regions harbouring NMPs move into the neural tube they downregulate Brachyury but maintain Sox2 expression and consolidate neural identity ., By contrast , NMPs that enter the primitive streak delaminate basally , downregulate Sox2 and acquire expression of the paraxial mesoderm marker Tbx6 22 en route to somite formation ., Strikingly , in embryos lacking Tbx6 , paraxial mesoderm cells express Sox2 and transdifferentiate into neural cells , providing additional support for the inter-relationship between spinal cord and somitic mesoderm 22–24 ., As yet , however , the existence of NMPs has only been revealed in vivo and the inaccessibility of this population makes them difficult to study ., The region occupied by NMPs is exposed to Wnt and Fgf ligands 16 ., These signals are required for body axis elongation 16 and both Wnt and Fgf signalling have been implicated in mesoderm and neural induction 22 , 25–31 ., In vivo and in vitro evidence has suggested that Wnt signalling is responsible for posteriorising tissue by inducing posterior Hox genes 29 , 32 , 33 ., Together , the data suggest that the generation of posterior neural tissue and paraxial mesoderm proceeds by Wnt and Fgf signalling inducing a neuromesodermal bipotential intermediate ., To test this idea , we developed an efficient in vitro differentiation method for spinal cord and paraxial mesoderm from mouse and human pluripotent stem cells ., We show that carefully timed and calibrated pulses of Wnt and Fgf signalling generate a population of cells that transiently coexpress Sox2 and Brachyury in which the expression of posterior Hox genes are induced ., Transcriptome analysis is consistent with the equivalence of these cells to the NMPs found in vivo ., In vivo grafting and directed in vitro differentiation confirm the ability of NMPs to assume spinal cord or paraxial mesoderm cell fates ., We further show that Brachyury is not required for the production of posterior neural cells or for the induction of posterior Hox genes , hence separating the posteriorising and mesoderm inducing functions of Wnt signalling ., Taken together the data define a means to generate posterior neural and paraxial mesodermal tissues in vitro and illustrate how the directed differentiation of stem cells provides novel insight into developmental mechanism ., To identify conditions for the generation of posterior neural cells from monolayers of mouse ES cells ( mESCs ) , we cultured mESCs in serum free media containing bFgf for 3 days ( D1–D3 ) and then transferred these to media lacking bFgf for an additional 2 days 1 ( Figure 1A ) ., This resulted in the induction of a post-implantation epiblast-like intermediate by D2 , indicated by the downregulation of the “naïve” pluripotency marker Zfp42 ( Rex1 ) and the upregulation of the epiblast marker Fgf5 ( Figure 1F ) 34 ., At this stage , Pou5f1 , which is expressed in both mESCs and epiblast-like cells , is maintained ( Figure 1F ) 34 ., In all experiments a Shh agonist , SAG , was added at D3 in order to generate a predictable ventralised identity for subsequent comparisons ., The transcriptome of cells was then analysed at D5 by mRNA-seq ., Consistent with previous studies 35–39 , cells in these conditions had acquired an anterior neural identity ( NA ) , exemplified by the expression of Otx1 and Otx2 2 ., The presence of SAG induced the expression of ventral neural markers ( Figure S1A ) ., Addition of retinoic acid ( RA ) and SAG to differentiating mESCs at D3 downregulated anterior neural markers ( e . g . Otx2 , Six3 , Lhx5 ) and instead genes typical of hindbrain identity , including Hoxa2 , Hoxb2 , Mafb , Epha4 and Ephb2 were expressed ( Figure 1B ) 40 ., However markers of spinal regions of the neural tube , such as the 5′ Hox genes Hoxc6 , Hoxc8 and Hoxc9 were not detected ( Figure 1B ) 5 , 9 ., Changing the timing or concentration of RA used in these experiments did not result in the efficient induction of more posterior spinal cord identity 29 ., To recapitulate the sequence of signalling events that generate the spinal cord , we seeded mESCs into serum free media containing bFgf ., At D2 , Wnt signalling was induced by the addition of the Wnt agonist CHIR99021 ( CHIR ) ., bFgf and Wnt agonist were removed at D3 and cells exposed to media containing RA and SAG until D5 ., Examination of gene expression profiles indicated that cells subjected to the FGF/CHIR/RA regime expressed genes characteristic of the spinal cord including high levels of 5′ Hox genes Hoxb6 , Hoxb8 , Hoxc6 , Hoxc8 , Hoxc9 and low levels of the anterior neural and brainstem markers Otx2 and Mafb ( Figure 1B ) ., Together , the data suggested that a brief pulse of Wnt signalling between D2–D3 was sufficient to posteriorise differentiating mESCs ., We termed the neural cells generated in this regime NP cells and cells that display anterior and brainstem identity NA and NH , respectively ( Figure 1A ) ., We confirmed the posteriorisation and neural identity of Np cells using qRT-PCR and immunostaining ( Figure S1C–D ) ., Analysis of the time course of Hox gene expression in NH and NP cells indicated that their temporal sequence of induction matched the in vivo time course 40: Hoxb1 was induced within 12 h of exposure to Wnt signalling , whereas more 5′ Hox genes were induced later ( Figure 1C ) ., Notably the more posterior Hox genes , e . g , Hoxc6 and Hoxc9 were not induced in NH cells ., In Np cells Hoxc6 , Hoxc8 and Hoxc9 were strongly induced at D4 ( Figure 1C ) ., Delaying the addition of CHIR to differentiating mESCs until D3 resulted in a concomitant shift in the timing of Hox gene induction ( Figure S2A–C ) ., Furthermore , in agreement with studies indicating that RA represses the most posterior Hox genes 41 , exposure of cells to FGF/CHIR without subsequent addition of RA induced Hoxc10 characteristic of the lumbar spinal cord ( Figure S2E ) ., Finally we passaged NH and NP cells at D5 and allowed them to differentiate until D8 , at which point we assayed the expression of genes expressed in motor neurons ( MNs ) ., Both NH and NP cells adopted a neuronal morphology and expressed the neuronal marker class III β-tubulin ( Tuj1 ) ., The NH cells acquired a posterior hindbrain MN identity evident by the coexpression of Hoxb4 and the cranial motor neuron marker Phox2b 42 ( Figure 1D ) ., In the case of NP cells however , only a few Hoxb4 expressing cells were detected ( Figure S1D ) and most of the β-tubulin expressing neurons acquired a Hoxc6 and Hoxc9 identity characteristic of neurons of the brachial and thoracic spinal cord , respectively 43 ( Figure 1E , G ) ., Moreover NP cells expressed Olig2 , a marker of somatic motor neuron progenitors , as well as the differentiated MN markers Hlxb9 and Islet1/2 3 ( Figure 1E ) ., Taken together these data indicate that similar to the situation in vivo 44 and in embryoid bodies 29 exposure of monolayers of differentiating ESCs to a combination of Wnt , Fgf and RA signalling generates spinal cord cells ., To address how the combination of Wnt and Fgf signalling induces spinal cord identity we examined gene expression in differentiating ESCs at D2 . 5 and D3 ( Figure 2A ) ., ESCs that had been exposed to Fgf/Wnt signalling for 12 h ( D2 . 5 ) and 24 h ( D3 ) induced the expression of Cdx2 and the mesoderm transcription factors Brachyury and Tbx6 ( Figure 2B ) ., Recombinant Wnt3a protein had a similar activity to CHIR in these assays ( Figure S2F ) ., By contrast , ESCs cultured in the absence of Wnt agonist , expressed significantly lower levels of these genes ( Figure 2B ) ., These data suggest that Wnt signalling , in combination with Fgf , is initiating a mesodermal transcriptional program ., This is consistent with the loss of mesoderm in mouse embryos lacking Wnt3a 45 and the induction of Brachyury by β-catenin 28 ., It was also noticeable that the levels of Sox2 mRNA were transiently reduced in D2 . 5 and D3 NP cells treated with FGF/CHIR ( Figure 2B ) ., We therefore assayed Sox2 and Brachyury proteins by immunostaining in D3 NP and NA cells ., Strikingly , the level of Sox2 protein was similar in NP and NA cells , consistent with the long half-life of Sox2 protein 46 ., Moreover , ∼80% of NP cells coexpressed Brachyury and Sox2 ( Figure 2C ) whereas only a small number of NA cells expressed Brachyury ., These data suggest that the exposure to bFgf and Wnt signalling induces a cell identity reminiscent of the dual-fated neuromesodermal progenitors present during axial elongation in the CLE 15 , 44 ( Figure S4E ) ., If D2–D3 NP cells represent NMPs , they should form mesoderm ., To test this , we transferred cells at D3 into media containing Wnt agonist but lacking bFgf ., In these conditions ( termed Meso ) the expression of Sox1 , Sox2 and Brachyury were downregulated and several genes characteristic of paraxial mesoderm , including Tbx6 and Msgn1 47 , were significantly upregulated ( Figure 2D ) ., Immunostaining revealed that more than 90% of cells in this condition expressed Tbx6 protein at D5 ( Figure 2E ) ., By D8 Desmin , the intermediate filament protein of muscle sarcomeres 48 and the muscle transcription factor MyoD were highly expressed ( Figure 2E ) ., Thus the continued exposure of cells to Wnt signalling induces a paraxial mesodermal identity that differentiates to a muscle-like identity ., This provides evidence that ESCs exposed to Wnt and bFgf at D2–D3 represent bipotential neuromesodermal cells that can differentiate into either mesoderm or neural tissue ., We next tested the in vivo potential of NMP cells ., For these experiments we took advantage of the chick ., Cells with NMP-like behaviour have been identified in chick 16 and chick embryos provide an accessible and experimentally tractable vertebrate host for grafts of mouse ESCs 49 ., We grafted small groups of DiI labelled D3 NA cells , not exposed to Wnt signalling , or D3 NMPs , exposed to Fgf/Wnt signalling for 24 h , into the caudal lateral epiblast of Hamburger-Hamilton ( HH ) stage 8–9 chick embryos ( Figure 2F ) ., Analysis of embryos 24 h later revealed efficient incorporation and migration of the NMP cells to both the neural tube and the somites ( Figure 2G ) ., Transplanted cells from a single graft contributed to multiple anterior-posterior levels and most embryos showed contribution to both spinal cord and somites ( Figure 2G–J ) ., In several embryos grafted cells were also observed in the tail bud of the embryo as well as the neural tube and somites ( Figure 2J ) ., Contribution to endoderm was not observed ., By contrast , transplanted NA cells showed somewhat lower rates of engraftment and contributed only to the neural tube and not to somites ( Figure 2K–L ) ., These data confirm the bipotency of the in vitro derived NMP cells and demonstrate that similar to in vivo NMPs 15 44 they contribute to both neural and paraxial mesoderm lineage ., Single cell and clonal analysis , in vivo and in vitro , will be necessary to test the potency of individual cells and to understand the molecular mechanism by which neural and/or mesodermal progeny are generated from NMP ., We took advantage of the in vitro differentiation to analyse the transcriptional programmes that generate each of the neural and mesodermal lineages ( Figure 3A ) ., Principal component analysis of the transcriptomes indicated that each differentiation pathway could be clearly distinguished ( Figure 3B ) ., Strikingly , the first principal component ( PC ) appeared to represent developmental time and the second PC the tissue identity of the differentiated cells ., The data revealed a set of genes that distinguished NP , NA , NH cells and Meso cells ( Figure S3 and Tables S2 , S3 ) ., These included the upregulation of Mafb and Phox2b in NH samples and the upregulation of posterior Hox genes , notably Hoxc6 , Hoxc8 and Hoxc9 in NP samples ., By contrast , the induction of genes such as Tbx6 , Hes7 and Hoxc8 and Hoxc9 in D5 cells subjected to mesodermal conditions confirmed the posterior paraxial identity of these cells ., Moreover , the analysis indicated a bifurcation in the transcriptional programmes that generate anterior neural and brainstem cells from those that produce posterior neural and paraxial mesodermal cells ., It was notable that gene expression typical of paraxial mesoderm was evident at D4 of NP differentiation suggesting a gradual separation of neural and mesodermal identity ., Together these data provide a molecular correlate to the distinct cellular origins of anterior and posterior neural tissue 15 and identifies the NMP state as the branch point in the developmental trajectories ., We identified genes upregulated in NMPs compared to mESCs at D1 and NA cells at D3 ., Comparing these to genes induced in D5 neural and mesodermal cells revealed a large intersection ., Thus , in part , NMPs have a transcriptional programme that is a combination of neural and mesodermal gene expression ., In addition however , a set of ∼240 genes appeared uniquely upregulated in NMP cells ( Table S1 ) ., These included the transcription factors Brachyury , Nkx1 . 2 ( also known as Sax1 ) , which is expressed in the stem zone of midgestation embryos 50 , 51 Mixl1 52 , Wnt3a and Cdx2 which are expressed in the primitive streak and nascent mesoderm 32 , 33 ., In addition Follistatin , which plays a key role in neural induction by blocking TGFβ signalling 53 and components of the Fgf signalling pathway , which is implicated in mesoderm induction 16 , are upregulated in NMPs ( Figure 3C ) ., Together these data support the idea that exposure of differentiating ESCs to Fgf/Wnt signalling between D2 and D3 induces a bipotential neuromesodermal population equivalent to that found in vivo in the CLE 15 , 16 , 18 , 22 and that the balance and timing of these two signals influences the further differentiation of these cells into neural or mesodermal tissues ., The activation of Wnt signalling in differentiating mouse epiblast stem cells ( EpiSCs ) leads to a modest induction of Brachyury/Sox2 coexpressing cells , suggestive of NMP identity 19 ., To improve the efficiency of this induction we adapted our mESC protocol to take account of the more advanced developmental state of EpiSC compared to mESCs ( Figure 1F ) ., Accordingly , we exposed EpiSCs to a range of CHIR ( Wnt ) and bFgf concentrations and assayed the expression of Sox2 and Brachyury ( Figure S4A ) ., Maximal proportions of Sox2/Brachyury coexpressing cells resulted from 3 µM CHIR and 20 ng/ml bFgf ( hereafter referred to as FGF/CHIR ) ( Figure 4A , Figure S4A ) ., Assaying a broader panel of genes supported the idea that FGF/CHIR was inducing NMP identity ., The expression of the pluripotency factor Nanog was undetectable and the majority of the Sox2 expressing cells expressed minimal levels of Oct4 , suggesting that they had exited pluripotency ( Figure 4B , Figure S4D ) ., Moreover , the acquisition of Brachyury/Sox2 coexpression coincided with an upregulation of Wnt3a , Cdx2 and Nkx1 . 2 as well as trunk Hox genes ( Figure 4B ) , characteristic of embryo and mESC derived NMPs ., Consistent with this , the paraxial/somitic mesoderm markers Tbx6 and Meox1 and the neural factor Sox1 were expressed in these conditions ( Figure S5A ) ., Immunostaining indicated that by D3 of differentiation Tbx6 and Sox2 expression were mutually exclusive ( Figure S5B ) ., By contrast the expression of genes characteristic of anterior neural plate ( e . g . Otx2 and Six3 ) and endoderm ( Foxa2 ) 54 were largely absent in FGF/CHIR conditions ( Figure S5A ) ., Collectively , these data indicate that , similar to mESCs , stimulation of Wnt and Fgf signalling in mouse EpiSCs leads to the induction of an NMP state ., The developmental potential of differentiated mouse EpiSCs has previously been tested by transplantation into mouse embryos 19 , 55 ., We therefore grafted EpiSC-derived NMPs constitutively expressing GFP into the NSB of E8 . 5 embryos ., After 48 h in culture , we observed extensive incorporation of GFP expressing cells ( 15/15 embryos ) ( Figure 4C–D ) ., Sections from these embryos revealed integration of transplanted cells into the somites and presomitic mesoderm of host embryos ( 10/10 ) and neural tube ( 4/10 ) ( Figure 4D ) ., We did not observe contributions to endoderm or other tissues ., Antibody staining for paraxial mesoderm ( Tbx6 ) , somite/dermomyotome ( Pax3 ) , neural ( Sox2 ) and floor plate ( Foxa2 ) markers confirmed that the engrafted cells had acquired the marker expression of their host environment ( Figure 4E ) ., Moreover , examination of the rostral limit of labelling using the somite level as a reference revealed that grafted EpiSC derived NMPs behaved similarly to homotopic grafts of microdissected E8 . 5 NSB cells 56 ., Strikingly , few cells grafted into the node of E7 . 5 embryos showed any incorporation ( 2 out of 8 embryos had 8–10 incorporated cells/embryo ) , suggesting that these conditions produce a population incompatible with gastrulation-stage development ., Similarly , cells differentiated for 24 h in FGF/CHIR did not incorporate into the NSB of E8 . 5 embryos ( n\u200a=\u200a5 ) ( Figure 4D ) ., Collectively , these results suggest that 48 h treatment of EpiSCs with FGF/CHIR results in coexpression of Brachyury/Sox2 ( up to 90% , Figure S4B ) and generates NMPs that functionally resemble their in vivo counterparts ., The resemblance of mouse EpiSCs to human embryonic stem cells ( hESCs ) prompted us to ask whether an analogous FGF/CHIR treatment regimen was sufficient to generate human NMPs ., Treatment of three independent hESC lines with CHIR and bFgf from D0–D3 downregulated NANOG and OCT4 and upregulated the suite of NMP expressed genes—BRACHYURY , NKX1 . 2 and CDX2—similar to mouse ESCs and EpiSCs ( Figure 5C , Figure S4D ) ., SOX2 expression was maintained in this population and up to ∼80% of cells co-expressed SOX2 and BRACHYURY ( Figure 5B , Figure S4B ) ., We also observed the spontaneous upregulation of paraxial mesoderm/somite markers ( TBX6 , MSGN1 , MEOX1 ) ., ( Figure S5C ) ., By contrast , the expression of a lateral plate ( KDR ) and an endoderm ( FOXA2 ) marker were minimal ( Figure S5C ) ., Thus FGF/CHIR treated hESCs appear to adopt an NMP identity and are likely to represent the in vitro correlates of the SOX2 and BRACHYURY co-expressing cells found in the caudal epiblast of human embryos 8 ., Consequently we dubbed these cells hNMPs ., To test the potency of hNMPs , we treated hESCs with FGF/CHIR for 72 h to drive the generation of BRACHYURY+/SOX2+ cells and then re-plated them for a further 48 h in serum free media without additional factors to promote the induction of spinal cord identity ( Figure 5A ) ., We termed these cells NP and compared them to neural cells derived from hESCs using a dual SMAD inhibition protocol involving Nodal and BMP inhibitors ( SB/LDN ) 57 ., Both conditions induced neural identity , exemplified by increased levels of SOX2 , TUBB3 and PAX6 ( Figure 5D ) ., As expected , neural cells generated using dual SMAD inhibition expressed the anterior marker OTX2 but lacked expression of HOX genes ( Figure 5D ) ., By contrast , neural cells derived from NMPs expressed SOX1 and the posterior HOX genes HOXC6 , HOXC8 and HOXC9 but not OTX2 ( Figure 5D ) ., A similar expression profile was obtained after treatment with RA and dual Shh agonists SAG and purmorphamine ( Pur ) ., This also induced expression of the motor neuron progenitor marker OLIG2 ( Figure S5D ) ., Antibody staining verified HOXC8 expression in NP conditions and revealed that the majority of HOXC8+ cells co-expressed SOX2 , confirming their neural identity ( Figure 5E , F ) ., Treatment of neural cells for 48 h with FGF/CHIR following 72 h dual SMAD inhibition did not result in HOXC8 induction suggesting that posteriorisation is necessary before or concomitant with neural induction ( Figure S5E ) ., Together these data suggest that neural differentiation of hNMPs generates spinal cord progenitors similar to mNMPs ., We next tested whether hNMPs differentiate into mesoderm by culturing them in the presence of CHIR alone ., This resulted in the expression of paraxial/somitic mesoderm markers TBX6 , MSGN1 and MEOX1 ( Figure S5D ) , but little if any expression of KDR , a lateral plate mesoderm marker ( Figure S5D ) ., Taken together these findings provide evidence of a human NMP population that gives rise to spinal cord and paraxial mesoderm derivatives but not anterior neurectoderm or lateral plate mesoderm ., Moreover , a similar set of developmental cues induces and directs NMPs in human and mouse , consistent with a similar ontogeny of trunk tissues in these species ., The ability to generate NMPs in vitro allows experimental investigations of trunk development that are challenging or impossible in vivo ., For example , although the requirement for Brachyury in mesoderm formation is well-established 58–60 , the truncation of embryos lacking Brachyury has complicated analysis of its role in the elaboration of spinal cord identity ., In zebrafish , a non-autonomous role for Brachyury orthologues has been identified 60 ., It is unclear whether in mammals Brachyury is required directly to maintain NMPs and therefore generate spinal tissue or indirectly via Wnt induction to establish a mesodermal niche that signals to generate or maintain posterior neural tissue ., To address this we took advantage of Brachyury null mESCs ( BTBR10 ) derived from embryos lacking Brachyury 61 ., Assaying Brachyury null cells at D3 of differentiation indicated that , in contrast to wild-type ESCs , Tbx6 expression was not upregulated by exposure to FGF/CHIR signalling , whereas Cdx2 and Hoxb1 expression were induced ( Figure 6B ) ., This is consistent with the lack of posterior mesoderm induction in Brachyury mutant embryos and prompted us to address the fate of Brachyury mutant cells that would normally form mesoderm ., In wild type cells exposed to Meso conditions , Tbx6 was highly expressed at D5 ( Figure 6D ) , as were Desmin and MyoD at D8 ( Figure 6E ) ., By contrast Brachyury null cells subjected to the same conditions failed to differentiate into paraxial mesoderm as indicated by the absence of Tbx6 ( Figure 6D ) ., Instead these cells expressed Sox1 , Sox2 and posterior Hox genes ( Hoxc6 and Hoxc9 ) at D5 ( Figure 6C ) and differentiated into β-Tubulin expressing neurons ( Figure 6E ) ., These data indicate that within mouse NMPs , Brachyury not only specifies mesodermal identity via mechanism ( s ) in addition to its direct stimulation of Wnt signalling , but also represses neural identity ., In the absence of Brachyury , NMPs adopt a neural differentiation route ., Thus the induction of posterior neural tissue is not dependent on Brachyury ., Moreover the data separate the mesoderm inducing and posteriorising activity of Wnt signalling and provide evidence that posteriorisation of the CNS is not dependent on mesoderm derived signals ., What could be responsible for the induction of posterior Hox genes ?, Analysis of the transcriptome data revealed the induction in NMPs of the Cdx genes Cdx1 , 2 and 4 , which have been implicated in the regulation of Hox gene expression ( Figure 6F ) 29 , 62 ., Induction of both Cdx1 and Cdx2 were detectable within 12 h of FGF/CHIR exposure and the levels of all three genes increased further at D3 and D4 of NP differentiation and at D5 of Meso differentiation ., Moreover , the induction of Cdx2 by Fgf/Wnt signalling was maintained in Brachyury null ESCs ( Figure 6B ) ., Thus the induction of Cdx proteins by Fgf/Wnt signalling represents a good candidate for the posteriorisation of NMPs ., Moreover the temporal accumulation of Cdx levels following Wnt exposure might provide a timing mechanism for the progressive induction of increasingly more posterior Hox genes ., We describe the in vitro generation of bipotential neuromesodermal progenitors from both mouse and human pluripotent stem cells that are capable of producing posterior neural tissue and paraxial mesodermal tissue ., This recapitulates the behaviour of NMPs residing in the CLE and NSB 15 , 22 ( Figure 6G ) ., Moreover , we provide evidence that Wnt signalling has two distinct functions in NMPs , initiating a mesodermal differentiation programme by regulating Brachyury expression and independently posteriorising these cells ., It is also likely that Brachyury maintains NMPs during axis elongation by forming a positive feedback loop with Wnt gene expression as has been previously shown 60 ., Strikingly , a neuromesodermal precursor is also present in ascidian embryos 63 ., Similar to vertebrates , the induction of these cells depends on the timing of Wnt and Fgf signalling 64 , 65 ., Moreover the mesoderm and posterior nervous system of many arthropods , including short germband insects , arises from a shared progenitor population that is exposed to Wingless signalling and expresses Cdx 66 ., Thus molecular and cellular features of the development of the neural and mesodermal components of the trunk appear to be evolutionarily conserved across bilaterian embryos ., This emphasizes the distinct developmental origins of cells that form anterior and posterior regions of bilaterian embryos , suggesting an explanation as to why it has proved difficult to generate spinal cells and skeletal muscle from ESCs ., More generally , the ability to produce and manipulate NMPs in vitro has the potential to increase the efficiency with which cell types derived from posterior neural and paraxial mesodermal tissue can be generated and analysed ., Animal experiments were performed under the UK Home Office project licenses PPL80/2528 and PPL60/4435 , approved by the Animal Welfare and Ethical Review Panel of the MRC-National Institute for Medical Research and MRC Centre for Regenerative Medicine and within the conditions of the Animals ( Scientific Procedures ) Act 1986 ., Human Embryonic Stem Cell UK Steering Committee approval has been obtained ( ref . SCSC14-09 ) ., The mouse ES cell lines , HM1 67 and BTBR10 68 were maintained in ES cell medium 69 with 1000 U/ml LIF ( Chemicon ) on mitotically inactive primary mouse embryo fibroblasts ., To initiate differentiation , ES cells were removed from feeders by dissociation using 0 . 05% trypsin and then plated onto tissue culture plates for two short successive periods ( 20–30 mins ) to remove feeder layers ., To induce differentiation , the cells were plated on CellBINDSurface dishes ( Corning ) precoated with 0 . 1% gelatin ( Sigma ) at a density of 5×103 cells cm−2 in ‘N2B27’ medium ., This medium comprised Advanced Dulbeccos Modified Eagle Medium F12 ( Gibco ) and Neurobasal medium ( Gibco ) ( 1∶1 ) , supplemented with 1×N2 ( Gibco ) , 1×B27 ( Gibco ) , 2 mM L-glutamine ( Gibco ) , 40 µg/ml BSA ( Sigma ) , 0 . 1 mM 2-mercaptoethanol ., Cells were grown in N2B27 supplemented with 10 ng/ml bFgf ( R&D ) for 3 days ( D1–D3 ) and then were transferred into serum free media without bFgf ( D3–D5 ) ., To induce ventral hindbrain identity NPCs ( NH ) 100 nM RA ( Sigma ) and 500 nM SAG ( Calbiochem ) was added from D3–D5 ., Spinal cord identity ( NP ) was induced by the addition of 5 µM CHIR99021 ( Axon ) or 100 ng/ml Wnt3a ( R&D ) from D2 to D3 followed by 100 nM RA , 500 nM SAG from D3–D5 ., To induce mesodermal differentiation the cells were treated with CHIR99021 from D2–D5 ., To induce terminal differentiation , cells were trypsinised and plated as single-cell suspension on plates coated with Matrigel ( BD Biosciences ) at a density of 1×105 cells cm−2 in N2B27 medium supplemented with bFgf ( 10 ng/ml ) ., The next day bFgf was removed and cells were left to differentiate for an additional 3 days ., The mouse EpiSC line R04-GFP 55 was routinely maintained in N2B27 supplemented with Activin A ( 20 ng/ml; R&D Systems ) and bFgf ( 10 ng/ml; Peprotech ) as previously described 70 ., For differentiation of EpiSCs into NM progenitors approximately 1500–2000 cells/cm2 were plated on fibronectin ( Sigma ) -coated wells in N2B27 medium supplemented with CHIR99021 ( 3 µM; Signal Transduction Division , Dundee ) and bFgf ( 20 ng/ml ) ., For grafting experiments the initial plating density was 2500 cells/cm2 and cells were plated on either fibronectin or gelatin ., Human ESC lines MasterShef 5 and 7 ( a gift of Prof . Harry Moore , University of Sheffield ) and a Sox2GFP reporter line ( a gift of Dr Andrew Smith , University of Edinburgh ) were cultured in Essential 8™ medium on Geltrex™-coated plates ., For hNMP differentiation cells were pre-treated for 1 h with ROCK inhibitor Y-27632 ( 10 µM; Calbiochem ) , dissociated with accutase and plated at approximately 10 , 000 cells/cm2 ( Sox2-GFP hESCs ) or 80 , 000 cells/cm2 ( MasterShef5 and 7 hESC lines ) on fibronectin-coated wells in N2B27 medium supplemented with 3 µM CHIR99021/20 ng/ml bFgf and Y-27632 ( 10 µM ) ., The medium was replaced the following day with fresh N2B27 containing the same components minus the ROCK inhibitor ., For directed differentiation of hESCs , cultures were differentiated in the presence of CHIR99021/bFgf for 72 h as described above ., For neural/spinal cord differentiation 72 h CHIR99021/bFgf-differentiated cells were treated with Accutase ( Sigma ) and transferred onto Geltrex ( Life Technologies ) -coated plates either in N2B27 alone or N2B27 supplemented with RA ( 0 . 1 µM; Sigma ) , SAG ( 0 . 5 µM; Calbiochem ) and purmorphamine ( 1 µM; Calbiochem ) for 48 h ., For mesodermal differentiation 72 h CHIR99021/bFgf differentiated cells were cultured in N2B27 supplemented with CHIR99021 ( 3 µM ) for a further 48 h ., For dual SMAD inhibition Sox2-GFP hES cells were plated at 10 , 000 cells/cm2 on Geltrex™-coated wells in N2B27 supplemented with LD | Introduction, Results, Discussion, Materials and Methods | Cells of the spinal cord and somites arise from shared , dual-fated precursors , located towards the posterior of the elongating embryo ., Here we show that these neuromesodermal progenitors ( NMPs ) can readily be generated in vitro from mouse and human pluripotent stem cells by activating Wnt and Fgf signalling , timed to emulate in vivo development ., Similar to NMPs in vivo , these cells co-express the neural factor Sox2 and the mesodermal factor Brachyury and differentiate into neural and paraxial mesoderm in vitro and in vivo ., The neural cells produced by NMPs have spinal cord but not anterior neural identity and can differentiate into spinal cord motor neurons ., This is consistent with the shared origin of spinal cord and somites and the distinct ontogeny of the anterior and posterior nervous system ., Systematic analysis of the transcriptome during differentiation identifies the molecular correlates of each of the cell identities and the routes by which they are obtained ., Moreover , we take advantage of the system to provide evidence that Brachyury represses neural differentiation and that signals from mesoderm are not necessary to induce the posterior identity of spinal cord cells ., This indicates that the mesoderm inducing and posteriorising functions of Wnt signalling represent two molecularly separate activities ., Together the data illustrate how reverse engineering normal developmental mechanisms allows the differentiation of specific cell types in vitro and the analysis of previous difficult to access aspects of embryo development . | Stem cells are providing insight into embryo development and offering new approaches to clinical and therapeutic research ., In part this progress arises from “directed differentiation” – artificially controlling the types of cells produced from stem cells ., Here we describe the directed differentiation of mouse and human pluripotent stem cells into cells of the spinal cord and paraxial mesoderm ( the tissue that generates muscle and bone that is normally found adjacent to the spinal cord ) ., During embryo development , spinal cord and paraxial mesoderm arise from a shared group of precursors known as neuromesodermal progenitors ( NMPs ) ., We show that signals to which NMPs are exposed in embryos can be used to generate NMPs from pluripotent stem cells in a dish ., We define conditions for the conversion of these NMPs into either spinal cord or mesoderm cells ., Using these conditions , we provide evidence that the decision between spinal cord and mesoderm involves a gene , Brachyury , that promotes mesoderm production by inhibiting spinal cord generation ., Together the data illustrate how mimicking normal embryonic development allows the generation of specific cell types from stem cells and that this can be used to analyse cells that are otherwise difficult to study . | biology and life sciences, cell differentiation, developmental biology, neuronal differentiation | Timed pulses of WNT and FGF signaling convert human and mouse pluripotent stem cells into neuromesodermal progenitors that can be directed to differentiate into spinal cord and paraxial mesoderm cells |
journal.pcbi.1006552 | 2,018 | Modeling and subtleties of K-Ras and Calmodulin interaction | Ras proteins are well-known small GTPases involved in the regulation of key signal transduction pathways ., Cycling from the inactive ( GDP-bound ) to the active ( GTP-bound ) state , they faithfully respond to extracellular signals due to their tight regulation by GTP-exchange factors ( GEFs ) and GTPase activating proteins ( GAPs ) ., In the GTP-bound form , two regions of the protein change conformation ( switch I and II domains ) allowing its binding with different effector proteins and consequently the activation of diverse signal transduction pathways ., Among those , the best characterized are the RAF1/MEK/ERK and the phosphatidylinositol-3-kinase ( PI3K ) /AKT 1 , which are known to regulate proliferation , differentiation and survival ., Activating point mutations render Ras proteins that are always found in the GTP-bound state independently of the extracellular signals and are crucial steps in the development of the vast majority of cancers 2 ., Ras genes were the first oncogenes identified in human cancer cells , and nowadays they are well established as the most frequently mutated oncogenes in human cancer 3 ., Three different genes code for a total of four different Ras isoforms named H-Ras , N-Ras , K-RasA and K-RasB ( herein after referred to as K-Ras ) ., K-Ras is the most frequently mutated oncogene in solid tumors and its oncogenic mutations occur mostly in pancreatic ductal adenocarcinomas ( 95% ) , colon ( 40% ) and adenocarcinomas of the lung ( 35% ) 3–5 ., Although they all have a highly conserved globular domain ( from residue 1 to 165 ) containing the guanosine nucleotide and effector binding sites ( Switch I and Switch II ) , the last C-terminal residues of Ras proteins , named the hypervariable region ( HVR ) , which contains the membrane targeting signals , are not conserved among the different isoforms ., H- and N-Ras achieve high-affinity hydrophobic membrane binding mainly through lipid modifications ., By contrast , K-Ras has , adjacent to the farnesylated cysteine Cys185 , a stretch of lysine residues—known as the polybasic domain—that promotes an electrostatic interaction with the negatively charged phospholipids 6 , 7 , which confines K-Ras almost entirely to non-raft microdomains within the plasma membrane 8 ., The different membrane anchors interact with lipids and proteins of the plasma membrane and , together with the hypervariable region ( HVR ) , drive the Ras isoforms into spatially and structurally distinct nanodomains , of which each then contains a cluster of molecules ( nanocluster ) 9–11 ., Importantly , the nanodomains that are occupied by the three isoforms of Ras do not show any overlap ., Furthermore , not only are the different Ras isoforms laterally segregated , but inactive GDP-loaded Ras occupies nanodomains that are spatially distinct from those occupied by the active GTP-loaded form ., Formation of these nanoclusters is essential for activation of mitogen-activated protein kinases ( MAPKs ) , because they constitute exclusive sites in the plasma membrane for Raf-1 recruitment and ERK activation 12–14 ., Because oncogenic mutations of K-RAS give rise to an always GTP-bound protein that constitutively binds to effectors , positive or negative physiologic regulation of oncogenic K-RAS was not initially expected ., In recent years , interaction of K-Ras with proteins , which are not effectors but which may function as allosteric regulators or scaffolds , have been described and some proved to be crucial to fully display K-RAS oncogenic phenotype ., Galectin-3 15 , calmodulin ( CaM ) 16 , phosphodiesterase δ 17 , 18 , nucleophosmin , nucleolin 19 and heterogeneous nuclear ribonucleoprotein A2/B1 ( hnRNPA2/B1 ) 20 have been shown to interact with K-Ras and modulate its downstream signaling ., The mechanism by which these proteins modulate K-Ras signaling is diverse: phosphodiesterase δ by binding to the farnesyl group facilitates the diffusion of K-Ras from endomembrane to the cytoplasm , ultimately favoring its correct relocalization to the plasma membrane and consequently enhances its signaling 18; Galectin-3 regulates K-Ras nanoclustering at the plasma membrane and also enhances its signaling 15; and , hnRNPA2/B1 favors the interaction of K-Ras with PI3K 20 ., In contrast , while K-Ras interaction with CaM has been known for many years , there is still some controversy regarding the consequences of this interaction ., Our group demonstrated that CaM interaction with K-Ras inhibits K-Ras signaling to Raf/MEK/ERK 16 and inhibits K-Ras phosphorylation at Ser181 in the HVR 21 ., Interestingly , CaM also binds to PI3K enhancing its activity 22 , and the existence of a complex containing K-Ras , CaM and PI3K has been proposed 23 ., CaM is a small ( 148 amino acids ) and well conserved Ca2+-binding protein 24 ., The crystal structure of CaM in the Ca2+-bound form shows a dumbbell-shaped molecule with two globular domains arranged in a trans configuration ., These domains are connected by a long extended central α-helix , the middle portion of which is highly mobile and acts as a flexible tether ., Each domain consists of two helix-loop-helix motifs ( EF hands ) , with each binding one molecule of Ca2+ ., Ca2+ binding changes the orientation of the two EF hands of each domain , inducing the appearance of hydrophobic patches that interact with proteins known as CaM-binding proteins ( CaMBPs ) 25 ., Binding of CaM to CaMBPs modulates the function of these proteins and , in consequence , affects many aspects of cell regulation ., The carboxyl-terminal lobe binds Ca2+ with high affinity ( Kd 10−7 M ) , whereas the amino-terminal lobe binds it with lower affinity ( Kd 10−6 M ) ., The fact that the Kd values fall within the range of intracellular Ca2+ concentration exhibited for most cells ( 10−7–10−6 M ) makes it a good sensor for changes in Ca2+ intracellular levels 26–28 ., The CaM binding domain of some of the CaMBPs with high affinity for CaM ( nM range ) consists of a 20-amino acid sequence that has an amphipathic α-helix conformation 29 ., CaM binding domains with lower affinity for CaM ( μM range ) have also been described 30 ., Some proteins like MARCKS and CAP-23/NAP-22 use the myristoyl group to interact with CaM 31 , 32 ., As well as K-Ras , diverse Ras superfamily GTPases like Kir/Gem 33 , Ric 34 , Rin 35 , Rab3A 36 , and RalA 37 have been shown to bind to CaM ., Biochemical data indicate that at least two different regions in the K-Ras molecule are important for K-Ras/CaM interaction: the hypervariable region and the α-helix between amino acids 151 and 166 38 ., Within the hypervariable region , both the hydrophobic farnesyl group and the positive-charged amino acids were essential for the interaction between K-Ras and CaM ., Consistently , K-Ras S181D mutant , which mimics phosphorylation of Ser-181 of K-Ras , also completely abolished binding to CaM ., Although the presence of the farnesyl group increases the affinity of purified K-Ras to CaM , full length non-farnesylated K-Ras still has micromolar affinity for CaM 39 ., Accordingly to the above mentioned , the NMR data of this complex show that the N-terminal lobe of CaM interacts with the globular domain of K-Ras and the C-terminal lobe of CaM interacts with the HVR 40 ., But controversial data exist regarding how CaM interaction with K-Ras could modulate K-Ras activity ., While some data indicate that CaM could extract K-Ras from membranes in vitro , most probably by interacting with the farnesyl group 41 , 42 , it is not clear if in vivo this hydrophobic group would always be available for CaM to interact with ., In fact , our group has demonstrated that K-Ras and CaM colocalize mainly in the plasma membrane , suggesting that in vivo interaction does not directly lead to K-Ras internalization 38 ., CaM could be modulating interaction of K-Ras within the plasma membrane , with effectors , scaffolds or with different lipids , ultimately regulating K-Ras signaling from the plasma membrane ., Thus , modelling of K-Ras/CaM interaction is important to decipher the cellular role of this interaction ., To mimic the situation of K-Ras bound to the membrane , thus with farnesyl group hindered between the phosphoslipids , we aimed to model the interaction between a full length non-farnesyated K-Ras and CaM ., CaM and K-Ras have been widely studied computationally 43 , 44; in fact , CaM is one of the most studied proteins with molecular dynamics ( MD ) due to its high degree of flexibility ., These systems have also been joined to a certain degree 45 , but up to date no simulations of the whole proteins have been performed ., Thus , we decided to carry on conventional MD ( cMD ) on a system with both proteins in order to determine which the details of the interaction are ., Furthermore , in order to increase the exploration of the conformational space of the K-Ras/CaM system , scaled MD ( sMD ) a recently developed methodology that proved to be effective to sample wider conformational areas faster than cMD 46 , was used ., In order to computationally study the interaction between K-Ras and CaM , a system with both proteins had to be prepared ., Since NMR experimental data regarding the interaction between these two proteins has already been published 40 , we decided to mimic the experimental settings: oncogenic K-Ras ( G12D mutation ) full-length without post-translational modifications paired with holo-CaM ., Prior to a simulation between the proteins , a system composed of GTP-bound K-Ras with a fully extended HVR was prepared ., This system was used to determine whether the HVR could be found in an extended conformation in several frames or if it would be mostly bent to interact with the globular domain ., Fifty nanoseconds of cMD were performed and the provided trajectories were analyzed by measuring the distance between residues 161 ( from the α-helix 5 ) and 178 ( from the HVR ) ., The HVR presented an extended conformation most of the time , showing great motility ( Fig 1A ) ., Interestingly , other groups have seen similar behavior when simulating K-Ras in its active state , reporting that the HVR does not significantly interact with the globular domain 47 ., Since the binding of these two proteins does not seem to be mediated by the common binding mechanism of CaM ( where it wraps its lobes around a single structure , such as an α-helix ) , we decided to use the structure of CaM with PDB code 2MGU ., This structure presents its lobes rather extended , which could fit with a model in which the N-lobe of CaM interacts with the globular domain of K-Ras and the C-lobe interacts with the HVR ., The peptide present in the structure was replaced by the HVR of K-Ras with Modeller , and subsequently rotated to fit the model ( see Methods for more details ) ., Last , the globular domain of K-Ras was attached to obtain the system with both proteins ( Fig 1B ) ., Once the system was prepared , a total of 6 cMD and 4 sMD simulations were carried out , each of them with a total length of 50 ns ., The trajectories were visually analyzed in order to determine which simulations had stablished a proper interaction between the two proteins , and in which K-Ras/CaM had fallen apart ., Interestingly , in 9 out of 10 simulations the proteins interacted throughout most of the simulation length , even with the additional energy boost of the sMD simulations ( Fig 2 ) ., Furthermore , the N-Terminal domain of CaM remained close to the α-helix 5 of K-Ras in most of the simulations ., The end of the HVR maintained a close contact with the C-Terminal lobe of CaM , while the polybasic domain of K-Ras interacts with the linker region of CaM ., The energy of the system was determined by performing a MMPB/GBSA analysis ., The dynamics were considered stable if the last 5 ns did not present significant deviations ., If any of the simulations were not stable enough , they were extended until stability was reach ., The energy profile was similar for PB and GB ., The interaction presented between -60 and -100 kcal/mol for GB and between -40 and -120 kcal/mol for PB both for cMD and sMD ( Fig 3 ) ., The last ns of each simulation were used to calculate the contribution of each residue to the binding energy through a “per residue” analysis ., The residues of CaM were studied in order to find matches with the experimental data available ., Two thresholds were imposed to consider a residue as actively participating in the interaction between K-Ras and CaM: the first was a requisite of at least -0 . 7 kcal/mol of average contribution to the binding , whereas the second was its presence in at least 3 of the simulations ., Up to twelve residues matched with the experimental data available , many of which are negatively charged residues ( 78 to 84 ) that can interact with the polybasic domain of K-Ras ( Fig 4A ) ., Intriguingly , certain residues of CaM whose surroundings are modified when interacting with K-Ras ( experimentally ) do not present a significant implication in the interaction between both proteins in the simulations ( Fig 4A ) ., The presence of changes in nearby residues when binding to other proteins can explain why there are NMR shifts assigned to those residues while no energy contribution is seen in our simulations ( Fig 4B ) ., With all things considered , the model can be considered robust enough to analyze the residues of K-Ras that participate in the interaction , some of which have not been described yet ., After analyzing the residues of CaM , we focused on the residues of K-Ras relevant for the interaction ., A threshold of -1 kcal/mol of average was imposed to the residues that participated in the interaction ., Also , their participation had to be present in at least 5 simulations ., In concordance with the experimental data , most of the residues responsible for the interaction were found within the HVR ., However , 5 residues were identified in the globular domain ., Furthermore , most of them presented energy values below -3 . 5 kcal/mol , being arginine 135 the most significant residue in terms of average energy ( Fig 5A ) ., When visualized , the simulations revealed that the selected residues of the globular domain were , in fact , closely interacting with CaM ., The arginines from the α –helix 5 formed hydrogen bonds with the EF hand of the N-Terminal domain , while arginine 135 and proline 140 interact with one of the four α helixes present in the N-Terminal lobe of CaM ( Fig 5B–5E ) ., To further confirm the results obtained with the performed simulations , a different methodology was used: the umbrella sampling ., This kind of simulation allows a more progressive scenario for the proteins to adapt , as more time is given to position them nearer ., To perform this simulation , a restriction was added to maintain the mass center of the α-helix 5 of K-Ras and the mass center of the N-Terminal lobe of CaM at a prefixed distance ., Then , the restricted groups were slowly approached , at a rate of 0 . 5 Å per step , remaining for 0 . 5 ns at each distance before closing the gap between them ., The initial distance was set at 20 Å , while the last step was set at 5 Å ., Once the simulations were performed , structures from the US with the mass centers maintained at 5 , 6 , 7 and 8 Å of distance were obtained and 10 ns of cMD were calculated ., All these simulations presented high interaction between K-Ras and CaM , with the N-Terminal lobe of CaM wrapping the α-helix 5 of K-Ras and the HVR embedded in the C-Terminal lobe of CaM ( Fig 6A ) ., A MMGBSA analysis was also performed to determine the binding energy of the proteins and analyze the stability of such interaction ( Fig 6B ) ., Values around -150 kcal/mol were obtained for all four simulations , exceeding the values seen in simple cMD or sMD ( whose values were around -100 Kcal/mol ) ., A “per residue” analysis was also performed so as to determine if the residues described as relevant with the previous methodology were still actively participating in the binding ., Since these residues had more time to accommodate and orientate in a favorable angle for the interaction , only those residues actively interacting in the four simulations with an average energy below -1kcal/mol were selected ., Even though according to the US simulations some residues selected with the initial model were not relevant for the interaction , most of them matched ., Furthermore , only one of the studied residues of the globular domain did not surpass the thresholds , which backs up the idea that the globular domain is playing a part in the interaction ( Fig 6C ) ., Taking into account all the data provided by the simulations performed , it seems the globular domain interacts with CaM , specifically through residues R135 , P140 , R161 , R164 and , to a minor extent , K165 ., With the purpose of verifying the obtained results with experimental data , three mutants of the globular domain ( 1–166 aa ) were obtained through point mutation ., The mutants were designed according to the results of the simulations: R135E , R161E and R164E ., The corresponding GST-K-Ras mutants were expressed in bacteria , affinity purified and then its binding to CaM determined by Surface Plasmon Resonance ( SPR ) ., Biotinilated CaM was attached to a chip with streptavidin and the GST-tagged globular domain of K-Ras ( either wild-type or mutant ) was injected as an analyte ., A control flow cell with no CaM , was also injected with the globular domain as a blank ., To discard that the binding was due to the GST tag , recombinant GST was injected in all flow cells and no binding was observed ., The injection of the globular domain of K-Ras led to an increase in the Resonance Units ( RU ) of the flow cells with CaM , which stemmed from the binding of the injected protein to CaM ., The mutants also showed binding with CaM , but to a lower extent ., An affinity study was performed to determine the KD , and the results reflected that the wild-type globular domain presented a lower KD than any mutant ., All the mutations led to an increase in the KD , that is , in a reduction of the affinity with CaM ( Fig 7 ) ., Thus , our experimental data support the results obtained through computational simulations , where these 3 residues were identified as key players in the interaction of the globular domain of K-Ras with CaM ., Even though it has been almost twenty years since the discovery of the interaction between K-Ras and CaM 16 , its subtleties have remained elusive ., In the present work , thanks to the use of computational aided techniques , we have shed some light upon this interesting matter ., Since these proteins have never been simulated computationally , we looked for a system with validated experimental results to be able to contrast the data generated ., Thus , we used full-length K-Ras without its farnesylation , as this system had already been experimentally studied with NMR 40 ., Even though we are aware of the relevance of the post-translational modifications of K-Ras for the interaction with CaM , it’s unlikely that CaM can initially interact with the farnesyl group , as it will be attached to the plasma membrane ., Thus , our simulations could mimic a situation in which the farnesyl group is not available for the interaction , as it would be hidden within the membrane ., We used CaM with an extended conformation , with each lobe interacting with different domains of K-Ras , rather than wrapping around an specific region , similar to the interaction with other proteins in which the lobes of CaM interact with different regions 25 ., It could be observed that the interaction between the proteins was stable , as it was maintained throughout most of the simulations ., Furthermore , several residues of CaM matched the experimental data , despite the motility present in the system and the energy boosts provided in the sMD simulations ., Our group had previously described the participation of the globular domain and specially α-helix 5 of K-Ras in the interaction with CaM 38 ., In the present work , we have taken another step forward to model this interaction and have confirmed and identified new residues of the globular domain that are implicated in said interaction ., Arginine 161 , arginine 164 ( α-helix 5 ) and , to a minor extent , arginine 135 ( α-helix 4 ) seem to be responsible of the interaction with the N-terminal lobe of CaM , since single point mutations in any of those residues lead to an increase in the experimental values of KD between the globular domain of K-Ras and CaM ., Interestingly , previous publications have highlighted the relevance of residue 135 in Ras signaling , as mutations in this residue led to enhanced binding with C-Raf RBD 48 ., This phenomena may be explained by the diminished interaction with CaM , as the binding of this protein to K-Ras is known to diminish MAPK pathway signaling 16 ., While we previously showed that the K-Ras switch II mutant , R68D/R73D , had a compromised interaction with CaM 38 , our present model does not predict direct interactions of these two arginines with CaM ., The most plausible explanation is that the substitution of the two positive residues by negative ones induces a conformational change in K-Ras , and indirectly , an increase in the negative charge density in the surface of CaM interaction that prevents the binding with this acidic protein ., As for the HVR , the simulations we have performed here proven to fit the available data ., The highly negatively charged linker region of CaM is attracted to the polylysine domain of K-Ras , where they interact through electrostatic couplings ., This fact has already been described experimentally by other groups 39 ., As for the last C-Terminal residues of K-Ras , they are embedded by the C-Terminal lobe of CaM ., However , it must be considered that this interaction may vary after the post-translational modifications , as the–AAX residues are removed and the farnesyl group is attached ., Interestingly , in most of the simulations ( six out of nine ) cysteine 185 is not embedded within the C-terminal lobe of CaM , which would fit with a model in which the farnesyl group of this residue would be attached to the plasma membrane ., Our simulations can help to understand why the phosphorylation of K-Ras leads to the abrogation of this interaction with CaM 49 ., As shown in the performed MD , the polybasic region of K-Ras plays an important role in the interaction with CaM , creating electrostatic interactions with the acidic linker region ., Thus , the addition of a phosphate group , highly negatively charged , is bound to have a negative impact on the K-Ras/CaM interaction ., Our model may also provide one of the reasons why CaM does only interact with K-Ras when bound to GTP ( its active state ) 16 ., Since the α-helixes of K-Ras are oriented towards the membrane when bound to GDP 50 , the N-Terminal lobe of CaM would not be able to reach its interaction zone with the globular domain of K-Ras , as it would be covered by the PM ., When active , K-Ras would expose its α-helix 4 and 5 , giving the N-Terminal lobe of CaM a chance to interact with it ., However , lack of interaction of full-length GDP-loaded K-Ras has also been described in the absence of lipid membranes ., In this case the proposed autoinhibitory effect of the HVR could prevent CaM binding 44: the globular domain would be inaccessible for CaM due to its binding with the HVR when K-Ras is in its GDP bound state , but would become reachable when GTP is loaded and the HVR is released ., In fact , our simulations support the idea that , when bound to GTP , the HVR of K-Ras is not stably interacting with the globular domain ., The study we performed of the dynamism of the HVR revealed that , even though there are some interactions between these groups , they are neither stable nor prolonged through much more than a few nanoseconds , thus giving to CaM the opportunity to interact with the HVR ., However , it must be considered that experimental data show that CaM fails to interact with the purified inactive globular domain of K-Ras 16 , so , despite our model being able to provide some explanations , a few details remain elusive ., Beyond the ins and outs of the interaction , the biological significance of such binding is becoming more interesting day after day ., Although other interaction models between CaM and K-Ras may be feasible , especially with cytosolic K-Ras , our simulations would support a model in which K-Ras and CaM would interact at membrane level without indispensably inducing an extraction of K-Ras from the membrane ( Fig 8 ) , a fact that has been previously described by our group 38 , and lately supported by recent publications that demonstrate that CaM can bind to K-Ras even when attached to nanodiscs emulating diverse types of PM 45 ., In fact , CaM interaction with K-Ras may be modulating K-Ras clustering and signaling from the PM ., For instance , CaM is thought to form a ternary complex with K-Ras and PI3K , which would enhance K-Ras signaling through AKT signaling while diminish it through Raf 23 ., Our simulations provide interesting data suggesting that , while keeping one of its lobes interacting with K-Ras ( probably the C-Terminal due to the interaction of the linker region with the polybasic domain ) , CaM could use its other lobe to interact with PI3K ., Also regarding K-Ras signaling , several authors have described the relevance of K-Ras dimerization in the activation of downstream effectors ., While dimerization through α-helix 1 and β sheets 1/2 would inhibit the binding of effectors such as Raf or PI3K , due to the overlapping interaction surfaces , dimerization of K-Ras using the α-helixes 3/4/5 and β sheet 2 has been proposed to promote Raf dimerization and hence its activation 51 , 52 ., As shown in the present work , CaM would also attach to the region of α-helixes 4 and 5 of the globular domain of K-Ras ., Interestingly , this region has recently been described as relevant for proper K-Ras dimerization , as arginine 161 forms a salt bridge when forming the dimer 53 ., As stated above , according to our model the region used by CaM to interact with K-Ras may overlap with the one used to form K-Ras dimers ., Moreover , not only do these interactions share the surfaces by which they interact but also certain residues used in the interaction such as arginine 161 ., Thus , one can conclude that CaM’s interaction with K-Ras would most probably interfere with K-Ras dimerization , and consequently this would be another mechanism ( besides inhibiting phosphorylation 21 ) by which CaM is negatively regulating K-Ras-Raf-ERK signaling ., All in all , we can affirm that our simulations ( and later experimental validation ) propose a reliable model in which residues R135 , R161 and R164 play a significant role in the interaction of the globular domain of K-Ras with CaM , while the polybasic domain of the HVR interacts with the acidic linker region of CaM ., The joining of the proteins was performed in several steps ., In order to have a model to work with , we first examined the original structure or the NMR structure of CaM with the HIV-1 matrix peptide ( PDB code 2MGU ) ., The peptide presented a certain degree of homology with the HVR of K-Ras , and we took advantage of that fact by replacing the existing peptide with a fragment of K-Ras ( residues 165 to 188 ) ., To this end , the peptide was replaced using the program modeller ( https://salilab . org/modeller/ ) ., The best structure generated by modeller was selected to perform the following simulations ., The HIV-1 matrix peptide was replaced by the K-Ras peptide ., However , the homology between the sequences did not match the real orientation of the interaction between K-Ras and CaM ., Thus , once the HIV-1 matrix peptide was replaced by the K-Ras fragment , another program was developed to rotate it ., This software creates a vector between two given atoms ( one from the CaM and another from the K-Ras fragment ) and increases the module ( the distance between those atoms ) ., Afterwards , it performs a rotation on its axis ( rotating the whole K-Ras fragment ) and decreases the module ( diminishes the distance between the selected atoms ) ., Even though several combinations were tried , the distances were finally set to 10 Å and 5Å and the rotation angle was fixed at 180° ., The system was then minimized in a multi-step manner , applying the same restraints as in the simulations with K-Ras ., The minimized complex was heated up to 300 k in a step wise manner , at a rate of 30 K every 20 ps ., The protein backbone atoms were restrained with a force constant of 0 . 5 kcal/mol·Å ., Additionally , 200 ps of simulation at constant pressure ( NPT ensemble ) were performed without any restraint in order to allow density equilibration ., Then , a short MD simulation of 2 ns length within the NVT assembly was done to allow small structural readjustments ., The final structure after this process was used as a reference to add the full-length oncogenic K-Ras ( mutation G12D ) ( PDB code 4DSN ) ., The lacking residues ( a majoritarian part of the globular domain , residues 1–164 ) were added by merging the two systems ., This step was done by superimposing the residues 165 to 168 of K-Ras and removing the leftover atoms of the K-Ras peptide ., The final complex was placed in a cubic periodic box filled with TIP3P water molecules , imposing a minimal distance of 15 Å between the protein and the box walls ., Water molecules closer than 2 . 2 Å were removed and neutralizing counter-ions ( sodium ions ) were added at positions of lowest electrostatic potential ., Minimization was carried in a multistep procedure:, 1 ) Full complex restraint , both K-Ras and CaM were restrained with a 10 kcal/mol Å constant , including GTP and calcium ions;, 2 ) CaM and globular domain of K-Ras restricted while releasing the lateral chains of the HVR;, 3 ) Release of the lateral chains of CaM;, 4 ) Progressive release of the HVR and CaM by diminishing the restriction constants from 10 to 0 . 5 kcal/mol Å;, 5 ) Progressive release of the globular domain of K-Ras ( in the same way as with HVR and CaM ) ;, 6 ) Minimum restraint on all backbones ( 0 . 1 kcal/mol Å ) ;, 7 ) No restraints minimization ., These minimizations were performed through 5000 steps of the conjugate gradient algorithm keeping fixed the selected parts of the system fixed with the indicated restriction constants , except for the last minimization , which was carried out for 10000 steps ., First , the cMD simulations were carried out ., To do so , the minimized systems were heated up to 300 K in increments of 30 degrees per step of 20 ps ., Afterwards , 200 ps of simulation at NPT ensemble were performed ., Also , a short MD simulation of 2 ns length within the NVT assembly was carried out ., The MD simulations of the systems were performed in a multi-step procedure ( each step of 10 ns ) ., The temperature was regulated by using the Langevin thermostat with a collision frequency ɣ of 2 . 0 ps-1 ., All bonds involving hydrogen atoms were constrained to their equilibrium value using the SHAKE algorithm , allowing the use of a 2 fs integration time step in all of the simulations ., Non-bonded interactions were truncated at a cut-off of 10 Å , and long range electrostatic interactions were treated with the particle-mesh Ewald method ., A total of 6 molecular dynamics simulations using different sets of initial velocities aimed at providing a better sampling 54 were performed of at least 50 ns each one ., As for the sMD simulations , the initial coordinates were taken from the first 5 ns of the cMD simulation ., A λ factor of 0 . 8 was applied , and a total of 4 simulations of at least 50 ns were produced ., In order to determine the binding energy between the proteins , the AmberTools module of AMBER was used with Molecular Mechanics Poisson Boltzmann ( Generalized Born ) Solvation Area ., To perform the calculations , structures f | Introduction, Results, Discussion, Methods | K-Ras , one of the most common small GTPases of the cell , still presents many riddles , despite the intense efforts to unveil its mysteries ., Such is the case of its interaction with Calmodulin , a small acidic protein known for its role as a calcium ion sensor ., Although the interaction between these two proteins and its biological implications have been widely studied , a model of their interaction has not been performed ., In the present work we analyse this intriguing interaction by computational means ., To do so , both conventional molecular dynamics and scaled molecular dynamics have been used ., Our simulations suggest a model in which Calmodulin would interact with both the hypervariable region and the globular domain of K-Ras , using a lobe to interact with each of them ., According to the presented model , the interface of helixes α4 and α5 of the globular domain of K-Ras would be relevant for the interaction with a lobe of Calmodulin ., These results were also obtained when bringing the proteins together in a step wise manner with the umbrella sampling methodology ., The computational results have been validated using SPR to determine the relevance of certain residues ., Our results demonstrate that , when mutating residues of the α4-α5 interface described to be relevant for the interaction with Calmodulin , the interaction of the globular domain of K-Ras with Calmodulin diminishes ., However , it is to be considered that our simulations indicate that the bulk of the interaction would fall on the hypervariable region of K-Ras , as many more interactions are identified in said region ., All in all our simulations present a suitable model in which K-Ras could interact with Calmodulin at membrane level using both its globular domain and its hypervariable region to stablish an interaction that leads to an altered signalling . | K-Ras is one of the most mutated oncogenes in human cancer ., Although several studies validate K-Ras protein as good candidate for direct therapeutic targeting , pharmacologic targeting has not been successful ., During the last years increasing evidences demonstrate that oncogenic K-Ras activity can be modulated in vivo by dimerization , nanoclustering at the plasma membrane or interaction with non-effector proteins , consequently opening new therapeutic strategies ., We have previously demonstrated that Calmodulin , an ubiquitous Ca2+-binding protein , is one of this K-Ras interacting proteins and that it negatively modulates K-Ras signaling ., Although experimental data were available showing the relevant regions for this interaction , a model of K-Ras and Calmodulin interaction was missing ., In the present work by using different computational modeling techniques we obtained a model for this interaction that agrees with the experimental data ., We believe the present model will help to better understand K-Ras regulation , and to design new inhibitors ., For instance , base on our model , we can predict that the interaction can take place at the plasma membrane , and that since the surface of K-Ras that interact with Calmodulin is the same that it uses for dimerization , that Calmodulin could be inhibiting K-Ras dimerization . | protein interactions, bioelectrochemical analysis, molecular dynamics, biochemical analysis, simulation and modeling, membrane proteins, bioassays and physiological analysis, cellular structures and organelles, research and analysis methods, chemical properties, physical chemistry, dimerization, proteins, chemistry, cell membranes, biochemistry, biochemical simulations, amperometry, cell biology, biology and life sciences, physical sciences, computational chemistry, computational biology | null |
journal.pgen.1003894 | 2,013 | Genome-Wide High-Resolution Mapping of UV-Induced Mitotic Recombination Events in Saccharomyces cerevisiae | Recombination occurs in both meiotic and mitotic cells ., In budding yeast , there are about 100 meiotic crossovers per cell 1 ., Although mitotic recombination events in S . cerevisiae are about 105-fold less frequent than meiotic exchanges 2 , homologous recombination ( HR ) is important for the repair of double-stranded DNA breaks ( DSBs ) that occur spontaneously or that are induced by DNA damage ., Yeast strains that lack HR grow more slowly than wild-type strains , and are sensitive to DNA damaging agents 3 ., In HR events in diploid cells , the broken chromosome is repaired utilizing an intact sister chromatid or homolog as a template ., Most organisms also have a pathway termed “non-homologous end-joining” ( NHEJ ) in which the broken ends are re-joined by a mechanism that does not require sequence homology ., In diploid cells of S . cerevisiae , HR is much more important than NHEJ for repair of DNA damage 4 ., We will first discuss pathways of HR , followed by a description of UV-induced DNA damage , and the recombinogenic effects of this damage ., DSBs can be repaired by a number of different HR pathways 5 ., For all of these pathways , the broken DNA ends are processed by 5′ to 3′ degradation , followed by invasion of the processed chromosome end into either a sister chromatid or a homolog ( Figure 1 ) ., In the synthesis-dependent strand annealing ( SDSA ) pathway , after strand invasion and DNA synthesis , the invading broken end is displaced and reanneals to the other broken end ., The resulting product has a region of heteroduplex DNA and mismatches within the heteroduplex can be repaired to yield a gene conversion event unassociated with a crossover ( Figure 1A ) ., Alternatively , the broken ends can both engage in pairing with the intact chromosome resulting in a double Holliday junction ( Figure 1B ) ., This structure can be resolved to yield a crossover or non-crossover ., As in the SDSA pathway , mismatches within the heteroduplex region can be repaired to generate a conversion event ., Lastly , invasion of one broken end can result in the generation of a replication structure that duplicates sequences from the other chromosome from the point of invasion to the end of the chromosome ( break-induced replication , BIR; Figure 1C ) ., One consequence of mitotic recombination is to cause loss of heterozygosity ( LOH ) for markers near the initiating lesion ( gene conversions ) or extending distal from the initiating lesion to the end of the chromosome ( crossovers and BIR events ) ., In Figure 2 , we show the repair of DSBs in diploid mitotic cells by HR involving the homolog ., In Figure 2A , we show the repair of a single broken chromatid ( G2 event ) using the homolog as a template ., The red and black colors indicate that the two homologs have single-nucleotide polymorphisms ( SNPs ) that allow the detection of recombination events ., Figure 2A shows a crossover between chromatids 2 and 3 ., If chromatids 1 and 3 segregate into one daughter cell ( D1 ) , and 2 and 4 segregate into the other ( D2 ) , a reciprocal pattern of LOH would be observed ., Segregation of unrecombined chromatids 1 and 4 into one cell and the recombined chromatids 2 and 3 into the other would not lead to LOH ., These two patterns of segregation are equally frequent in yeast 6 ., Our previous studies 2 , 7 showed that most ( 80% ) crossovers are associated with gene conversion events ( indicated by boxes in Figure 2 ) ., In Figure 2B , we show a conversion event unassociated with a crossover which produces an interstitial LOH event in one of the daughter cells ., The conversion events shown in Figure 2A and 2B are termed “3∶1” events since three of the chromatids have one type of SNP and one has the other within the boxed region ., A BIR event produces a region of LOH that extends to the telomere in one but not both daughter cells ( Figure 2C ) ., The 3∶1 conversion events shown in Figures 2A and 2B are expected from the repair of a single DSB generated in S or G2 of the cell cycle ., In addition , since the chromosome with the DSB acts as a recipient of information derived from the intact chromosome , these conversion events have the pattern expected if the recombinogenic DSB was on the black chromosome 4 ., We observed previously , however , that over half of the mitotic conversion events had a different form from that shown in Figures 2A and 2B ., In Figure 2D , we show a conversion event unassociated with a crossover in which both daughter cells have an interstitial region of LOH that is homozygous for the same SNPs; these events are called “4∶0” conversions ., We interpret 4∶0 events as resulting from the repair of two broken sister chromatids in which the DSBs are located at the same positions ., One simple mechanism to obtain this pattern of breakage is that the recombinogenic DSB is generated in G1 , the broken chromosome is replicated , and the two resulting broken chromatids are repaired in G2 ( Figure 2D ) ., The alternative model in which the DSB is generated and repaired in G1 is ruled out because such events would not be associated with LOH for markers located distal to the conversion event 8 ., If the two broken chromatids are repaired to generate conversion tracts of the same lengths , a 4∶0 event is generated ., If one conversion tract is longer than the other , repair of two broken sister chromatids can also generate hybrid 3∶1/4∶0 conversion tracts 2 , 7 ., Our previous studies indicated that most spontaneous crossovers had conversion events consistent with a G1-initiated DSB rather than a G2-initiated DSB 9 , 10 , and spontaneous events resembled those induced by gamma rays in G1-synchronized yeast cells 11 ., UV results in DNA lesions that are both mutagenic and recombinogenic ., The primary types of lesions caused by low dosages of UV-C ( ∼254 nm ) are pyrimidine dimers including cyclobutane dimers ( CPDs ) and ( 6-4 ) photoproducts ( 6-4 PPs ) 3 ., Although CPDs can be reversed in yeast by the action of photolyase , the repair of most lesions in wild-type cells likely reflects nucleotide excision repair ( NER ) ., In NER , multiple proteins act to excise a short oligonucleotide containing the damaged bases ., The resulting 30-nucleotide gap is filled in by DNA polymerase delta and/or epsilon 12 , and the remaining nick is sealed by Lig1p ., In yeast , as in many other organisms , UV-induced lesions are more quickly repaired in transcribed genes than in non-transcribed regions 3 ., Although most UV-induced lesions are removed quickly by this error-free process , a small fraction of the 30-nucleotide gaps are expanded by the action of Exo1p , resulting in large RPA-coated gaps 13 , 14 ., These RPA-coated regions recruit Mec1p/Ddc2p and the 9-1-1 complex , followed by subsequent recruitment of other components of the DNA damage checkpoint 15 ., In addition to checkpoints triggered by the action of Exo1p , if unrepaired lesions persist into the S-phase , single-stranded regions may also be generated during the re-start of blocked replication forks ., Strong activation of Mec1p by UV is observed in S-phase cells , presumably by this mechanism 16 ., Although it is clear from many previous studies that UV greatly elevates the frequency of mitotic recombination in yeast 17–23 , the recombinogenic mechanism is not well understood ., There are two types of models ., First , it is possible that the recombinogenic lesion is generated by NER ., Consistent with this model , Galli and Schiestl ( 1999 ) 20 observed that UV of G1-synchronized cells was not recombinogenic unless the cells were allowed to replicate ., They concluded that the recombinogenic lesion was likely to represent an NER-associated gap that was replicated to produce the recombination-stimulating DSB ., This model predicts that the gene conversion events associated with UV-treatment of G1-synchronized cells would be exclusively 3∶1 conversion events ( Figure 2A ) ., In a preliminary study 7 , however , we found that about half of the observed UV-induced conversions were 3∶1 events and about half were 4∶0 events ( Figure 2D ) ., This observation is inconsistent with the simplest form of the model proposed by Galli and Schiestl ., An alternative model is that the unexcised dimers and other DNA lesions are the recombinogenic lesion ., For example , replication forks stalled at an unexcised dimer may engage in replication re-start or be broken ., Although both re-start and the repair of an S-phase DSB would be expected to involve an interaction with the sister chromatid 24 , some fraction of these events could involve the homolog , resulting in LOH ., Kadyk and Hartwell ( 1993 ) 21 showed that UV stimulates recombination between both sister-chromatids and homologs in NER-proficient cells ., In rad1/rad1 ( NER-deficient ) diploids , conversions , but not crossovers , were stimulated by UV in a replication-dependent manner 21 ., One complication in interpreting this result is that Rad1p is involved with multiple recombination-related reactions 25–27 in addition to its role in NER ., Regardless of this ambiguity , it is likely that unexcised dimers are recombinogenic ., The summary of studies performed thus far is that some fraction of UV-induced recombination events reflects lesions resulting from NER and another fraction reflects unexcised dimers ., In the experiments described below , we examine mitotic crossovers and gene conversion events induced by UV in diploid cells ., In G1-synchronized cells treated with high doses of UV , most of the events reflect the repair of two broken sister chromatids whereas at low doses , most events reflect repair of a single broken chromatid ., We also show that UV induces crossovers more efficiently than BIR events ., We mapped the distribution of about 100 UV-induced LOH events selected on chromosome V and about 400 unselected LOH events throughout the genome ., We found that the unselected events were widely distributed throughout the genome with no very strong hotspots ., The ribosomal RNA gene cluster , however , was significantly “cold” for crossovers compared to the rest of the genome ., In order to determine different types of mitotic recombination and to determine whether the conversion events are of the 3∶1 or 4∶0 configuration , we used a method of identifying recombination events that allows the recovery of both daughter cells with the recombinant chromosomes ., The system used in the present study ( Figure, 3 ) is similar to that employed previously 2 , 28 ., Near the telomere of chromosome V , one homolog ( shown in black in Figure 3A ) has an insertion of SUP4-o , an ochre-suppressing tRNA gene ., The diploid is also homozygous for the ade2-1 ochre mutation ., Diploids homozygous for the ade2-1 mutation and zero , one or two copies of SUP4-o form colonies that are red , pink , and white , respectively 28 ., In most of the experiments described below , G1-synchronized diploid cells were plated and immediately irradiated with UV ., If the resulting DNA damage induces a crossover between the heterozygous SUP4-o gene and the centromere of chromosome V before the first cell division , a red/white sectored colony will be formed ( Figure 3A ) ., Since formation of a sectored colony requires a crossover , followed by the segregation pattern in which each daughter cell receives one recombined chromosome and one unrecombined chromosome ( Figures 2A and 3A ) , only half of the crossovers induced in the first division following irradiation result in LOH ., If the UV-induced DNA damage is not repaired in the first cell cycle but persists into subsequent cell cycles , a pink/white/red sectored colony could be produced ( Figure 3B ) ., As described below , most of the events induced by UV treatment in G1-synchronized cells generate a red/white sectored colony rather than a tri-colored colony ., Neither gene conversion events unassociated with a crossover nor BIR events on chromosome V result in a red/white sectored colony ., As will be shown below , such events can be detected as unselected events in cells that have a selected crossover on chromosome V . The transition between heterozygous markers and homozygous markers in the sectored colony locates the position of the crossover ., To detect the position of the selected crossover on chromosome V and to detect unselected LOH events throughout the genome , we used a diploid strain ( PG311 ) derived from mating two sequence-diverged haploid strains: W303a and YJM789 2 , 7 , 29 ., These two strains differ by about 52 , 000 SNPs ., We detect LOH using microarrays that examine 13 , 000 of these SNPs 7 , allowing mapping of most events to a resolution of about 1 kb ., Each SNP is represented by four 25-bp probes , two with W303a sequences ( Watson and Crick ) and two with YJM789 sequences ., At the hybridization temperature optimized for the whole probe set , W303a genomic DNA hybridizes strongly to W303a oligonucleotides with very weak cross-hybridization to the corresponding YJM789 oligonucleotides , and vice versa for YJM789 genomic sequences ., Genomic DNA is isolated from each sector of red/white sectored colonies , labeled with Cy5-tagged nucleotides , and competitively hybridized to the SNP microarray with genomic DNA from the untreated strain labeled with Cy3-tagged nucleotides ., By assaying the ratio of hybridization of the differentially-tagged samples to each oligonucleotide 7 , we can readily map LOH events ., The transition between heterozygous and homozygous markers should be located near the site of the recombinogenic DNA lesion ., Figure 4 shows the analysis of one red/white sectored colony ( 59RW ) ., In this figure , we show the normalized ratio of hybridization of genomic sequences to W303a- and YJM789-specific oligonucleotides on chromosome V with red lines and black lines , respectively; CEN5 is located near coordinate 152 kb ., In the top part of Figure 4A , we depict the pattern of hybridization of genomic DNA isolated from the red sector ., The ratio of hybridization is about 1 for all SNPs from coordinate 105 kb to the right telomere , indicating that SNPs in this region are heterozygous ., In the red sector , SNPs centromere-distal to coordinate 105 kb on the left arm are homozygous for the W303a-derived SNPs whereas the genomic DNA from the white sector becomes homozygous at approximately the same position for YJM789-derived SNPs ., In Figure 4B , the same recombination event is depicted at higher resolution; each square and diamond shows the level of hybridization to an individual YJM789-specific or a W303a-specific SNP , respectively ., As shown in this figure , the red sector has a single transition between heterozygous and homozygous SNPs whereas the white sector has three transitions ., The pattern of these transitions indicates that the crossover is associated with a 3∶1/4∶0 hybrid conversion tract ., Most of our experiments involve UV treatment of G1-synchronized cells with 15 J/m2; the experimental parameters used for each experiment are in Table S1 ., PG311 is hemizygous at the MAT locus ( MATa/MATα::NAT ) , allowing its synchronization in G1 using the alpha pheromone 11 ., The synchronized cells were plated onto solid medium and immediately irradiated at doses varying between 1 and 15 J/m2 ., Even at the maximum dose of UV , cell viability was 70% ., No sectored colonies were observed in cells that were not treated with UV ., Based on our earlier study of spontaneous crossovers in the same strain 2 , the rate of crossovers in untreated cells is 1 . 1×10−6/division in the 120 kb interval between CEN5 and the SUP4-o marker ., Relative to this rate , UV treatment stimulated sector formation by factors of 1500 ( 1 J/m2 ) , 1600 ( 5 J/m2 ) , 5000 ( 10 J/m2 ) , and 8500 ( 15 J/m2 ) ., The strong stimulation of mitotic crossovers by UV is consistent with previous studies 23 ., In some studies 2 , 4 , 28 , the frequency of mitotic recombination events is higher in diploids that express both mating types than in diploids that express only one mating type ., Consequently , we compared the frequency of red/white sectored colonies in G1-synchronized cultures of PG311 and PSL101 ( the MATa/MATα progenitor of PG311 ) ., Because PSL101 cannot be synchronized in G1 using alpha pheromone , both strains were synchronized in G1 by growing the cells into stationary phase ( Text S1 ) ., After treatment of the G1-synchronized cells with 15 J/m2 of UV , 0 . 4% ( 0 . 2–0 . 9% , 95% confidence limits ) of the PG311 colonies formed red/white sectors compared to 0 . 6% ( 0 . 4–1% ) of the PSL101 colonies ., Although the confidence limits are wide , these results indicate that mating type heterozygosity does not have a large effect on the frequency of UV-induced mitotic crossovers in our system ., In addition to red/white sectored colonies , in the irradiated samples , we also observed pink/red and pink/white/red colonies ., Such colonies could represent non-reciprocal recombination events ( for example , BIR events ) , persistence of recombinogenic DNA damage beyond the first cell cycle , or an artifact ( two closely-located independent cells ) ., To exclude sectors formed artifactually , we micromanipulated individual G1-irradiated ( 15 J/m2 dose ) single cells to specific positions on plates with solid medium , and monitored their subsequent development to form sectored or unsectored colonies ., From a total of 970 isolated irradiated single cells , we observed eleven sectored colonies of the following types: seven red/white colonies , two pink/red colonies , and two pink/white/red colonies ., From our SNP microarray analysis of the LOH patterns on chromosome V in these colonies ( described in Text S1 and Figure S1 ) , we found that all seven of the red/white colonies represented crossovers induced during the first cell cycle ., The two pink/red sectored colonies reflected chromosome loss , resulting in a monosomic red sector and a pink sector ., Only one of the pink/white/red colonies was a consequence of a UV-induced recombination event in the second division ( Figure 3B , Figures S1 and S2 ) ., In summary , of the nine sectored colonies in which sectoring reflected a UV-induced crossover , eight occurred prior to the first cell division and only one occurred after the first cell division , indicating that most UV-induced DNA lesions are rapidly repaired ., We used SNP microarrays to analyze 47 sectored colonies of G1-synchronized cells treated with 5 , 10 or 15 J/m2 of UV ( Tables S2 and S3 ) ., 80% of the colonies were from cells treated with 15 J/m2 ., Nine of these colonies were derived from the single-cell experiments described above ., 45 of the 47 sectored colonies examined had patterns of LOH on chromosome V consistent with a reciprocal crossover on the left arm of chromosome V . In one of the two exceptional colonies , there was a loss of one copy of chromosome V . In the other colony , there were two independent conversions that resulted in LOH events that were unassociated with a crossover ., These two sectored colonies were not used in our subsequent analysis of selected events on chromosome V , although data from these colonies were used to analyze unselected recombination events ., In addition to the selected LOH events on chromosome V , we observed an average of eight unselected LOH events per sectored colony ., As described below , our analysis of the 45 selected and 381 unselected events ( 300 gene conversion events unassociated with crossovers , 60 crossovers , and 21 BIR events ) allowed us to determine several important features of the UV-induced recombination events:, 1 ) the patterns of gene conversion in selected and unselected recombination events ,, 2 ) the lengths of gene conversion tracts associated or unassociated with crossovers , and, 3 ) the locations of selected and unselected recombination events induced by UV ., Since the frequency of selected sectored colonies in cells irradiated with 15 J/m2 was about 1% , and the selected interval on chromosome V is about 1% of the genome , we expect about one unselected crossover per irradiated cell , roughly the observed frequency ( 60 unselected crossovers/47 sectored colonies ) ., One interpretation of our observation of frequent DSCBs in G1-irradiated cells is that the repair of two very closely-spaced single-stranded DNA lesions induced by 15 J/m2 results in DSCBs in the G1-synchronized cells , whereas SCBs reflect DNA lesions on one strand ., Thus , the productions of DSCBs by this mechanism would be proportional to the square of UV dosage , whereas the frequency of SCBs would be linearly proportional to the UV dosage ., By this model ( details to be discussed below ) , one might expect that a low dose of UV should have a relatively higher frequency of SCBs ., Consequently , we examined the frequency and types of recombination events induced in G1-synchronized cells by 1 J/m2 ., As expected , the frequency of red/white sectored colonies was reduced in cells irradiated with 1 J/m2 relative to cells irradiated with 15 J/m2 ( 1 . 6×10−3/division versus 9 . 4×10−3/division ) ., Ten sectored colonies were examined by whole-genome microarrays ., Only four unselected events were observed ., This frequency ( 0 . 4 events/sectored colony ) was about twenty-fold less than that observed in samples irradiated with 15 J/m2 ( 8 events/sectored colony ) ., Consequently , in the additional thirty-six sectored colonies examined , we used microarrays specific for detecting LOH on chromosome V . The depictions of the LOH events in the 1 J/m2 irradiated samples that had the same patterns as observed for the 15 J/m2 samples are shown in Table S2; the numbers of samples with specific classes of events are shown in parentheses in this table ., Patterns of LOH that were unique to the 1 J/m2 samples are shown in Table S6 ., The coordinates for these LOH events are shown in Table S7 ., The distribution of the LOH events on chromosome V for the 1 J/m2 samples was not significantly different from that observed for the 15 J/m2 samples or the spontaneous events using the same “binning” procedure and statistical test described above ., The median length of conversion events associated with crossovers on chromosome V in cells irradiated with 1 J/m2 was 4 . 3 kb ( 2 . 3 kb–8 . 2 kb; 95% confidence limits ) kb ., In cells irradiated with 15 J/m2 , the median length of conversion tracts associated with crossovers on chromosome V was 6 . 7 ( 4 . 2–13 kb ) ., The distributions of tract lengths analyzed by the Mann-Whitney test showed that these distributions were not significantly different ( p\u200a=\u200a0 . 12 ) ., A striking difference was observed in the distributions of events diagnostic of SCBs and DSCBs in cells irradiated with 1 and 15 J/m2 of UV ., Of selected events on chromosome V in cells irradiated with 1 J/m2 , we observed 5 crossovers unassociated with conversion , 31 SCB events , and 10 DSCB events ., In contrast , in cells irradiated with 15 J/m2 , most of the selected events on chromosome V were DSCB events ( Figure 8 ) ., By the Fisher exact test , the difference in the numbers of SCB and DSCB events induced by the two different UV treatments is very significant ( p<0 . 0001 ) ., The conclusion that G1-synchronized cells have different recombinogenic DNA lesions induced by different UV doses will be discussed further below ., As noted in previous studies , UV very effectively induces mitotic recombination in yeast 7 , 10 , 20 , 21 , 23 , 38 ., In experiments involving heteroallelic recombination in synchronized cells , UV is somewhat more recombinogenic in G1-synchronized cells than in G2-synchronized cells 18 , 21; our results support these observations ., Kadyk and Hartwell ( 1992 ) 24 concluded that DSBs induced by X-rays in G2-synchronized cells were repaired primarily by sister-chromatid recombination , whereas X-ray treatment of G1-synchronized cells effectively stimulated recombination between homologs ., Our previous interpretation of both spontaneous and DNA damage-induced crossovers is also consistent with this conclusion 2 , 7 , 29 ., We argue that most spontaneous crossovers between homologs are initiated by a DSB in G1 in one chromosome , and replication of the broken chromosome produces two sister chromatids broken at the same position ., Since these lesions cannot be repaired by sister-chromatid recombination , they are repaired by recombination with the homolog ., Although it is likely that DSBs formed in S or G2 are primarily repaired by sister-chromatid recombination , some DSBs generated in G2 are repaired by interaction with the homolog 11 ., As observed in our previous studies 2 , 7 , 29 , the mitotic conversion tracts are long compared to those observed in meiosis , and the tracts associated with crossovers are longer than the tracts unassociated with crossovers ., Most of the conversion events are explicable as a consequence of repair of one broken chromatid or two sister chromatids broken at the same position by the standard HR pathways shown in Figure 1 , with only conversion-type MMR and not restoration-type MMR ., About 15% of the conversion events , however , are more complex , requiring “patchy” repair of mismatches within a heteroduplex ( mismatches corrected by both conversion-type repair and restoration-type repair within one heteroduplex ) , and/or branch migration of the Holliday junction ., The fraction of complex conversion tracts in the current study is similar to those observed in our previous studies 7 , 29 ., Although these events ( described in detail in Text S1 and Figures S5 , S6 , S7 , S8 , S9 , S10 , S11 , S12 , S13 , S14 , S15 ) are explicable by modifications of the standard models shown in Figure 1 , it is possible that some of these conversion events involve a substantially different mechanism such as multiple template switching events during BIR ., In this context , template switching during BIR has been observed in experiments in which linear DNA fragments are transformed into yeast 39 ., In addition to the complex tracts , it is possible that the very long conversion tracts reflect BIR rather than mismatch repair in a heteroduplex; 16% of the conversion events unassociated with crossovers are greater than 10 kb in length , and the longest exceeds 50 kb ., Finally , it should be pointed that , although single BIR events would not be expected to generate crossovers , a model for production of a crossover by a double BIR event is shown in Figure S4 of Lee et al . ( 2009 ) 2 ., A central issue is the nature of the recombinogenic DNA damage generated by UV ., Based on the mechanism of NER and on the observation that unrepaired pyrimidine dimers block replication , there are two obvious potential sources of DSBs 40 ., First , if a DNA molecule with an unrepaired gap resulting from NER is replicated before filling-in of the gap and ligation , the net result would be a pair of sister chromatids with a single DSB ( Figure 9A ) ., Alternatively , if a replication fork encounters an unrepaired UV-induced lesion , breakage of the fork could also result in a single broken chromatid ( Figure 9B ) ., Based on the observation that UV treatment of G1- or G2-synchronized cells was not recombinogenic unless cells were allowed to divide , Galli and Schiestl ( 1999 ) 20 suggested that cell division was required to convert DNA lesions to recombinogenic lesions , consistent with both of the possibilities described above; their assay detected only intrachromatid deletions ., Kadyk and Hartwell ( 1993 ) 21 found that unrepaired UV lesions stimulate gene conversion events between homologs , but have little effect on mitotic crossovers ., This conclusion may be affected by the use of the rad1 mutation to prevent dimer excision , since rad1 strains have reduced frequencies of crossovers in some assays 41 ., Both of the models discussed above predict that UV-induced DNA damage in G1-synchronized cells would produce primarily gene conversion events involving a single broken chromatid ( SCBs ) ., In our study , about two-thirds of the conversion events in which cells were irradiated with 15 J/m2 reflect two broken sister chromatids , but only one-quarter of the conversions reflect two broken sister chromatids in cells irradiated with 1 J/m2 ( Figure 8 ) ., Thus , there is a qualitative change in the nature of the DNA lesion with increasing UV dose ., In addition , since our single-cell experiments demonstrate that UV-induced lesions are recombinogenic during the first division following treatment , DSCBs cannot be explained as reflecting the segregation of a chromosome with an unrepaired G2-associated DSB from the previous division ., We suggest that most DSCBs are a consequence of a DSB in G1 ., Although UV damage is generally regarded as an agent that produces DNA nicks rather than DSBs , a gel-based detection of the conversion of a circular chromosome to a linear chromosome indicated that a dose of 40 J/m2 produces 5 to 10 DSBs in G2-synchronized cells 42 ., There are several related mechanisms by which NER could produce a DSB in G1 cells ., First , the excision tracts resulting from removal of two closely-opposed dimers could result in very short ( <6 bp ) unstable duplex regions between the repair tracts , resulting in a DSB ( Figure 9C ) ., A second model is that , following the removal of two closely-opposed dimers by NER , one or both of the resulting short gaps is expanded by Exo1p ( Figure 9D ) ., A third related model is that the excision tract generated by NER is expanded into a large single-stranded gap that is cleaved by an endonuclease to yield the DSB ( Figure 9E ) ., Based on our results and those of others , it is likely that UV produces a variety of recombinogenic lesions ., In our experiments , at a low dose of UV ( 1 J/m2 ) , we observed primarily SCBs , consistent with the two models shown in Figures 9A and 9B ., At 15 J/m2 , we observed DSCBs more frequently than SCBs ., This observation supports models shown in Figure 9C and 9D that require closely-opposed lesions , and argues against the model shown in Figure 9E in which the relative fraction of DSCB and SCB events would be expected to be independent of the density of the NER tracts ., It can be calculated that diploid cell irradiated with 15 J/m2 have about 7500 dimers/genome 43; if these dimers are distributed randomly , we expect about 35 closely-opposed ( separated by ≤75 bases ) dimers , enough to explain the detected DSCB events ., It is likely that the number of closely-spaced dimers is greater than that determined by this calculation ., Lam and Reynolds ( 1987 ) 43 found that the fraction of dimers located within 15 base pairs of each other is greater than expected from a random distribution , and this fraction is somewhat independent of UV dose ., These dimers may be responsible for the DSCB events detected in strains treated with the 1 J/m2 UV dose ., In summary , we suggest that low doses of UV primarily result in SCBs as a consequence of replication of a chromosome with a NER-generated DNA gap in one strand , or an unrepaired dimer resulting in breakage of one arm of the replication fork ., In contrast , we suggest that high doses of UV often result in DSCB events as a consequence of a G1-generated DSB , reflecting cellular enzymes acting on closely-opposed dimers ., Although this explanation seems straightforward , we cannot exclude more complex explanations of our data ., For example , it is possible that the very large number of UV-induced lesions at high doses may overwhelm the DNA repair systems , resulting in changes in the use of repair pathways ., In addition , we stress that our analysis based on interhomolog recombination does not yield an estimate of the relative frequencies of UV-induced recombinogenic lesions produced in G1 , S , and G2 , since most recombinogenic lesions produced in S and G2 are likely repaired by sister-chromatid recombination 24 , a mechanism that does not lead to LOH 7 ., The UV-induced recombination events were broadly distributed thro | Introduction, Results, Discussion, Materials and Methods | In the yeast Saccharomyces cerevisiae and most other eukaryotes , mitotic recombination is important for the repair of double-stranded DNA breaks ( DSBs ) ., Mitotic recombination between homologous chromosomes can result in loss of heterozygosity ( LOH ) ., In this study , LOH events induced by ultraviolet ( UV ) light are mapped throughout the genome to a resolution of about 1 kb using single-nucleotide polymorphism ( SNP ) microarrays ., UV doses that have little effect on the viability of diploid cells stimulate crossovers more than 1000-fold in wild-type cells ., In addition , UV stimulates recombination in G1-synchronized cells about 10-fold more efficiently than in G2-synchronized cells ., Importantly , at high doses of UV , most conversion events reflect the repair of two sister chromatids that are broken at approximately the same position whereas at low doses , most conversion events reflect the repair of a single broken chromatid ., Genome-wide mapping of about 380 unselected crossovers , break-induced replication ( BIR ) events , and gene conversions shows that UV-induced recombination events occur throughout the genome without pronounced hotspots , although the ribosomal RNA gene cluster has a significantly lower frequency of crossovers . | Nearly every living organism has to cope with DNA damage caused by ultraviolet ( UV ) exposure from the sun ., UV causes various types of DNA damage ., Defects in the repair of these DNA lesions are associated with the human disease xeroderma pigmentosum , one symptom of which is predisposition to skin cancer ., The DNA damage introduced by UV stimulates recombination and , in this study , we characterize the resulting recombination events at high resolution throughout the yeast genome ., At high UV doses , we show that most recombination events reflect the repair of two sister chromatids broken at the same position , indicating that UV can cause double-stranded DNA breaks ., At lower doses of UV , most events involve the repair of a single broken chromatid ., Our mapping of events also demonstrates that certain regions of the yeast genome are relatively resistant to UV-induced recombination ., Finally , we show that most UV-induced DNA lesions are repaired during the first cell cycle , and do not lead to recombination in subsequent cycles . | null | null |
journal.ppat.1000093 | 2,008 | Quorum Sensing Coordinates Brute Force and Stealth Modes of Infection in the Plant Pathogen Pectobacterium atrosepticum | Quorum sensing ( QS ) is a population density-dependent regulatory mechanism , utilising freely diffusible chemical signal molecules , which controls a wide range of phenotypes in many different bacteria 1 ., The best-studied QS systems are those utilising N-acyl-homoserine lactone ( AHL ) signal molecules , synthesised by LuxI homologues ., AHL concentration increases with bacterial population growth until , at high cell density , a threshold level of signal is reached ., This is detected by AHL binding to receptor proteins , LuxR-family transcriptional regulators , resulting in altered gene expression 2 ., QS plays an essential role in the pathogenesis of many bacterial pathogens of both plants and animals ., Amongst the best studied AHL QS systems are those of the soft rotting enterobacterial plant pathogens Pectobacterium atrosepticum ( Pba ) and Pectobacterium carotovorum subsp ., carotovorum ( Pcc; formerly Erwinia carotovora subsp . atroseptica and E . c . subsp . carotovorum respectively ) 3 ., These pathogens cause disease primarily through the coordinate and prolific production of a variety of plant cell wall degrading enzymes ( PCWDEs ) , which are secreted to the extracellular environment through the Type I ( protease ) and Type II ( pectinases and cellulases ) secretion systems 4 ., However , they also possess a Type III secretion system ( T3SS ) with cognate effector ( DspA/E ) and helper/harpin proteins ( HrpN/HrpW ) , which is required for full virulence 5 ., While the role of the T3SS in the soft rotting pathogens remains to be elucidated , in the closely-related E . amylovora , DspA/E has been reported to interact with leucine-rich repeat receptor-like protein kinases ( LLR-RLKs ) of apple plants , implying a role in the manipulation of host defences 6 ., QS in pectobacteria has been reported to regulate PCWDEs 7 , the Type III secreted harpin HrpN 8 , and other virulence factors , including Nip and Svx 9–11 , a very small number of virulence regulators ( expR , rsmA and virR ) 12–14 , and the antibiotic carbapenem 15 ., These are controlled by the AHL , N- ( 3-oxohexanoyl ) -L-homoserine lactone ( OHHL ) , synthesised by ExpI ., Different strains of pectobacteria possess up to three homologues of LuxR 16 including: VirR , which plays a central role in the repression of QS-regulated virulence factors 12; CarR , which regulates the production of carbapenem 15; and ExpR , which activates transcription of the global repressor , rsmA , in the absence of AHL 13 ., Until now , studies on QS in pectobacteria have largely been in vitro and have examined its role in the regulation of targeted virulence factors , particularly PCWDEs ., Such virulence factors are thought to operate as part of a necrotrophic mode of action ( where the invading organism causes death of host tissue and colonises dead substrate ) ., As a consequence , this group of pathogens have been termed “brute force” in line with this physical attack on plant cell walls ., This is in contrast to pathogens such as Pseudomonas syringae , which are hemibiotrophic ( requiring living host tissue as part of the infection process , during which they actively manipulate host defences ) and , due to their ability to manipulate plant defences as part of the infection process , have been termed “stealth” pathogens ., In the pectobacteria , it has been hypothesised that QS acts to delay the onset of PCWDE production until sufficient numbers of cells are present to overcome plant defences , which are induced by the formation of cell wall breakdown products 17 , 18 ., However , in previous work we showed that premature addition of OHHL to potato plants infected with low numbers of Pba induced early disease development 19 , suggesting that this hypothesis may be an over-simplification of a more complex process ., In addition , the full genome sequence of Pba strain SCRI1043 ( Pba1043 ) has revealed many additional putative virulence determinants , including coronafacoyl-amide conjugates and homologues of the hemolysin-co-regulated protein ( Hcp ) and Rhs accessory element VgrG 20 ., In Pseudomonas syringae , coronafacoyl-amide conjugates promote disease development and , together with the T3SS , may act to suppress salicylic acid-based defences as part of this process 21 , 22 ., Hcp and VgrG have been associated with virulence in animal pathogens and are potential effector proteins delivered through a Type VI secretion system ( T6SS ) 23–26 ., Hcp and VgrG homologues were recently detected in the secretome of Pba1043 and over-expression of hcp1 increased Pba virulence , suggesting that this and other hcp family members are virulence determinants 27 ., The presence of such determinants in Pba suggests that the pectobacteria may also act in a stealth-like manner by manipulating resistance during the infection process ., However , whether these determinants are produced and act independently , or together with PCWDEs as part of a coordinated assault on the plant , is unknown ., We developed a whole genome microarray for Pba1043 and report its use to study gene expression from an expI mutant of Pba1043 grown in planta , to determine global effects of QS on gene regulation during potato infection , with particular emphasis on the relationship between PCWDEs and possible stealth mechanisms ., The expI gene and ExpI product , OHHL , are required for full virulence in Pba and Pcc 7 , 12 , 28 ., The virulence of an expI ( ECA0105 ) mutant was significantly reduced on both potato stems and tubers and was restored following complementation with the expI gene in trans ( Fig . S1 ) ., To confirm that virulence could be restored in planta by the presence of OHHL , the expI mutant strain was inoculated at low cell densities into an OHHL-producing transgenic potato plant 19 , where virulence was restored compared to inoculation on a non-transgenic control plant ( Fig . 1A ) ., This supports previous work where the presence of OHHL in these transgenic plants induced early disease development from low cell densities ( 102 cells per inoculation site ) of both WT and expI mutant strains 19 ., A luminescence-based assay was used to monitor OHHL production during growth of the expI mutant and wild type strains in potato tubers ( Fig . 1B ) ., Both strains grew at comparable rates over a 120 h infection time course and reached similar population levels ., Although the wild type and expI mutant strains would be expected to show differences in growth in natural condition during the course of disease development , the relatively short infection time ( 120 h ) and the method of inoculation ( onto the cut tuber surface ) , may account for the results observed ., In the wild type , the level of OHHL rose sharply over the first 16 hours in line with log phase growth , before reaching a plateau at a concentration of approximately 80 µg/ml ., At this plateau , bacterial cell density was approximately 5 . 0×106 cfu/ml , which was similar to previous reports 15 ., In the expI mutant , OHHL production remained at background levels ., Based on the above data , time points at 12 and 20 hours post inoculation ( hpi ) , i . e . just prior to and just following maximum OHHL synthesis in planta , were selected to study transcriptional changes during QS ( Fig . 1B ) ., Differential expression of genes ( pelA ECA4067 , pelC ECA4069 , celV ECA1981 , prtW ECA2785 , pehA ECA1095 , ECA2220 , svx ECA0931 and nip ECA3087 ) previously shown to be under QS control 9 , 10 , 12 was investigated using quantitative real-time PCR ( qRT-PCR ) at 12 and 20 hpi in the expI mutant and wild type strains ., In all cases , significant up-regulation of these genes was observed in the wild type only ( Table S1 ) ., cDNA from the wild type and expI mutant at 12 and 20 hpi was hybridised to the Pba microarray ., 1167 coding sequences ( CDSs ) ( approx . 26% of the genome ) showed statistically significant differences ( P≤0 . 05 ) in expression between the expI mutant and wild type ( Table S2 ) ., 498 CDSs showed reduced transcript abundance ( 421 at 12 hpi , 169 at 20 hpi , 92 at both time points ) and 687 CDSs exhibited increased transcript abundance ( 551 at 12 hpi , 180 at 20 hpi , 44 at both time points ) in the expI mutant compared to the wild type ., Microarray comparison of mutant and wild type cDNAs from cells in buffer solution prepared for tuber inoculation following overnight growth in LB to stationary phase ( zero time-point ) , was consistent with there being no overall transcriptional difference ( P≤0 . 05 ) between the strains prior to plant inoculation ( data not shown ) ., Only 16% of CDSs within the horizontally-acquired islands 20 showed differential gene expression , suggesting that such CDSs are less likely to have been incorporated into the QS regulon than those on the chromosome backbone ., qRT-PCR was used to study a number of genes in the expI mutant and wild type to examine differential gene expression , either to verify changes observed in the microarray or to examine the effects of a mutation in expI on additional genes ( Table S1 ) ., Importantly , qRT-PCR analysis of selected genes after growth of the expI mutant and wild type in vitro revealed the same pattern of expI-dependence as observed in vivo , and these changes could be fully complemented by the addition of exogenous OHHL ( Fig . 2 ) ., The role of QS in pathogenesis of pectobacteria has been intensively studied in vitro , particularly for its ability to co-ordinately up-regulate PCWDEs 7–10 , 12 , 14 , 28 ., Previous work based on enzyme plate assays observed that all major groups of PCWDEs , including pectate lyases ( Pel ) , cellulases ( Cel ) , protease ( Prt ) , pectin lyase ( Pnl ) polygalacturonase ( Peh ) and pectin methyl esterase ( Pme ) were under QS control 8 ., In this study , we found that genes encoding all these groups showed lower transcript abundance in the expI mutant compared to the wild type at both 12 and 20 hpi ( Table 1 , Table S1 ) ., The major pectate lyases PelA , PelB ( ECA4068 ) and PelC ( ECA4069 ) , as well as CelV ( ECA1981 ) , a putative cellulase ECA2220 , PrtW ( ECA2785 ) and PehA ( ECA1095 ) have previously been associated with QS in pectobacteria , either through transcriptional or proteomic analyses 7 , 10 , 12 , 28 ., However , in addition the transcription ( using microarray and/or qRT-PCR analyses ) of genes encoding other PCWDEs and their isoforms , including “minor” pectate lyases ( PelZ ECA4070 , Pel-3 ECA1094 , PelB and PelW ECA2402 ) , CelB ( ECA2827 ) and CelH ( ECA3646 ) , PehN ( ECA1190 ) , PmeB ( ECA0107 ) and Pnl ( ECA1499 ) was found to be expI-dependent ., These results confirm previous observations of QS regulatory control in vitro and validate our in planta approach ., Other genes previously shown to fall under QS control in vitro , including svx , nip and a gene of unknown function ( ECA3946 ) , as well as three regulators ( expR ECA0106 , rsmA ECA3366 and virR ECA1561 ) involved in the production of PCWDEs 9 , 10 , 12–14 , also showed reduced transcript abundance in the expI mutant compared to the wild type strain ., This again justifies our approach in assessing the genome-wide effects of QS regulation during the potato interaction ., While 1167 genes , representing a variety of processes , were found to be differentially expressed in the microarray experiment ( Table S2 ) , we focus predominantly on those that display reduced transcript levels in the mutant ( as these are presumably induced directly or indirectly by QS ) , and which also have a known or putative role in virulence ( Table 1 ) ., To successfully cause disease Pba must secrete a multitude of PCWDEs and other proteins , many of which are under QS control ., We observed that both Type I and Type II secretion systems ( T1SS and T2SS , respectively ) , which can be considered as ‘accessory virulence factors’ , are modulated by QS ( Table 1 ) ., Prior to this study the secretion systems responsible for the delivery of these virulence factors had not been reported as QS-regulated , and this observation indicates a novel facet to QS control of pathogenesis in pectobacteria ., The T2SS is well characterised in pectobacteria and is responsible for secretion of many key virulence factors , e . g . Pel , Cel and Svx 10 ., The T2SS of pectobacteria is encoded by a cluster of 15 out genes ( ECA3098-3110 and ECA3113-3114 ) 20 , 29 , of which six ( outMLHGFD ) ( by microarray analysis ) exhibited reduced transcript abundance levels in the expI mutant ( Table 1 ) ., Analysis of these and seven other out genes by qRT-PCR confirmed that all were expressed at a lower level in the expI mutant ( Table S1 ) , implying that expression of the Out T2SS is up-regulated by QS in vivo ., Similar QS modulation of out expression was also demonstrated by both qRT-PCR ( Table S1 ) and the use of an outD-gusA reporter fusion in vitro ( Fig . 3 ) ., In the latter experiment , expression of outD ( ECA3109 ) was reduced in the expI mutant and restored to wild type levels by the exogenous addition of OHHL , confirming QS modulation of out gene expression ., Regulation of the major secreted protease , PrtW , is QS-dependent in Pba 10 ., Secretion of Prt by the PrtDEF T1SS is well-characterised in Dickeya dadantii ( formerly Erwinia chrysanthemi ) 30 and , by analogy , PrtW is expected to be secreted by the T1SS encoded by the neighbouring prtDEF ( ECA 2781-2783 ) genes in Pba ., To support this , the microarray data indicated that transcription of the T1SS genes prtDF was reduced in the expI mutant ., Use of qRT-PCR confirmed that expression of all three T1SS genes , prtDEF , was significantly reduced in the expI mutant compared to the wild type ( Table S1 ) ., QS-dependence of the T1SS and T2SS is a logical accompaniment to the simultaneous QS-dependent induction of their substrates , presumably allowing the systems to cope efficiently with the greatly increased quantity of these substrates ., Examples of QS-modulated secretion have been reported previously in other pathogens , e . g . the Xcp T2SS of Pseudomonas aeruginosa and the Lip T1SS of Serratia marcescens 31 , 32 , although this is the first time that QS-dependant secretion systems have been described in pectobacteria ., As well as physically attacking the plant cell wall through the action of PCWDEs , in Pba1043 the Type III secretion system ( T3SS ) is also necessary for full virulence 5 ., The T3SS is found in many Gram-negative pathogens of both animals and plants and is used to translocate effector proteins into host cells , where they manipulate host defences ., Helper proteins ( or harpins ) are secreted to the extracellular environment , and may assist in effector translocation 33 ., We observed that expression of the T3SS structural , putative effector and helper genes , and Type III-associated regulators were all modulated by QS ., In Pba1043 , and other pectobacteria , the T3SS is encoded by the hrp cluster , composed of around 40 CDSs ., These CDSs encode components of the structural apparatus , as well as the putative effector DspA/E ECA2113 , and helpers HrpN ECA2103 and HrpW ECA2112 ., The Pba1043 hrp cluster also contains a group of CDSs ( ECA2104-ECA2110 ) , which includes a number of lipoproteins , that appear to be absent in closely-related species 5 ., ECA2104 shows homology to vgrG and is described below ., In the microarray experiment , two CDSs hrpE ECA2097 , associated with the Type III structural apparatus , and a putative lipoprotein ( ECA2108 ) exhibited decreased transcript abundance in the expI mutant ( Table 1 ) ., qRT-PCR analysis of these and an additional 17 CDSs subsequently confirmed that CDSs encoding the Type III structural apparatus , the putative effector dspE , helpers hrpN and hrpW , regulators hrpL , hrpS and hrpY , and all CDSs between ECA2104 and ECA2110 were significantly reduced in the expI mutant compared to the wild type , predominantly at 12 h ( Table S1 ) ., Either positive or negative QS regulation of the T3SS has been observed in other pathogens , e . g . Pseudomonas aeruginosa 34 , Vibrio harveyi 35 , enteropathogenic E . coli 36 , Ralstonia solanacearum 37 , and QS regulation of hrpN has been shown in Pba 8 ., However , this is the first published evidence that QS plays a role in regulating the entire T3SS and its effectors in the enterobacterial plant pathogens , indicating that co-ordinated physical ( PCWDEs ) and stealth ( T3SS ) attacks may be necessary for successful disease development ., Recently , a novel T6SS was described and implicated in pathogenicity in Vibrio cholerae and P . aeruginosa 23 , 38 ., In V . cholerae , the system is encoded by the VAS locus , genes VCA0107-VCA0123 ., This locus is one member of a group of conserved gene clusters that are conserved in several pathogens ., In both V . cholerae and P . aeruginosa , the T6SS is required for secretion of HcpA and VgrG proteins , although whether these represent putative effectors or simply secreted components of the secretion machinery is not yet clear 23 , 38 ., In Pba1043 , the locus ECA3445–ECA3427 is predicted to encode a VAS-like T6SS and its putative substrates , since these genes encode proteins very similar to those encoded by VCA0107-VCA0123 and is similarly arranged on the chromosome ., Microarray analysis indicated that 11 of the 18 genes were expressed at significantly lower levels in the expI mutant ( Table 1 ) , and so transcription of the T6SS also appears to be QS-dependent ., The modulated genes included Pba homologues of VCA0120 , VCA0116 and VCA0110 , which are required for Type VI secretion in V . cholerae and/or P . aeruginosa 23 , 38 ., Moreover , the expression of several predicted T6SS substrates , i . e . encoded by hcpA and vgrG-like genes , was also found to be QS-dependent ( Table 1 ) ., There are seven hcpA homologues in Pba , three of which ( ECA4275hcp1 , ECA3428hcp2 and ECA2866hcp3 ) are highly similar 20 , 27 ., ECA3428 and ECA4275 are sufficiently similar that it was not possible to design probes specific to each locus ., Nevertheless , the probe detecting expression of both these genes showed decreased transcript abundance in expression in the expI mutant , indicating QS-dependent regulation ., Expression of ECA2866 and four other homologues ( ECA0456hcp4 , ECA3672 , ECA0176 and ECA4277 ) was also decreased ( Table 1 ) ., A combination of microarray analysis and qRT-PCR indicated reduced transcript abundance in the expI mutant of all five vgrG homologues , ECA2867 , ECA3427 , ECA2104 , ECA4142 and ECA4276 in the Pba1043 genome ( Table S1 ) ., Previous work showed that Hcp1-4 and a VgrG homologue ( ECA3427 ) were found in the secretome of Pba1043 ., Over-expression of Hcp1 increased virulence , suggesting that this and related proteins are virulence factors in pectobacteria 27 ., The VAS-like T6SS genes , ECA3445-ECA3427 , appear to constitute an operon that may extend for a further seven CDSs ( ECA3426-ECA3420 ) ., Of these , expression of six was reduced in the expI mutant ( Table S2 ) , raising the possibility that they may encode T6SS-dependent effectors ., As the T6SS is clearly important for virulence in other pathogens , and a predicted substrate ( Hcp1 ) affects virulence in Pba , we investigated whether the putative T6SS plays a role in virulence in Pba ., Mutants in ECA3438 and ECA3444 , when tested in potato stem and tuber virulence assays , both showed significantly reduced virulence compared with the wild type ( Fig . 4 ) ., In tuber tests , complementation of the mutants in trans was shown to return virulence to wild type levels ( Fig . 4B ) ., Our results indicate , for the first time in any pathogen , a role for QS in the regulation of the T6SS and its putative substrates ., It also demonstrates that the T6SS in Pba plays a role in pathogenesis , which appears to act in conjunction with PCWDE , the T3SS and other virulence determinants during the QS process ., Microarray analysis revealed the QS-dependent differential expression of at least 79 CDSs with either known or putative regulatory functions ( Table S2 ) ., Twelve CDSs , five of which showed enhanced ( hexA ECA3030 , kdgR ECA2425 , phoPpehR ECA2445 , rdgA ECA2435 and rsmA ) and seven of which showed reduced ( aepA ECA1022 , expA ECA2882 , expR , hexY ECA0809 , hor ECA1931 , rexZ ECA4123 and virR ) transcript abundance in the expI mutant , are known to regulate PCWDEs production and are required for full virulence in pectobacteria 4 , 12–14 , 39 ( Table 1 ) ., However , only three ( expR , rsmA and virR ) have previously been shown , in vitro , to fall under QS control 12–14 ., Three CDSs ( hrpL ECA2087 , hrpY ECA2089 and hrpS ECA2090 ) involved in the regulation of the T3SS in pectobacteria and other phytopathogens 40 also showed decreased transcript abundance in the expI mutant ., As all 15 of these CDSs are QS-dependent , this places QS at the apex of a regulatory hierarchy controlling both PCWDEs and the T3SS with its cognate effector proteins ., Other QS-controlled regulators are also likely to be important during interaction with the plant ( see below ) ., Although QS is central to pathogenesis , elucidating the hierarchical relationships between “subordinate” regulators presents a particular challenge due to the lack of data on such relationships in this particular strain ., Several virulence regulators in pectobacteria are known to operate though the Rsm system , which plays a major role in controlling virulence 14 ., While not investigated as part of this work , it is highly likely that at least some of the regulators identified in this study operate through this system ., Nevertheless , we have still been able to add considerable new information to existing regulatory models 41 and propose an extended model for virulence in the pectobacteria ( Fig . 5 ) ., In addition to regulators previously characterised in pectobacteria , differential expression of 18 further CDSs were found that are similar to a diverse range of transcriptional regulators in other bacteria ( Table S2 ) ., These include CDSs with putative regulatory functions in nitrogen signal transduction and assimilation ( citB ECA2578 , glnB ECA3254 , nac ECA4483 ) , hydrogenase activity ( hypA ECA1235 ) , oxygen sensing ( fnr ECA2207 ) , defence against superoxides and other stress responses ( ohrR ECA3168 , phoB ECA1110 , recX ECA3368 , rseB ECA3282 , rseC ECA3281 ) , motility ( flgM ECA1700 , fliZ ECA1740 ) and survival in soil ( sftR ECA4305 ) ( Table S2 ) ., Three of these additional regulators ( fliZ , ohrR and rscR ) have been implicated in virulence in other bacterial pathogens ( Table 1 ) 42–44 ., However , it does not necessarily follow that homologous regulatory proteins in bacteria are responsible for regulation of homologous processes 45 ., Many CDSs encoding putative regulators of unknown function were shown to be regulated by QS ., These CDSs thus represent novel candidates for virulence factors ., Expression of one such CDS , ECA1562 , subsequently named virS , was enhanced in the expI mutant at 12 and 20 hpi and is thus proposed to be repressed by QS ( Table 1 ) ., VirS is a predicted TetR-family transcriptional regulator whose target ( s ) is unknown , although its closest reported homologue is a TetR family regulator , TvrR , implicated in virulence in the plant pathogen Pseudomonas syringae pv ., tomato 46 ., virS is located adjacent to the gene encoding a key QS-controlled regulator , VirR ( ECA1561 , 12 ) ., However , inactivation of virS does not affect transcription of virR ( data not shown ) ., In order to determine whether virS plays a role in virulence , a defined virS mutant was constructed and tested in stem and tubers virulence assay ., The virS mutant showed significantly reduced lesion formation compared with the wild type ( Fig . 4 ) and is thus a novel virulence factor in Pba ., In tuber tests , complementation of the mutant in trans returned virulence to wild type levels ( Fig . 4B ) ., The precise role of virS in planta is under investigation ., The microarray data revealed a small reduction in expression of genes cfa2 ( ECA0607 ) and cfa8A ( ECA0601 ) in the expI mutant compared to the wild type ( Table 1 ) ., These genes are of particular interest as they are part of a cluster responsible for the synthesis of coronafacic acid ( CFA ) which , in Pseudomonas syringae , is a component of the phytotoxin coronatine 47 ., We showed previously that mutations in this cluster ( cfa6 ECA0603 and cfa7 ECA0602 ) significantly reduce pathogenicity of Pba1043 on potato stems 20 ., Transcriptional changes in cfa2 , cfa6 and cfa7 , compared to a QS up-regulated ( pelA ) control , were thus examined at 12 and 20 hpi using qRT-PCR ., At both time-points , pelA , the cfa genes , and the cfl ( ECA0609 ) gene ( involved in the formation of coronafacoyl conjugates by ligation of amino acids to CFA ) showed reduced expression in the expI mutant , indicating that they are all under QS control ( Table S1 ) ., Salicylic acid ( SA ) and jasmonic acid ( JA ) are signalling molecules that play major roles in the activation of plant defences against pathogen attack 48 ., CFA and its amino acid conjugates appear to act as structural and functional analogues of JA and its conjugates 49 ., Recent work by Uppalapati et al . 21 showed that Pseudomonas syringae DC3000 mutants lacking CFA and/or coronatine were impaired in their ability to persist in tomato plants at the later stages of infection , and that the ability to persist coincided with the activation of JA-based , and concomitant suppression of SA-based , defences ., It is hypothesised that , through this suppression of SA-mediated defences , coronafacoyl conjugates may aid P . syringae to enter the necrotrophic phase of infection and promote disease symptoms ., It would appear therefore that Pba , through QS , synthesises CFA and coronafacoyl conjugates co-ordinately with multiple PCWDEs , the T3SS and T6SS in a synchronised assault on the plant as it progresses from biotrophy to necrotrophy ., Although the effect of Pba-encoded CFA conjugates on plant defences has yet to be determined , such a two-pronged attack may be necessary for Pba to establish disease ., It will be interesting to determine whether QS plays a similar role in P . syringae and related pathogens ., QS regulation in pectobacteria was observed originally in vitro through dramatic impacts on PCWDE production , pathogenesis and ( in Pcc ) carbapenem antibiotic production 16 ., The microarray analysis of global gene expression in planta presented here indicates a far broader physiological impact of QS , uncovering effects on the expression of many other genes associated with pathogenesis , and on other physiological processes not necessarily connected to plant pathogenesis ., As QS is AHL concentration-dependent , its impact is likely to be greatest towards the latter stages of infection , where large quantities of PCWDEs are induced to attack plant cells and the characteristic soft rot disease symptoms occur 4 ., Correspondingly , we find that production of the T1SS and T2SS , which are involved in the secretion of PCWDEs to the extracellular environment , are also under QS control ., A very small number of virulence regulators have previously been shown to fall under QS control ., Our study has added over 70 other regulators to this list , including the major known virulence regulators associated with PCWDE production in pectobacteria ., An important inference from these microarray analyses is that the QS control system occupies a critical position in the regulatory hierarchy and that multiple downstream regulators , some which may operate through the Rsm system 14 , are under QS control ., Furthermore , QS is seen to have both positive ( activation ) and negative ( repression ) effects on its downstream targets ., Our knowledge of the hierarchical chain of command in control of the complex regulatory systems of PCWDE and other virulence factors is fragmentary , in part because the current literature describes experimental data derived from multiple strains of Pba , Pcc , and Dickeya spp ., It may not therefore be completely legitimate to assume that the identified regulators play conserved roles in these different bacterial strains 45 ., Nevertheless , while accepting this caveat , our results are consistent with the notion that , within the infected potato plant , QS acts as a key “master regulator” sensory system in this phytopathogen ( Fig . 5 ) ., Additionally , regulators associated with virulence in other bacteria , and many novel putative regulators have also been identified; including VirS , which has been associated with virulence in this study ., In addition to the T1SS and T2SS , we have identified a T6SS in Pba and shown , for the first time in a plant pathogen , that it has a role in virulence ., Moreover , we have described the first example of a QS-controlled T6SS in any pathogen ., The precise functional roles of Hcp , VgrG and other possible Type VI substrates is unknown , but their proposed functions as effector proteins may be important for manipulating host defences whilst PCWDEs mount a simultaneous physical attack on plant cell walls ., This does appear to be the case for both the T3SS and associated effectors , and coronafacoyl-amide conjugates , which are similarly QS-dependent , and consequently may suppress or otherwise manipulate defences ., This has important implications for the infection process in pectobacteria , as it suggests that these pathogens do not infect merely by “brute force” , where the action of PCWDEs alone is sufficient to overwhelm plant defences and break down plant cell walls towards the end of infection ., It seems increasingly likely that , in conjunction with PCWDEs , the production of virulence determinants that actively suppress plant defences , may be necessary to facilitate the transition from biotrophy to necrotrophy during disease development ., Pba1043 20 , and strains with mutations in expI , ECA3438 , ECA3444 and virS were used in this study ., The expI mutant was derived from phage M1-mediated transduction of expI::mTn5gusAgfp from mutant MC3 into the wild type strain 10 ., Mutants ECA3438 and ECA3444 were isolated from a mutation library of Pba1043 5 ., For inactivation of virS , 1085 bp of virS and surrounding regions were PCR-amplified using primers SC51 ( ATTTGGATCCGTTGTTCCTGTTCTGTCG ) and SC52 ( TATATCTAGAGTTTACTGAGCAAGCGACG ) and cloned into pBluescript-II KS+ using BamHI-XbaI sites ., The KnR cassette from pACYC177 ( NEB ) was cloned into the NsiI site in the middle of virS ., The resulting virS::KnR fragment was then cloned into the suicide vector , pKNG101 50 , generating the marker-exchange plasmid ., The plasmid was introduced into Pba1043 by conjugation and transconjugants , resulting from integration of the suicide plasmid into the chromosome by homologous recombination , were selected by ability to grow on minimal medium containing 0 . 2% glucose+streptomycin ., Following overnight growth in the absence of antibiotic selection , exconjugants , in which resolution of the plasmid from the chromosome leaving only the disrupted allele had occurred , were selected by ability to grow on minimal medium containing kanamycin+10% sucrose as sole carbon source and inability to grow on streptomycin ., The disruption of the locus was confirmed by PCR analysis and DNA sequencing ., All strains were maintained on Luria Bertani ( LB ) agar supplemented with kanamycin ( 50 µg/ml ) and , unless stated otherwise , were cultured in 10 ml LB broth at 27°C overnight with aeration ., Mutations were transduced into a clean Pba1043 background using phage M1 51 ., Pathogenicity tests were performed both on potato stems and tubers 19 ., Approx ., 102 and 104 cells per inoculation site were used for stems and tubers , respectively ., Complementation of mutant strains was carried out in trans | Introduction, Results/Discussion, Materials and Methods | Quorum sensing ( QS ) in vitro controls production of plant cell wall degrading enzymes ( PCWDEs ) and other virulence factors in the soft rotting enterobacterial plant pathogen Pectobacterium atrosepticum ( Pba ) ., Here , we demonstrate the genome-wide regulatory role of QS in vivo during the Pba–potato interaction , using a Pba-specific microarray ., We show that 26% of the Pba genome exhibited differential transcription in a QS ( expI- ) mutant , compared to the wild-type , suggesting that QS may make a greater contribution to pathogenesis than previously thought ., We identify novel components of the QS regulon , including the Type I and II secretion systems , which are involved in the secretion of PCWDEs; a novel Type VI secretion system ( T6SS ) and its predicted substrates Hcp and VgrG; more than 70 known or putative regulators , some of which have been demonstrated to control pathogenesis and , remarkably , the Type III secretion system and associated effector proteins , and coronafacoyl-amide conjugates , both of which play roles in the manipulation of plant defences ., We show that the T6SS and a novel potential regulator , VirS , are required for full virulence in Pba , and propose a model placing QS at the apex of a regulatory hierarchy controlling the later stages of disease progression in Pba ., Our findings indicate that QS is a master regulator of phytopathogenesis , controlling multiple other regulators that , in turn , co-ordinately regulate genes associated with manipulation of host defences in concert with the destructive arsenal of PCWDEs that manifest the soft rot disease phenotype . | Many Gram-negative bacteria use a population density-dependent regulatory mechanism called quorum sensing ( QS ) to control the production of virulence factors during infection ., In the bacterial plant pathogen Pectobacterium atrosepticum ( formerly Erwinia carotovora subsp . atroseptica ) , an important model for QS , this mechanism regulates production of enzymes that physically attack the host plant cell wall ., This study used a whole genome microarray-based approach to investigate the entire QS regulon during plant infection ., Results demonstrate that QS regulates a much wider set of essential virulence factors than was previously appreciated ., These include virulence factors similar to those in other plant and animal pathogens that have not previously been associated with QS , e . g . , a Type VI secretion system ( and its potential substrates ) , shown for the first time to be required for virulence in a plant pathogen; and the plant toxin coronafacic acid , known in other pathogens to play a role in manipulating plant defences ., This study provides the first evidence that Pectobacterium may target host defences simultaneously with a physical attack on the plant cell wall ., Moreover , the study demonstrates that a wide range of previously known and unknown virulence regulators lie within the QS regulon , revealing it to be the master regulator of virulence . | microbiology/cellular microbiology and pathogenesis, genetics and genomics/functional genomics | null |
journal.pgen.1004525 | 2,014 | 8.2% of the Human Genome Is Constrained: Variation in Rates of Turnover across Functional Element Classes in the Human Lineage | “What proportion of the human genome is functional ? ” remains a contentious question 1–3 ., In great part this reflects the use of definitions of ‘function’ that differ from the traditional definition that is based on fitness and selection ( see e . g . 4 for a discussion ) ., For instance , equating functionality with annotation by at least one of the ENCODE consortiums biochemical assays 5 results in approximately 80% of the human genome being labeled as functional 1 , 6 ., While this approach has the advantage of being empirical , it makes the definition of functionality dependent on the choice of experiments and details such as P value cutoffs ., It is also questionable whether , for instance , introns should be classified as functional based merely on their transcription 2 , 4 ., By contrast , evolutionary studies often equate functionality with signatures of selection ., While it is undisputed that many functional regions have evolved under complex selective regimes including selective sweeps 7 or ongoing balancing selection 8 , 9 , and it appears likely that loci exist where recent positive selection or reduction of constraint has decoupled deep evolutionary patterns from present functional status 10 , 11 , it is widely accepted that purifying selection persisting over long evolutionary times is a ubiquitous mode of evolution 12 , 13 ., While acknowledging the caveats , this justifies the definition of functional nucleotides used here , as those that are presently subject to purifying selection ., This is of course not useful as an operational definition , as selection cannot be measured instantaneously ., Instead , most studies define functional sites as those subject to purifying selection between two ( or more ) particular species ., Studies that follow this definition have estimated the proportion of functional nucleotides in the human genome , denoted as αsel 14 , 15 , between 3% and 15% ( 3 and references therein , 16 ) ., Since each species lineage gains and loses functional elements over time , αsel needs to be understood in the context of divergence between species ., The divergence influences the estimate of αsel in two ways ., On the one hand , constrained sequence between closely related species , including lineage-specific constrained sequence , is harder to detect than more broadly conserved sequence because of a paucity of informative mutations , which reduces detection power ., On the other hand , estimates of constraint between any two species will only include sequence that was present in their common ancestor and that has been constrained in the lineages leading up to both extant species genomes , with the consequence that turnover of functional sequence leads to diminishing αsel estimates as the species divergence increases ., Assuming that the first effect can be controlled for , higher estimates of sequence constraint that are obtained between more closely related species 15 , 17 are thus indicative of the turnover of functional sequence 15 ., Here we understand turnover to mean the loss or gain of purifying selection at a particular locus of the genome , when changes in the physical or genetic environment , or mutations at the locus itself , cause the locus to switch from being functional to being non-functional or vice versa ., Two previous studies have made quantitative estimates of the overall rate of turnover ( 15 , 17 , reviewed in 3 ) ., The estimate by Smith et al . ( 2004 ) 17 was derived from an analysis of point mutations in alignments across a 1 . 8 Mb genomic region ., While a high rate of turnover was inferred , the authors emphasised the preliminary nature of their work as a consequence of the limited amount of data available to them at that time ., Later , Meader et al . ( 2010 ) 15 performed genome-wide analysis with a neutral indel model ( see 18 , here referred to as NIM1 ) to estimate the fraction , termed αselIndel , of human sequence that was constrained with respect to insertions or deletion mutations ( indels ) ., This study also found a high rate of turnover , and estimated using two ad hoc heuristic approaches that 6 . 5–10% of the human genome is functional ., Extrapolations using these data subsequently suggested that 10–15% of the human genome is presently functional 3 ., NIM1 is a quantitative model describing the distribution of distances between neighbouring indels ( intergap segments; IGSs ) in neutrally evolving sequence , which provides an excellent description of the observed frequency of medium-sized IGSs ., However , across whole genome alignments longer IGSs are strikingly overrepresented compared to this expectation under neutrality , presumably as a result of the presence of functional genomic segments under purifying selection in which indel mutations are unlikely to become fixed ., By quantifying this overrepresentation it is possible to estimate αselIndel , the fraction of nucleotides contained within these functional segments ., The model ( which also accounts for G+C content and sex chromosome-dependent mutational biases ) performs well for simulated data , and accurately identifies coding regions and ancestral repeats as highly conserved and neutrally evolving , respectively 15 , 18 ., However , some concerns about the models derivation and the quality of whole-genome alignments we used were subsequently brought to our attention , which motivated us to initiate this study ., Here we present improved methods for the estimation of αselIndel and the inference of functional turnover , building on our previous approaches 15 , 18 ., We apply these improved approaches to pairwise alignments between the genomes of diverse eutherian mammals , and we estimate that 7 . 1–9 . 2% of the human genome is presently subject to purifying selection , equating to 220–286 Mb of constrained sequence ., We also take advantage of the additional high-quality eutherian genome sequences that have become available since our previous study to provide improved estimates of the rate of turnover of functional sequence in these species ., Improvements in biological and biochemical annotation of genomic sequence mean that we can investigate turnover rates within particular classes of functional elements , such as coding sequences , DNase 1 hypersensitivity sites ( DNase HSs ) , transcription factor binding sites ( TFBSs ) , enhancers , promoters , and long noncoding RNAs ( lncRNAs ) ., We find striking differences between the functional element classes; in particular constrained coding sequences are much more evolutionary stable than constrained noncoding sequences , and lncRNAs show the most rapid rate of turnover of all the noncoding element types ., We observe a strong negative correlation between estimates of αselIndel and the divergence of the two species being compared ( Figure 1 ) , consistent with substantial turnover of functional sequence and thus with earlier conclusions 15 , 17 , and inconsistent with simulation results under a scenario in which turnover is absent ( Figure 1A ) ., To exclude the possibility that technical artefacts are driving this observation , we investigated ENCODE annotations in lineage-specific NIM1-constrained sequence ., Specifically , we identified NIM1-constrained sequence that was not identified as pan-mammalian conserved by either the PhastCons 12 or GERP++ algorithms 19 , and found that such sequence is enriched for biochemically annotated sequences ( DNase HSs , TFBSs , and enhancers defined by the ENCODE consortium 5 ) ( Figure 2; Figure S4 ) ., This is expected if functional elements , including these ENCODE functional classes , have been subject to evolutionary turnover , but is not expected if technical artefacts were causing the observations in Figure, 1 . Furthermore , using low-frequency polymorphic indels from the 1000 Genomes project we could exclude the possibility that lower mutation rates in ENCODE functional regions were causing the observations ., We therefore conclude that observations in Figure 1 reflect turnover of functional elements ., A more detailed discussion on this issue is provided in Text S6 and Text S7 ., To help describe and interpret the observations of turnover ( Figure 1 ) we propose a time-homogeneous model for sequence turnover on a genomic scale ., We apply this model to specific sequence classes , such as protein coding genes or TFBSs , allowing us to discuss the rates of turnover for particular types of functional element ., The model assumes that within a particular functional class both the total amount a of functional sequence and the rate b of turnover per nucleotide ( nt ) are constant , and that the turnover rate is the same for all nts in a class ., Under this model the total amount of functional nts in any class remains constant over time , but the amount that is currently functional and retains homology to functional nts in the ancestral species at divergence d ( i . e . , the amount that was constrained and has not turned over in the course of evolution to the present ) is ., We estimate the parameters a and b by fitting the model to observations using weighted linear regression ( Materials and Methods ) ., Instead of the rate parameter b , we , equivalently , often refer to the turnover half life , d1/2 , which is defined as the divergence at which half the functional sequences in the class is expected to have turned over and is calculated as loge ( 2 ) /b ., We express this divergence in time units corresponding to one expected nucleotide substitution per site in neutrally evolving sequence ( ‘divergence time’ ) ., To convert this divergence to years , we apply a substitution rate of 2 . 2×10−9 per site per year 20 ., This will be a more appropriate value for the human lineage , on which we focus , than on rodent lineages whose per-year substitution rate are substantially higher ., The model is time-reversible , so that the same expression describes the amount of mutually constrained sequence between two extant species at divergence d , where d is calculated by adding the divergences along the two branches to their last common ancestor ., Similarly , to convert d ( in years ) to the age of the most recent common ancestor , it should be divided by, 2 . To calculate the divergence time we use ancestral repeats ( ARs , sequence derived from transposable elements whose insertion predates the species last common ancestor ) as a proxy for neutrally evolving sequence , because they virtually all show the patterns of indel mutation expected under neutral evolution 18 ., Our estimates of divergence using either ARs or synonymous sites as neutral proxy are concordant , hence our results are insensitive to the choice of putatively neutral sequence ( Figure S5 ) ., We next used NIM1 to estimate the fraction of constrained sequence within coding and noncoding sequences ( Materials and Methods ) ., Within protein coding sequence selective constraint is pervasive , as expected ( Figure 1B ) : 80–88% of human or mouse annotated coding sequence has been under selective constraint with respect to indels across eutherian evolution; slightly lower proportions were estimated under the NIM2 and for dog annotated coding sequences ( Figure S6; Text S8 ) ., In contrast to protein coding sequence , estimates for the extent of constraint in noncoding sequence show a pronounced drop-off with increasing divergence ( orange filled circles in Figure 1B ) , an observation compatible with turnover occurring predominantly within the noncoding functional fraction of the genome ., When applying the time-homogeneous turnover model to these data , we estimate the turnover rate parameter b for noncoding sequence at 2 . 48 turnover events per neutral substitution ( 2 . 26–2 . 71 , 95% confidence interval ) , equivalent to a turnover half life d1/2 of 0 . 28 ( 0 . 25–0 . 31 ) in units of divergence time , or 127 My ( 116–139 My ) in natural time units ., The present estimate represents a slower turnover rate than a previous estimate of d1/2\u200a=\u200a0 . 19 ( 86 My ) made by Ponting et al . ( 2011 ) 3 with data from Meader et al . ( 2010 ) 15 ., We observe a low yet significantly non-zero rate of turnover in coding sequence , b\u200a=\u200a0 . 24 ( 0 . 14–0 . 33 ) events per neutral substitution , corresponding to d1/2\u200a=\u200a2 . 9 ( 2 . 1–5 . 0 ) , or in natural units 1300 My ( 950–2250 My ) ., These estimates represent an average across the undoubtedly variable rates of turnover across different types of protein coding gene sequence ., Nevertheless , under this simple model , we find that protein coding sequence is relatively evolutionarily stable , showing long-term conservation , so that assuming that protein coding sequences exhibit no turnover will often be justified ( e . g . 3 ) ., By contrast , present-day constrained noncoding sequence is less stable , being relatively rapidly gained and lost in a lineage-specific manner ., We next investigated whether various classes of functional element , identified in human primarily by the ENCODE project 5 , exhibit contrasting levels of constraint , and whether these constrained element classes show a propensity to turn over at different rates ., Of the functional classes we considered , promoters , untranslated regions ( UTRs ) , DNAse HSs and TFBSs , enhancers and un-annotated sequences ( defined as sequences not within 50 bp of ENCODE DNAse HSs , TFBS loci , lncRNAs from 21 , Ensembl coding sequence , or UTRs ) all show intermediate levels of turnover ( Figure 3; Figure S7 , Figure S8 ) ., LncRNA sequences show the highest level of turnover ( Figure 3; Figure S8 ) , and an even higher rate of turnover was inferred when the ENCODE-defined lncRNAs were used rather than the set from 21 ( Figure S9 ) ., The fraction of sequence that the model inferred to be under present day constraint also varied across these categories , with intermediate fractions inferred for UTRs , DNAse HSs and TFBSs , and lower fractions for lncRNAs and enhancers ., As expected , the lowest fractions were observed for un-annotated sequence; nevertheless , in absolute terms the amount of constrained sequence in this category is considerable ( 70 Mb , 45–85 Mb ) ( Figure 3 ) ., Constrained sequence in this category may represent lineage-specific functional sequences that were not identified by the ENCODE project , for instance because of their function in tissues or developmental stages not investigated by ENCODE ., Finally , transposable element-derived sequences show very small amounts of constraint , and as a result our methods have little power to detect turnover in this class ., We next examined how constrained sequence in the human genome is distributed cumulatively for selected functional element categories ., We do this by fitting the functional turnover model to the observed data and extrapolating to the present day ., In this way we also infer the reciprocal quantities of sequence that , when comparing to another species or human ancestor at a particular divergence , are presently functional in human yet have lost ( or not gained ) constraint in the lineage leading to the ancestor or other species ( Figure 4 ) ., We stress that this inference relies on the parsimonious yet not formally justified assumption that the total quantity of functional sequence in genomes remains constant over time and therefore across species , and within functional categories ., With these caveats we estimate that 8 . 6 Mb ( 26% ) of constrained coding sequence has lost constraint ( and thus has turned over ) since the divergence of humans from monotremes approximately 228 million year ago ( AR divergence time 1 . 00 ) , while 200 Mb ( 79% ) of the constrained noncoding human genome is inferred to have lost constraint over the same period ., DNAse HSs cover more indel constrained sequence at all divergence ranges than all other annotated noncoding sequence combined , implying that DNAse HSs are an abundant and informative biochemical marker of functionality outside protein coding regions ., Enhancers also show a marked contribution towards the constrained human genome , while TFBSs , promoters , UTRs and lncRNAs contribute considerably less sequence once their overlap with other annotations is removed ., Finally , about a quarter of sequence inferred to be presently under constraint is not present in any of the annotation categories we considered ., In Figure 4 we sum up the quantities of constrained sequence estimated from independent NIM1 runs for different annotation types ., If we make the assumption that the exponential decay model of functional sequence applies outside of the range of divergences we examined , then by extrapolating back to zero divergence we can estimate the total proportion of human genomes that is under present-day purifying selection with respect to indels ., We perform this extrapolation across different annotation sets ( Table S6 ) ., Although there is some variation in these estimates , we quote the estimate derived separately across multiple different annotation categories , namely coding sequence , DNase HSs , TFBS , Enhancers , unannotated sequence , and other sequence ( the latter consisting of promoter , UTR and lncRNA sequences ) ., This is because this estimate allows the rate of turnover to vary across each annotation type , and thus is likely to be more accurate than the estimates that assume a single rate of turnover across the whole genome , or the whole noncoding genome ., We therefore estimate that 8 . 2% of the human genome ( 253 Mb; 95% CI 7 . 1%–9 . 2% , 220–286 Mb ) is presently under purifying selection with respect to indels ., The question of what fraction of the human genome sequence are mutations preferentially purged owing to their deleterious effect has remained contentious ever since the first estimate was made in 2002 22 ., At that time it was not well appreciated that the amount of human constrained sequence that is also constrained in mouse is a minority ( 69 Mb; this study ) of all human constrained sequence , owing to the relatively rapid gain and loss of functional sequence in their two lineages since their last common ancestor ., We find that NIM1-constrained sequence lacking evidence for pan-mammalian conservation is enriched for sequences with experimental evidence for biochemical activity , and we provide a detailed argument indicating that this is incompatible with the notion of technical artefacts causing the observed signature of turnover ( Text S6 ) ., Extensive simulations indicating that estimates of constrained sequence are consistent across the divergence range we investigate further support this conclusion ., Our estimate that 7 . 1–9 . 2% of human genomes is subject to contemporaneous selective constraint considerably exceeds previous estimates and falls short of others 3 , 23 ., We have shown that our methods previous estimates for specific species pairs , as well as the calculation that suggested 10–15% of the human genome is currently under negative selection were inflated 3 , in large part owing to inaccuracies in whole genome alignments upon which our estimates were based ., The problems associated with using whole-genome alignments could be circumvented entirely by instead using polymorphism data within a single species ., However , this approach is technically highly challenging , and results have so far been controversial 16 , 24 , 25; in addition this approach is not informative about functional turnover ., Other published estimates 12 , 18 , 26 are lower because they , by design , were not sensitive to lineage-specific constrained sequence ., Our current estimates have their own particular caveats ., While our results show that turnover is a real and substantial effect , simulations show that NIM1 underestimates the true amount of mutually constrained sequence to an extent that shows some dependence on the divergence ., While simulations and theory indicate that point estimates of constraint remain conservative , the possibility of an upward bias in the inferred rate of turnover cannot be excluded , which in turn could lead to upwardly biased extrapolations of present-day constraint ., In addition , the assumptions of the turnover model , in particular that all elements within a class are subject to the same rate of turnover , clearly are only approximately valid ., These potential sources of error are not reflected in our confidence estimates ( Table S6 ) ., Our estimate that 7 . 1%–9 . 2% of the human genome is functional is around ten-fold lower than the quantity of sequence covered by the ENCODE defined elements 1 , 5 , 6 ., This indicates that a large fraction of the sequence comprised by elements identified by ENCODE as having biochemical activity can be deleted without impacting on fitness ., By contrast , the fraction of the human genome that is covered by coding exons , bound motifs and DNase1 footprints , all elements that are likely to contain a high fraction of nucleotides under selection , is 9% ., While not all of the elements in these categories will be functional , and functional elements will exist outside of these categories , this figure is consistent with the proportion of sequence we estimate as being currently under the influence of selection ., As expected , turnover has occurred least in protein coding sequence , and thus has been most concentrated on noncoding sequence ( Figure 4 ) ., For example , of the 43 . 5 Mb of sequence annotated by the ENCODE project as being within a human TFBS peak and that we find to be constrained ( 19 . 3% of the total extent of ENCODE TFBS peaks ) , only a third ( 30 . 6%; 13 . 3 Mb ) is identified by NIM1 as being constrained in both human and mouse ., A slightly higher proportion ( 45 . 6%; 19 . 8 Mb ) is constrained in human and dog , presumably reflecting these species lower divergence ., These estimates are in good agreement with previous experimental findings: for instance 23–41% of TF binding events have been found to be conserved across human , dog and mouse for four liver TFs 27 , while for two additional liver TFs , 7–14% of TF binding events are shared between human and mouse , and 15–20% between human and dog 28 ., The phenomenon of turnover is well supported by both anecdotal evidence 27–29 and by broader studies of particular classes of elements , mostly TFBSs and enhancer elements 30–32 ., The class of functional element inferred to turnover fastest was that of lncRNAs , again consistent with observations that most human lncRNAs are primate-specific and only 19% of lncRNAs are conserved over more than 90 My 33 ., What our approach cannot clarify is to what extent the observed turnover at the sequence level amounts to different sequences encoding equivalent function 29 , 30 , or species-specific functional change 16 , 31 , 34 ., Several lines of evidence , both from anecdotal 29 and broader 30 , 31 studies of TFBSs , indicate that a large fraction of sequence changes involving TFBSs preserve function ., For example , some deeply conserved transcription factors have species-specific binding sites in the vicinity of orthologous genes 27 , 28 implying that despite their sequence divergence , the different DNA binding sites confer equivalent functions ( on orthologous genes ) in different lineages ., Comprehensive studies of human and mouse embryonic heart enhancers found these to be weakly conserved 35 , 36 , despite human enhancers sequences largely driving expected tissue-specific expression in mouse embryonic heart tissue 36 ., Another study found that two mammalian hypothalamic enhancers have no homolog across non-mammalian vertebrates , yet are still able to drive specific expression patterns in zebrafish neurons 37 ., These findings are consistent with gene expression evolution being shaped predominantly by stabilizing selection on the expression level 38 , while evolution on the sequence level may involve an interplay between fixation of weakly deleterious mutations through drift , and weak positive selection on compensatory mutations 39 ., However , not all TFBS turnover events are neutral or nearly neutral on the level of gene expression , and the fraction of such events that change gene expression may be substantial 31 ., More generally , lineage-specific sequence is clearly a likely substrate for lineage-specific biology 16 , 34 , although adaptations to pre-existing functional sequence remain an alternative plausible mode for creating species-specific change 40 ., Nevertheless , the sheer ubiquity of sequence turnover , and the clear potential for substantial regulatory change resulting from it , suggests that many aspects of noncoding human biology will not be fully recapitulated by orthologous sequence in eutherian model organisms , including mouse ., Thus , our findings could provide a more quantitative basis for assessing the relevance of model organisms to specific questions of human biology ., We restricted our analyses to genome assemblies that have been sequenced at relatively high coverage , not using for example the 2-fold coverage assemblies of mammalian genomes 41 , to minimize the impact of sequencing and assembly errors ., From the UCSC Genome Informatics website ( http://genome . ucsc . edu/ ) , we acquired softmasked versions of the following genome assemblies: human ( hg19 ) , mouse ( mm10 , mm9 , and mm8 ) , rat ( rn5 ) , cattle ( bosTau7 ) , dog ( canFam2 ) , horse ( equCab2 ) , guinea pig ( cavPor3 ) , rabbit ( oryCun2 ) , bushbaby ( otoGar3 ) , panda ( ailMel1 ) , and rhino ( cerSim1 ) ., We also acquired a Ferret genome assembly ( M_putorius_furo_v1 ) produced by the Broad Institute ., We softmasked the ferret genome assembly using RepeatMasker with carnivore repeat libraries 42 ., When available , whole genome pairwise alignments were downloaded from the UCSC Genome Informatics website ( http://genome . ucsc . edu ) ., Otherwise , we constructed alignments following UCSCs protocol 43 ., Initial alignments were constructed with LASTZ ( http://www . bx . psu . edu/miller_lab/ ) , a derivative of BLASTZ 44 , and these alignments were subsequently chained and netted using tools from UCSC ( Table S1 for alignment parameterisations ) ., We trimmed each of the whole genome alignments once we found that UCSC alignments contained a minority of poorly aligning sequence ( Figure S1 , Table S2 ) ., Each alignment was rescored to generate a new substitution matrix using a log-odds ratio approach as described previously 45 ., We did not impose symmetry on the scoring matrixes with respect to strand or species ., We then used the generated substitution matrix , with gap penalties derived from the original alignments , to discard ( “trim” ) the maximal non-positively scoring terminal segments of the alignment blocks and any non-positively scoring inter-gap segments ., Trimming removes terminal and internal alignment segments that are more likely to have arisen under a model of independent evolution than of evolution from a common ancestor ., Subsequent analyses were carried out following the discarding of all trimmed sequence ., We also excluded alignments that were led by sequence not mapped to chromosomes ., We did not exclude non-reciprocally aligning sequence or sequence that lay within known indel hotpot locations as we found removing such sequence had relatively small effects on estimates of αselIndel ( Table S3 ) ., The neutral indel model of Lunter et al . ( 2006 ) 18 ( NIM1 ) estimates the genomic fraction ( αselIndel ) of sequence constrained with respect to indels between a species pair ., The model examines the distribution of IGSs from a set of whole genome pairwise alignments using a regression approach over a range of medium IGS lengths to estimate the parameters of a predicted geometric distribution of IGSs in neutral sequence ., αselIndel in bp is then estimated by summing up the quantity x - 2K over all the long IGSs inferred to be in excess of predictions under neutral evolution ., Here where x is the length of the overrepresented IGS , and K is the estimated mean spacing between indels ( “neutral overhang” ) ., 20 equally populated G+C content bins are analysed separately to account , in part , for mutational variation that correlates with G+C content ., The X chromosome is also analysed separately ., A detailed description of the model is given in the original publications 15 , 18 ., However , two theoretical issues of the model have not been described previously ., These are: ( A ) that thresholding biases the expected lengths of the neutral overhang and , ( B ) that neutral segments are depleted from the background distribution due to the presence of constrained segments , changing the expected neutral distribution of IGS lengths; resolution of the two issues is described in Text S1 ., Our implementation of the NIM1 differs from that of the preceding studies in the manner in which we calculate the bounds of the estimates ., The previous approaches constructed the upper and lower bound estimates based on the uncertainty in the degree of clustering of functional elements ., The lower bound estimate was derived assuming that functional elements are unclustered ( each overrepresented IGS contributes x - 2K bp towards the αselIndel estimate ) , while the upper bound was derived assuming a high degree of clustering ( each overrepresented IGS contributes x - K bp ) ., In our revised approach , we construct a 95% confidence interval around the lower x - 2K bp estimate ., The impact of this change on αselIndel estimates can be seen in the simulation study ( Table S5 ) ., We made this conservative modification to the NIM1 for five reasons: Firstly , the previous upper bound estimate assumes an unrealistically high degree of clustering of functional elements ., Secondly , only our modified estimate is always conservative under all the simulation scenarios , whereas the previous implementation of the NIM1 sometimes overestimates the true value of αselIndel ( Table S5 ) ., Thirdly , altering the clustering of functional elements in the simulations actually has only a minor effect on the estimated quantities of constrained sequence ( Figure S11 ) ., Fourthly , in addition to the clustering of functional elements , other parameterisations also influenced αselIndel estimates ( Table S5 ) , yet the uncertainty in the values of these parameters was not also incorporated into the NIM1 estimate ., Instead , we now choose to incorporate the full extent of uncertainty into the simulations ., Finally , by providing a 95% confidence interval for the αselIndel estimate of NIM1 , we have an estimate that is directly comparable to the NIM2 estimates ., We have described above how NIM1 is used to estimate the fraction αselIndel of constrained bases within a genome G consisting largely of neutrally evolving sequence ., To estimate αselIndel within a subset S⊆G that is not dominated by neutrally evolving sequence , for instance when estimating αselIndel within coding sequence , we instead estimate αselIndel within the subsets G and G\\S; the difference between the resulting estimates is the estimate of αselIndel within S . We extracted ancestral repeat ( AR ) alignments from the trimmed whole genome alignments using RepeatMasker annotations to identify transposable element and repeat-derived sequence 42 ., We then calculated the substitution rate for the alignments using the HKY85 model applied in the PAML package BASEML 46 ., We also estimated synonymous substitution rates ( dS ) across protein coding regions for some species pairs ., Estimates of dS for a species pair were made by calculating the median dS of all one-to-one gene orthologs in the Ensembl Compara database with dS<1 ., Nucleotide substitution rates in AR sequences are very similar to estimates of the synonymous substitution rate ( dS ) ( Figure S5 ) , hence our results appear insensitive to the choice of neutral sequence standard ., The time-homogeneous turnover model makes the following assumptions: for a particular class of functional elements , both the total amount of functional seq | Introduction, Results, Discussion, Materials and Methods | Ten years on from the finishing of the human reference genome sequence , it remains unclear what fraction of the human genome confers function , where this sequence resides , and how much is shared with other mammalian species ., When addressing these questions , functional sequence has often been equated with pan-mammalian conserved sequence ., However , functional elements that are short-lived , including those contributing to species-specific biology , will not leave a footprint of long-lasting negative selection ., Here , we address these issues by identifying and characterising sequence that has been constrained with respect to insertions and deletions for pairs of eutherian genomes over a range of divergences ., Within noncoding sequence , we find increasing amounts of mutually constrained sequence as species pairs become more closely related , indicating that noncoding constrained sequence turns over rapidly ., We estimate that half of present-day noncoding constrained sequence has been gained or lost in approximately the last 130 million years ( half-life in units of divergence time , d1/2\u200a=\u200a0 . 25–0 . 31 ) ., While enriched with ENCODE biochemical annotations , much of the short-lived constrained sequences we identify are not detected by models optimized for wider pan-mammalian conservation ., Constrained DNase 1 hypersensitivity sites , promoters and untranslated regions have been more evolutionarily stable than long noncoding RNA loci which have turned over especially rapidly ., By contrast , protein coding sequence has been highly stable , with an estimated half-life of over a billion years ( d1/2\u200a=\u200a2 . 1–5 . 0 ) ., From extrapolations we estimate that 8 . 2% ( 7 . 1–9 . 2% ) of the human genome is presently subject to negative selection and thus is likely to be functional , while only 2 . 2% has maintained constraint in both human and mouse since these species diverged ., These results reveal that the evolutionary history of the human genome has been highly dynamic , particularly for its noncoding yet biologically functional fraction . | Nearly 99% of the human genome does not encode proteins , and while there recently has been extensive biochemical annotation of the remaining noncoding fraction , it remains unclear whether or not the bulk of these DNA sequences have important functional roles ., By comparing the genome sequences of different species we identify genomic regions that have evolved unexpectedly slowly , a signature of natural selection upon functional sequence ., Using a high resolution evolutionary approach to find sequence showing evolutionary signatures of functionality we estimate that a total of 8 . 2% ( 7 . 1–9 . 2% ) of the human genome is presently functional , more than three times as much than is functional and shared between human and mouse ., This implies that there is an abundance of sequences with short lived lineage-specific functionality ., As expected , most of the sequence involved in this functional “turnover” is noncoding , while protein coding sequence is stably preserved over longer evolutionary timescales ., More generally , we find that the rate of functional turnover varies significantly across categories of functional noncoding elements ., Our results provide a pan-mammalian and whole genome perspective on how rapidly different classes of sequence have gained and lost functionality down the human lineage . | genomics, functional genomics, genome evolution, genetics, biology and life sciences, comparative genomics, computational biology | null |
journal.ppat.1006339 | 2,017 | Promiscuous signaling by a regulatory system unique to the pandemic PMEN1 pneumococcal lineage | Streptococcus pneumoniae ( pneumococcus ) is one of the most important community acquired human pathogens , and is responsible for an estimated 850 , 000 deaths annually in children under the age of 51 ., Pneumococcus colonizes the nasopharynx of young children at very high rates , and is asymptomatic in most cases 2 , 3 ., However , it can also disseminate from the nasopharynx into tissues leading to diseases such as otitis media , pneumonia , bacteremia , meningitis , and inflammation of the heart 4–6 ., The pneumococcal molecules responsible for this transition from a commensal to a pathogen are not well understood ., Here we characterize a novel quorum sensing ( QS ) system ( TprA2/PhrA2 ) that limits pneumococcal disease , without affecting nasopharyngeal colonization ., At the genomic level , there is extensive diversity among pneumococccal lineages ., These genomic variations contribute to the differences in colonization and virulence potential 7 ., Only half of the pangenome is shared across all strains ( core set ) , while the other half is unevenly distributed amongst isolates 8 , 9 ., The Pneumococcal Molecular Epidemiology Network ( PMEN ) has grouped strains of multi locus sequencing type ( MLST ) 81 into the PMEN1 lineage ( also known as Spain23F-1 and SPN23F ) 10 ., Over the past 30 years , PMEN1 has distinguished itself by its worldwide distribution , multi-drug resistant profile , and emergence of vaccine-escape strains ., Historically , the PMEN1 lineage was responsible for the Spanish epidemic of the 1980s and has since spread to North and South America , Europe , Asia , Africa , and Australia 2 , 10 ., Most PMEN1 isolates are resistant to penicillin , chloramphenicol , and tetracycline , and many isolates have additional resistances to fluoroquinilones and macrolides 11 , 12 ., PMEN1 isolates are predominantly of serotype 23F , but there are also capsular switches to other serotypes , some of which represent vaccine-escape isolates 13 ., Further , the PMEN1 lineage has impacted the genome content of the pneumococcal population by virtue of its high frequency of DNA donation , including genes for drug-resistance , to other pneumococcal lineages 14 ., The PMEN1 genome encodes an integrative conjugative element ( ICESp23FST81 ) 13 , 15 , 16 ., As described by Croucher and colleagues upon sequencing of the first PMEN1 genome , this ICE encodes drug resistance determinants , a complete lanthionine-peptide gene cluster and a regulator-peptide pair , which in this study we have identified as the TprA2/PhrA2 QS system ., Quorum sensing systems serve as a critical , decision-making process in the response of bacteria to the environment , and their ability to colonize and/or disseminate to tissues ., The best characterized kind of QS machinery is the two component system , where the signal is sensed by a surface-localized histidine kinase and transferred to a cytosolic response regulator 17 ., Streptococci , enterococci and bacilli have been shown to encode a second kind of QS characterized by the emerging RRNPP ( Rgg/Rap/NprR/PlcR/PrgX ) superfamily of transcriptional regulators and their cognate peptides 18 ., In these systems , the secreted peptide is exported from the producer cell , processed , and imported into the cytosol of producing or neighboring cells , where it interacts with the RRNPP regulator 18 ., RRNPP-peptide systems have been shown to regulate virulence , biofilm formation , and the production of bacteriocins 19–21 ., In pneumococcus , the majority of characterized peptides signal via two component systems 17 ., These peptides regulate competence and class II bacteriocin production 22 , 23 ., The first RRNPP-peptide pair was recently characterized in the pneumococcus strain D39 24 ., It is composed of the TprA regulator and its cognate peptide PhrA ., PhrA alleviates gene inhibition leading to the expression of physiologically important genes 24 ., PhrA levels are repressed by glucose and activated by galactose , consistent with activity in the upper respiratory track where galactose is a major source of energy 25 ., In this study we characterize the TprA2/PhrA2 QS system , a novel pneumococcal RRNPP-peptide pair , highly expressed in middle ear effusions ., TprA2/PhrA2 is present almost exclusively in PMEN1 isolates where it restrains dissemination ., Unlike other lineages , the PMEN1 strains encode both the TprA/PhrA and the TprA2/PhrA2 signaling systems ., Extracellular PhrA2 leads to induction of TprA in PMEN1 cells as well as in D39 cells ., Thus , horizontal acquisition of TprA2/PhrA2 has provided the PMEN1 lineage with a QS system and associated regulon , as well as the molecular machinery to regulate a widespread cell-cell communication system and in doing so , influence not only its own gene expression but also that of other strains ., Genes enriched in the PMEN1 strains may provide this lineage with exclusive phenotypic properties , explaining its prevalent occurrence and rapid spread ., We performed a comparative genomic screen to search for genes that are present in the majority of the PMEN1 isolates , but absent in other pneumococcal lineages ., The analysis was performed on 60 pneumococcal genomes , selected to capture the diversity in the pneumococcal population ( S1 Table , labeled “To establish PMEN1 enrichment” ) ., We employed RAST 26 to annotate the whole genome sequences ( WGS ) into 125 , 612 coding sequences ( CDSs ) , and organized these into 3 , 571 clusters of homologous sequences as previously described 27 ., The screen identified a genomic region present only in the PMEN1 strains ., This region encodes a transcriptional regulator ( tprA2 ) on the opposite strand of a small peptide ( phrA2 ) and three ABC transporters ., Immediately downstream are three genes lcpA , lcpM , and lcpT ., LcpA encodes a putative 71aa peptide with the full size weight of 7 . 5kDa , which we predict is a lanthionine containing peptide ., Lanthionine and methyllanthionine are usually formed by the dehydration of threonines or serines , and subsequent cyclization to cysteine ( lcpA encodes for serine , threonines , and cysteines ) 28 ., Cyclization is performed by lanthipeptide synthetases , of which there are four known classes 29 ., The lcpM gene downstream of LcpA is consistent with class II synthetases ( CDD score: LanM-like e-value 0e+00 30 ) ., Finally , the lcpT encodes a transporter with a C39 peptidase domain , which we predict is involved in LcpA cleavage and export ( Fig 1 , S2 Table ) ., We performed a detailed assessment on the phylogenetic distribution of the QS-Lcp genes in the pneumococcus species and the Streptococcus genus ., First , for the assessment of the distribution of TprA2 in the PMEN1 lineage , we searched for this gene in 215 PMEN1 isolates ., To this end we used either polymerase chain reaction ( PCR ) or genomic data assembled by Croucher and colleagues 13 ., The tprA2 gene was present in 212 isolates ., It was either disrupted or deleted in the genomes of strains 111 ( ERS004810 ) , 11933 ( ERS005313 ) and HKP38 ( ERS004775 ) ( genome data was confirmed by PCR ) ., Next , we broadened our search into the non-redundant database , which revealed that tprA2 was present in only one strain outside the PMEN1 lineage ( GA13494 ) 15 ( Fig 2 ) ., Finally , we expanded our search for tprA2 in related streptococcal species , specifically S . pseudopneumoniae , S . mitis , S . oralis , and S . infantis ( S1 Table labeled “Distribution with Streptococcus sp” ) ., We found one occurrence in S . mitis and one in S . infantis , but these species did not encode the downstream lcpAMT locus ( Fig 2 ) ., These phylogenetic analyses demonstrate that the QS system and lcpAMT are present in >98% of the PMEN1 isolates and are rare outside this lineage ., This distribution suggests these genes were acquired via horizontal gene transfer by a PMEN1 ancestral strain ., To determine whether QS-Lcp genes are active during infection , we measured their gene expression during middle ear infection ., We utilized the nCounter NanoString technology since this allows for an automated , highly sensitive enumeration of pathogen’s mRNA transcripts in the infected host tissue ., Our probes capture tprA2 , lcpA , lcpM , and lcpT ., Further , since we were unable to design a probe for the short coding sequence of phrA2 , we used ABCATPase as a proxy since it is present on the same transcript ( S1 Fig ) ., For normalization we used probes to gyrB and metG , and normalized to the geometric mean of these housekeeping genes ., The PMEN1 strain PN4595-T23 31 was inoculated transbullarly into the chinchilla ., We isolated RNA from effusions of the chinchilla middle ears at 48h post-transbullar inoculation ., All five genes were expressed in middle ear effusions ( Fig 3 ) ., The average counts for ABCATPase and lcpA were comparable to those of psaA ( 56 , 036 counts ) , which has been shown to be highly expressed in vivo 32 , consistent with high levels of QS-Lcp in vivo ., To evaluate whether these genes were induced in the middle ear relative to growth in rich media , we calculated the ratio of the average number of transcripts between middle ear effusions and in vitro planktonic cultures ., The gene expression levels of ABCATPase , lcpA , lcpM and lcpT were 69 , 108 , 93 and 45-fold higher in vivo relative to planktonic cultures , respectively ., From these in vitro and in vivo measurements we infer that the QS-Lcp system is both induced and highly expressed during infection ., The expression of sensory peptides can be cell-density dependent ( reviewed in detail in 33 ., Using quantitative real time PCR ( qRT-PCR ) we found that phrA2 is regulated in a density-dependent manner ., Expression of phrA2 increases at higher cell density , as observed by measuring gene expression at lag , early-log , mid-log and stationary phase ( Fig 4 , solid bars ) ., Further , when a lag phase culture was left to grow for one hour , the levels of phrA2 expression increased 3 fold ., When the same culture was exposed to cell-free supernatant from a wild-type high-density culture , the levels of phrA2 expression increased 8 fold ., Yet , when it was exposed to the cell-free supernatant from a ΔphrA2-ABC high-density culture , the levels of phrA2 did not increase ( Fig 4 , striped bars ) ., Thus , the wild-type cells but not the ΔphrA2-ABC mutant , secrete a molecule that induces expression of phrA2 in the population ., These data are consistent with secretion and autoinduction of PhrA2 ., To identify the TprA2 regulon , we compared the gene expression levels of the wild-type ( WT ) PMEN1 strain PN4595-T23 and the isogenic tprA2 deletion mutant ( ΔtprA2 ) , utilizing a pneumococcal gene array ( S5 Table ) 34 ., The expression of the phrA2-ABC and lcpAMT genes were >30-fold higher in ΔtprA2 relative to the WT strain ., These results were verified , using independent biological replicates , by both qRT-PCR and nanoString technology ( Table 1 ) ., These findings suggest that TprA2 is a negative regulator of these neighboring genes ., To confirm the role of TprA2 , we generated a complemented strain ( ΔtprA2::tprA2 ) where tprA2 was inserted into the ΔtprA2 strain at a distant chromosomal location , under the influence of the constitutive erythromycin-resistance gene promoter ( ermB ) ., We measured gene expression of tprA2 , phrA2 , ABC transporter ATPase , and lcpA in the WT , ΔtprA2 and ΔtprA2::tprA2 strains ( Fig 5 ) ., The tprA2 gene was expressed in the ΔtprA2::tprA2 strain , and its expression level was higher than in the WT ., Further , low levels of phrA2 , ABC transporter ATPase , and lcpA were re-established in the complement strain ., These findings strongly support our conclusion that the gene product of tprA2 is a negative regulator of phrA2 and lcpAMT ., The TprA2 regulator displays sequence similarity to the Bacillus sp ., transcription factor , PlcR and to the pneumococcal TprA , which are regulated by extracellular forms of the C-terminal heptapeptides from their cognate peptides 24 , 35 ., Given that TprA2 is part of the PlcR family , we hypothesized that the C-terminal heptapeptide of PhrA2 would encompass a functional peptide capable of influencing TprA2 activity ., Thus , we utilized synthetic peptides corresponding to the seven terminal residues of PhrA2 ( sequence: VDLGLAD ) and a scrambled control ( sequence: DAGVLDL ) ., Addition of the PhrA2 peptide , but not the scrambled peptide to planktonic culture led to a significant increase in expression levels of tprA2 , phrA2 , ABC transporter ATPase and adjacent lcpAMT genes ( Fig 6 ) ., The PhrA2 peptide up-regulates its own production demonstrating autoinduction of this density-dependent system ., We also observed an increase in the levels of tprA2 suggesting that TprA2 serves as a negative regulator of its own expression ., The induction of gene expression by the synthetic peptide explains the observation that supernatant from a high-density WT culture , but not a ΔphrA2-ABC , can induce gene expression ( Fig 4 ) ., Further , cell-free supernatant from a PhrA2 overexpressing strain increases levels of phrA2 and lcpA by over 5 fold when compared to media alone ( S2 Fig ) ., These findings strongly support a model in which the phrA2 gene product is exported ., We investigated the regulation of the TprA2/PhrA2 system in vivo to verify whether our in vitro finding were relevant to the in vivo environment ., We analyzed WT , ΔtprA2 , and ΔtprA2::tprA2 ., Three chinchillas were independently inoculated with each strain , middle ear effusions were extracted 48 hours post-inoculation , and bacterial mRNA for tprA2 , ABCATPase , lcpA and lcpM was quantified using nanostring technology ., As observed in vitro , deletion of tprA2 led to increase expression of ABCATPase ( on the same transcript as phrA2 ) and lcpM ( Fig 7 ) ., LcpA values were also higher in this mutant , but display elevated inter-animal variability such that the change was not statistically significant ., The modest fold increase is consistent with our observation that the TprA2-regulon in the WT is highly expressed in vivo , such that complete removal of the negative regulator has a moderate effect ., In contrast , overexpression of tprA2 in the complement strain led to a decrease in the levels of ABCATPase and lcpA ., Together , these findings suggest TprA2 is negative regulator of its neighboring genes in vivo ., To assess the in vivo role of the QS-Lcp region we made use of two pneumococcal infection models ., To study colonization of the nasopharynx and spread to the lungs we utilized a murine model where animals are inoculated intranasally and disease progresses causing pneumonia or sepsis or both 36 , 37 ., To study middle ear disease we utilized the chinchilla otitis media model ., The murine model revealed that TprA2 protects against lung disease ., We did not observe infection in mice inoculated with PN4595-T23 strains , thus we generated the parallel mutants in another naturally occurring PMEN1 strain with a type 3 capsule ( SV36 ) ., Cohorts of ten BALB/c mice were infected with SV36 , SV36ΔtprA2 or SV36ΔphrA2-ABC and observed over 4 days ., The bacterial titers in the nasal lavages were similar for all three strains when tested at 48 hours post-inoculation ( Fig 8B ) ., Notably , SV36ΔtprA2 displayed a statistically significant increase in mortality ( Fig 8A ) ., TprA2 is a negative regulator of lcpAMT ( Fig 5 ) ., To test whether overexpression of lcpAMT in the SV36ΔtprA2 was associated with the increase virulence of this strain , we tested a double mutant with deletions in tprA2 and lcpAMT and observed that it restored the wild-type phenotype ., These results strongly suggest that LcpA is a virulence determinant , and that TprA2 can modulate virulence by controlling levels of lcpAMT ., Finally , to study middle ear disease , bacteria were inoculated directly into the middle ear of chinchillas ., The overall mortality was the same for all three strains , perhaps reflecting differences in peripheral disease progression from the chinchilla middle ear versus the murine nasopharynx ( Fig 8C ) ., Further , we observed a trend toward increased middle ear disease in the ΔtprA2 ( Fig 8D ) , and the ΔtprA2 displayed the highest lung dissemination ( S3 Table ) , consistent with our finding that lcpAMT plays a role in virulence ., In conclusion , our findings suggest that TprA2 controls lcpA expression and in doing so can promote commensalism over dissemination ., TprA2 shares moderate homology to TprA , another streptococcal transcription factor that belongs to the recently characterized TprA/PhrA system , where TprA inhibits expression of PhrA and downstream lantibiotic genes 24 ., Unlike tprA2 , which occurs rarely outside the PMEN1 lineage , tprA has a wide distribution in pneumococci ., Using a set of highly curated WGSs , with representatives of the major lineages of S . pneumoniae , we found that tprA was present in over 90% of the isolates in our set ( Fig 2 , all tprA genes displayed > = 86% similarity ) ., The prominent exception is a set of strains in a basal pneumococcal branch associated with unencapsulated strains and conjunctivitis infections 38 , 39 ( Fig 2 ) ., Hoover and colleagues first characterized the TprA/PhrA system , and also reported a wide distribution ( approximately 60% ) in pneumococcal strains 24 ., PMEN1 strains are notable in that they code for both the TprA2/PhrA2 and TprA/PhrA QS systems ., In the PMEN1 strain PN4595-T23 , the TprA and TprA2 protein sequences share approximately 60% identity ., We searched the genomes of 55 streptococcal strains , identified 48 sequences to construct a phylogenetic tree of these regulators using maximum likelihood , and found that the tprA2 and tprA homologues are separated into two distinct branches ( Fig 9A ) ., Their cognate peptides in PMEN1 , PhrA2 and PhrA share only 28% identity over the full length , but display very high similarity at their C-termini ., To analyze the extent of conservation of the C-terminal residues , we generated a consensus logo from the six PhrA2 sequences and the thirty-six PhrA sequences ., The C-terminal residues are either identical or share similar charge in 6/7 residues; but can be distinguished by position -3 that codes for a conserved leucine in PhrA2 and a lysine in PhrA ( Fig 9A and 9B ) ., The sequence separation between the QS components suggests that the tprA2/phrA2 genes did not originate from a recent duplication within PMEN1 , and is consistent with acquisition of TprA2/PhrA2 by horizontal gene transfer ., The co-occurrence of both QS systems in the PMEN1 strains led us to investigate whether PhrA2 and PhrA peptides can exert regulatory effects on their non-cognate QS systems , TprA/PhrA and TprA2/PhrA2 respectively ., To test this , we measured how the addition of synthetic peptides to the extracellular milieu affects gene expression of the non-cognate regulon ., Addition of synthetic PhrA2 ( VDLGLAD ) , but not the scrambled peptide , induced gene expression of the TprA regulon ( tprA , phrA , and the TprA-associated lanA , lanM , and lanT ) at levels similar to those induced by cognate PhrA ( LDVGKAD ) itself ( Fig 10A ) ., In contrast , neither the addition of synthetic PhrA nor the addition of the scrambled peptide had any effect on expression of the tprA2 , phrA2 , or lcpA genes in the TprA2/PhrA2 regulon ( Fig 10B ) ., These findings suggest that PhrA2 regulates gene expression of the TprA regulon , and PhrA has no effect on the TprA2 regulon ., The unidirectional influence of PhrA2 gene expression upon TprA/PhrA led us to investigate whether the PMEN1 peptide could influence gene expression in non-PMEN1 cells ., We used strain D39 as a representative of the non-PMEN1 strains since TprA/PhrA system has been previously described in D39 ., Hoover et al . have demonstrated that phrA is under catabolite repression ., The gene encoding phrA is expressed in galactose and repressed in glucose , and the phrA promoter region contains a cre ( catabolite response element ) site for CcpA catabolite repression 24 , 40 ., In contrast , we have not identified a cre site in the phrA2 promoter region ., Therefore , to maximally discern the input through PhrA2 in our experiment , we used a D39-derived strain with a deletion of phrA and grew it in chemically-defined medium with galactose as the sole sugar ., We found that exogenous PhrA2 interacts with the TprA regulon in non-PMEN1 strains ., Specifically , D39ΔphrA cultures were exposed to treatments with synthetic PhrA2 , PhrA , and scrambled peptides for an hour and gene expression of tprA and lanA was measured relative to no treatment ., Treatment with PhrA2 significantly induced expression of tprA and lanA by 11-fold and 2-fold , respectively ( Fig 11 ) ., Treatment with scrambled peptide showed no induction of gene expression in D39ΔphrA ., The extent of lanA induction by PhrA is lower in the D39ΔphrA strain than in experiments with the WT strain ( Fig 10A ) , we presume this difference is due to the absence of phrA-autoinduction in the mutant strain ., These findings suggests that PhrA2 can be internalized by strains outside the PMEN1 lineage and induce changes in their gene expression ., Our findings demonstrate that acquisition of the TprA2/PhrA2 QS system by horizontal gene transfer into the PMEN1 lineage has endowed these strains with a virulence determinant and a mechanism to regulate its expression and thereby control disease ., PMEN1 ( ST81 ) lineage is postulated to have evolved from an ancestor in 1967 , and by the end of 1990s it represented an estimated 40% of penicillin resistant strains in US 14 , 41 ., These strains display very high rates of carriage 2 , 3 , 41 , 42 ., PMEN1 also displays very high rates of disease 2 , 3 , 43 ., Is the prevalence of PMEN1 in invasive disease a function of its carriage rates or does it reflect a propensity to cause disease ?, Multiple studies have shown that sequence types vary regarding their propensity to cause disease 44–47 and Sjostrom et al . show that PMEN1 displays a low propensity to cause invasive disease 47 ., Thus , high rates of PMEN1 invasive disease in the population likely reflect high carriage rates , and not heightened virulence potential ., In this context , it is possible that acquisition of the TprA2/PhrA2 by PMEN1 strains contributes to its low proclivity to cause invasive disease ., TprA2/PhrA2 may provide PMEN1 strains with the means to manipulate gene expression in neighboring strains from other lineages in multi-strain infections ., We show that synthetic C-terminal PhrA2 can stimulate expression of the TprA/PhrA system as well as its associated lantibiotic biosynthesis cluster in distantly related strain D39 ( Figs 11 and 12 ) ., We have observed that the expression of PMEN1-phrA2 is six fold that of D39-phrA in rich media , thus exemplifying a condition where PMEN1-phrA2 expression is high when D39-phrA is low ( S3 Fig ) ., We are currently investigating this interaction in physiologically relevant conditions ., The activation of phrA in response to galactose has led to the conclusion that TprA/PhrA may promote colonization in the nasopharynx where free sugars are rare and pneumococci survive by breaking down host mucins to free complex sugars , most prominently galactose 24 ., However , experiments with TprA/PhrA in the murine model demonstrate that this system is a virulence determinant in multiple models of pneumococcal disease ( personal communication , Motib and Yesilkaya ) , in this manner , PhrA2 may trigger a virulence regulon in neighboring strains ., We propose that PhrA2 signaling across systems is physiologically relevant in multi-strain infections ., We conclude that PhrA2 peptide is secreted by PMEN1-cells , since cell-free culture supernatants reiterate the function of extracellular addition of synthetic PhrA2 ., We predict that export occurs via the Sec secretion system , consistent with other peptides from the PlcR family of regulator-peptide pairs 48–50 ., Import must occur via a relatively widespread transporter , given that PhrA2 can influence D39 gene expression ., Further , the high sequence similarity between the functional C-termini of PhrA and PhrA2 suggests common import machinery ., The oligopeptide permease amiACDEF has been shown to be required for import of processed PhrA , and its homologues are required for import of PlcR-associated peptides in other species 48–50 ., Thus , amiACDEF is a high value candidate for a PhrA2 importer ., Sequence comparisons suggest that LcpA is a bacteriocin , however its function remains unknown ., We propose that its effect on virulence is not the result of bacteriocidal activity given that mouse experiments where performed with single strains ., However , we cannot exclude the possibility that an interaction between LcpA and the natural microbiome of the mouse influences the outcome of the infection ., The function of LcpA is under investigation ., We have identified and characterized a new quorum sensing system from the emerging RRNPP family ., TprA2/PhrA2 consists of a negative regulator of a lanthionine containing peptide and a cognate activating peptide ., Our findings suggest that this system has provided PMEN1 with the ability to control LcpA virulence and perhaps influence its propensity to cause invasive disease ., Finally , to our knowledge this is the first example of a gene transfer event that has integrated with an ancestral regulatory networks to control inter-strain gene regulation ., Laboratory animals were maintained in accordance with the applicable portions of the Animal Welfare Act and the guidelines prescribed in the DHHS publication , Guide for the Care and Use of Laboratory Animals ., The Office of Laboratory Animal Welfare ( OLAW ) Assurance of Compliance number is A3693-01 ., All chinchilla experiments were conducted with the approval of the Allegheny-Singer Research Institute ( ASRI ) Institutional Animal Care and Use Committee ( IACUC ) A3693-01/1000 ., Research grade young adult chinchillas ( Chinchilla lanigera ) weighing 400–600 grams were acquired from R and R Chinchilla Inc . , Ohio ., Animals were maintained in BSL2 facilities and all experiments were done while chinchillas were under subcutaneously injected ketamine-xylazine anaesthesia ( 1 . 7mg/kg animal weight for each ) ., For virulence studies , chinchillas ( a minimum of 10 in each cohort ) were infected with 100 CFUs/ear by transbullar inoculation within each middle ear ., During the course of the experiment ( 10 days ) , animals with severe acute infection perished; animals showing prolonged signs of discomfort were administered with pain relief ( Rimadyl , 0 . 1ml of 50mg/mL ) ) ., Animals with severe signs of pain and illness were euthanized by administering an intra-cardiac injection of 1mL potassium chloride after regular sedation ., All experiments involving mice were performed with prior approval of and in accordance with guidelines of the St . Jude Institutional Animal Care and Use Committee ., The St Jude laboratory animal facilities have been fully accredited by the American Association for Accreditation of Laboratory Animal Care ., All mice were maintained in BSL2 facilities and all experiments were done while the mice were under inhaled isoflurane ( 2 . 5% ) anesthesia ., Mice were monitored daily for signs of infection ., This work was approved under the IACUC protocol number 538-100013-04/12 R1 ., Mice were monitored for disease progression and euthanized via CO2 asphyxiation ., We performed a comparative genomic analysis of PMEN1 and non-PMEN1 strains to identify genes unique to the PMEN1 lineage 27 ., To this end , we used a set of 60 curated pneumococcal whole-genome sequences ( WGS ) , including four from the PMEN1 lineage ( S1 Table ) ., The set of 60 genomes includes the 44 genomes used for the first large-scale pneumococcal pangenome study 8 , additional genomes from PCV-7 immunized children 51 , as well as genomes from non-encapsulated strains 52 ., Together these strains reflect a large variety of multilocus sequence types ( MLSTs ) and serotypes , as well as strains isolated from different disease states and geographic locations ., To determine the distribution of tprA2 across pneumococcal strains we searched for this gene in the genome sequence of 215 PMEN1 isolates 13 ., A few genomes displayed disruption in the tprA2 locus , so the sequences were confirmed by PCR ., Primers to tprA2 and gapdh ( positive control ) were used to amplify these respective genes from genomic DNA ., The genomes from strains 111 ( ERS004810 ) , 11933 ( ERS005313 ) and HKP38 ( ERS004775 ) display substantial differences in the locus encoding TprA2/PhrA2 ., To search for cre sites we inspected the 190 basepairs upstream of phrA2 and before the start of tprA2 ., We searched for the cre site motif from L . lactis ( WGWAARCGYTWWMA ) , and allowed for up to three discrepancies as has been observed in a subset of S . pneumoniae cre 40 , 53 ., Wild-type S . pneumoniae strains PN4595-T23 ( GenBank ABXO01 ) and SV36 ( GenBank ADNO01 ) , graciously provided by Drs ., Alexander Tomasz and Herminia deLancastre , were used as PMEN1 representatives 31 ., Strains 111 ( ERS004810 ) , 11933 ( ERS005313 ) and HKP38 ( ERS004775 ) were shared by Drs ., Julian Parkhill and Stephen Bentley , and originally obtained from Drs ., Lesley McGee , Mark can der Linden , So Hyun Kim and Jae Hoon Song ., For growth on solid media , S . pneumoniae ( PN4595-T23 ) and isogenic mutants were streaked on TSA II plates with 5% sheep blood ( BD BBL , New Jersey , USA ) ., For growth in liquid culture , colonies from a frozen stock were grown overnight on TSA plates , inoculated into Columbia broth ( Remel Microbiology Products , Thermo Fisher Scientific , USA ) , and incubated at 37°C and 5% CO2 without shaking ., Columbia broth contains 10mM glucose ., Experiments in chemically defined media ( CDM ) were performed utilizing previously published recipe 40 , and galactose was used at a final concentration of 55mM ., Growth in CDM was initiated by growing a pre-culture for 9 hours and back dilution to OD600 0 . 1 to initiate a culture ., All deletion mutant strains were generated by site-directed homologous recombination where the target region was replaced with the spectinomycin-resistance gene ( aadR ) or kanamycin-resistance gene , as previously described 27 54 ., Briefly , ~2kb of flanking region upstream and downstream of the deletion target were amplified from the parental strain by PCR using Q5 2x Master Mix ( New England Biolabs , USA ) generating flanking regions , and the spectinomycin resistant gene was amplified from the plasmid pR412 ( provided by Dr . Donald Morrison ) ., Assembly of the transforming cassette was achieved either by sticky-end ligation of restriction enzyme-cut PCR products or by Gibson Assembly using NEBuilder HiFi DNA Assembly Cloning Kit ., The resulting construct was transformed into PN4595-T23 and confirmed using PCR and DNA sequencing ., Complement strains were made by generating a cassette where ~100bp of the 5’UTR and the CDS of the gene to be complemented were fused at the 3’ end of an antibiotic selection cassette lacking a transcription terminator ., This cassette was introduced in the genome of the strain at one of the two regions: the intergenic region between the orthologues of spr_0515 and spr_0516 , an inert genomic region that has been successfully employed in other constructs in the lab , or the bga region a commonly employed site for complementation 55 ., After subsequent transformation , qRT-PCR ( LightCycler480 , Roche Life Sciences , USA ) was done to verify the levels of expression of the complemented gene ., Primers used to generate the constructs are listed in S4 Table ., For all bacterial transformations , about 1μg of transforming DNA was added to the growing culture of a target strain at OD600 of 0 . 05 , supplemented with 125μg/mL of CSP2 ( sequence: EMRISRIILDF | Introduction, Results, Discussion, Materials and methods | Streptococcus pneumoniae ( pneumococcus ) is a leading cause of death and disease in children and elderly ., Genetic variability among isolates from this species is high ., These differences , often the product of gene loss or gene acquisition via horizontal gene transfer , can endow strains with new molecular pathways , diverse phenotypes , and ecological advantages ., PMEN1 is a widespread and multidrug-resistant pneumococcal lineage ., Using comparative genomics we have determined that a regulator-peptide signal transduction system , TprA2/PhrA2 , was acquired by a PMEN1 ancestor and is encoded by the vast majority of strains in this lineage ., We show that TprA2 is a negative regulator of a PMEN1-specific gene encoding a lanthionine-containing peptide ( lcpA ) ., The activity of TprA2 is modulated by its cognate peptide , PhrA2 ., Expression of phrA2 is density-dependent and its C-terminus relieves TprA2-mediated inhibition leading to expression of lcpA ., In the pneumococcal mouse model with intranasal inoculation , TprA2 had no effect on nasopharyngeal colonization but was associated with decreased lung disease via its control of lcpA levels ., Furthermore , the TprA2/PhrA2 system has integrated into the pneumococcal regulatory circuitry , as PhrA2 activates TprA/PhrA , a second regulator-peptide signal transduction system widespread among pneumococci ., Extracellular PhrA2 can release TprA-mediated inhibition , activating expression of TprA-repressed genes in both PMEN1 cells as well as another pneumococcal lineage ., Acquisition of TprA2/PhrA2 has provided PMEN1 isolates with a mechanism to promote commensalism over dissemination and control inter-strain gene regulation . | Streptococcus pneumoniae ( pneumococcus ) , an important human pathogen , exhibits a dual lifestyle featuring asymptomatic colonization of the host on the one hand as well as infliction of severe local and systemic disease on the other ., In pneumococcal strains , differences in gene possession often lead to varied phenotypic outcomes ., Epidemiologically , pandemic strains of the PMEN1 lineage show high prevalence in disease as well as carriage , posing an interesting question on the composition and function of the genomic toolkit that leads to their widespread success ., Here , we characterize TprA2/PhrA2 sensory system , a genomic region acquired exclusively by the PMEN1 strains ., The system consists of a regulator-peptide pair that was horizontally acquired into PMEN1 along with its regulatory circuitry ., The regulatory peptide PhrA2 is receptive to cell density of PMEN1 cells and is an example of elegant communication signaling between bacterial cells ., The regulatory influence of PhrA2 extends beyond PMEN1 cells such that it controls genes of a widespread signaling system and virulence regulon in non-PMEN1 strains ., This work contributes to the knowledge of peptide-communication signals in pneumococcus and further adds a novel mechanism by which an ecologically successful linage may modify the transcriptomic and functional landscape of a multi-strain pneumococcal community . | chinchillas, medicine and health sciences, pathology and laboratory medicine, pneumococcus, ears, gene regulation, pathogens, rna extraction, microbiology, vertebrates, animals, mammals, middle ear, regulator genes, gene types, extraction techniques, bacteria, bacterial pathogens, research and analysis methods, genomics, medical microbiology, gene expression, streptococcus, microbial pathogens, comparative genomics, head, rodents, anatomy, genetics, biology and life sciences, computational biology, amniotes, organisms | null |
journal.pgen.1002976 | 2,012 | Recovery of Arrested Replication Forks by Homologous Recombination Is Error-Prone | Maintenance of genome stability requires the faithful and accurate replication of the genetic material ., Genome instability is a hallmark for most types of cancer and it is strongly associated with predisposition to cancer in many human syndromes ( for a review , see 1 , 2 ) ., Genome instability is manifest at two levels: at the nucleotide level , resulting in base-substitutions , frame-shifts or in micro-insertions/deletions ( referred to herein as mutations ) ; and at the chromosomal level , resulting in duplications , deletions , inversions and translocations ( referred to herein as gross chromosomal rearrangements or GCRs ) ., Genome instability during cancer development and in other human genomic disorders may be consequences of failures in chromosome replication ( for a review , see 3 , 4 ) ., Origin spacing has recently been shown to cause chromosomal fragility at some human fragile sites 5 , 6 ., Impediments to replication fork movements per se may also cause genome instability 7–9 ., Indeed , both slowing down and blockages to fork progression can lead to chromosomal fragilities or GCRs in human cells and yeast models 10–14 ., However , how a blocked replication fork leads to genetic instability remains poorly understood ., In eukaryotes , DNA replication is initiated at numerous origins along linear chromosomes , and impediments to fork progression appear unavoidable during each S-phase ( for a review , see 9 , 15 ) ., Impediments to fork progression can be caused by DNA lesions , by non-histone proteins tightly bound to DNA , by sequence-caused secondary structures such as cruciform structures and possibly G-quadruplexes , by nucleotide pool imbalance and by conflicts with transcription machinery ( for a review , see 16 , 17 ) ., In case of failures in fork progression , DNA replication can be completed either by the recovery of the arrested fork by fork-restart mechanisms , or as a result of the progression of a converging fork which can be ensured by activation of dormant origins 7 , 15 , 18 ., Fork restart is presumably essential in unidirectional replication regions , such as the rDNA locus , in regions of low densities of origins , such as some human fragile sites , and when two converging forks are both impeded 5 , 19 , 20 ., To ensure faithful and complete DNA replication , cells coordinate DNA synthesis restart with specific pathways , including DNA replication checkpoint and homologous recombination mechanisms 17 ., The integrity of replication forks is guaranteed by the DNA replication checkpoint that maintains the replisome in a replication-competent state to keep DNA polymerases at the site of nucleotide incorporation 21 ., It remains unclear how the DNA replication checkpoint modulates replisome activities to maintain its function 21 , 22 ., The DNA replication checkpoint also regulates nuclease activities ( e . g . Exo1 or Mus81 ) which contribute to preserving the integrity of stalled forks 23 , 24 ., If replisome function is lost or the replisome dissociates at broken replication forks , the resumption of DNA synthesis appears to require the replisome to be rebuilt ., In E . coli , restart of a collapsed fork involves homologous recombination and the PriA helicase that allows replisome components to be loaded de novo on joint-molecule structures 25 , 26 ., In eukaryotes , the restart of collapsed or broken replication forks is dependent upon homologous recombination , but the mechanism of origin-independent loading of the replisome remains to be described 20 , 27–30 ., It has been proposed that the repair of a double-strand break ( DSB ) by recombination ( break-induced replication , BIR ) in budding yeast similarly involves the assembly of a replication fork ( for a review , see 30–32 ) ., When BIR occurs outside S-phase , recombination-dependent replication fork assembly can synthesise hundreds of kilobases ( Kb ) ., However , this DNA synthesis is highly inaccurate due to frequent template switching of nascent-strands and frame-shift mutations 33 , 34 ., We previously reported a system that displays replication fork arrest at a specific locus in the fission yeast S . pombe ., The system is a polar replication fork barrier ( RFB ) regulated by the Rtf1 protein binding to its RTS1 binding site 35 ., The RTS1-RFB causes fork arrest because of a non-histone protein complex binding to the DNA ., As proposed for other polar RFBs , the RTS1-RFB is thought to block fork progression by directly ( contact between proteins and the replisome ) or indirectly ( topological constraint ) affecting the replicative helicase activity and thereby preventing DNA unwinding 36 , 37 ., Recovery of the arrested fork occurs by a DSB-independent mechanism and involves the recruitment of recombination proteins at the RTS1-RFB site ., We proposed that recombination proteins associate with unwound nascent strands that then anneal with the initial template to allow DNA synthesis to restart 11 , 20 ., The causative protein barrier then has to be removed either by DNA helicase or by the recombination machinery itself to allow fork-progression to resume 38–40 ., Occasionally , the unwound nascent strand can mistakenly anneal with a homologous template in the vicinity of the collapsed fork , resulting in the restart of DNA synthesis on non-contiguous template ., This incorrect template switch of nascent strands results in inversions and iso-acentric and dicentric chromosomes in ∼2–3% of cells/generation 11 , 20 ., Error-free template switching between sister-chromatids provides an efficient mechanism for filling-in single-stranded gaps left behind damage-induced stalled forks 41 ., Inverted chromosome fusions in yeast and rare-genome rearrangements in human genomic disorders , may both be consequences of template switching between ectopic repeats associated with impeded replication forks 8 , 14 ., Here , we used the RTS1-RFB to investigate the consequences of fork collapse on genome instability ., We report that recovery from a collapsed fork is associated with a high frequency of instability , with a single fork arrest increasing the rates of mutation , deletion and translocation by 10 , 40 and 5 fold , respectively ., We show that genetic instability associated with fork arrest is dependent on homologous recombination ., Fork-arrest-induced GCRs ( deletion and translocation ) result from inappropriate ectopic recombination at the site of the collapsed fork ., We also demonstrate that restoration of fork progression by homologous recombination results in error-prone DNA synthesis due to frequent replication slippage between short tandem repeats ., We investigated the molecular mechanisms of this replication slippage and found that post-replication repair , including ubiquitination of PCNA or trans-lesions-synthesis ( TLS ) DNA polymerases , is not involved in fork-arrest-induced replication slippage ., Micro-deletions/insertions flanked by micro-homology associated with copy number variations ( CNVs ) in cancer cells or in response to replication stress may therefore be scars left following the restoration of forks progression by homologous recombination ., We generated fork arrest constructs by manipulating the polar RTS1-RFB ( Figure 1A ) ., We introduced the RTS1 sequence on the centromere-proximal ( cen-proximal ) side of the ura4 locus , 5 kb away from the strong replication origin ( ori ) 3006/7 on chromosome III ., This created the t-ura4<ori locus , in which “t” and “ori” refer to the telomere and the origin 3006/7 , respectively; and “<” and“ >”refer to the RTS1-barrier and its polarity that is whether it blocks replication forks travelling from the ori 3006/7 towards the telomere or forks travelling from the telomere towards the ori 3006/7 , respectively ., We previously confirmed that forks moving from ori 3006/7 towards the telomere ( tel ) are efficiently blocked by the RTS1-RFB at the t-ura4<ori locus 35 ., In this model system , fork arrest is activated by inducing the expression of rtf1+ gene that is under control of the thiamine repressible promoter nmt41 ., Thus , the RTS1-RFB is inactivated by adding thiamine to the media and it is activated in thiamine-free media ., Efficient induction of Rtf1 expression requires incubation for 12–16 hours in thiamine-free media ., Replication intermediates were analysed by native 2-dimensional gel electrophoresis ( 2DGE ) ., In conditions of Rtf1 expression , more than 95% of replication forks were blocked by the RTS1-RFB at the t-ura4<ori locus ( see black arrow on Figure 1B , t-ura4<ori ON ) ., Arrested forks were not detected without Rtf1 induction ( Figure 1B , t-ura4<ori OFF ) 20 ., The RTS1 sequence was inserted on the tel-proximal side of ura4 creating the t<ura4-ori locus ., 2DGE analysis of this construct revealed a strong fork arrest signal on the descending large Y arc ( Figure 1A and 1B , t<ura4-ori ON ) ., The ura4+ gene , used in this system as a reporter to score genetic instability , is located behind the arrested fork when the RTS1-RFB is active at the t<ura4-ori locus and ahead of the arrested fork at the t-ura4<ori locus ., This explains the distinct position of the arrested fork signal on the Y arc ., Inversion of the RTS1 sequence at the tel-proximal side of ura4 created the t>ura4-ori locus and no fork arrest signal was detected for this construct by 2DGE when Rtf1 was expressed ( Figure 1A and 1B , t>ura4-ori ON ) ., Thus , RTS1 behaves as a polar RFB at the ura4 locus , and replication across this locus is strongly unidirectional due to the relative positions of the origins 42 ., Introducing a second RTS1 sequence , such that the two RTS1 sequences are inverted repeats ( IRs ) , created t>ura4<ori and t<ura4>ori loci ( Figure 1A and 1B , t>ura4<ori and t<ura4>ori ON ) ., Given the orientation of the polar RTS1-RFB in the t<ura4>ori strain , converging forks cannot be blocked ., Whereas block of converging forks can virtually occur in the t>ura4<ori strain , 2DGE in this construct revealed that forks arrested on the cen-proximal side of ura4 were efficiently recovered by recombination before forks are arrested on the tel-proximal side ., Indeed , joint-molecules ( JMs ) resulting from recombination between RTS1 repeats were detected by 2DGE ( see red arrows on Figure 1B , t>ura4<ori and t<ura4>ori ON ) ., Resolution of these JMs gives rise to chromosomal rearrangements 20 ., In the absence of homologous recombination ( i . e . in a rad22-d mutant ) , JMs were not detected and termination signals accumulated ( see green arrow on Figure 1B , t>ura4<ori rad22-d strain ) ., Similarly , termination signals accumulated in the rad22-d t-ura4<ori strain ( see green arrow on Figure 1B , t-ura4<ori rad22-d ) , showing that , when arrested forks are not restarted by homologous recombination , the RTS1-RFB behaves as a hot spot for replication termination 20 ., We investigated fork-arrest-induced genome instability by selecting for cell resistance to 5-FOAR , the result of loss of ura4+ function ., Inducing fork-arrest at t-ura4<ori increased ura4 loss 3 fold ( Table 1 ) ., Rtf1 expression in the t-ura4-ori and t>ura4-ori strains did not cause site-specific fork-arrest at ura4 as assessed by 2DGE and did not increase the rate of ura4 loss ., Thus , ura4 loss results from the RTS1-RFB activity and not simply from the presence of RTS1 and/or Rtf1 expression ( Table 1 ) ., To investigate the nature of this genetic instability , primers were designed to amplify the ura4 coding sequence and , as a control , the essential rng3 gene , mapping 30 kb tel-proximal to ura4 , that should not be rearranged ( Figure 2A and 2B ) 35 ., The absence of ura4 amplification was classified as a deletion event; sequencing of amplified ura4 sequence was used to identify point mutation events ( Figure 2B ) ., A single arrested fork at the t-ura4<ori locus was sufficient to increase the rate of genomic deletion up to 40 times over spontaneous events ( i . e . in the t-ura4-ori strain , p\u200a=\u200a0 . 006 ) ( Figure 2C and Figure S1A ) ., Fork-arrest-induced deletion was recombination-dependent ., Spontaneously ( i . e . when the RTS1-barrier was inactivated ) , the rate of genomic deletion in rad22-d or rhp51-d strains was higher than that in the wild-type strain ( Figure S1B ) ., Nonetheless , no further increase in the rate of genomic deletion was observed in the surviving rad22-d or rhp51-d cells upon activation of the RTS1-barrier ( Figure S1B , t-ura4<ori ) ., Frequent spontaneous genomic deletion in the rad22-d or rhp51-d strains is consistent with previous reports showing that mutations in recombination genes are associated with an increase level of GCRs 14 , 43 , 44 ., Deleting the natural RTS1 sequence from chromosome II abolished deletion events at collapsed forks , indicating that fork-arrest-induced deletion was also mediated by inter-chromosomal recombination ( Figure 2C and t-ura4<ori RTS1-d on Figure S1A ) ., Thus , these data are consistent with the view that homologous recombination makes a major contribution to suppressing genome instability , but can occasionally drive non allelic recombination events leading to GCRs 35 , 45 ., We detected no fork-arrest-induced deletion in the t<ura4-ori strain , in contrast to the t-ura4<ori strain ( Figure S1A and Figure 2C ) ., The ura4 marker is located behind and ahead of collapsed forks in the t<ura4-ori and t-ura4<ori strains , respectively ( Figure 1A ) ., Therefore , replicated regions , located behind collapsed forks , do not display instability , and fork-arrest-induced deletion occurs within unreplicated regions immediately in front of arrested forks ., Overall , our data establish that genomic deletion at collapsed forks results from inappropriate recombination between ectopic sequences during the process of fork recovery by recombination proteins ., Inverted repeats ( IRs ) are structural elements often associated with genome rearrangements 11 , 14 , 46 , 47 ., We investigated the effects of IRs in the vicinity of the RTS1-RFB on fork-arrest-induced genomic deletion ., We first compared the t>ura4<ori strain ( IRs flanking ura4 ) to the t-ura4<ori strain ( no IRs near the RTS1-RFB ) ., The rate of fork-arrest-induced genomic deletion was 200 times higher in the t>ura4<ori than that in the t-ura4<ori strain ( p\u200a=\u200a0 . 009 , Figure 2C and Figure S1A ) ., Thus , intra-chromosomal ectopic recombination permitted by the RTS1 sequence on the tel-proximal side of ura4 accounted for 99 . 5% of the genomic deletions observed in the t>ura4<ori strain ( Figure 2C , compare with t-ura4>ori ) ., Preventing inter-chromosomal recombination by deleting RTS1 from the chromosome II ( t>ura4<ori RTS1-d ) abolished 90% of deletion events ( Figure 2C and Figure S1A ) ., Thus , genomic deletions induced by fork-arrest near IRs are due to inter- and intra-chromosomal recombination events ., In support of this , stimulation of fork-arrest-induced deletion by IRs is mediated by homologous recombination ., Indeed , the rate of genomic deletion was not increased upon induction of the RTS1-RFB in the surviving population of t>ura4<ori rad22-d and rhp51-d strains ( Figure S1B ) ., These data indicate that IRs favour genomic deletion at collapsed forks by promoting inappropriate inter- and intra-chromosomal recombination during fork recovery by recombination proteins ., We verified that our data were not influenced by the orientation of IRs or by rare blocking of converging forks in the t>ura4<ori strain ., We analysed the t<ura4>ori construct in which RTS1 repeats are in the opposite orientations relative to the t>ura4<ori construct , such that forks converging towards ura4 cannot be blocked ( Figure 1 ) ., The rate of fork-arrest-induced genomic deletion was 1 , 000 times higher in the t<ura4>ori than that in the t<ura4-ori strain , that does not contain IRs near the RTS1-RFB ( p\u200a=\u200a0 . 008 , Figure 2C and Figure S1A ) ., Thus , intra-chromosomal recombination , permitted by the RTS1-RFB sequence on the cen-proximal side of ura4 , accounted for nearly 100% of the genomic deletions observed in the t<ura4>ori strain ( Figure 2C , compare with t<ura4-ori ) ., Preventing inter-chromosomal recombination by deleting RTS1 from the chromosome II ( t<ura4>ori RTS1-d ) abolished 90% of deletion events ( Figure 2C and Figure S1A ) ., Importantly , the deletion rates for the t<ura4>ori and t>ura4<ori strains were not significantly different ( Figure 2C ) , showing that IRs cause genomic deletion at collapsed forks irrespective of their orientations and independently of blockage of converging forks ., Fork-arrest at t>ura4<ori results in translocations between ectopic RTS1 repeats on chromosomes II and III 35 ., We investigated the influence of IRs on fork-arrest-induced translocation ., We designed primers to amplify the predicted translocation junction between chromosomes II and III ( TLII and TLIII on Figure 2A and 2B ) ., A single arrested fork at the t-ura4<ori locus was sufficient to increase the translocation rate to 5 times higher than the spontaneous rate ( p\u200a=\u200a0 . 002 , Figure 2D and Figure S1C ) ., The translocation rate for the t>ura4<ori construct ( containing IRs ) was 1 , 500 fold higher than that for the t-ura4<ori strain that does not contain IRs near the RTS1-RFB ( p\u200a=\u200a0 . 009 , Figure 2D and Figure S1C ) ., Thus , intra-chromosomal recombination accounted for nearly 99% of translocation events observed in the t>ura4<ori construct ( Figure 2D and Figure S1C , compare with t-ura4<ori ) ., No translocation events were detected when inter-chromosomal recombination was prevented by deleting RTS1 from the chromosome II ( t>ura4<ori RTS1-d on Figure 2B ) ., Therefore , as reported for genomic deletions , fork-arrest-induced translocation associated with IRs is due to inter- and intra-chromosomal ectopic recombination ., No translocations were detected in the t<ura4>ori strain ( data not shown ) , so we cannot formally exclude the possibility that fork-arrest-induced translocations in the t>ura4<ori strain was caused by blocking of converging forks ., However , as no translocation event occurred in the absence of Rad22Rad52 or Rhp51Rad51 , it is most likely that translocations occur during fork recovery by recombination ( Figure 2B and 35 ) ., Overall , our data indicate that recovery of a single collapsed fork causes translocations and IRs near the site of fork-arrest stimulate translocations by promoting inappropriate inter- and intra-chromosomal recombination ., Fork-arrest-induced GCRs are caused by inter- and intra-chromosomal recombination ., We noticed a slightly greater contribution of intra- than inter-chromosomal recombination ( Figure 2C ) ., This is consistent with ectopic recombination preferentially occurring at the most proximal homologous sequence , as previously reported 48 ., Nonetheless , the rate of fork-arrest-induced deletion in the t>ura4<ori strain ( 8 . 4 10−7 ) was not simply the sum of the rates of intra-chromosomal recombination events ( 9 . 9 10−8 in the t>ura4<ori RTS1-d strain ) and inter-chromosomal recombination events ( 4 10−9 in the t-ura4>ori strain ) ., Similar reasoning can be applied for the t<ura4>ori strain ., Thus , independent intra- and inter-recombination events cannot themselves explain high rate of GCRs induced by fork arrest near IRs ., Therefore , we infer that there is interplay between inter- and intra-chromosomal recombination such that fork-arrest-induced GCRs may involve recombination between three homologous sequences ( tri-parental recombination ) ., To confirm that fork-arrest-induced GCRs are the result of inappropriate ectopic recombination during fork recovery , we analysed the involvement of the RecQ helicase Rqh1 ., We previously reported that Rqh1 limits inappropriate template switching of stalled nascent strands without affecting the efficiency of fork restart 20 ., In the t-ura4<ori construct ( in which only inter-chromosomal recombination is possible ) , fork-arrest-induced deletion and translocation rates were 31 and 109 times higher in the rqh1-d strain than that in the wild-type control , respectively ( p\u200a=\u200a0 . 0003 , Figure 2D–2E and Figure S1C ) ., For the t>ura4<ori construct ( containing IRs near fork-arrest ) , fork-arrest-induced deletion and translocation rates were 5 times higher in the rqh1-d than that in the wild-type control ( p\u200a=\u200a0 . 0007 , Figure 2D–2E and Figure S1C ) ., Thus , Rqh1 limits GCRs at collapsed forks by preventing inappropriate ectopic recombination during the process of fork recovery by recombination proteins ., We analysed the effects of collapsed forks on the mutation rate ., We sequenced the ura4 coding sequence from 5-FOAR isolated cells and identified base-substitutions , frame-shifts and small insertions and duplications between short tandem repeats ( Table 2 ) ., A single collapsed fork in the t-ura4<ori strain increased the overall mutation rate up to 10 times over spontaneous events ( Figure 3A , p\u200a=\u200a0 . 003 ) ., Similar increases in the overall mutation rate were found for the strains with IRs near the arrested fork and those with RTS1 deleted from chromosome II ( Figure 3A and Figure S2A ) ., Thus , fork-arrest-induced mutation is not mediated by inappropriate ectopic recombination ., Induction of the RTS1-RFB in the t<ura4-ori strain did not increase the mutation rate of the ura4 gene ., Thus , as for GCRs , replicated regions behind arrested forks are not prone to mutation ., This observation rules out the hypothesis that fork-arrest-induced mutation is a consequence of the accumulation of damaged single-stranded DNA behind collapsed forks ( see discussion ) ., Our data suggest that recovery from collapsed forks results in error-prone DNA-synthesis ., We then analysed the spectra of mutations found in the ura4 ORF by sequencing the PCR products ., The rates of base-substitutions and frame-shifts were not significantly increased by the RTS1-RFB activity over spontaneous events ( i . e . compare to t-ura4-ori strain , Figure 3C and Table 2 ) ., In contrast , the rate of deletions and duplications ( Del/Dup ) flanked by short homology was increased by 7 times over spontaneous events in the t-ura4<ori strain , but not in the t<ura4-ori strain ( Figure 3C and Table 1 ) ., These data further confirm that fork-arrest does not promote mutation events behind collapsed forks ., We used reversed mutation assays to test if fork-arrest at the RTS1-RFB specifically induced Del/Dup mutations ., We made use of strains harbouring a single mutation within the ura4 ORF: either a single base-substitution or a −1 frame-shift in homo-nucleotide ( Figure S2B ) ., We also studied strains harbouring either a duplication of 20 or 22 nt flanked by 5 or 4 bp of micro-homology , respectively ( defined as ura4-dup20 and ura4-dup22 , Figure S2B ) ., These non-functional ura4− alleles were inserted in front of the RTS1-RFB in the t-ura4<ori configuration and we then tested whether fork arrest could restore a functional ura4+ gene ., Activation of the RTS1-RFB at ura4 increased the frequency of Ura+ revertants up to 15 and 7 times in strains harbouring ura4-dup22 and ura4-dup20 , respectively ( Figure 3D and Figure S2B ) ., Thirty Ura+ colonies were studied by PCR and all gave a product of the same size as the wild-type ura4+ gene: they had therefore lost the duplication ( Figure 3E and data not shown ) ., Sequencing the full ura4 ORF confirmed that Ura+ revertants contained an intact ura4+ sequence , showing that the reversion of these alleles was due solely to the precise deletion of 20 or 22 nt ( Figure 3F and data not shown ) ., In contrast , activation of the RTS1-RFB did not increase the frequency of Ura+ revertants of strains harbouring ura4 alleles with a single base-substitution or a −1 frame-shift ( Figure 3D and Figure S2B ) ., Thus , collapsed forks tend to induce deletion events between short tandem repeats rather than base-substitution or frame-shift mutations ., Among Del/Dup events , deletions represented the two-third of events in the t-ura4<ori strain ( Table 2 ) ., The median size of Del/Dup events was 24 and 22 nt respectively , and Del/Dup occurred between short direct repeats 1 to 10 nt long ( Figure S3 ) ., Thus , the ura4-dup20 and ura4-dup22 alleles used in the reverse mutation assay were representative of the Del/Dup events observed ., Del/Dup flanked by micro-homology result from intra-molecular template switching mechanisms in which nascent strands dissociate from the template and misalign with the template when restarting the elongation step ., This leads to loop formation , either in the nascent strand or in the template , resulting in duplication or deletion events , respectively 49 ., Consequently , we will hereafter refer to Del/Dup as replication slippage ., Replication slippage was observed all along the ura4 ORF and up to 1 . 2 kb ahead of the arrested fork , even if a hot spot of deletion was present 500 bp away from the RTS1-RFB ( Figure 3G and Figure S3B ) ., Thus , our data suggest that the DNA synthesis is prone to replication slippage at least for the first 1 , 200 nt synthetized during the recovery of collapsed forks ., Inaccuracy of DNA synthesis on further distances was not directly addressed ., To confirm that replication slippage occurs as forks recover , and not behind the fork in the DNA already replicated , we inserted the ura4-dup20 or the ura4-dup22 allele either behind ( in the t<ura4-ori configuration ) or in front of the RTS1-RFB ( in the t-ura4<ori configuration ) ( Figure 4 ) ., This allows the analysis of the same event of replication slippage behind and ahead of collapsed forks ., In the t-ura4<ori configuration , induction of the RTS1-RFB resulted in a 8 and 16 fold increases in the replication slippage frequency for the ura4-dup20 and ura4-dup22 alleles , respectively ( Figure 4A and 4B ) ., Similar increases in the rate of replication slippages were observed ( Figure 4C ) ., In contrast , in the t<ura4-ori background , the frequency of replication slippage was induced by only 2–3 fold by the RTS1-RFB ( Figure 4B–4C ) ., These data confirm that DNA located ahead of collapsed forks is more prone to replication slippage than replicated DNA adjacent to arrested forks , further evidence that replication slippage arises during fork recovery ., Replication slippage occurs in DNA in front of ( and not behind ) the arrested fork , this DNA being replicated only after restart of the fork ., Thus , a defect preventing fork recovery would be expected to abolish the error-prone DNA synthesis during restart ., We analyzed fork-arrest-induced mutation in recombination mutants in which collapsed forks at the RTS1-RFB cannot recover , resulting in cell death ., Induction of the RTS1-RFB did not increase the overall mutation rate in the surviving populations of t>ura4<ori or t-ura4<ori rad22-d and rhp51-d strains ( Figure 3B ) ., In addition , only 7% of mutation events in the survivors of the rad22-d t-ura4<ori strain were Del/Dup mutations , compared to 40% in the wild-type strain ( Figure 3C and Table 1 ) ., We currently cannot assess mutation events associated with defects in fork recovery because this appears to be lethal in the absence of recombination ., Nevertheless , our data are consistent with fork-arrest-induced replication slippage being dependent on homologous recombination ., The rad22-d and rhp51-d strains are themselves spontaneously mutagenic ., Consequently , any small increase in the fork-arrest-induced mutation rate might be masked by the high frequency of spontaneous 5-FOAR cells in rad22-d and rhp51-d strains ., We therefore used a more specific mutation assay , based on the ura4-dup20 allele , to determine the rate of replication slippage induced by the RTS1-RFB over spontaneous events ., Strains carrying mutations in recombination genes grow slowly , so replication slippage was scored as a function of the number of generations following thiamine removal ( i . e . generations subject to fork arrest at ura4 ) ( Figure 4D and 4E ) ., In the wild-type strain , fork arrest at the RTS1-RFB resulted in a 10 fold-increase in the frequency of replication slippage , as expected ., In recombination mutants ( rad50-d , rhp51-d and rad22-d ) , fork-arrest at the RTS1-RFB increased the frequency of replication slippage by only 2 times over spontaneous events: therefore , replication slippage occurs less frequently in survivors from recombination mutants than those from the wild-type strain ( Figure 4D–4F ) ., Based on 2DGE analysis , fork-restart is severely impaired in the absence of Rad22Rad52 ( Figure 1B and 20 ) , such that even the two-fold induction in replication slippage by fork arrest in the rad22-d strain was surprising ., The rad22-d strain accumulates suppressors involving the Fbh1 helicase that limits Rhp51Rad51- dependent recombination at blocked replication forks 50 , 51 ., Therefore , we analyzed replication slippage in the rad22-d rhp51-d double mutant in which no homologous recombination event occurs ., In this background , there was no detectable fork-arrest-induced replication slippage ( Figure 4D–4F ) ., Thus , complete defect in fork restart results in a complete abolition of fork-arrest-induced replication slippage in the surviving population ., Overall , our data establish that replication slippage results from inaccurate DNA synthesis during the restart of collapsed forks by recombination ., We investigated the effects of replication stress , other than the replication block imposed by the RTS1-RFB , on replication slippage ., Strains harbouring ura4− alleles ( base-substitutions , −1 frame-shift , and ura4-dup20 ) were exposed to replication-blocking agents or UV-C-induced DNA damages and the frequency of Ura+ revertants was scored ., Three hours of treatment with either the topoisomerase I inhibitor camptothecin ( CPT ) or mitomycin C ( MMC ) , an inter-strand cross-linking agent ( ICls ) , increased the frequency of Ura+ revertants by 3 to 4 fold in the ura4-dup20 strain ( Figure 5A and 5B ) ., At equivalent survival ( 70–90% ) , DNA-damages induced by a dose of 100 J/m2 of UV-C did not increase the frequency of Ura+ revertants in the ura4-dup20 strain ., Increasing the UV-C dose ( 150 J/m2 ) resulted in an increased reversion effect ., The other ura4 alleles exhibited an opposite behaviour pattern ., As expected , UV-C-induced DNA damages , but not CPT or MMC treatment , increased the frequency of Ura+ revertants of the base-substitution and the −1 frame-shift mutants ( Figure 5A ) ., Thus , replication slippage , unlike other point mutations , appears to be a mutation event specifically induced by replication stress ., Hydroxyurea ( HU ) that prevents the bulk of dNTP synthesis during S-phase by inhibiting the ribonucleotide reductase , results in a slow-down of fork progression which did not induce replication slippage ( Figure 5A ) ., In contrast , CPT and MMC treatments that lead to replication stress by causing fork collapse induced replication slippage ., Homologous recombination is repressed during HU treatment and recombination proteins are recruited to collapsed but not stalled forks 52–54 ., Consistent with this , we found that the rad22-d mutant is highly sensitive to acute exposure to CPT , but not to HU ( Figure S4 ) ., Thus , acute exposure to HU results in stalled forks that recover without recombination , whereas recombination may be required for restarting forks that have collapsed due to CPT or MMC treatment ., We confirmed that CPT-induced replication slippage results from collapsed forks and was thus S-phase specific: the ura4-dup20 strain was synchronized in early S-phase by HU treatment and released into S-phase with or without CPT ., HU-synchronization and release into DMSO ( used as vehicle for CPT ) did not induce replication slippage ., In contrast , the release of cells into S-phase in the presence of CPT stimulated replication slippage up to 12 fold ( Figure 5C ) ., These data in | Introduction, Results, Discussion, Materials and Methods | Homologous recombination is a universal mechanism that allows repair of DNA and provides support for DNA replication ., Homologous recombination is therefore a major pathway that suppresses non-homology-mediated genome instability ., Here , we report that recovery of impeded replication forks by homologous recombination is error-prone ., Using a fork-arrest-based assay in fission yeast , we demonstrate that a single collapsed fork can cause mutations and large-scale genomic changes , including deletions and translocations ., Fork-arrest-induced gross chromosomal rearrangements are mediated by inappropriate ectopic recombination events at the site of collapsed forks ., Inverted repeats near the site of fork collapse stimulate large-scale genomic changes up to 1 , 500 times over spontaneous events ., We also show that the high accuracy of DNA replication during S-phase is impaired by impediments to fork progression , since fork-arrest-induced mutation is due to erroneous DNA synthesis during recovery of replication forks ., The mutations caused are small insertions/duplications between short tandem repeats ( micro-homology ) indicative of replication slippage ., Our data establish that collapsed forks , but not stalled forks , recovered by homologous recombination are prone to replication slippage ., The inaccuracy of DNA synthesis does not rely on PCNA ubiquitination or trans-lesion-synthesis DNA polymerases , and it is not counteracted by mismatch repair ., We propose that deletions/insertions , mediated by micro-homology , leading to copy number variations during replication stress may arise by progression of error-prone replication forks restarted by homologous recombination . | The appropriate transmission of genetic material during successive cell divisions requires the accurate duplication and segregation of parental DNA ., The semi-conservative replication of chromosomes during S-phase is highly accurate and prevents accumulation of deleterious mutations ., However , during each round of duplication , there are many impediments to the replication fork machinery that may hinder faithful chromosome duplication ., Homologous recombination is a universal mechanism involved in the rescue of replication forks by rebuilding a replication apparatus at the fork ( by mechanisms that are not yet understood ) ., However , recombination can jeopardize genome stability because it allows genetic exchanges between homologous repeated sequences dispersed through the genome ., In this study , we employ a fission yeast-based arrest of a single replication fork to investigate the consequences of replication fork arrest for genome stability ., We report that a single blocked fork favours genomic deletions , translocations , and mutations; and this instability occurs during fork recovery by recombination ., We also report that a single arrested fork that resumes its progression by recombination is prone to causing replication slippage mediated by micro-homology ., We propose that deletions/duplications observed in human cancer cells suffering from replication stress can be viewed as scars left by error-prone replication forks restarted by recombination . | cellular stress responses, microbiology, mitosis, model organisms, dna replication, dna, dna synthesis, mycology, chromosome biology, schizosaccharomyces pombe, biology, molecular biology, yeast, cell biology, nucleic acids, genetics, yeast and fungal models, dna recombination, molecular cell biology, genetics and genomics | null |
journal.pcbi.1000258 | 2,009 | Evolutionary Sequence Modeling for Discovery of Peptide Hormones | G protein coupled receptors ( GPCRs ) probably represent the largest gene family , making up 3% of the mammalian genome 1 ., These proteins are made up of several subfamilies , including Class A rhodopsin-like , Class B secretin-like , Class C metabotropic glutamate/pheromone-like , and other nonmammalian receptors ., Within each class , there is a very large number of smaller subclassifications , such as a family of receptors for peptide hormones within rhodopsin-like receptors ., There are approximately 1 , 000 GPCRs , the vast majority of which are olfactory receptors , with more than 650 GPCRs in the rhodopsin family alone 2 ., A large number of these receptors have been identified only by computational methods , while others have been cloned and transfected into cells; however , the cognate neurotransmitter and the receptor functions for many GPCRs are currently unknown ., Any receptor for which the native neurotransmitter is unknown is considered an orphan receptor ., Of all the orphan receptors that remain , some percentage represents receptors for peptide hormones ., This large family of proteins is important not only from a basic science perspective , but because of their extracellular sites of action and importance as first messengers for cellular signaling , GPCRs have become a primary target for drug development ., In fact , over 30% of all pharmaceuticals act either as agonists or antagonists of GPCRs 3 ., Many pharmaceutical companies are identifying , cloning , and patenting new orphan GPCRs , with the hope that orphan receptors will ultimately lead to new drug development and new pharmaceutical agents ., Although the identification of putative GPCRs can be accomplished relatively easily , the discovery of the endogenous ligands that activate these receptors is far more difficult ., These ligands can exist as small molecules , lipids , peptides , or proteins 4 , 5 ., Many , such as ATP , may have important functions other than activating a GPCR ., Even within a class of hormones , there are seldom obvious clues that identify a new candidate ., This is particularly true within the family of peptide hormones , as they are processed from a larger species known as preprohormones 6 ., Peptide hormones , or neuropeptides , are a string of amino acids ranging from approximately 3 to 50 residues ., They are found within a larger protein ( a preprohormone ) , and the production of the actual hormone usually follows specific rules ., Preprohormones are secreted proteins , and each has a signal sequence that is necessary for the transport of the protein out of the Golgi complex into a secretory vesicle for processing and secretion where the signal sequence is removed , revealing the prohormone 7 ., In general , hormones are surrounded by a pair of basic residues , i . e . Arg-Arg , Arg-Lys , Lys-Arg , or Lys-Lys , which are found directly adjacent to the putative hormone ., These double basic residues act as recognition sites for processing enzymes , usually serine proteases that cleave the prohormone to liberate the active peptide 7 , 8 ., In many cases , there is more than a single active peptide within one precursor protein 6 ., Even with these common features , the identification of a peptide hormone from a DNA or protein sequence is very difficult ., Even though all of the GPCRs are obviously related based upon DNA or protein sequence , the neuropeptides that bind to the receptors are only obviously related within discrete families of prohormones ., For instance , the family of opioid-like peptides has four members ., These prohormones , proopiomelanocortin ( POMC ) , proenkephalin , prodynorphin , and pronociceptin ( proN/OFQ ) , share similar genomic structures and a very slight similarity of protein sequence , most notably the Y ( F ) GGF of enkephalin , β-endorphin , dynorphin , and N/OFQ 9 , 10 ., However , if one were to conduct a BLAST search in Genbank for DNA sequences similar to proenkephalin , one would not find any other neuropeptide ., Simple search strategies within Genbank are not adequate for identifying novel neuropeptides , especially those not belonging to known neuropepeptide families ., There is an additional feature of neuropeptides that may more clearly differentiate them from other types of molecules ., Neuropeptides are usually well conserved among various species ( rat , mouse , human ) , while the intervening sequences , presumably because they are simply discarded , are not well conserved 11 ., Here we describe a novel Hidden Markov Model ( HMM ) -based computational framework , the Match Profile HMM ( MPHMM ) method for neuropeptide identification based upon an approach that models spatial structure along the genomic sequence simultaneously with the temporal evolutionary path structure across species , and show how such models can be used to discover new functional molecules via cross-genomic sequence comparisons ., This computational tool was used to identify a novel prohormone , NPQ , containing up to four potential neuropeptides 12, The extended list of matches , the GUI SequenceMatcher , and the HIGHER tools will be made are available at http://www . cslu . ogi . edu/people/sonmezk/hormone ., Initially , we will enable the visualization of our ENSEMBL and CELERA runs via the GUI ., The next version will allow evolutionary HMM searches specified by the user ., The HIGHER codebase will also be made available at the website once it is ready for release ., We have presented a computational framework that is capable of accounting for protein structure and cross-species evolutionary divergence simultaneously ., By aligning low-level evolutionary HMM modules within a high-level functional-element grammar , it is possible to build precise models of the effects of evolutionary pressures on genomic structures ., In particular , we have applied this technique to modeling of prohormones across species with the goal of identifying novel prohormones and associated peptide hormones based on their evolutionary divergence profiles and genomic structures ., This technique has resulted in high accuracy detection in a known dataset and led to putative hormones in a set of hypothetical proteins ., Biochemical validation of the findings has resulted in the initial characterization of the prohormone preproNPQ , containing four potential previously undiscovered neuropeptides ., In order to determine if the putative transcript named preproneuropeptide Q ( preproNPQ ) is found in the brain , we performed PCR using rat , human and mouse specific primers with their corresponding cDNAs ., The sequences of the primers used were: Rat Forward Primer 5′-GAAGGGGCCGAGCATCCTGG-3′ and Reverse Primer 5′-CACCAGTAAAAGCGTCTGTCTTC-3′; Mouse Forward Primer 5′-GGACAGGGTCGGAACATGAAG-3′ and Reverse Primer 5′-GTGTTTTCACCAGTTGAAGAGTC-3′; Human Forward Primer 5′-ACGCAGAACATGAAGGGACTCAGA-3′ and Reverse Primer 5′-CCAGTATATTTTCACCAGTTAAGC-3′ ., Advantage Genomic Polymerase Mix enzyme ( BD Biosciences Clontech , CA ) was used for PCR , according to manufacturers instructions ., Approximately 200–300 ng cDNA was used for each 50 ml reaction , along with 10 mM of specific forward and reverse primer , 2 . 2 ml magnesium acetate and dNTPs ( 10 mM ) ., The annealing temperature was set at 53°C , and after 25 cycles of amplification , the PCR products were run on a 1 . 5% agarose gel and visualized using ethidium bromide ., A positive control PCR reaction was also performed at the same time , using rat brain cDNA and specific primers for the prepronociceptin gene , and the reaction product was run on the gel ., In order to determine if the preproNPQ transcript could be detected in various human tissues , we used Ambions First Choice Human Blot ( a nylon membrane bound with 3 mg RNA from various human tissues , Ambion Inc , TX ) ., The blot was prehybridized and probed with human NPQ cDNA prepared using the above preproNPQ human primers and the human DNA clone in pOTB7 vector from ATCC ( Cat # 6710068 , Manassas , VA ) ., This clone contained the putative sequence for human preproNPQ , and the primers were used to isolate a 370 bp preproNPQ sequence that was used as the cDNA probe for hybridization to the RNA ., Random-prime labeling of approximately 20–30 ng DNA was performed using 32P-dCTP and Klenow DNA polymerase , and after purifying the labeled probe on a G-50 column , the labeled DNA probe was hybridized to the nylon membrane overnight at 42°C ., The membrane was washed and exposed to film . | Introduction, Results, Discussion, Materials and Methods | There are currently a large number of “orphan” G-protein-coupled receptors ( GPCRs ) whose endogenous ligands ( peptide hormones ) are unknown ., Identification of these peptide hormones is a difficult and important problem ., We describe a computational framework that models spatial structure along the genomic sequence simultaneously with the temporal evolutionary path structure across species and show how such models can be used to discover new functional molecules , in particular peptide hormones , via cross-genomic sequence comparisons ., The computational framework incorporates a priori high-level knowledge of structural and evolutionary constraints into a hierarchical grammar of evolutionary probabilistic models ., This computational method was used for identifying novel prohormones and the processed peptide sites by producing sequence alignments across many species at the functional-element level ., Experimental results with an initial implementation of the algorithm were used to identify potential prohormones by comparing the human and non-human proteins in the Swiss-Prot database of known annotated proteins ., In this proof of concept , we identified 45 out of 54 prohormones with only 44 false positives ., The comparison of known and hypothetical human and mouse proteins resulted in the identification of a novel putative prohormone with at least four potential neuropeptides ., Finally , in order to validate the computational methodology , we present the basic molecular biological characterization of the novel putative peptide hormone , including its identification and regional localization in the brain ., This species comparison , HMM-based computational approach succeeded in identifying a previously undiscovered neuropeptide from whole genome protein sequences ., This novel putative peptide hormone is found in discreet brain regions as well as other organs ., The success of this approach will have a great impact on our understanding of GPCRs and associated pathways and help to identify new targets for drug development . | Peptide hormones , or neuropeptides , are made up of a string of amino acids ranging from approximately 3 to 50 residues ., These peptides are processed from a larger protein called a prohormone and activate a class of proteins called G-protein-coupled receptors ( GPCRs ) ., Neuropeptides signal neurons and other cells leading to changes in cellular biochemistry and potentially gene expression ., There are a number of “orphan” GPCRs , i . e . , receptors that have been discovered either by genomic sequence or by cloning , in which its respective peptide hormone is unknown ., We have devised a computational method that models patterns in protein sequence simultaneously with evolutionary differences across species in order to identify previously unknown peptide hormones ., We have used this computational methodology to identify a previously unknown putative prohormone that contains up to four potential neuropeptides , and we have characterized this prohormone with respect to location in rat brain and various human tissues ., This computational technique will be useful for the identification of additional neuropeptides and help to characterize orphan GPCRs ., Because roughly half of all pharmaceuticals act through activation or inhibition of GPCRs , this technique should lead to the identification of additional pharmaceutical targets and ultimately clinically used drugs . | computational biology/sequence motif analysis, computational biology/comparative sequence analysis, evolutionary biology/genomics, computational biology/evolutionary modeling, neuroscience/neuronal signaling mechanisms, molecular biology/mrna transport and localization | null |
journal.pgen.1000650 | 2,009 | Multiple Organ System Defects and Transcriptional Dysregulation in the Nipbl+/− Mouse, a Model of Cornelia de Lange Syndrome | Cornelia de Lange Syndrome ( CdLS; OMIM#122470 ) is characterized by developmental abnormalities of the cardiopulmonary , gastrointestinal , skeletal , craniofacial , neurological , and genitourinary systems 1–3 ., The clinical presentation ranges from subtle dysmorphology to conditions incompatible with postnatal life ., Common structural birth defects observed in CdLS include upper limb reduction ( significant in just under half of cases ) , cardiac abnormalities ( especially atrial and ventricular septal defects ) , and craniofacial dysmorphia ( including dental and middle ear abnormalities , occasional clefting of the palate , and highly characteristic facies ) 2–8 ., Other findings include small head size , lean body habitus , hirsutism , ophthalmologic abnormalities , pre- and postnatal growth retardation , and structural abnormalities of the gastrointestinal tract ( duodenal atresia , annular pancreas , small bowel duplications ) 2 , 3 , 9–11 ., Physiological disturbances in CdLS include moderate to severe mental retardation 12 often accompanied by autistic behaviors 13 , and severe gastrointestinal reflux 14 ., Although prevalence has been estimated at between ∼1/10 , 000 and 1/50 , 000 births 8 , 15 , wide phenotypic variability in the syndrome makes it likely that large numbers of mildly-affected individuals are not being counted ., A genetic basis for CdLS was uncovered in 2004 with the demonstration that many affected individuals carry mutations in Nipped-B-like ( NIPBL ) , so named for its homology to the Drosophila gene , Nipped-B 16 , 17 ., Heterozygous NIPBL mutations are found in about 50% of individuals with CdLS 18 ., As many of these mutations are expected to produce absent or truncated protein , haploinsufficiency is the presumed genetic mechanism 19 ., NIPBL/Nipped-B protein is found in the nuclei of all eukaryotic cells , where it interacts with cohesin , the protein complex that mediates sister chromatid cohesion 20 , 21 ., The NIPBL ortholog in fungi plays a role in loading cohesin onto chromosomes , and a role in unloading has been suggested as well ., The fact that a minority of cases of mild CdLS result from mutations in the SMC1L1/SMC1A ( ∼5%; OMIM 300590 ) and SMC3 ( 1 case; OMIM 610579 ) genes , which encode two of the four cohesin structural components , supports the view that CdLS is caused by abnormal cohesin function 22 , 23 ., Consistent with the hypothesis that cohesin plays important roles during embryonic development , it was found that mutations in the cohesin regulatory protein ESCO2 cause Roberts-SC phocomelia syndrome , another multi-organ systems birth defects syndrome 24 , 25 ., In mice , deletion of the cohesin regulators PDS5A and PDS5B also produces a wide variety of developmental defects , some of which overlap with CdLS 26 , 27 ., In addition , there has recently been a report of one family showing atypical inheritance of CdLS , in which both affected and unaffected siblings harbor a missense mutation in the PDS5B gene , raising the possibility of some genetic association between PDS5B and CdLS 27 ., How alterations in cohesin function give rise to pervasive developmental abnormalities is largely unknown ., Cohesin is involved in sister chromatid cohesion and DNA repair in many organisms , but observed alterations in cohesion and repair in individuals with CdLS are mild at best 28 , 29 ., More recently , observations in model organisms and cultured cells have suggested that cohesin plays important roles in the control of transcription reviewed in 18 ., In Drosophila , for example , changes in levels of Nipped-B or cohesin structural components alter the expression of developmental regulator genes , such as homeodomain transcription factors 30–33 ., Such effects on gene expression , which have been proposed to reflect the disruption of long-range promoter-enhancer communication , occur with small changes in Nipped-B or cohesin levels that do not produce cohesion defects; they can also occur in postmitotic cells , in which chromosome segregation is presumably not an issue 18 ., Studies using Drosophila cell lines have demonstrated that cohesin and Nipped-B binding are concentrated near the promoters of active transcriptional units 34 ., In mammalian cells , cohesin often binds , in an NIPBL-dependent manner , to sites occupied by the transcriptional insulator protein CTCF , where it plays a significant role in CTCF function 35–37 ., Recently , NIPBL has also been shown to bind and recruit histone deacetylases to chromatin 38 ., These observations suggest that cohesin and NIPBL may interact in multiple ways with the transcriptional machinery ., As a first step toward understanding the molecular etiology of CdLS , we generated a mouse model of Nipbl haploinsufficiency , which replicates a remarkable number of the pathological features of CdLS ., Cellular and molecular analysis of mutant cells and tissues revealed widespread , yet subtle , changes in the expression of genes , some of which are found in genomic locales in which transcription is known to be controlled through long-range chromosomal interactions ., We propose that the aggregate effects of many small transcriptional changes are the cause of developmental abnormalities of CdLS , and present evidence that one set of transcriptional changes may explain the notably lean body habitus of many individuals with CdLS ., Two mouse ES cell lines bearing gene-trap insertions into Nipbl were obtained and injected into C57BL/6 blastocysts to produce chimeras ( see Materials and Methods ) ., Male chimeras were bred against both inbred ( C57BL/6 ) and outbred ( CD-1 ) mice ., For only one cell line ( RRS564 , which contains a beta-geo insertion in intron1 , and is predicted to produce a truncated transcript with no open reading frame; Figure S1 ) was ES cell contribution to the germline obtained ( as scored by coat color; Table 1 ) ., Whereas Mendelian inheritance predicts that half the germline progeny of chimeric mice should be heterozygous ( Nipbl+/− ) for the gene trap insertion , the observed frequency was much lower ., When chimeras were bred against CD-1 females , 22 out of 113 germline progeny ( 19% ) carried the mutant allele ( Table 1 ) ., With C57BL/6 females only one out of 18 germline progeny carried the mutation ( 5 . 5% ) , and this animal , although male , did not produce any progeny when subsequently mated ., In view of these data , it was decided that further analysis of the Nipbl− allele would take place through outcrossing onto the CD-1 background ., As shown in Table 1 , when the Nipbl+/− offspring of chimera by CD-1 crosses ( N0 generation ) were bred against wildtype ( CD-1 ) females , 17% of surviving adult progeny carried the mutant allele ., When animals of this “N1 generation” were again outcrossed against CD-1 , 18% of surviving progeny ( N2 ) carried the mutant allele ., Similar survival ratios were observed for subsequent generations of outcrossing ., The data imply that 75–80% of Nipbl+/− mice die prior to genotyping ( typically done at 4 weeks of age ) , a fraction that remains stable as the mutant allele is progressively outcrossed onto the CD-1 background ., To determine whether lethality occurs in utero , we examined litters for Nipbl+/− embryos just before birth ( gestational days E17 . 5 and E18 . 5 ) ., With no visible marker available for the ES-cell derived progeny of chimeras , this test was carried out with progeny of the N0 generation , in which the Mendelian expectation for the mutant allele is 50% ., Mutants were found to comprise 41% ( 30 out of 67 ) of progeny , a frequency not significantly different from expected for this sample size ( Table 1 ) ., These data imply that most mutants die at or after birth ., To evaluate the extent to which Nipbl+/− mice provide a good model for CdLS , we performed an analysis in which we examined these animals for a number of different structural phenotypes analogous to common clinical findings observed in CdLS ( summarized in Table S1 ) ., Among the most common clinical features of CdLS are small body size , often evident before birth; heart defects; and upper limb abnormalities ranging from small hands to frank limb truncations 2–5 , 7 , 8 , 39 ., As shown in Table 2 , Nipbl+/− embryos examined shortly before birth ( E17 . 5–E18 . 5 ) were 18–19% smaller than wildtype littermates ( P<0 . 001 ) , a reduction not accompanied by decreased placental size ., Nipbl+/− embryos at earlier stages were also noted to be slightly smaller than littermates ( data not shown ) ., Nipbl+/− embryos did not display limb or digit truncations , or obvious loss of any other bony elements ., However , upon staining embryonic skeletons , we observed delays in ossification of both endochondral and membranous bones of Nipbl+/− embryos ., As shown in Figure 1A–1D , delayed ossification of the skull and digits was apparent between E16 . 5 and E18 . 5 ., Measurement of long bones and digits at E17 . 5 revealed , in addition to a symmetrical reduction in bone length ( consistent with smaller body size ) , a significant decrease in the relative extent of ossification ( Figure 1E ) ., Otherwise , the patterning of cartilaginous elements was relatively normal , although some subtle differences in morphology were consistently observed , e . g . the shape of the olecranon process of the ulna was consistently abnormal in Nipbl+/− mouse embryos ( Figure 1F–1G ) ., Interestingly , dys- and hypoplastic changes of the ulna are common findings in CdLS 40 ., Among the cardiac defects that occur in CdLS , atrial and ventricular septal defects are especially common 2 , 5 , 7 ., Atrial septal defects , which were typically large , were observed in about half of Nipbl+/− mouse embryos , ( Figure 1H–1K; Table S1 ) , and could be detected as early as E15 . 5 , shortly after atrial septation normally finishes ., A reduction in atrial size was also seen in some mutants , but was not a consistent finding ., No defects were detected in the atrioventricular valves or septum , outflow tract , or pulmonary vasculature ., However , many mutant embryos displayed subtle abnormalities of the ventricular and interventricular myocardium , including abnormal lacunar structures and disorganization of the compact layer , especially near the apex ( data not shown ) ., Significantly , no histological or functional cardiac abnormalities were detected among mutant mice that survived the perinatal period ( data not shown ) ., This implies that the cause of perinatal mortality is either cardiac , or correlates strongly with the presence of cardiac structural defects ., Histological examination of other organ systems in late embryonic mutant mice revealed no obvious anatomical abnormalities of the lungs , diaphragm , liver , stomach , spleen , kidney or bladder ., Brains of neonatal Nipbl+/− mice displayed relatively normal gross anatomy , although a single mutant was observed to have a large brainstem epidermoid cyst ( not shown ) ., Most Nipbl+/− mice that survived the perinatal period reached adulthood , and appeared to have a normal lifespan ., However , marked decrease in the body size of mutant mice was evident at birth and throughout all ages ( Figure 2A and 2B ) ., Indeed , the 18–19% weight difference between mutant and wildtype mice observed before birth ( Table, 2 ) widens to 40–50% by postnatal weeks 3–4 ( Figure 2C–2E; this finding has remained consistent over 6 generations data not shown ) ., To investigate early postnatal growth of Nipbl+/− mice in more detail , litters fathered by N1 and N2 generation animals were subjected to daily weighing from shortly after birth until sexual maturity ( 5–6 weeks of age; Figure 2F ) ., Most mutant mice exhibited failure to thrive during the first weeks of life , with many undergoing several days of wasting followed by death ( Figure 2F , inset ) ., By 3 weeks of age , the average weight of surviving mutants was only 40% of wildtype , but after weaning this pattern abruptly changed: mutants ( even ones that had already begun to show wasting ) underwent rapid catch-up growth ( Figure 2F ) , such that by 9 weeks of age they had reached 65–70% of wildtype weight ., These observations suggest that , in addition to being intrinsically small , Nipbl+/− mice may have difficulty with suckling , or may receive inadequate nutrition from milk ., Remarkably , the weights of children with CdLS also fall further behind age norms during the first year of life , but show significant catch-up growth later on 11 ., The distinctive craniofacial features of CdLS , including microbrachycephaly , synophrys , upturned nose , and down-turned lips , play an important role in clinical diagnosis 3 , 6 ., Micro-CT analysis was used to assess whether Nipbl+/− mice also display consistent craniofacial changes ., Analysis of the skulls of 63 adult mice showed significantly smaller size ( microcephaly ) among all mutants ( N\u200a=\u200a23 ) , as well as a variety of significant shape changes ( Figure 3 ) ., The latter included foreshortening of the anterior-posterior dimensions of the skull ( i . e . brachycephaly ) and an upward deflection of the tip of the snout ( Figure 3B–3E ) ., The upturned nares ( Figure 3C and 3E ) reflect reduced size of the ethmoid and sphenoid bones , which produces a sunken midface ., Together , these shape changes in the basicranium and face are consistent with a greater reduction in the size of chondrocranial , as opposed to dermatocranial , elements within the skull ., In addition , an 8% average decrease in bone thickness was also observed ( ANOVA , df\u200a=\u200a47 , F\u200a=\u200a18 . 6 , p<0 . 01 ) ., Neurological abnormalities in CdLS include mental retardation , abnormal sensitivity to pain , and seizures 41 ., Although Nipbl+/− mice have not been subjected to intensive long-term neurological or behavioral tests , several distinctive behaviors were observed: Repetitive circling ( Videos S1 , S2 , S3 ) was noted in 20% ( 34/173; 15 females and 19 males ) of adult Nipbl+/− mice ( >5 weeks of age ) , across all generations examined ( N0–N4 ) ., Repetitive behaviors—including twirling in place 42—are common symptoms in children with CdLS ., In addition , 30% ( 4/13; all males ) of Nipbl+/− mice were noted to adopt opisthotonic postures in response to administration of a normal anesthetic dose of avertin ( see Materials and Methods ) , strongly suggesting seizure activity ., Seizures are also common in individuals with CdLS 43 , 44 ., We also observed that 15% of Nipbl+/− adult mice ( 24/158; 11 females and 13 males ) displayed reflexive hindlimb clasping when suspended by their tails ( Videos S4 , S5 ) , whereas only 2% ( 6/268 ) littermates showed the same behavior ( Table S1 ) ., Hindlimb clasping has been observed in several mouse models of neurological disorders , including Retts syndrome 45–47 , mucolipidosis type IV 48 , infantile neuroaxonal dystrophy and neurodegeneration with brain iron accumulation 49 , and Huntingtons disease 50–53 ., Histological examination of mutant brains revealed the presence of all major brain structures , grossly normal lamination of the cerebral and cerebellar cortices , but an overall reduction in brain size , consistent with a 25% reduction in endocranial volume observed with micro-CT ( Figure 4A , two-tailed T-test , df\u200a=\u200a28 , T\u200a=\u200a5 . 7 p<0 . 01 ) ., Absence or reduction in size of the corpus callosum was occasionally observed in Nipbl+/− mice ( Figure 4B ) ., Obvious patterning defects were noted only in the midline cerebellum , where lobe IX displayed specific reductions ( Figure 4C ) ., Interestingly , midline cerebellar hypoplasia is one of the few consistently-reported changes in brain anatomy in CdLS 54–56 ., Children with CdLS display a range of ophthalmological abnormalities including ptosis , microcornea , nasolacrymal duct obstruction , strabismus , blepharitis and conjunctivitis 57–59 ., We noted that 22% of Nipbl+/− mice exhibited one or more gross ophthalmological abnormalities ( Table S1 ) ., Most frequently observed was ocular opacification , observed in 14% of animals ( Figure 4D ) ; opacities were often evident as early as three weeks of age ., In several cases , this condition was associated with marked periorbital inflammation , and progressed to permanent closure of the eyelids ( not shown ) ., Histological analysis revealed inflammatory and fibrotic changes within the corneal epithelium and stroma ( Figure 4E ) , consistent with repeated abrasion or injury ., Such injury might arise from neglect due to abnormalities in corneal sensation , from abnormal production or composition of tear fluid , or secondary to periorbital inflammation or infection ( e . g . blepharitis; cf . Table S1 ) ., Some degree of hearing loss is observed in almost all individuals with CdLS , and this may play a role in the marked speech disability often seen in this syndrome 60 , 61 ., To assess hearing in Nipbl+/− mice , we measured auditory brainstem evoked responses ( ABR 62 ) ., Abnormalities were found in the majority of mutant mice examined ( Table S1 ) ., In a few cases , markedly increased thresholds to stimulation were observed ( Figure 4F ) ., More commonly , stimulus thresholds were within normal limits , but the relative intensities of the components of the ABR were altered ., In particular , mutant mice displayed a characteristic reduction in the amplitude of the third peak ( at about 3 msec following stimulus ) , a latency consistent with an abnormality in the auditory nerve and/or early brainstem neural pathways ( Figure 4G ) ., The Nipbl564 gene-trap mutation is expected to produce a truncated message lacking all but the first exon ( Figure S1 ) ., Therefore , the level of full-length Nipbl mRNA in Nipbl+/− mice should provide an indication of the activity of the wildtype allele ., To measure this level , we used an RNase protection assay based on hybridization to sequences found in exons 10 and 11 ., Total RNA was analyzed from two tissues: adult liver and E17 . 5 brain , using age-matched littermate controls ., As shown in Figure 5 , Nipbl levels in mutants , as a percentage of wildtype levels , were 72–82% in adult liver , and ∼70% in embryonic brain ., When western blotting was used to quantify levels of NIPBL protein in Nipbl+/− embryo fibroblasts ( MEFs ) , a reduction to about 70% of wildtype levels was observed ( Figure S2 ) ., The observation that Nipbl+/− mice exhibit only a 25–30% decrease in transcript and protein expression , rather than an expected decrease of 50% , is consistent with Nipbl gene being autoregulatory ., An alternative explanation is that the mutant allele is “leaky” , i . e . alternative splicing around the gene trap cassette produces some wildtype message ., We favor the former explanation because , in both Drosophila and man , the evidence indicates that null mutation of a single allele of Nipped-B/NIPBL produces only a 25–30% drop in transcript levels , the same decrease we observe in Nipbl+/− mice 31 , 63 , 64 ., Thus , even if the Nipbl allele studied here is not null , it is probably quite close to being so ., More importantly , the degree of decrease in Nipbl expression in Nipbl+/− mice is comparable to that which causes CdLS in man ., Overall the data from multiple species strongly argue that pervasive developmental abnormalities result from remarkably small changes in NIPBL levels ., There has been one report of precocious sister chromatid separation ( PSCS ) in cell lines derived from individuals with CdLS 28 , which was not seen in a second study 29 ., We found no statistically-significant elevation of PSCS in cultured Nipbl+/− MEFs ( Figure S3 ) , Nipbl+/− embryonic stem cells ( data not shown ) , or adult B-lymphocytes ( Figure S3 ) ., These results suggest that cohesion defects in the Nipbl heterozygotes , if present , are very subtle; they are also in accord with findings in Drosophila , where PSCS is seen only when both alleles of Nipped-B are mutated 31 ., To investigate whether heterozygous loss of Nipbl leads to alterations in transcription , we turned to expression profiling of tissues and cells from Nipbl+/− mice ., Because such mice display pervasive developmental abnormalities , transcriptome data can be expected to reflect not only the direct consequences of reduced Nipbl function , but also a potentially large number of transcriptional effects that are secondary consequences of abnormal morphology and physiology ., In an effort to minimize the detection of such secondary effects , we focused on profiling samples in which frank pathology was not seen , or had yet to develop by the time of profiling ., The samples chosen for analysis were embryonic day 13 . 5 ( E13 . 5 ) brain , and cultures of fibroblasts derived from E15 . 5 embryos ( mouse embryo fibroblasts; MEFs ) ., Although mature brain appears to be functionally abnormal in Nipbl+/− mice ( see above ) , at E13 . 5 it at least appears anatomically normal ., Cultured MEFs were chosen because they are established with similar efficiency from both mutant and wildtype embryos; exhibit similar morphology and growth characteristics in culture; and by virtue of being maintained ex vivo , are freed of the secondary influences of any systemic metabolic or circulatory derangements within Nipbl+/− embryos ., Transcriptome analysis was performed using Affymetrix microarrays ., MEF RNA samples were obtained from 10 mutant and 9 wildtype embryos taken from three litters ( 19 separate microarrays ) ; brain RNA was analyzed from 10 mutant and 11 wildtype embryos from two litters ( 21 separate microarrays ) ., Gene expression changes were detected in both comparisons ., In the brain ( Table S2 ) , 1285 probe sets , corresponding to 978 genes , displayed statistically significant differences in expression between wildtype and mutant mice ( per-probe-set false discovery rate of Q<0 . 05 ) ., By and large , the effects were small: 97 . 5% of changes were within 1 . 5-fold of wildtype expression values; >99 . 6% were within 2-fold ., The single largest statistically-significant change was 2 . 5-fold ., Genes encoding products of virtually all structural and functional categories could be found among those affected , with no dramatic enrichment of any particular functional sets ( by Gene Set Enrichment Analysis 65; data not shown ) ., In cultured Nipbl+/− MEFs , 89 probe sets , corresponding to 81 genes ( Table S3 ) , displayed statistically-significant ( Q<0 . 05 ) differences in expression between wildtype and mutant mice ., Again , effects were small: 89% of changes were within 1 . 5-fold of wildtype , and 99% were within 2-fold ., The single largest statistically-significant change was 2 . 1-fold ., The lower number of transcriptional changes identified in MEFs versus brain may not be biologically meaningful , as MEFs happened to display a somewhat higher average within-sample variance than E13 . 5 brain , making it more difficult for small changes to be judged significant ., As with embryonic brain , transcriptional effects in MEFs involved genes that encode a wide variety of proteins ., Although automated analyses failed to single out any particular functional class as being highly overrepresented , manual curation revealed significant changes in the expression of a number of genes implicated in adipogenesis ( Figure 6A ) ., For example , Cebpb and Ebf1—which encode transcriptional factors central to the process of adipocyte differentiation 66–68—were both down-regulated in Nipbl+/− MEFs , as were Fabp4 and Aqp7 , well-known adipocyte markers 69 , 70 ., Other genes down-regulated in Nipbl+/− MEFs ( Table S3 ) could also be found , through literature searches , to exhibit expression positively correlated with adipocyte differentiation , including Adm , Lpar1 , Osmr , and Ptx3 69 , 71 , 72 ., Several additional genes ( Amacr , Avpr1a , Il4ra , Prkcdp , S100b ) down-regulated in Nipbl+/− MEFs can be inferred , from publicly-available expression data , to be enriched in pre-adipocytes and/or brown or white adipose tissue 73–75 ., Conversely , Lmo7 , which is normally down-regulated during late adipogenic differentiation 71 , was found to be up-regulated in Nipbl+/− MEFs ., Furthermore , we noted that genes such as Cebpa and Cebpd ( transcriptional activators of adipocyte differentiation 66 , 76 ) , Il6 ( a cytokine stimulator of adipocyte differentiation that controls adiposity in man 77 , 78 ) and Socs3 ( an intracellular signaling regulator induced by Il6 79 ) , were also down-regulated in the MEF samples , but at false-discovery rates slightly too high to permit their inclusion in Table S3 ( Q\u200a=\u200a0 . 065 , 0 . 085 , 0 . 075 , and 0 . 17 , respectively ) ., Together , these data raise the possibility that Nipbl+/− mice are specifically impaired in adipogenesis ., Support for this idea was obtained by weighing intrascapular fat dissected from adult mutant and wildtype littermates 80 ., As shown in Figure 6B , both brown and white fat are substantially depleted in Nipbl+/− mice ., To correct for the fact that mutant mice are generally smaller than their wildtype littermates , we normalized fat measurements to brain weight ( which scales with overall body size ) ., As shown in Figure 6C , even by this measure , Nipbl+/− mice displayed a significant , substantial reduction in body fat ., As mentioned earlier , lean body habitus is also a characteristic of CdLS ., To investigate whether the reduction in body fat in Nipbl+/− mice reflects an intrinsic defect in the differentiation potential of mutant fibroblasts , we studied adipogenic differentiation in vitro ., It is known that embryonic fibroblasts can be converted , in large numbers , to adipocytes by treatment with agents such as glucocorticoids , PPAR-γ agonists , isobutylmethylxanthine and insulin , which stimulate the activity of a core network of pro-adipogenic transcription factors ( C/EBPα , C/EBPβ , C/EBPδ , PPARγ; 81 , 82 ) ., In response to such agents , we observed no significant difference between Nipbl+/− and wildtype MEFs in terms of the number of adipocytes or adipocyte colonies produced ( data not shown ) ., However , when we omitted these pharmacological agents , and measured the ( much lower ) level of spontaneous adipogenic differentiation that occurs in MEF cultures 83 , we observed a substantially-lower level in mutant cultures ( Figure 6D–6F ) ., The observation that Nipbl+/− MEFs are impaired in spontaneous , but not induced , adipogenesis implies that their primary defect does not lie downstream of the targets of pharmacological inducers ., Of the 80 genes ( not counting Nipbl itself ) with significant differential expression in Nipbl+/− MEFs ( Table S3 ) , 20% ( 16/80 ) are also found among the 978 genes whose expression was altered in Nipbl+/− embryonic brain ( Table S2 ) ., Using a more stringent false discovery rate cutoff of Q<0 . 02 for both samples , we find that 23% ( 9/40 ) of differentially expressed MEF genes are among the 560 that are differentially expressed in brain ., These data suggest that common transcriptional targets exist in the two tissues ., Further support for this idea is obtained by correlating fold-increase or -decrease of affected transcripts ., In this case a less conservative approach to false discovery is justified ( the goal is to estimate overall correlation between samples , not implicate individual genes ) , so the log-fold changes for all probe sets that exhibited differential expression exceeding an arbitrary t-statistic threshold ( t>2 ) in both tissues were plotted against each other ( shown in Figure 7 ) ., The data are clearly strongly correlated ( R\u200a=\u200a0 . 77 ) , suggesting that at least some of the transcriptional effects of Nipbl deficiency are shared across tissues ., Among the genes in which expression changes contributed substantially to the correlation are four members of the protocadherin β cluster ( Pcdh17 , Pcdh20 , Pcdh21 , Pcdh22; all down-regulated ) , Lpar1 ( also down-regulated; encoding the lysophosphatidic acid receptor ) , Vldlr ( down-regulated; encoding a receptor involved in both lipid metabolism and cerebral cortical development ) , and Stag1 ( up-regulated; encoding SA1 , a cohesin component ) ., Interestingly , in Drosophila , inhibition of Nipped-B expression also leads to up-regulation of the ortholog of Stag1 31 ., Recently , STAG1 up-regulation has also been seen in lymphoblastoid cell lines of individuals with CdLS 64; Table S5 ., Among the most significant changes common to mutant MEF and brain samples were decreases in expression of transcripts from the 22-gene Pcdhb ( protocadherin beta ) cluster on chromosome 18 ( Table S2 and Table S3 , Figure 7 ) ., As shown in Figure 8A , affected transcripts included Pcdhb7 , 16 , 17 , 19 , 20 , 21 and 22 , which lie predominantly at the 3′ end of the cluster ., This observation raised the possibility that the transcriptional effects of Nipbl might be related to the physical locations of genes ., However , as genes at the 5′ end of the Pcdhb cluster tend to be expressed at lower levels than those at the 3′ end , lower signal-to-noise ratios might have made small changes in expression at the 5′ end more difficult to detect ., To resolve this issue , and to provide independent confirmation of microarray data , quantitative RT-PCR was used to measure transcripts levels at multiple locations throughout the Pcdhb cluster ( Figure 8B ) ., For these experiments , brain mRNA was prepared at a later developmental stage ( E17 . 5 , when most Pcdhb transcripts are more highly expressed ) from 13 independent samples ( 7 mutant and 6 wildtype embryos ) ., Robust RT-PCR signals were obtained for 14 of 15 transcripts tested ( Pcdhb2 , 3 , 4 , 5 , 7 , 8 , 9 , 10 , 13 , 14 , 16 , 17 , 19 , and 22; but not Pcdhb1 ) ., As shown in Figure 8B , the data support the microarray results from the earlier embryonic stage , and indicate that most transcriptional changes in Nipbl+/− brain indeed occur preferentially at the 3′ end of the cluster ( Pchdb13 , 14 , 15 , 16 , 17 , 19 , 22 ) ., Additionally , they suggest that at least one 5′ gene , Pcdhb2 , may also be affected ., A more revealing analysis of the data can be obtained by correlating Pcdhb transcript levels in each tissue sample , regardless of genotype , against Nipbl transcript levels within that sample ( i . e . treating Nipbl expression as a quantitative trait; Figure 8C , Figure S4 ) ., This approach offers greater discriminatory power because Nipbl expression in individual samples varies significantly , even within mutant and wildtype groups , and occasionally overlaps between the two groups ., Indeed , the results of the analysis indicate that Pcdhb expression correlates strongly with Nipbl transcript level , lending support to the view that Pcdhb transcription is directly affected by the amount of NIPBL present in cells ., In Figure 8C , the results of such correlations for all 13 tested Pcdhb transcripts are summarized by plotting the slopes of regression lines ( the sensitivity of each transcripts expression to Nipbl level ) against gene location , with error bars reflecting the strength of correlation for each gene ., The results strongly suggest a continuum of sensitivity to Nipbl across the entire Pcdhb cluster , with genes at both the 5′ and 3′ ends being the most sensitive , and those in the middle being least affected ., We show here that mice heterozygous for a gene-trap mutation upstream of the first coding exon of Nipbl displayed many features of human CdLS , including pre- and postnatal growth retardation , cardiac septal defects , delayed bone development , lean body habitus , microbrachycephaly with characteristic craniofacial changes , behavioral disturbances , ophthalmological abnormalities , cerebellar hypoplasia , and hearing deficits ( Figures 1–4 , Table S1 , Videos S1 , S2 , S3 , S4 , S5 ) ., These phenotypes remained stable through many generations of outcrossing , and occurred in the context of modest ( 25–35% ) re | Introduction, Results, Discussion, Materials and Methods | Cornelia de Lange Syndrome ( CdLS ) is a multi-organ system birth defects disorder linked , in at least half of cases , to heterozygous mutations in the NIPBL gene ., In animals and fungi , orthologs of NIPBL regulate cohesin , a complex of proteins that is essential for chromosome cohesion and is also implicated in DNA repair and transcriptional regulation ., Mice heterozygous for a gene-trap mutation in Nipbl were produced and exhibited defects characteristic of CdLS , including small size , craniofacial anomalies , microbrachycephaly , heart defects , hearing abnormalities , delayed bone maturation , reduced body fat , behavioral disturbances , and high mortality ( 75–80% ) during the first weeks of life ., These phenotypes arose despite a decrease in Nipbl transcript levels of only ∼30% , implying extreme sensitivity of development to small changes in Nipbl activity ., Gene expression profiling demonstrated that Nipbl deficiency leads to modest but significant transcriptional dysregulation of many genes ., Expression changes at the protocadherin beta ( Pcdhb ) locus , as well as at other loci , support the view that NIPBL influences long-range chromosomal regulatory interactions ., In addition , evidence is presented that reduced expression of genes involved in adipogenic differentiation may underlie the low amounts of body fat observed both in Nipbl+/− mice and in individuals with CdLS . | Cornelia de Lange Syndrome ( CdLS ) is a genetic disease marked by growth retardation , cognitive and neurological problems , and structural defects in many organ systems ., The majority of CdLS cases are due to mutation of one copy of the Nipped B-like ( NIPBL ) gene , the product of which regulates a complex of chromosomal proteins called cohesin ., How reduction of NIPBL function gives rise to pervasive developmental defects in CdLS is not understood , so a model of CdLS was developed by generating mice that carry one null allele of Nipbl ., Developmental defects in these mice show remarkable similarity to those observed in individuals with CdLS , including small stature , craniofacial abnormalities , reduced body fat , behavioral disturbances , and high perinatal mortality ., Molecular analysis of tissues and cells from Nipbl mutant mice provide the first evidence that the major role of Nipbl in the etiology of CdLS is to exert modest , but significant , effects on the expression of diverse sets of genes , some of which are located in characteristic arrangements along the DNA ., Among affected genes is a set involved in the development of adipocytes , the cells that make and accumulate body fat , potentially explaining reductions in body fat accumulation commonly observed in individuals with CdLS . | genetics and genomics/disease models, developmental biology/developmental molecular mechanisms, molecular biology/chromatin structure, genetics and genomics/gene expression | null |
journal.pntd.0006395 | 2,018 | Bacterial and protozoal pathogens found in ticks collected from humans in Corum province of Turkey | Ticks are important vectors of a variety of diseases all over the world , including Turkey ., They may transmit different kind of pathogens including bacteria , viruses , and protozoa affecting humans , domestic and wild animals 1 , 2 ., Turkey is composed from a mosaic of habitats for ticks due to its diverse climate , vegetation , and large variety of wild and domestic animals 1 , 3 ., Today , 48 tick species are known from this country , 31 of which have been found infesting humans 3 ., Nineteen tick-borne diseases ( TBDs ) have been detected either in animals or humans in Turkey 1 ., From 2002 to 2015 , a total of 9 , 787 human cases of Crimean Congo hemorrhagic fever ( CCHF ) have been reported , 469 of which resulted in death 4 ., Lyme borreliosis were reported in Turkey 5 , while the sero-prevalence of Borrelia burgdorferi in humans was 4% 6 ., Between 2005 and 2011 , 4 , 824 human cases with tularemia were reported to the Ministry of Health 7 ., Anaplasmosis is known from farm animals 8 , while in humans , sero-positivity was 10 . 62% 9 ., Ehrlichiosis and hepatozoonosis have been diagnosed in dogs 10 , 11 ., The sero-prevalence for bartonellosis was 18 . 6% in cats 12 , 6% in human blood donors 13 , and 22 . 2% in cattle breeders and veterinarians 14 ., Rickettsiosis was reported in Thrace and East Mediterranean regions of Turkey 15 , 16 , the most prevalent being the Mediterranean Spotted Fever ( MSF ) 17 ., Q fever cases in humans were reported from the Black Sea region of Turkey 18 ., Babesiosis in animals is highly prevalent in Turkey , but there are no reports about clinical cases in humans 1 ., Toxoplasmosis is one of the more common parasitic zoonosis worldwide , and in Turkey the prevalence in humans was found to vary between 13 . 9% and 76 . 6% 19 ., Between the years 1988–2010 , 50 , 381 cases of cutaneous leishmaniasis were reported to the Turkish Ministry of Health 20 ., According to recent studies , ticks can be also possible vectors of toxoplasmosis and leishmaniasis 21 , 22 ., The first CCHF cases in Turkey were observed in the province of Tokat which is a neighboring province of Corum; both cities are located in Kelkit Valley where the main vector , Hyalomma marginatum is prevalent 1 , 4 ., Recently , 327 cases of CCHF and other TBDs such as rickettsial infections were reported from Corum 3 , 23–27 ., The present study aims to investigate the human infested ticks species distribution; to determine their broad-ranges pathogens like Rickettsia spp ., , Anaplasma spp ., , Ehrlichia spp ., , Coxiella burnetii , Borrelia burgdorferi sensu lato , Francisella tularensis , Bartonella spp ., , Leishmania spp ., , Toxoplasma gondii , Babesia spp ., , Theileria spp ., , Hepatozoon spp ., , and Hemolivia mauritanica in Corum province of Turkey ., This study was carried out in the province of Corum ( 40° 33′ 00′′ N , 34° 57′ 14′′ E ) , which is located in Central Anatolia region of Turkey ( Fig 1 ) ., It has a surface area of 12 , 820 km2 , a population of 527 , 220 people , 152 , 244 of which live in the country site and another 374 , 926 in urban centers ., The mean altitude is 801 m , the mean annual precipitation 429 mm , and the mean temperature 10–11°C ., Due to the influences of the Black Sea and continental climates , the summers are hot and dry , while the winters are cold and rainy ., Wild animals such as deer , boar , bear , badger , fox , rabbit , wolf , marten , squirrel and beaver are abundant throughout the province ( Special Provincial Administration , Anonymous , 2009 ) , while in rural areas farm animals are bred ., From March to November 2014 specimens were collected from patients who applied to the Emergency Service of the Hitit University Research and Training Hospital with a tick infestation ., Ticks were morphologically identified under the stereomicroscope ( Leica MZ16 , Germany ) using standard taxonomic keys 28–30 ., Individual ticks were mechanically homogenized by crushing with liquid nitrogen using disposable micro pestle and the DNA was extracted using the Tissue and Bacterial DNA Purification Kit ( EURx DNA , Gdansk , Poland ) according to the manufacturer’s protocols ., All Polymerase Chain Reaction ( PCR ) amplifications were conducted with final volumes of 25 μl with 2 . 5 μl of template DNA , while negative and positive controls for each pathogen were used ., With the exception of Francisella tularensis and protozoa , ticks were molecularly screened for pathogens by real-time-PCR using Evagreen master mix ( Biotium , State , USA ) , while suspected samples were subjected to PCR ., For the detection of F . tularensis and Leishmania a real-time-PCR taqman probe was used ., For the identification of Babesia , the conventional PCR was used ., All positive samples were sequenced ., The primers BJ1 and BN2 amplifying Babesia spp ., , detected also Theileria spp ., , Hepatozoon spp ., and H . mauritanica ., The PCR methods , target genes and primer sequences used can be seen in Table 1 31–41 ., PCR positive samples were purified and sequenced in one direction at a commercial sequencing service provider ( Macrogen , Netherlands ) ., Nucleotide sequences were analyzed using nucleotide Blast ( National Centre for Biotechnology Information , www . blast . ncbi . nlm . nih . gov/Blast ) ., Representative nucleotide sequences from this study were submitted to GenBank under accession numbers MF383491-MF383615 and MF494656-MF494660 ., A phylogenetic tree was constructed using the MEGA5 . 1 program ., A total of 322 ticks were collected from humans and identified as Hyalomma marginatum ( n = 164 , 50 . 9% ) , Hyalomma excavatum ( n = 5; 1 . 5% ) , Hyalomma aegyptium ( n = 1; 0 . 31% ) , Hyalomma spp ., ( n = 46; 14 . 3% ) , Haemaphysalis parva ( n = 41; 12 . 7% ) , Haemaphysalis punctata ( n = 6; 1 . 8% ) , Haemaphysalis sulcata ( n = 1; 0 . 31% ) , Rhipicephalus turanicus ( n = 34; 10 . 5% ) , Rhipicephalus bursa ( n = 3; 0 . 93% ) , Dermacentor marginatus ( n = 17; 5 . 2% ) and Ixodes ricinus ( n = 4; 1 . 24% ) ., Overall , 37 . 2% of the examined ticks were infected with at least one pathogen; 3 . 7% of which with two different pathogens ., The infection rate was 100% in Dermacentor spp ., , 89% in Haemaphysalis spp ., , 75% in Ixodes spp ., , 37% in Hyalomma spp ., and 27% in Rhipicephalus spp ., A total of 17 microorganism species were identified ( Table 2 ) ., The most prevalent Rickettsia spp ., being R . aeschlimannii ( 19 . 5% ) , R . slovaca ( 4 . 5% ) , R . raoultii ( 2 . 2% ) , R . hoogstraalii ( 1 . 9% ) , R . sibirica subsp ., mongolitimonae ( 1 . 2% ) , R . monacensis ( 0 . 31% ) , and Rickettsia spp ., ( 1 . 2% ) ., In addition , the following pathogens were identified: Borrelia afzelii ( 0 . 31% ) , Anaplasma spp ., ( 0 . 31% ) , Ehrlichia spp ., ( 0 . 93% ) , Babesia microti ( 0 . 93% ) , Babesia ovis ( 0 . 31% ) , Babesia occultans ( 3 . 4% ) , Theileria spp ., ( 1 . 6% ) , Hepatozoon felis ( 0 . 31% ) , Hepatozoon canis ( 0 . 31% ) , and Hemolivia mauritanica ( 2 . 1% ) ., Table 3 shows the presence of bacterial pathogens according to the tick species , while in Table 4 the distribution of protozoan pathogens can be seen ., All samples were negative for Francisella tularensis , Coxiella burnetii , Bartonella spp ., , Toxoplasma gondii and Leishmania spp ., Recently , a lot of attention is being given to ticks and tick-borne diseases in Turkey , were many individuals died as a result of CCHF 1 , 3 , 4 ., Table 5 summarizes the studies done on ticks and their pathogens in the seven main regions of Turkey ( Fig 2 ) 8 , 12 , 14 , 24–27 , 42–83 ., In Corum province , 10 tick species infesting humans were identified , the most prevalent being H . marginatum , Hae ., parva , R . turanicus and D . marginatus ., Similar results from the same region has been obtained by Keskin et al . , 84 , 85 , who , in addition to the tick species found in the present study , also reported the infestation of humans with Haemaphysalis erinacei taurica and Ixodes laguri ., In their study the most prevalent tick species isolated from humans were H . marginatum , D . marginatus , R . turanicus and R . bursa ., The differences could be explained with the changes in tick abundance according to climatic conditions , host factors , socio-demographic factors , deforestation , as well as agricultural and wildlife management 86 ., In the present study all D . marginatus specimens were infected with at least one pathogen , while the infection rate was high also in Haemaphysalis spp ., Orkun et al . who investigated tick pathogens in Ankara province found high infection rate of Rickettsia spp ., , Babesia spp ., , and Borrelia spp ., in the same tick species 26 ., Rickettsia spp ., was identified as the most prevalent tick-borne pathogen in this study ( 31% ) ., Other studies reported an average infection rate of 41 . 3 in Istanbul 24 , while in Yozgat province the rate was 10 . 5% 56 , and in Ankara province 27 . 2%26 ., Rickettsia aeschlimannii is commonly transmitted by Hyalomma and Rhipicephalus spp ., 2 ., In Turkey , R . aeschlimannii was detected in H . marginatum , H . aegyptium , H . excavatum , R . bursa and R . turanicus ticks 24 , 26 , 56 , 87 , 88 ., In our study , this pathogen was found in all tick species examined with the exception of H . excavatum and R . bursa ., To the best of our knowledge , this is the first report that R . aeschlimannii was found in Haemaphysalis spp ., , Dermacentor spp ., , and Ixodes spp ., ticks , indicating that the pathogen might be transmitted also by other tick species ., According to nucleotide Blast and phylogenetic analysis ( ompA ) ( Annex 1 ) , R . aeschlimannii strains in our study is closely related with R . aeschlimannii isolate BB-35/Camli-H . marg ( 99–100% identity , accession number KF791251 ) ., Rickettsia aeschlimannii was the most prevalent ( 19 . 5% ) pathogen among Rickettsia-positive ticks in this study ., In an investigation which was performed in 2009 in Corum province , R . aeschlimannii was found in 5% of the ticks 87 , while in Ankara and Yozgat provinces , where similar climatic conditions prevail , this pathogen was detected in 4 . 7% and 4 . 3% , respectively of ticks examined 26 , 56 ., It was reported that R . aeschlimannii infections exhibited symptoms similar to Mediterranean spotted fever ( MSF ) 89 , and potentially lead to severe symptoms resembling to those of viral hemorrhagic fever 17 ., Accordingly , R . aeschlimannii infection should be included in the differential diagnosis , especially in endemic regions of MSF ., Rickettsia slovaca is usually transmitted by Dermacentor ticks and is associated with symptoms characterized by inoculation eschar on the scalp , necrosis erythema and cervical lymphadenopathy 2 , 24 , 56 , 88 , 90 ., This disease is either called tick-borne neck lymphadenopathy ( TIBOLA ) or Dermacentor-borne necrosis erythema and lymphadenopathy ( DEBONEL ) 90 ., Incidence of R . slovaca infections is likely underestimated ., Parola et al . reported that in 49 out of 86 ( 57% ) TIBOLA/DEBONEL cases the etiologic agent was R . slovaca 90 ., Throughout Europe , Dermacentor marginatus and Dermacentor reticulatus ticks are responsible from transmission of this pathogen 90 ., In our study , in addition to Dermacentor spp ., ticks , this pathogen was for the first time also detected in H . marginatum , Hyalomma spp ., nymphs and Hae ., parva ( Table 3 ) ., Nucleotide Blast and phylogenetic analysis ( ompA , ) of R . slovaca Corum strains were 99% identical to R . slovaca isolate BB-51/Akyurt-D . marg ( accession number KF791235 ) ( Annex 1 ) , while the gltA gene of R . slovaca Corum strains ( Annex 2 ) , showed a 99% identity to R . slovaca strain PotiR30 ( accession number DQ821852 ) ., In the present study R . slovaca was detected in 4 . 6% of the ticks ., In similar studies conducted earlier , R . slovaca was found in 0 . 3% of ticks in Corum 87 , in 4 . 8% in Yozgat province 56 , and in 9 . 4% in Ankara province 26 ., Similar to R . slovaca , R . raoultii is also the etiological agent of TIBOLA/DEBONEL and this Rickettsia seems to be less pathogenic and less frequent than R . slovaca 90 ., Parola et al reported that in 7 out of 86 ( 8% ) TIBOLA/DEBONEL cases the etiologic agent was R . raoultii 90 ., Dermacentor ticks are known vectors of R . raoultii 24 , 56 , 88 ., In the present study , in addition to Dermacentor spp ., , R . raoultii was also found in H . marginatum and Hyalomma spp ., nymphs ( Table 3 ) ., The nucleotide Blast and phylogenetic analysis of gltA gene of Corum R . raoultii strains ( Annex 2 ) share a 99% sequence identity to R . raoultii clone Ds1 ( accession number KF003009 ) and accordingly to ompA genes ( Annex 1 ) ., In addition , a 99% similarity was found to R . raoultii strain WB16/Dm Monterenzio ( accession number HM161789 ) ., Rickettsia raoultii was detected in 2 . 2% of the ticks examined ., Earlier studies from Corum reported that the percentage was 0 . 3% 27 and in Yozgat province 0 . 4% 56 , while this rickettsia was not detected in ticks from the Ankara region 26 ., In Corum province , the rate of R . slovaca and R . raoultii in ticks infesting humans increased in comparison to 2009 , and it seems that these pathogens are extending their vector diversity ., Rickettsia hoogstraalii has an unknown pathogenicity and it is transmitted by Hae ., Parva 26 , 56 , 88 , however , we found it in Hae ., parva and Hae ., punctata ticks ., The nucleotide Blast and phylogenetic analysis of gltA gene of Corum R . hoogstraalii strains ( Annex 2 ) have a 99% similarity to R . hoogstraalii strain RCCE3 with accession number EF629539 ., In our study the prevalence of R . hoogstraalii was 1 . 9% , while in Yozgat was 0 . 87% 56 , and in Ankara 13% 26 ., Rickettsia sibirica subsp ., mongolitimonae , symptoms are characterized by fever , eschar and lymphadenopathies 91 and it is transmitted by ticks such as Hyalomma asiaticum , Hyalomma truncatum , H . excavatum and R . bursa 2 , 91–93 ., We found this pathogen in H . marginatum , H . excavatum , R . bursa , and Hae ., parva ticks ., To the best of our knowledge this is the first detection of this pathogen in Hae ., parva ticks ., Nucleotide Blast and phylogenetic analysis of R . sibirica subsp ., mongolitimonae Corum strains ( ompA ) ( Annex 1 ) , showed a 99% identity to R . sibirica subsp ., mongolitimonae Bpy1 ( accession number KT345980 ) ., In this study this Rickettsia species was detected earlier in 1 . 2% of the ticks , while it was reported in 0 . 3% of H . marginatum ticks in Corum 87 and in 0 . 25% of ticks in Tokat province 71 ., Rickettsia monacensis infection shows flu-like symptoms , eschar and rash , the main vector of this pathogen being Ixodes ricinus 91 ., In Anatolian region of Turkey this tick species is rare 3 ., The ompA genes of Corum R . monacensis , which was detected also in our study in I . ricinus ticks , showed 99% identity with R . monacensis isolate Est1623 ( accession number KT119437 ) ( Annex 1 ) ., In previous studies this pathogens was not found in the Ankara and Yozgat provinces 26 , 56 , whereas the infection rate was 30 . 5% in ticks infesting humans in Istanbul 24 Ehrlichia spp ., effect both humans and animals such as dogs and domestic ruminants with symptoms like fever , malaise , leucopenia , thrombocytopenia , and abnormal liver function 94 ., The vectors of this pathogen are Amblyomma , Dermacentor , Rhipicephalus , Ixodes and Haemaphysalis ticks 2 , 94 ., In this study , Ehrlichia spp ., were detected in 0 . 93% of H . marginatum , Hyalomma spp ., nymphs and Hae ., parva ., Nucleotide Blast and phylogenetic analysis of groEL genes of Corum Ehrlichia spp ., strain ( Annex 3 ) was 99% identical to Ehrlichia ewingii isolate AaFT81 GroEL ., In Turkey , bovine anaplasmosis was detected in I . ricinus ticks which were collected from cattle in the cost of Black Sea 67 ., In the present study , Anaplasma spp ., was found in Hae ., parva ticks ., Nucleotide Blast and phylogenetic analysis of groEL genes of Corum Anaplasma spp ., strain shared an 81% identity to Anaplasma phagocytophilum isolate Omsk-vole52 with accession number KF745743 , ( Annex 3 ) ., Coxiella burnetii is the etiological agent of Q-fever with flu-like symptoms and is considered as a zoonotic disease ., The role of ticks in the transmission of C . burnetii to humans is low 95 ., In present study this pathogen was not detected in ticks infesting humans ., Borrelia afzelii is the pathogenic agent of Lyme disease transmitted mainly by ticks belonging to the genus Ixodes ., This pathogen is known from Europe , western parts of the former USSR and Northern Africa 2 ., We detected it in one I . ricinus specimen ., According to flagelline gene sequence analyses B . afzelii Corum strain was 100% identical to B . afzelii strain S60 with accession number KM198345 ( Annex 4 ) ., Orkun et al . reported the presence of Borrelia burgdorferi sensu stricto in 3 . 5% of Hyalomma spp ., and Hae ., parva in Ankara province 26 ., Lyme disease pathogens are prevalent in Istanbul region which has a moderate and wet climate and the infection rate in ticks collected from different areas was 38 . 7% 47 ., Francisella tularensis is the causative agent of tularemia a severe zoonotic diseases affecting animals and humans ., This pathogen was isolated from the bird-rabbit tick , Haemaphysalis leporispalustris 95 and from Dermacentor reticulatus infesting red foxes 96 ., In Turkey , tularemia cases were generally transmitted as water-borne but there are few tick-borne cases 46 , 57 , 97 ., F . tularensis was neither found in ticks collected from several barns , cattle and people 98 , nor in the ticks of the present study ., Bartonella spp ., are zoonotic vector-borne infection agents of humans ., One of them , B . henselae is the pathogenic agent of cat-scratch disease , the main vector being the cat flea ( Ctenocephalides felis ) 12 , however a direct link between tick bites , B . henselae and disease symptoms was reported in humans 99 ., In the present study B . henselae was not detected in any of the ticks examined ., Babesia spp ., are the pathogenic agents of babesiosis in humans and animals , which are considered as emerging diseases worldwide 86 ., In Europe , infection rates of Babesia spp ., in ticks ranges from 0 . 9 to 20% 100 ., B . microti is pathogenic to humans causing malaria-like symptoms ., The geographical distribution of this pathogen is USA , Canada , and Europe while the main vector is Ixodes spp ., 2 , 100 ., In USA , the prevalence of B . microti in ticks was 8 . 4% 101 , while in ticks collected from vegetation in Poland was 2 . 8% 102 ., In addition to Ixodes spp ., , B . microti was also detected in 0 . 7% of Dermacentor reticulatus in Switzerland 39 ., In Turkey , B . microti was for the first time detected in one I . ricinus tick collected from a ruminant 63 ., In Sinop province of Turkey , the sero-prevalence of B . microti in humans was 6 . 23% 64 , while in the present study , the prevalence of B . microti in H . marginatum ticks was 0 . 93% ., According to 18SrRNA gene nucleotide Blast and phylogenetic analysis , B . microti Corum strains were 100% identical to B . microti isolate RUS/Nov15-2950-Ipr with accession number KX987864 ( Annex 5 ) ., This is the first report showing the presence of B . microti in H . marginatum infesting humans , which is the most prevalent tick species in Corum province and is the main vector for B . microti ., Babesia occultans is a bovine parasite with high prevalence in South Africa , the vectors being Hyalomma spp ., 2 ., In Turkey , presence of B . occultans was reported by Aktas et al . in H . marginatum and R . turanicus collected from the vegetation , agricultural fields and grazing cattle , with a prevalence rate of 7%; transstadial and transovarial transmission of B . occultans were also demonstrated 103 ., Orkun et al . reported this pathogen in 0 . 6% of H . marginatum infesting humans 26 ., In our study B . occultans was present in 3 . 4% of H . marginatum , strongly supporting the presence of this pathogen not only in ticks infesting animals but also humans ., The 18SrRNA genes of Corum B . occultans strains showed a 99% similarity to B . occultans isolate Trender1with accession number KP745626 ( Annex 5 ) ., Babesia ovis is the causative agent of sheep babesiosis and mainly prevalent in Africa , Asia , and Europe , the vectors of this pathogen are R . bursa and R . turanicus 2 ., In Turkey , in ticks collected from sheep and goats the prevalence was 16 . 37% 79 ., B . ovis was detected by us in one R . bursa infesting a patient ., According to 18SrRNA gene nucleotide Blast and phylogenetic analyses ( Annex 5 ) , B . ovis Corum strains was 99% identical to B . ovis isolate tick20 . 3D with accession number KT587794 ( Annex 5 ) ., Recent studies show that ticks collected from cats and dogs may be responsible for the transmission of Toxoplasma gondii 21 ., Leishmania infantum was also found on ticks infesting dogs 22 ., In our study , these agents could not be detected ., Hepatozoon canis and Hepatozoon felis are the causative agents of hepatozoonosis in dogs and cats ., These pathogens are transmitted by Rhipicephalus sanguineus , Hae ., longicornis , and R . turanicus 2 ., In Turkey , H . canis and H . felis were for the first time identified in R . sanguineus ticks removed from dogs 83 , while H . canis infection was also reported in dogs 104 ., We demonstrated the presence of H . canis in D . marginatus and of H . felis in R . turanicus ., The 18SrRNA genes of Corum H . canis strain showed a 99% similarity to H . canis isolate 204B/13b ( accession number KP216425 ) , while the Corum H . felis strain showed a 99% similarity to H . felis , clone 8533 , accession number KC138533 ( Annex 5 ) ., Theileria spp ., are the pathological agents of theileriosis of ruminants , equids and felids , the vectors being ticks from the genera Hyalomma and Rhipicephalus 1 , 2 ., A transstadial but not transovarial transmission was reported in these ticks 105 ., In our study Theileria spp ., was demonstrated in Hyalomma spp ., infesting humans and the prevalence rate was 1 . 6% ., According to 18SrRNA genes , the Corum strain of Theileria spp showed a 92% similarity to Theileria youngi ( accession number AF245279 ) ( Annex 5 ) ., Hemolivia mauritanica is a pathogen of tortoises and transmitted by H . aegyptium 106 ., In the present study , this pathogen was found only in Hyalomma spp ., nymphs infesting humans and the prevalence rate was 2 . 1% ., According to 18SrRNA genes , Corum H . mauritanica strains showed a 99% similarity to H . mauritanica isolate SY-45-10 ( accession number KF992707 ( Annex 5 ) ., In conclusion , ticks in Corum province carry a large variety of human and zoonotic pathogens ., There are indications showing that there is an increase in the rate of ticks carrying spotted fever group and lymphangitis-associated Rickettsiae , while Ehrlichia spp ., and Anaplasma spp ., were reported for the first time in the region ., To the best of our knowledge B . microti was detected for the first time in H . marginatum infesting humans ., The presence of pathogens such as B . occultans , B . ovis , Hepatozoon spp ., , Theileria spp ., and H . mauritanica show the role of ticks for diseases of veterinary importance ., Pathogens are detected not only in ticks known as vectors but in a variety of other ticks , indicating wider vector diversity ., Patients with a tick bite history in Corum region should be followed not only for CCHF but also for other pathogens of medical importance . | Introduction, Methods, Results, Discussion | Tick-borne diseases are increasing all over the word , including Turkey ., The aim of this study was to determine the bacterial and protozoan vector-borne pathogens in ticks infesting humans in the Corum province of Turkey ., From March to November 2014 a total of 322 ticks were collected from patients who attended the local hospitals with tick bites ., Ticks were screened by real time-PCR and PCR , and obtained amplicons were sequenced ., The dedected tick was belonging to the genus Hyalomma , Haemaphysalis , Rhipicephalus , Dermacentor and Ixodes ., A total of 17 microorganism species were identified in ticks ., The most prevalent Rickettsia spp ., were: R . aeschlimannii ( 19 . 5% ) , R . slovaca ( 4 . 5% ) , R . raoultii ( 2 . 2% ) , R . hoogstraalii ( 1 . 9% ) , R . sibirica subsp ., mongolitimonae ( 1 . 2% ) , R . monacensis ( 0 . 31% ) , and Rickettsia spp ., ( 1 . 2% ) ., In addition , the following pathogens were identified: Borrelia afzelii ( 0 . 31% ) , Anaplasma spp ., ( 0 . 31% ) , Ehrlichia spp ., ( 0 . 93% ) , Babesia microti ( 0 . 93% ) , Babesia ovis ( 0 . 31% ) , Babesia occultans ( 3 . 4% ) , Theileria spp ., ( 1 . 6% ) , Hepatozoon felis ( 0 . 31% ) , Hepatozoon canis ( 0 . 31% ) , and Hemolivia mauritanica ( 2 . 1% ) ., All samples were negative for Francisella tularensis , Coxiella burnetii , Bartonella spp ., , Toxoplasma gondii and Leishmania spp ., Ticks in Corum carry a large variety of human and zoonotic pathogens that were detected not only in known vectors , but showed a wider vector diversity ., There is an increase in the prevalence of ticks infected with the spotted fever group and lymphangitis-associated rickettsiosis , while Ehrlichia spp ., and Anaplasma spp ., were reported for the first time from this region ., B . microti was detected for the first time in Hyalomma marginatum infesting humans ., The detection of B . occultans , B . ovis , Hepatozoon spp ., , Theileria spp ., and Hemolivia mauritanica indicate the importance of these ticks as vectors of pathogens of veterinary importance , therefore patients with a tick infestation should be followed for a variety of pathogens with medical importance . | Ticks are important vectors for different kind of pathogens , both of medical and veterinary importance , while tick-borne diseases ( TBDs ) are increasing all over the world ., In Turkey , many important human and zoonotic TBDs such as , Lyme borreliosis , rickettsiosis , anaplasmosis , ehrlichiosis , tularemia , bartonellosis , babesiosis , theileriosis , and hepatozoonosis have been reported ., Nonetheless , there is lack of research-based information concerning the epidemiology , ecology , and vector diversity of these tick-borne pathogens ., In this study , we aimed to investigate broad-range bacterial and protozoan vector-borne pathogens by PCR/RT-PCR and sequencing , those ticks infesting humans in the Corum province ., Spotted fever group rickettsiae and lymphangitis-associated rickettsiae , Borrelia afzelii , Anaplasma spp ., , Ehrlichia spp ., were detected ., Babesia microti was detected in Hyalomma marginatum infesting humans ., Interestingly zoonotic pathogens like Babesia ovis , Babesia occultans , Theileria spp , Hepatozoon felis , Hepatozoon canis , and Hemolivia mauritanica were also detected , showing the role of ticks for diseases also of veterinary importance ., This study provides important data for understanding the epidemiology of tick-borne pathogens and it is hoped that these results will challenge clinicians and veterinarians to unify their efforts in the management of TBDs . | taxonomy, invertebrates, medicine and health sciences, parasite groups, ixodes, pathology and laboratory medicine, pathogens, microbiology, vertebrates, animals, rickettsia, parasitology, parasitic protozoans, mammals, dogs, apicomplexa, phylogenetics, data management, protozoans, phylogenetic analysis, ticks, rickettsiales, bacteria, bacterial pathogens, infectious diseases, computer and information sciences, medical microbiology, microbial pathogens, ehrlichia, evolutionary systematics, disease vectors, arthropoda, arachnida, eukaryota, babesia, biology and life sciences, species interactions, evolutionary biology, amniotes, organisms | null |
journal.pcbi.1004272 | 2,015 | A Multiscale Model Evaluates Screening for Neoplasia in Barrett’s Esophagus | The incidence of esophageal adenocarcinoma ( EAC ) has increased dramatically over the past few decades in the US and other Western countries , prompting numerous epidemiological and clinical studies to characterize etiologic , genetic , and environmental factors that may contribute to this alarming trend 1 , 2 ., EAC arises primarily ( if not exclusively ) in Barrett’s esophagus ( BE ) , a metaplastic tissue alteration in the esophageal lining ., Screening is targeted toward identifying BE patients who are at the highest risk of developing dysplasia and cancer ., Although the risk of BE progressing to EAC is estimated to be low ( around 0 . 2–0 . 5% per year 3 ) , clinical evidence suggests that the risk of neoplastic progression in BE varies significantly between individuals depending on age , gender , race/ethnicity , gastroesophageal reflux disease ( GERD ) and whether or not dysplasia is present in BE ., High grade dysplasia ( HGD ) occurring in BE is generally non-invasive but carries a high risk of progression to EAC ., Low grade dysplasia ( LGD ) also occurs , but its clinical relevance is less certain ., Most patients diagnosed with HGD undergo endoscopic mucosal resection ( EMR ) or treatment with radio frequency ablation ( RFA ) to remove HGD tissue and , in the case of RFA , to reduce the amount of underlying metaplastic BE tissue ., Genetic and genomic studies , including longitudinal studies with multiple BE tissue samples from individual patients in the Seattle BE cohort 4 , also implicate specific genomic alterations in the neoplastic progression process ., Frequently observed alterations in BE include epigenetic silencing or loss of heterozygosity ( LOH ) of the P16INK4A and/or TP53 tumor suppressor genes 5–8 ., Whether these alterations necessarily lead to the clinical presentation of dysplasia and other cellular and architectural changes associated with this diagnosis is presently unknown ., However , our working hypothesis is that fields of HGD are comprised of clonal populations of premalignant cells that originate from distinct progenitors in the BE tissue ., Because dysplasia ( in particular HGD ) continues to be a widely used clinical predictor for progression to EAC , most BE patients are recommended to undergo periodic endoscopic surveillance with biopsies taken at specified spatial locations in BE to detect neoplastic changes ( dysplasia and/or cancer ) ., However , due to the large number of adults with BE in the general population ( ∼ 1–3% 9 , 10 ) , excessive or ineffective BE screening and surveillance that do not significantly reduce EAC incidence and mortality are a considerable public health concern ., To examine these issues , we developed a mathematical and computational framework that allows concurrent modeling of the BE-to-EAC progression and endoscopic screening for dysplasia and preclinical cancer prior to EAC diagnosis ., We will present the screening model as three cohesive modules ., First , we present the stochastic model for EAC at the cell level capturing key events of the random , GERD-dependent onset of BE , the initiation and stochastic growth of premalignant clones , malignant transformations in premalignant clones , and stochastic growth of malignant clones prior to ( symptomatic ) cancer detection ., This framework provides a bridge between the cell and population scales and has previously been described for modeling EAC incidence data in the US 11 , 12 ., The following two , novel parts of the model utilize and improve on this prior work and are essential for describing the screening process ., The second module is an explicit computational method to efficiently simulate the entire cell model in an individual BE patient until the time of a hypothetical screen ., This requires computation of the joint size distribution of premalignant and malignant clones in the BE tissue prior to development of an incident , symptomatic cancer ., The method captures the clonal progression of an idealized , 2D in silico tissue composed of intestinal crypts and can generate a variety of spatial patterns ( from circular to very diffuse shapes ) of both premalignant and malignant clones within the BE segment of a patient ., The third module simulates an endoscopic screen of a patient’s BE segment ., For a biopsy-based screen , the model mimics the Seattle standard protocol for screening patients with BE , probing the tissue every 1–2 cm with 4 quadrant biopsies for the presence of dysplasia and signs of invasive cancer ., We show that the efficacy of this protocol is highly variable and dependent on the sensitivity of detecting neoplastic abnormalities within a biopsy ., This sensitivity also affects the amount of dysplastic patients predicted to harbor undetected malignancy at time of screening ., The outcomes of biopsy-based screens are then compared with the model’s prediction for screening outcomes when using high-resolution imaging , a new screening technology not yet widely in use ., With information about the amount of small neoplasms that go undetected during biopsy-based screening , the model quantifies the potential advantages that image-based screening might offer ., Finally , this module simulates ablative treatment of BE patients with detected dysplasia during screening ., By explicitly modeling the curative effects of ablative treatment , we gain insights into the critical factors that may prevent treatment success ., The multistage clonal expansion for EAC ( MSCE-EAC ) cellular model assumes that the stepwise progression to cancer , formulated mathematically as a continuous-time Markov process , involves tissue alteration whereby part of the distal normal esophageal squamous epithelium ( with variable extent ) undergoes metaplastic transformation resulting in a columnar-lined epithelium called Barrett’s esophagus ( BE ) ., This tissue alteration provides a natural starting point for a cell-level description of the neoplastic progression to EAC ., Because gastroesophageal reflux disease ( GERD ) increases the risk of BE 13–15 , we assume that the rate of conversion of normal esophageal tissue to BE metaplasia is GERD-dependent ., Here we define symptomatic GERD ( sGERD ) patients as those with GERD symptoms occurring weekly or more frequently ., This represents an extension of an earlier model that did not include the effects of GERD 16 ., Specifically , we model the exponential BE rate , ν ( t ) , as a function of the prevalence of symptomatic GERD at age t , such that, ν ( t ) = ν 0 ( ( 1 - p s G E R D ( t ) ) + R R · p s G E R D ( t ) ) , ( 1 ), where psGERD ( t ) is the prevalence of GERD symptoms at age t and RR is the relative risk RR of GERD for BE ., The time-dependent cumulative distribution for BE onset is then given by, F B E ( t ) = Pr T B E ≤ t = 1 - e - ∫ 0 t ν ( s ) d s ., ( 2 ), See S1 Text for details on modeling psGERD ( t ) and S1 and S2 Figs for values of psGERD ( t ) and BE prevalence , FBE ( t ) , for males and females , respectively ., Once a tissue conversion occurs resulting in BE at exponentially distributed age TBE , the model continues as a multi-type branching process that includes stem cell counts of three different types: pre-initiated , initiated ( premalignant ) , and malignant ., This cellular description of carcinogenesis begins with the initiation of stem cells , enabling them to undergo clonal expansion ., In the current formulation of the model , initiation occurs as a result of two rate-limiting events ( e . g . , bi-allelic inactivation of a tumor suppressor , such as TP53 ) due to previous likelihood-based model selection 16 ., Once a stem cell is initiated , it undergoes clonal expansion through a stochastic birth-death-mutation ( b-d-m ) process with cell division rate αP and cell death/differentiation rate βP , so thus βP/αP is the asymptotic probability of extinction ., An initiated ( dysplastic ) cell may also undergo a transforming mutation with rate μ2 that generates an initiated cell and a malignant cell ., Malignant cells may undergo an independent clonal expansion with cell division and death rates αM and βM , respectively , allowing for stochastic growth and possibly extinction of the malignant tumor ., The inclusion of clinical , or symptomatic , detection of a malignant tumor occurs through a size-based detection process with parameter ρ ., Thus , the model captures two distinct clonal populations of cells—premalignant ( which we associate with dysplasia ) and malignant ( representing growth from early intramucosal to advanced carcinomas ) ., See Fig 2 for an illustrated realization of this MSCE-EAC stochastic process ., This cell-level description is linked to the population scale by means of the model hazard function , defined as the instantaneous rate of detecting cancers among individuals who have not been previously diagnosed with cancer , as previously shown 12 ., This quantity may be derived from the backward Kolmogorov equations for the stochastic multistage process described above and solved numerically via a system of coupled ordinary differential equations ( ODEs ) 17 ., See S1 Text for a mathematical derivation of these equations and how the hazard function can be obtained from their solutions ., Thus , one may infer rates of cellular processes from population level data as was previously done by likelihood-based calibration of the MSCE-EAC cellular kinetic parameters to incidence data , see 12 and S1 Text for more details ., For the illustrations and examples presented in this paper , we use the parameter estimates for biological rates employed in 12 , which are provided in Table, 1 . Our previous work with the stochastic , cell-level MSCE-EAC model did not directly calculate the expected number and sizes of independent focal lesions of each type in a patient’s BE segment at any given age in his/her lifetime , as depicted in Fig, 2 . However , this knowledge is clinically relevant for effectively monitoring progression to EAC in a BE patient ., In this module , we first describe the computational tool developed to obtain the MSCE-EAC stochastic realizations of the number and sizes of premalignant and malignant lesions in a BE patient at any given age ., Next , we use these model-derived outcomes of initiated stem cell numbers to simulate their spatial configuration as lesions in the BE tissue , which is important given the spatial nature of the biopsy screening protocols ., Because the mathematical complexity of this multistage model makes it difficult to derive tractable analytic size distributions for all cell types through time we resort to direct simulations of sample populations of individual lifetime trajectories to track clone number and sizes as progeny from certain cell types ., Recent advances in stochastic simulation allow further efficiency in computation of cell counts , enabling rapid model testing and examination of many possible scenarios ., See S1 Text for the full algorithm and implementation of the MSCE-EAC hybrid simulation of the number of clones and their sizes for all cell types present at time ts during a hypothetical screening ., We call this a ‘hybrid’ simulation because it employs stochastic simulation when necessary but also makes use of samples from analytical distributions when possible ., For the simulation of premalignant ( dysplastic ) clones , we employ two methods ., The first is an exact method , the stochastic simulation algorithm ( SSA ) , first described by Gillespie 18 , that simulates every jump in cell count and exponential waiting times between events ., The second is a highly efficient approximation to SSA called τ-leaping ., S1 Text explains these two methods and describes how the MSCE-EAC simulation uses them cooperatively in a highly efficient approach ., The accuracy of both the size distributions generated by the SSA and the τ-leaping method are shown in S3 Fig as Q-Q plots for the size distributions of non-extinct premalignant clones compared to the analytical distribution for an independent b-d-m process ., With cell module parameters as input , the MSCE-EAC hybrid algorithm simulates the multi-type branching process for an individual’s cellular progression from birth until time ( age ) ts , which can be repeated to generate ( synthetic ) data for a sample population ., In summary , for those individuals who are found to have BE by time of screening , each patient has a specific number of BE stem cells ( X ) , number of pre-initiated cells ( P* ) , a number of non-extinct premalignant ( P ) clones with respective sizes , a number of non-extinct malignant ( M ) clones with respective sizes and information about the parental P clones from which the M clones originated , and lastly whether the patient is a prevalent , clinical EAC case by time ts ., Note , the stochastic model captures the possibility that the ancestor premalignant clone may go extinct while the malignant clone is still growing at the time of screening ts ., Fig 3A shows the random trajectories for a simulated BE patient’s clones obtained via this algorithm for the five years of life prior to initial screening at age ts = 60 ., The MSCE-EAC tissue module computes the number of stem cells in each neoplastic clone and generates the shapes of these clones within a BE segment at any given age of a patient ., The MSCE-EAC screening module takes this information and performs an endoscopic screen on this realized BE segment ., Here we outline the methodology for generating model predictions related to three specific screening outcomes: ( 1 ) the probability that small cancers are missed during biopsy-based screening , ( 2 ) the potential gains in neoplasia detection probabilities if screening occurred via high-resolution tomographic imaging , and ( 3 ) the efficacy of ablative treatments that result in the curative depletion of metaplastic and neoplastic cell populations in BE in terms of the long-term impact on reducing EAC incidence ., These model predictions are described in Results ., The methods outlined in this section are implemented by the comprehensive MSCE-EAC screening model consisting of three modules: cell , tissue , and screening ., All necessary tools to employ this method , including examples of user inputs used in the upcoming Results , are available in documented R code at https://github . com/yosoykit/MSCE_EAC_Screening_Model ., In the current epidemiological literature , studies beginning with a biopsy-based index screen of BE patients ( i . e . , the screen when a patient is first diagnosed with BE ) provide widely variable estimates of the prevalence of HGD , ranging from ∼ 2 . 75–8 . 25% 25 , 30–34 ., To compare this with model-derived predictions , we simulated an index endoscopic screen on a sample population of patients with BE and computed the prevalences of both premalignancy ( HGD ) and screen-detected ( non symptomatic ) malignancy ., For an illustrative example of the MSCE-EAC screening model outputs , we simulated an index endoscopy for all males and females at screening time ( age ) ts = 60 in the year 1990 ( indicative of index screens from prospective studies that estimate the BE to EAC progression rate ) ., With the BE prevalence given in Eq ( 2 ) , these results focus on expected observations in output regarding the subpopulation of individuals found with BE , for whom the MSCE-EAC screening model provides screening results ( see Methods ) ., Because the detection of a neoplastic lesion may involve both premalignant and malignant cells transformed within the lesion , we first consider the ( random ) sum of the two cell types to determine the efficacy of the biopsy protocol to detect a neoplastic lesion in BE ., The biopsy sensitivity was varied from 10% to 95% , as seen in Figs 4 and 5 , to allow for systematic exploration of sensitivity effects ( see Methods ) ., If a neoplasm is detected on a biopsy , we doubled the biopsy sensitivity for malignant content because the biopsy is under closer inspection ., Along with difficulties in first detecting dyplasia present in BE during endoscopic screening , several studies suggest that many BE patients who are diagnosed with HGD without malignancy actually have an undetected cancer that was missed during biopsy screening 36 , 37 ., The MSCE-EAC screening model estimates the probability that a positive HGD patient actually harbors a synchronous , occult malignant clone that is not screen-detected either because it was completely missed in a biopsy sample ( e . g . see the small malignancy depicted in Fig 3B ) or because it was undetected in a biopsy for a particular biopsy sensitivity , perhaps due to insufficient histologic sectioning ., This is an interesting , clinically relevant feature of our modeling ., The model predicted the expected fraction of undetected EAC in BE patients diagnosed with HGD to be between 3 . 2%–14 . 2% for men and 4 . 3%–19 . 3% for women ( see Fig 5 ) ., We conclude that the higher probability of missed malignancy in women is due to the lower probability of finding any neoplasia ( due to smaller clone sizes , S4–S7 Figs ) in women during index endoscopy ( see Fig 4 ) ., These predicted ranges are compared with studies of HGD patients found with concurrent adenocarcinoma , which remained undetected even by rigorous biopsy protocols but are later discovered during resection of the esophagus 38–42 ., However , from these esophagectomy studies conducted over the past two decades , the reported prevalence of synchronous malignancy among HGD patients widely varies from 0–75% ., With strict adherence to the Seattle protocol , our model generated a lower estimate of concurrent EAC risk in HGD patients than most published studies , yet it is consistent with the most recent study by Konda et al . when biopsy sensitivity is low 41 ., It is also possible that the studies with high estimates of concurrent malignancy were biased because cancer was suspected in these patients indicating esophagectomy ., High-resolution imaging of BE ( a technology still in infancy and not yet widely utilized ) may provide a benefit through the early detection and endoscopic resection of small premalignant and malignant lesions ., The MSCE-EAC screening model can explore the potential quantitative improvements of screening for neoplasia when diagnosed via optical endomicroscopy compared with a less sensitive biopsy protocol ., To this end , we simulated the results from an optical coherence tomography ( OCT ) screen in which a positive detection of HGD and/or malignancy occurs if the geometric size of a clone on image is greater than a resolution area threshold , aOCT ( see Methods ) ., Assuming aOCT = 1mm2 and the same assumption for stem cell density σ that was used in previous results , the HGD prevalence ( excluding incident EAC cases ) rose to an expected 27 . 89% for the BE cohort used in the previous examples ( 1930 birth year , ts = 60 ) ., Therefore , for the range of probabilities of HGD detection shown in Fig 4 , the MSCE-EAC screening model estimated an expected 68 . 7% to 92 . 8% increase in HGD detection probability using a sensitive imaging technology for screening rather than biopsy-based screening ., This modeling exercise reinforces the conclusion that many neoplastic clones of detectable size are being missed with current biopsy protocol screening endoscopies ., As a third example demonstrating the utility of the MSCE-EAC screening model , we computed the projected cumulative hazard ΛMSCE ( t ) in Eq ( 3 ) after a single index screen of BE patients at time ts = 60 , removal of screen-detected EAC patients , and subsequent RFA treatment of HGD positive patients ., We explored RFA efficacy under various assumptions about the impact of ablation on cell counts , as specified by the ablation proportion vector ω ( see Methods ) ., When comparing to the background incidence ( in which no screening occurs ) , we predicted the effect on EAC cumulative incidence based on a range of RFA effectiveness assumptions ( See Fig 6 ) ., If patients that were positively detected with HGD at index screen ( 6% with 60% biopsy sensitivity ) receive RFA , the MSCE-EAC screening model predicted that by year 2030 , expected EAC cumulative incidence will be reduced by 17 . 1% if 50% of all BE cell types are effectively removed ( ω = { . 5 , . 5 , . 5 , . 5} ) and be reduced by 32 . 1% if 99% of all BE cell types are effectively removed ( ω = { . 01 , . 01 , . 01 , . 01} ) ., To explore the future influence of missed malignancies , the model predicted that if RFA removed all malignancies ( ω = {1 , 1 , 1 , 0} ) but left behind the HGD tissue , then treatment would only moderately reduce future EAC cumulative incidence by an expected 15 . 7% before 2030 ., However , removing the HGD tissue as well as preclinical malignancies ( ω = {1 , 1 , 0 , 0} ) during treatment would create a more significant average reduction in EAC cumulative incidence of an expected 38 . 7% ., The model’s predictions of the possible RFA effects on cell populations seem to support the hypothesis that the effectiveness of RFA is determined by its ability to ablate premalignant ( dysplastic ) tissue ., Interestingly , even the biopsy procedure on all BE patients offers a slight therapeutic effect ( EAC cumulative incidence will not return to background ) by the mere chance of endoscopically removing , at times , significant amounts of neoplastic tissue in a biopsy specimen , assuming no negative effects from wounding associated with tissue removal ., These results are clearly a simplification of a highly variable and complex clinical procedure , representing only a basic example , but the model is poised to incorporate realistic RFA touch-ups throughout surveillance , as it occurs in current practice , to give increasingly realistic projections ., Although few Barrett’s esophagus ( BE ) patients progress to EAC in their lifetime , the cancer burden is considerable due to generally poor treatment outcomes and survival ., EAC contributes approximately 4% to all male cancer deaths in the US 43 with a flattening but still increasing trend in mortality according to recent projections based on Surveillance , Epidemiology , and End Results ( SEER ) data 12 ., Because BE is an actionable EAC precursor with a considerable prevalence of 1–3% in the general population 9 , 10 ( translating into a large number of individuals ) and an annual risk of progressing to EAC of approximately 0 . 2–0 . 5% per year 3 , optimal surveillance for neoplastic alterations in BE and effective treatment strategies are a major challenge to clinicians given the current lack of evidence-based decision tools ., Thus we have developed a detailed multiscale model of EAC to better understand the natural history and impact of screening , intervention , and prevention of EAC ., The mathematical framework of our multistage clonal expansion for EAC ( MSCE-EAC ) screening model describes the step-wise progression and transformation from normal squamous esophageal tissue to a columnar crypt-structured metaplastic tissue in which clonal expansions of dysplastic and malignant cells can occur ., Because the description is fully stochastic , it affords predictions of important clinical endpoints that reflect the intrinsic ( inter-individual ) heterogeneity in the disease process that explains , at least in part , why some individuals progress to cancer in their life-time while others do not ., In contrast to earlier formulations of the multistage clonal expansion ( MSCE ) model for EAC 16 , 17 , which analyzed patterns of EAC incidence in the general population , the present model includes two novel modules for exploration of clinical endpoints before symptomatic detection of EAC ., The tissue module explicitly computes the number and sizes of neoplastic clones in a BE patient and quantifies their spatial structure within an idealized crypt-structured BE segment at time of screening ., With this patient-specific information , we then employ a screening module to perform a screen in silico at a specified screening age ., As our BE screening examples demonstrate , this model extension makes it possible to explicitly explore current BE screening efficacy while controlling the operational characteristics of the screening protocol ., We show that the detection of high grade dysplasia ( HGD ) or cancer using the standard ( Seattle ) biopsy protocol is strongly dependent on the minimum neoplastic tissue fraction needed to be detectable in the biopsy ., This sensitivity would be further affected by altering the spacing between biopsy levels and size of the biopsy forceps according to different protocols ., Additionally , our MSCE-EAC screening model predicts that over 10% of BE patients screened who receive a diagnosis of HGD with biopsy-based screening also harbor a missed preclinical malignancy with mid-range biopsy sensitivity ., We find that the overall efficacy of the biopsy protocol is highly uncertain due to variability in tissue sampling between practitioners and due to considerable uncertainties in the histological assessment of the biopsied tissues ., Our results also suggest that even the best current biopsy protocols may miss between 70%–90% of small HGD lesions that are detectable when using high-resolution optical coherence tomography ( OCT ) imaging at 1mm resolution ., While not yet widely available , high-resolution OCT allows a more complete ( wide-field ) examination of the BE segment ., Our results suggest that OCT could surpass the biopsy-based protocols in efficacy to detect neoplastic lesions ., However , because quantitative data with OCT are still lacking , the results remain speculative , but serve to demonstrate the potential gains of OCT screening over the standard biopsy protocol ., Finally , the present framework also allows for the modeling of treatment , such as radio frequency ablation ( RFA ) ., Ablation attempts to remove the intestinal metaplasia together with all neoplastic cells ., Assuming that ablation simply decimates the number of BE , dysplastic , and malignant crypts by specific fractions , we computed the residual cancer risk of EAC after RFA ( see Fig 6 ) ., This ‘decimation by fraction’ approach also lends itself to modeling the curative effect of multiple RFA ‘touch-ups’ delivered over a span of time to improve RFA efficacy ., From the results derived from simulating an ablative treatment on a population of BE patients found to be positive for HGD during screening , we found that it was crucial to ablate dysplastic and not only preclinical malignant tissue to achieve the most significant impact on future EAC incidence ., Although the example given in this study is somewhat simplistic and does not include the random spatial characteristics of the ablation process , the model framework can accommodate more complex assumptions regarding the biological effects of RFA , including random spatial effects of the ablation ‘burn’ and localized presence of intestinal metaplasia hidden beneath the neosquamous tissue after RFA treatment ., In summary , the MSCE-EAC screening model introduced in this paper offers a comprehensive multiscale method to model the neoplastic processes unfolding in BE together with a mechano-spatial modeling of the screening process and treatment ., Our results demonstrate the limitations of the standard biopsy-based protocol for the detection of HGD and early cancer due to a highly heterogeneous distribution of dysplastic precursors and malignant foci that can arise in dysplasia ., We further demonstrate that these limitations could be overcome by high-resolution OCT imaging which may provide additional biological details and insights into the cancer process , including the growth dynamics of neoplastic clones ( in particular their numbers and sizes over time ) , information that can easily be incorporated into the multiscale description of EAC development and screening presented here . | Introduction, Methods, Results, Discussion | Barrett’s esophagus ( BE ) patients are routinely screened for high grade dysplasia ( HGD ) and esophageal adenocarcinoma ( EAC ) through endoscopic screening , during which multiple esophageal tissue samples are removed for histological analysis ., We propose a computational method called the multistage clonal expansion for EAC ( MSCE-EAC ) screening model that is used for screening BE patients in silico to evaluate the effects of biopsy sampling , diagnostic sensitivity , and treatment on disease burden ., Our framework seamlessly integrates relevant cell-level processes during EAC development with a spatial screening process to provide a clinically relevant model for detecting dysplastic and malignant clones within the crypt-structured BE tissue ., With this computational approach , we retain spatio-temporal information about small , unobserved tissue lesions in BE that may remain undetected during biopsy-based screening but could be detected with high-resolution imaging ., This allows evaluation of the efficacy and sensitivity of current screening protocols to detect neoplasia ( dysplasia and early preclinical EAC ) in the esophageal lining ., We demonstrate the clinical utility of this model by predicting three important clinical outcomes: ( 1 ) the probability that small cancers are missed during biopsy-based screening , ( 2 ) the potential gains in neoplasia detection probabilities if screening occurred via high-resolution tomographic imaging , and ( 3 ) the efficacy of ablative treatments that result in the curative depletion of metaplastic and neoplastic cell populations in BE in terms of the long-term impact on reducing EAC incidence . | Endoscopic screening for detecting cancer and cancer precursors in Barrett’s esophagus ( BE ) is currently informed by repeated systematic biopsying of the metaplastic BE tissue ., Here we present a comprehensive multiscale model of the natural history of esophageal adenocarcinoma ( EAC ) , which describes the entire multistage process beginning with the conversion event of normal squamous esophageal tissue to BE metaplasia , the spatio-temporal formation of independent dysplastic and malignant clones at the cell level , and finally the appearance of symptomatic EAC in BE ., This model lends itself to a systematic exploration of the efficacy and sensitivity of current biopsy-based screening methods to detect neoplasia in BE patients , as well as alternative screening techniques based on high-resolution imaging of the BE tissue ., Moreover , the model can also be used to predict the impact of ablative treatments on the risk of occurrence or recurrence of dysplasia or cancer ., Due to the lack of studies that attempt to explicitly model the physical and biological dimensions of the screening process itself , our computational model provides a unique , publicly-available tool to improve understanding of factors that limit the efficacy of current screening protocols for BE patients . | null | null |
journal.pgen.1004551 | 2,014 | The Population Genetics of Evolutionary Rescue | The history of life is punctuated by periods of mass extinction ., It has become clear that we are now living through such a period: present species extinction rates are 100–1000 fold higher than background rates 1 , 2 ., It is also clear that this burst of species extinction largely reflects human activity , including the combined consequences of habitat destruction , pollution , and climate change e . g . , 3 , 1 , 4 , 5 ., Not surprisingly , present extinction rates— and the threat they pose to biodiversity— have received much attention over the last few decades ., Until recently , population genetics has had little to say about extinction ., Extinction is , however , partly a population-genetic phenomenon ., Theory as well as experiments with microbes suggest that some threatened species may be able to adapt to environmental change on a sufficiently fast time-scale to prevent their extinction ., This phenomenon , so-called evolutionary rescue , has been the focus of considerable empirical and , to some extent , theoretical work for an overview , see 6 and other papers in the special issue of the Proceedings of the Royal Society B ., Here we extend the population genetic theory of evolutionary rescue ., We focus on a sudden environmental change that is severe enough to lower the populations mean absolute fitness below one ., Consequently , the population cannot replace itself and begins to decline geometrically in numbers ., Unchecked , this decline will lead to extinction ., To survive , the population must adapt and it must do so quickly ., As Maynard Smith 7 emphasized , adaptation in a threatened population is unlike ordinary adaptation ., Instead , it is a race against extinction ., While a substantial literature considers the case in which adaptation involves a quantitative genetic ( polygenic ) response to selection , we consider the simple case in which adaptation involves evolution at a single locus ., This case appears to be important biologically , as responses to human-induced change— e . g . , insecticide resistance , industrial melanism , heavy metal tolerance , etc ., — often involve rapid change at single genes reviewed in 6 , 8 ., We further consider an abrupt change in the environment , which then remains in this new state for the period of time that we consider for gradual change in the environment , see reference 6 ., Finally , we focus on a particular regime in which the allele that might rescue a threatened population is initially rare , i . e . , either present in low copy number or appearing as a recurrent mutation ., If the allele were not rare , the population would suffer little risk of extinction in the first place ., Put differently , we restrict attention to that regime in which a species suffers a great risk of extinction ., Evolutionary rescue is characterized by a U-shaped curve of population size 9 , 10 ., As Figure 1 shows , when the environment changes at time t\u200a=\u200a0 , mean absolute fitness drops below one , and the population begins to decline in numbers ., Conditional on evolutionary rescue , mean absolute fitness will , at some point , rebound to exceed one and population size will begin to grow; this occurs at time ., Population size will then continue to increase until attaining some large stable value ., As Figure 1 also shows , the U-shaped curve for total population size is , in our scenario , the superposition of two curves: one that characterizes the geometric decline in number of individuals that carry the wildtype allele and the other that characterizes the increase in number of individuals that carry the beneficial allele ., Here our main goal is to better characterize evolutionary rescue mathematically ., In particular , we describe this U-shape curve analytically ., We focus on the behavior of the average size of a population through time conditional on evolutionary rescue occurring ., A complete solution to this problem has , to this point , proved elusive ., As we will see , part of the reason is that such a solution requires incorporating a subtle population-genetic effect ( familiar from the theory of genetic hitchhiking ) into this largely ecological problem ., We emphasize approximate results throughout ., Given the complexity of ecological problems— all else is rarely equal in the real world— we suspect that it is more important to obtain approximate results that are intelligible and somewhat robust to departures from assumptions than exact ones that are neither ., Our results are typically simple enough to allow intuitive interpretation ., We study the same model described in Orr and Unckless 11 ., Briefly , we consider a haploid model in which adaptation involves a rare beneficial allele at a single locus ., Mating is random and there is no population structure or migration ., The environment changes suddenly , altering the fitness of alleles; these new fitness values then remain constant through the time period studied ., We assume no clonal interference among beneficial alleles ., Time is discrete and measured in generations ., ( We will , however , make continuous time approximations when convenient . ), At time t\u200a=\u200a0 , a population of size N0 made up entirely ( or almost entirely ) of wildtype individuals experiences a sudden environmental change ., As the wildtype allele has absolute fitness 1-r in the new environment , the number of wildtype individuals decreases geometrically though time ., Following MacArthur and Wilson 12 , chapt ., 4 , Leigh 13 , Lande 14 , Orr and Unckless 11 and others , we assume a simple form of population regulation in which population size can grow exponentially until it hits a carrying capacity , K . A beneficial allele that increases absolute fitness to either resides at low frequency , , at t\u200a=\u200a0 or arises recurrently by mutation after the environmental change ., If the allele resides in the standing genetic variation , k copies are present at time t\u200a=\u200a0 ( ) ., As noted in the Introduction , we assume throughout our analysis that k is small ( though see Text S1 ) ., It seems likely that k might often be small in actual threatened populations as such populations often begin with fairly small sizes ., Evolutionary rescue , if it occurs , involves an increase in frequency of the beneficial allele before the population goes extinct ., Any allele that can cause evolutionary rescue must enjoy an absolute fitness greater than one , requiring s>r ( assuming that the product s * r is negligibly small ) ., One simplifying assumption that we make throughout is that the quantity s-r is small enough to justify Haldanes 2s ( in our case , 2 ( s-r ) ) approximation to the probability that a unique mutation escapes stochastic loss ., Some mutations that might save a population could be of large effect and would violate this assumption ., In such cases , it is straightforward to replace the approximate quantity 2 ( s-r ) in our calculations with the more exact one , 1-exp ( -2 ( s-r ) ) throughout ., The results will be more cumbersome and less intuitive but they generally do not change qualitatively ., While we are primarily interested in analytically characterizing evolutionary rescue , we check all of our approximate analytic results against computer simulations ., These simulations are described in Orr and Unckless 11 ., Briefly , these are exact stochastic ( forward ) Monte Carlo simulations that follow threatened populations of a given initial size through time ., Orr and Unckless 11 calculated the probability that newly-arising mutations cause evolutionary rescue ., Given a per gamete per generation rate of mutation , u , to a beneficial mutation of fitness effect s ( s>r ) , they showed that this probability is ( 1 ) Bell 15 derived essentially the same result ., An analogous calculation lets us find the probability that alleles from the standing genetic variation cause evolutionary rescue ., These alleles are present at time t\u200a=\u200a0 at frequency p0 ., So long as these alleles are rare and each copy enjoys an independent evolutionary fate , we have ( 2 ) Eq ., ( 2 ) is agnostic about the historical forces responsible for the presence of the allele at t\u200a=\u200a0 . The allele may , for instance , have been previously deleterious or previously neutral ., The total probability of evolutionary rescue from either new mutation or the standing genetic variation ( we ignore rare events wherein copies of both types of alleles contribute ) is or ( 3 ) It is easy to find the conditions under which evolutionary rescue is more likely to involve the standing genetic variation versus a new mutation , where both types of allele enjoy selective advantage s ., From Eq . s 1 and 2 , this occurs when ( 4 ) a result that seems not to have been noted in the literature ., New mutations are more likely than the standing variation to cause evolutionary rescue when the inequality is reversed ., Eq ., 4 is independent of both s and N0 ., Its dependence on p0 and u is intuitive – a higher initial frequency favors a role for standing genetic variation while a higher mutation rate favors a role for new mutation ., The effect of r is subtler ., A populations rate of decline affects both standing variation and new mutation in that it decreases the rate at which the number of mutant individuals can grow ( ∼1-r+s ) ., But r has a further effect on new mutations ., Each generation , it erodes the raw material— wildtype individuals— required for production of new mutations ., Thought of differently , Eq ., 4 reflects the fact that the expected number of copies of the beneficial allele in the standing variation is N0p0 while the expected cumulative number produced by new mutation before a population goes extinct is N0 u/r see reference 11 ., Given the shared factor of N0 , the relative magnitudes of p0 versus u/r determines which scenario involves the larger number of copies ., Although Eq ., 4 is obviously approximate , it agrees remarkably well with computer simulations ( Figure 2 ) ., If we were to assume that standing genetic variation segregates at the deterministic mutation-selection balance ( p0\u200a=\u200au/sd; where sd is the fitness cost of the mutation before the environmental change ) , Eq ., 4 suggests that standing genetic variation is more likely than new mutation to save the population when u/sd>u/r , i . e . , when sd<r ., We now consider the U-shaped curve in Figure, 1 . We would like to characterize this curve mathematically , tracking the average size of a population through time conditional on evolutionary rescue ., We first consider evolutionary rescue that involves a rare allele from the standing genetic variation ., We derive the average population size through time ( this section ) as well as properties of the average rescued population at the moment that it begins to rebound ( subsequent sections ) ., We then turn to the case in which evolutionary rescue involves new mutation , which is more complex ., The total size of a population at time t conditional on evolutionary rescue can be written , where is the number of mutant individuals conditional on the allele not having been lost ., Because accidental loss of a rare beneficial allele typically occurs early , this quantity can be interpreted , once t is appreciable , as the number of mutant individuals present conditional on evolutionary rescue ultimately occurring ., Put differently , once considerable time has passed , any beneficial allele that is still present has almost certainly escaped accidental loss ., Taking expectations , ( 5 ) The key to our approach involves finding ., If the mutant allele were to increase in frequency deterministically , loss would never occur and we would have ., Consequently , it might seem that the expected population size would be ( 6 ) where we use a continuous time approximation and that , in continuous time , ., We also assume that k is small enough that the initial number of wildtype individuals is ∼N0 ., Eq ., 6 essentially reflects the approach of Gomulkiewicz and Holt 9 , although their single-locus model was diploid and featured Malthusian fitness parameters ., While Eq ., 6 is adequate when both alleles are common , it can diverge dramatically from the correct solution when the rescuing allele is initially present in low copy number ( see below ) ., The source of the discrepancy is simple ., Loss of rare beneficial alleles is common and the above approach ignores this loss ., More subtly , loss of rare beneficial alleles affects not only the probability of evolutionary rescue but expected population size when the beneficial allele is not lost ., The point was seen by Maynard Smith 16 and emphasized by Maynard Smith and Haigh 17 in their classic analysis of genetic hitchhiking ., The key point is that successful alleles— those that sweep to fixation— are not a random sample of initially-rare beneficial alleles ., Instead , successful alleles are disproportionately those that rise by genetic drift to higher than expected copy number in the first few generations of their evolutionary histories 18 ., Such alleles have a greater chance of being successfully “grabbed” by natural selection ., Maynard Smith showed that this oversampling effect could be taken into account in otherwise-deterministic selection equations by a simple , albeit approximate , approach ., It is , he argued , as though the alleles that successfully fix began with higher copy number than they actually did , a finding that often plays a part in hitchhiking theory ., This point is also well known in the branching process literature , at least in certain limiting cases , e . g . , 19 This insight can be imported into our problem to find ., Here we follow Maynard Smiths 16 informal argument ., ( He considers a unique new mutation but , as we will see , his argument is trivially generalized to a rare allele from the standing variation . ), Maynard Smith noted that , with t appreciable enough that a mutant allele has either been lost or has reached large enough numbers that it will ultimately fix , we have , where the left hand side is the number of mutant individuals expected deterministically and ., Thus ( 7 ) In the case of a new mutation in a stable population , and , conditional on non-loss , the expected number of individuals that carry the beneficial allele at time t is larger by a factor of 1/ ( 2s ) than expected naively 17 ., Our problem differs from Maynard Smiths in two small ways ., First , our allele begins from the standing variation ., Second , our population is shrinking ., Both effects can be taken into account to calculate the appropriate in Eq ., 7 ., Given k copies at time t\u200a=\u200a0 , ( 8 ) where 2 ( s-r ) is the approximate probability of fixation of a unique copy of the beneficial allele with small selective advantage in a population that declines geometrically 20 ., For an allele that starts at low copy number , i . e . , k is small , Eq ., 8 is , to a good approximation , ., Thus Eq ., 6 becomes ( 9 ) where the last step is a continuous time approximation appropriate with small s-r ., Perhaps surprisingly , Eq ., 9 is independent of k ( for small k ) ., In words , the expected number of mutant individuals at time t conditional on ultimate fixation equals the deterministic expectation for a single new mutation normalized by its probability of fixation ., This result has a simple interpretation ., When starting with small k and conditioning on fixation , descendants of only one copy typically sweep to fixation ., ( This reflects the fact that the probability of fixation of each copy is generally small with weak selection , especially in a declining population . ), The expected number of copies present at time t conditional on ultimate fixation is therefore the same as that for a single mutation normalized by its probability of fixation ., Substituting in Eq ., 5 , the expected total size of a population at time t conditional on evolutionary rescue is ( 10 ) where we again use a continuous time approximation with in continuous time ., ( We also again assume that the initial number of mutant individuals is negligible compared to N0 . ), Eq ., 10 is one of our key results ., It lets us trace the expected size of a rescued population through time ., Figure 3 shows that Eq ., 10 performs very well when compared to exact computer simulations ., Figure 3 also shows that Eq ., 6 , which ignores Maynard Smiths oversampling effect , performs poorly when the desired beneficial allele is initially present as a small number of copies ., We can also find the variance in conditional on evolutionary rescue , at least roughly ., Given that the two types propagate independently , ., If we assume the wildtype population is large enough that it behaves approximately deterministically ( an assumption that will break down at some point ) , most of the variance in will reflect variance in Nmut ., Crudely , then , ., Theory extending Maynard Smiths insight shows that , conditional on fixation and with t large , the distribution of copy number for a beneficial allele that escapes loss is approximately exponential with the mean given in Eq ., 9 18 , 19 ., Thus ( 11 ) Figure 4 shows that this approximation performs well once t is appreciable ., The fit is less good early on ., Text S1 generalizes the results of this section when the number , k , of copies of the beneficial allele present at t\u200a=\u200a0 is not small ., We now characterize the average rescued population at the moment it begins to rebound in size , i . e . , the moment it hits its minimal size in Figure, 1 . When does this rebound occur ?, And what is the smallest size , , experienced by the average rescued population ?, To find the time the mean population begins to rebound , note from Eq ., 10 that ( 12 ) which equals zero when ( 13 ) As , Eq ., 13 represents a minimum , which can also be seen in Figure 3 ., Figure 5 also shows that Eq ., 13 is very accurate ., Eq ., 13 shows , not surprisingly , that the greater selective advantages of a mutant allele the shorter the time to rebound ., ( It may seem surprising that s alone , not s-r , appears . The reason is that the system of allele frequency changes depends only on the fitness difference between genotypes . ), The effect of r is harder to intuit ., Eq ., 13 shows that , all else equal , faster population decline increases the time until expected N rebounds ., To understand this , note that Eq ., 13 represents a time that is conditional on rescue ., With larger r the probability of such rescue decreases but— if rescue does occur— larger r slows the time to rebound ., The reason is that the absolute number of mutant individuals grows as ∼1+s-r , which is smaller for larger r ., It is also worth noting that the time until rebound is independent of copy-number , k , of the beneficial allele , so long as k is small ., This can also be seen in Figure 5 ., ( See Text S1 for results with somewhat larger k . ), Eq ., 13 also makes clear that the time until rebound is more sensitive to s than to N0 and r , which enter only logarithmically ., We can also find the minimum expected population size experienced during evolutionary rescue ., Substituting tmin into Eq ., 10 , we get ( 14 ) Figure 3 shows that Eq ., 14 performs well compared to simulations ., The size of the average population at its minimum will obviously affect a populations genetic diversity post-recovery as effective population size is sensitive to the smallest population size experienced through time ., Our approach in this section has involved characterizing the expected size of a population , , conditional on rescue ., Eq ., 13 , for instance , gives the time when hits its minimum; this is not necessarily identical to the average of the times when individual realizations hit their minimum ., Similarly , Eq ., 14 gives the minimum size of conditional on rescue; this is not necessarily identical to the average of the minima experienced in individual realizations ., ( The minimum of the average need not equal the average of the minima . ), Nonetheless , Figure 3 shows that our approach roughly captures the behavior of what might be loosely thought of as the “typical” rescued population ., We can also specify relations between the numbers of mutant and wildtype individuals present when the average population rebounds ., At tmin we find that ( 15 ) to the order of our approximations ., This result is independent of N0 and p0 , so long as k is small ., Simulations show that Eq ., 15 is reasonably accurate ., For example , with s\u200a=\u200a0 . 02 and r\u200a=\u200a0 . 001 , 0 . 005 or 0 . 01 , Eq ., 15 yields 0 . 053 , 0 . 333 and 1 . 00 , respectively ., The observed values were 0 . 054 , 0 . 316 and 0 . 923 given an initial population size of N0\u200a=\u200a10 , 000 and k\u200a=\u200a1 ( 100 , 000 successful realizations ) ., If we increase the initial population size to N0\u200a=\u200a100 , 000 ( with k\u200a=\u200a1 ) , the observed values were 0 . 054 , 0 . 347 and 1 . 05 ( 100 , 000 successful realizations ) ., If , on the other hand , we increase the number of copies of the mutant alleles segregating in the standing variation , i . e . , k\u200a=\u200a10 and N0\u200a=\u200a10 , 000 , the observed values are 0 . 051 , 0 . 331 , 0 . 951 ( 100 , 000 successful realizations ) ., We can also estimate the time needed for the average population to return to its pre-crisis carrying capacity , N0 ., This occurs when Eq ., 10 equals N0 , which is roughly ( 16 ) This approximation is crude in several ways ., First , we assume that the number of wildtype individuals is negligible by the time the population arrives at ., Second , we assume that the carrying capacity in the new environment is the same as in the old environment ., This need not be true and there is some experimental evidence that it is not , at least in the laboratory 21 ., Third , we assume , following MacArthur and Wilson 12 and Lande 14 and others , that the population maintains exponential growth until hitting carrying capacity ., Eq ., 16 would almost certainly be inappropriate under other forms of population regulation ., We now turn to characterizing the U-shaped curve when evolutionary rescue involves a new mutation ., This scenario is more complex than above as we must consider two dynamics: the time until a successful new mutation arises and the time then required for the allele to reach high frequency ., In particular , we modify the approach taken above to reflect a delay in the time required for a new mutation to appear that escapes stochastic loss ., During this delay the number of wild-type individuals continues to decline ., Consequently , the mean population size conditional on rescue will behave as: ( 17 ) where is the probability density of waiting times for the origin of a successful new mutation ., The quantity 2 ( s-r ) in the denominator again takes Maynard Smiths oversampling effect into account ., Orr and Unckless 11 showed that the distribution of waiting times until the appearance of a new mutation that escapes loss in a geometrically declining population is itself approximately geometric ( ∼ exponential ) : ( 18 ) In words , because the population declines geometrically , the supply of new mutations declines about geometrically and , consequently , rescue is more likely early than late ., Although improved solutions to the distribution of origination times of successful mutations are possible ( see Text S2 , Figure S1 ) , we rely here on Eq ., 18 , which is simple and often adequate ., From Eq . s 17 and 18 , we find that the expected population size is ( 19 ) Figure 6 shows that Eq ., 19 performs very well compared to simulations ., Figure 6 also shows when alleles derive from the standing variation ( Eq . 10 ) , allowing comparison between the two scenarios ., With new mutation , the U-shaped curved is stretched to the right: recovery takes longer than with standing variation , reflecting the waiting time required for the appearance of a successful new mutation ., Interestingly , the expected size of a rescued population under new mutation ( Eq . 19 ) is identical to that under standing variation ( Eq . 10 ) except for the factor of r/s in the second term of Eq ., 19 ., Because , with rescue , r/s<1 , the mean size of a rescued population is , at any moment , smaller with new mutation than with standing variation , reflecting the waiting time for the appearance of a successful new mutation ., By analogy to Eq ., 11 , and continuing to ignore the variance in numbers of wildtype individuals , the variance in population size at time t is ( 20 ) Simulations confirm that Eq ., 20 performs well when t is not very small ( not shown ) ., We can also solve for the time at which the mean population begins to rebound , conditional on rescue by new mutation ., It is: ( 21 ) As Figure 6 shows , this result is quite accurate ., Remarkably , tmin is independent of r , to the order of our approximations ., This is because r plays two opposing roles in the time to ENmin ., First , as r increases ( with s constant ) , the rate of increase of the mutant allele ( with fitness 1+s-r ) decreases ., This increases the time to ENmin ., Second , as r increases , the population declines faster , so that— conditional on rescue occurring— the new mutation must have arisen fairly soon after the environmental change ., As remarkably , tmin with new mutation is identical to that with standing variation except that the quantity s replaces r in the argument of the logarithm ( see Eq . 13 ) ., Because s>r , tmin for new mutation is larger than that for standing variation , again reflecting the additional waiting time for the appearance of a successful new mutation ., Finally , the mean population size at tmin with new mutation is ( 22 ) Not surprisingly , this is smaller than the mean size found for standing variation ( Eq . 14 ) ., Because rebound occurs later with new mutation , populations decline to a smaller average size before rebounding ., We can also estimate the time needed for the average population to return to its pre-crisis carrying capacity ., This occurs when Eq ., 10 equals N0 , which yields roughly ( 23 ) This equation is identical to that for the standing variation case ( Eq . 16 ) except , again , for the term of s/r in the argument of the logarithm; consequently , is longer for new mutation than it is for standing variation ., The same ( serious ) caveats apply to Eq ., 23 as to Eq ., 16 ., We have extended the population-genetic theory of evolutionary rescue ., In particular , we have considered a scenario in which the environment changes suddenly and the population attempts to adapt to this change via evolution at a single locus ., ( Though far from universal , considerable data indicate a role for single genes in response to , e . g . , human disturbance; see below . ), We further focus on a particular regime in which the desired beneficial allele is initially rare , i . e . , it is either present in few copies at time zero or it appears as a new mutation after time zero ., This regime is of special interest as adaptation is far from deterministic and the population is therefore seriously threatened by the environmental change and suffers a considerable probability of extinction ., If the beneficial allele were much more common , adaptation would be essentially deterministic and the population would suffer little probability of extinction ., Gomulkiewicz and Holt 9 considered this case in which natural selection at a single locus deterministically rescues a threatened population ., ( There is obviously a gray area between the rare and common allele regimes; see Text S1 for some analysis of this gray area . ), One of our most interesting findings involves comparing evolutionary rescue from a rare allele that resides in the standing genetic variation versus a new mutation ., Eq ., 4 shows that rescue from the standing variation is more likely than that by new mutation when the initial frequency , p0 , of the beneficial allele exceeds u/r ., Conversely , new mutation is more likely to be involved when p0<u/r ., These results are independent of s— given that both types of alleles share the same selective advantage— and reflect the relative expected number of copies of the desired allele that arise by new mutation before extinction versus that reside in the standing variation ., We have also derived an approximate equation for the U-shaped curve that characterizes evolutionary rescue ( Eqs . 10 and 19 ) ., This curve describes the trajectory of expected population size through time conditional on rescue ., When beneficial alleles are initially rare ( small k ) , derivation of this quantity requires taking into account a subtle population-genetic effect familiar from the theory of genetic hitchhiking: rare beneficial alleles that sweep to high frequency behave , in deterministic selection equations , as though they began at a higher frequency than they actually did ., The reason , first seen by Maynard Smith 16 , is that natural selection is more likely to “choose” alleles that accidentally drift to somewhat higher than expected frequencies early in their histories see also 18 ., Incorporating this oversampling effect into our calculations , we derive several key quantities that characterize the U-shaped curve of evolutionary rescue , including the time until the expected population size begins to rebound ( Eq . s 13 and 21 ) and the smallest expected population size experienced ( Eq . s 14 and 22 ) ., These quantities assume surprisingly simple forms and perform well when compared to exact computer simulations given small k ., The U-shaped curves of evolutionary rescue differ depending on whether rescue involves new mutation ( Eq . 19 ) or rare alleles from the standing variation ( Eq . 10 ) ., In particular , the waiting time until the mean population size begins to rebound is longer with new mutation than with standing variation ., Similarly , the minimum expected population size experienced during rescue is smaller with new mutation than with standing variation ., Both findings reflect the fact that the U-shaped curve for rescue is delayed , i . e . , stretched to the right , for new mutation relative to standing variation ., This , in turn , reflects that evolutionary rescue by new mutation involves a waiting time that does not appear with the standing variation— that required for the appearance of a successful mutation , i . e . , one that escapes accidental loss ., We also derive the variance in population size through time , albeit roughly ., Though beyond the scope of the current paper , our results have some implications for the genetic diversity of populations that experienced evolutionary rescue from a sudden environmental shock ., For example , as the minimum average population size is always smaller under evolutionary rescue from new mutation than from the standing genetic variation , populations that adapted via a new mutation will likely often experience more loss in diversity than populations that adapted from the standing variation ., ( It must be emphasized , however , that our results involve the expected size of rescued populations and there is often much variation about these expected values . ), Our analysis does feature several important limitations ., First , we focus on adaptation at a single locus ., While the frequency of single-locus adaptation ( or , more plausibly , adaptation that involves a major effect at some single locus ) remains somewhat uncertain , matters are clearer when considering sudden and dramatic environmental changes of the sort modeled here ., Many such changes , or at least those that have been analyzed genetically , involve responses to human disturbance , e . g . , antibiotic resistance , insecticide resistance , industrial melanism , etc ., These environmental challenges are often met via evolution at a single | Introduction, Results, Discussion | Evolutionary rescue occurs when a population that is threatened with extinction by an environmental change adapts to the change sufficiently rapidly to survive ., Here we extend the mathematical theory of evolutionary rescue ., In particular , we model evolutionary rescue to a sudden environmental change when adaptation involves evolution at a single locus ., We consider adaptation using either new mutations or alleles from the standing genetic variation that begin rare ., We obtain several results:, i ) the total probability of evolutionary rescue from either new mutation or standing variation;, ii ) the conditions under which rescue is more likely to involve a new mutation versus an allele from the standing genetic variation;, iii ) a mathematical description of the U-shaped curve of total population size through time , conditional on rescue; and, iv ) the time until the average population size begins to rebound as well as the minimal expected population size experienced by a rescued population ., Our analysis requires taking into account a subtle population-genetic effect ( familiar from the theory of genetic hitchhiking ) that involves “oversampling” of those lucky alleles that ultimately sweep to high frequency ., Our results are relevant to conservation biology , experimental microbial evolution , and medicine ( e . g . , the dynamics of antibiotic resistance ) . | Changes to an organisms environment may have such adverse effects on fitness that the population begins to decline in size ., To survive , the population must adapt before it goes extinct ., Such “evolutionary rescue” is characterized by a U-shaped curve: population size declines and then recovers as a beneficial allele increases in number ., Here we describe this U-shaped curve mathematically when the rescuing allele starts out rare ., We obtain several results ., First , we calculate when evolutionary rescue is more likely to come from new mutation than from the standing genetic variation ., Second , by describing the entire U-shaped curve mathematically , we derive the time until the average rescued population begins to rebound in size as well as the smallest average population size experienced before rescue ., We also find that evolutionary rescue from new mutation takes longer and involves a smaller minimum population size than rescue from the standing genetic variation . | evolutionary ecology, species extinction, ecology and environmental sciences, ecology, biology and life sciences, evolutionary adaptation, evolutionary theory, evolutionary biology, evolutionary processes, evolutionary genetics | null |
journal.pntd.0004661 | 2,016 | Projected Impact of Dengue Vaccination in Yucatán, Mexico | Dengue is currently the most important arboviral disease of humans and has an increasing global public health burden 1 ., Worldwide , the combined annual number of infections by the four dengue serotypes has been estimated to be close to 400 million , of which 96 million develop symptomatic disease 2 ., Globally , dengue incidence has consistently increased for the last five decades due to geographic expansion and transmission intensification in endemic tropical and subtropical regions 3–6 ., Since individuals may be infected multiple times with different viral serotypes , and because re-infection is associated with an increased risk for severe disease , dengue presents unique challenges for prevention and control 7 ., Vector control is the only option currently practiced to reduce dengue transmission , with most efforts targeting Aedes aegypti and Ae ., albopictus , but these programs provide limited protection and may not be sustainable ., Most communities undertaking vector control lack the budget , personnel , and expertise needed to effectively reduce mosquito populations ., Although the use of DDT as a vector control measure substantially reduced dengue transmission in the 1960’s and 70’s , vector control efforts in the post-DDT era have not been sufficient to prevent invasion of dengue into new regions 8–10 ., Vaccination may soon be available as an additional option for dengue intervention ., Six vaccines are in clinical development , but to date only the Sanofi-Pasteur vaccine , Dengvaxia , has completed phase III trials 11 ., Phase III trials conducted in Latin America estimated vaccine efficacy of 64 . 7% ( 95%CI 58 . 7 , 69 . 8 ) , while the estimate from a trial in South East Asia was 56 . 5% ( 95%CI 43 . 8 , 66 . 4 ) ., Pooled analysis of these two trials indicates vaccine efficacy is significantly higher for participants with pre-existing dengue neutralizing antibodies ( 81 . 9%; 95%CI 67 . 2 , 90 . 0 ) compared to those who were seronegative at the time of vaccination ( 52 . 5%; 95%CI 5 . 9 , 76 . 1 ) ., Vaccine efficacy against hospitalization for dengue in Latin America was 80 . 3% ( 95%CI 64 . 7 , 89 . 5 ) and in South East Asia was 67 . 2% ( 95%CI 50 . 3 , 78 . 6 ) , and vaccine efficacy estimates varied by serotype in both trials 12–14 ., Overall , these are promising results for Dengvaxia , but trial outcomes have been mixed ., Efficacy appears to decline in recipients that have not had a previous natural infection , including an apparent increased risk of hospitalization in pre-school age children ., Vaccine rollout plans will need to be carefully evaluated , particularly the age targeted for routine vaccination relative to the age by which people in the target region have typically had an infection 15–18 ., To accurately model the impact of potential vaccination strategies , we must account for the primary mechanisms that influence dengue transmission: mosquito population dynamics and behavior 10 , 19 , 20 , seasonal influences on these factors 21 , 22 , human movement and demography 23 , 24 , the build-up of strain-specific immunity in the population through time , and the immune response following re-exposure 6 , 15 , 19 , 25 ., Consistent with the global trend , dengue incidence and severity have increased significantly in Mexico over the past four decades , with transmission regularly reported in 28 of the 32 Mexican states ., Incidence is particularly well-documented in the state of Yucatán ., Dengue was reintroduced to Yucatán in 1979 after widespread DDT use in the region ceased; the virus had not been detected in the state for the previous two decades 26 , 27 ., We use an agent-based simulation model fitted to data on dengue occurrence to examine the possible effectiveness of deploying a Dengvaxia-like vaccine in Yucatán ., We compare potential vaccine rollout strategies under varying assumptions regarding the duration of vaccine-induced immunity ., In particular , we consider routine vaccination targeting different age groups ( 2 , 9 , or 16 year olds ) , with and without one-time catch-up campaigns , and with durable or waning vaccine-induced immunity ., Our transmission model extends previously published work that examined dengue vaccination in Thailand with a hypothetical vaccine 28 ., Uninfected mosquitoes acquire dengue virus by biting infectious humans ., Infected mosquitoes that survive the extrinsic incubation period ( EIP ) can infect new hosts on subsequent blood-feeding attempts ( Fig 1; see Section S1 in S1 Text ) ., Infected humans also incubate the pathogen before they become infectious ., Dengue virus comprises four serologically distinguishable lineages , called serotypes ., Infection produces lifelong immunity to the infecting serotype and induces temporary cross-protection against other serotypes , but can later enhance severity in a subsequent infection , an effect referred to as antibody-dependent enhancement ( ADE ) ., ADE also occurs in infants due to maternal antibodies 37 ., We represent infection with all four serotypes , with different disease outcome probabilities depending on number of previous infections ., Different disease outcomes ( asymptomatic , mild , and severe ) also influence transmission , since they have shorter ( asymptomatic ) to longer ( severe ) infectious periods ., In our model , seasonal rain reliability determines mosquito population size and temperature determines EIP ., With these seasonality drivers , after fitting we find the dengue basic reproduction number ( R0 ) falls below 1 for roughly four months each year , indicating that trans-seasonal maintenance via transmission in the local human population is unlikely ., The mechanism that causes dengue to re-emerge after the winter in Yucatán is unknown ., Plausible explanations include but are not limited to human movements from regions with on-going transmission , infected mosquitoes introduced via e . g . freight , and vertical transmission in the vector ., We address seasonal re-emergence using a small , fixed daily exposure probability to represent whatever the real processes are ., The serotypes for these exposures are based on the observed serotypes in Yucatán ., We classify infections as asymptomatic ( “inapparent” ) or symptomatic ( a “case” ) ., Cases are further separated into mild or severe , and for comparisons with empirical data , we assume that mild and severe cases can have different reporting rates ., We model the pathogenicity ( probability an infection is symptomatic ) as a reference probability combined with relative-risk factors for particular serotypes , past infection history , and age ., We use a similar approach for the probability of severe disease ., Finally , in our model infants ( people <1 year old ) may have maternal antibodies , which causes them to either resist infection or have enhanced pathogenicity ( see Section S1 in S1 Text ) ., We explicitly model individuals and their activity in a synthetic population within Yucatán , based on census and household data 38 , 39 , and national statistics on local economies and schools 40 , 41 ., Individuals have a fixed age , gender , and household , and may travel to school or work during the day ., We use gender to determine a mother-child relationship when an infant is exposed so that we can consider maternal antibodies , because the model population structure specifies cohabitation but not familial relationship ., For practical reasons , we do not consider changing population size or age structure: fixing these across all simulations dramatically reduces the model’s computational complexity ., Household composition and location , as well as associated schools or workplaces , are also static and identical across all simulations ., These locations and the distances between them provide all the spatial information in the model , as humans and mosquitoes strictly exist at and move between these places ., Using such locations for spatial distribution is convenient because household , school , and workplace data are available from national and international sources , and because it is a natural approximation of where people and mosquitoes interact ., Representing how acquired immunity changes over time in a population is critical for modeling long-term epidemic dynamics ., To avoid the complexity of a dynamic demographic model , instead of having individuals age , immune histories are annually shifted from younger to older individuals ., Thus , the population matures by accumulating immunity as epidemics occur , while maintaining a realistic household age distribution ., Because the size of age cohorts trends downward with increasing age , mature immune histories are implicitly lost due to mortality and replaced with newborns who have no acquired immunity , though they may have temporary maternally-derived immune responses ( see Section S2 in S1 Text ) ., We model the mosquito population in two parts: aggregate populations for uninfected ( susceptible ) mosquitoes at each location , and mobile agents for infected mosquitoes ., The aggregate mosquito populations have location-specific sizes drawn from an exponential distribution with a fitted mean ( see Section S5 in S1 Text ) that varies seasonally as a function of rainfall ( Fig 2 ) ., Upon infection , individual mosquitoes are separated from the aggregate population ., The mosquito’s age is determined by sampling from the total mosquito age distribution ., Her EIP is then drawn from the day’s EIP distribution ( based on effective temperature ) and added to her age to determine at what age she will become infectious ., Finally , the mosquito’s age-at-death is determined by sampling from the distribution of mosquito ages greater than or equal to the current age ( see Section S3 in S1 Text ) ., Mosquitoes that will die before becoming infectious cannot contribute to disease transmission and effectively die instantaneously ., Individual mosquitoes may move between houses , workplaces and schools; if they do move , they randomly select from adjacent locations weighted by the inverse distance-squared distance to those locations ., Simulation time has important features at several scales: daily day-night cycles , seasonal change in temperature and rainfall by day-of-year , annual population turnover , and multi-year eras ., Each day bridges a day-night cycle , 7 a . m . –7 a . m ., People go to work or school ( or stay home if not employed ) during the day and are at their homes during the night; there is no representation of weekends or holidays ., Individuals’ daily movement patterns sometimes change in response to disease ( see Section S2 in S1 Text ) , but not otherwise ., Mosquitoes are more likely to bite during the day than at night ., Seasonal effects on the mosquito population ( driven by precipitation patterns ) and on the EIP ( driven by temperature ) change daily , based on a time series generated from 35 years of historical temperature 42 and precipitation 43 data ., Each year , the human population ages on the same day of the year ( Julian day 99; i . e . , April 9 ) , which is roughly the nadir in transmission ., Leap-days are not modeled: each year is 365 days long ., We do not adjust the observed data to match this assumption ., At the multi-year scale , there are four time periods: ( 1 ) a priming period , to establish a stable distribution of acquired immunity in the population; ( 2 ) an intense vector control period , corresponding to the use of DDT in Yucatán ( 1956 to 1978 ) ; ( 3 ) a fitting period ( 1979 first recorded Yucatán dengue epidemic to 2013 last year with complete data ) ; and finally ( 4 ) a 20 year forecast period ( 2014 to 2033 ) where we consider vaccination strategies ., All periods ( priming , DDT era , fitting , and forecast ) are simulated with the same parameters in a particular run , except that during the DDT period , both mosquito populations and external introductions are reduced ., We represent vector control by reducing the mosquito populations by 77% and the rate of dengue introductions by 90% , based on the fraction of households that were treated with DDT and other insecticides in Yucatán and neighboring states , respectively , from 1956 to 1978 44 ., In our model , vaccination reduces , but does not eliminate , the probability of infection with dengue given exposure , and has no other direct effect on transmission or disease ., Vaccination status is added to an individual’s immune history but is distinct from immunity acquired by natural infection ., We consider vaccines that confer durable ( lifelong ) protection , as well as the possibility that vaccine-induced protection declines linearly ( “wanes” ) with the number of days since vaccination ., We consider three possible waning half-lives ., We assume vaccine efficacy ( VE ) values consistent with the the phase III trial results for Dengvaxia in Latin America 12 , 13 ( see Section S4 in S1 Text ) ., Trial results indicated that prior dengue infection approximately doubles the vaccine efficacy ., Table 1 gives the values used in the model ., Dengvaxia has a three dose schedule , each six months apart ., We also model a 3 dose regimen , with the first dose providing full efficacy , and all vaccinees receiving all three doses ., For scenarios where the vaccine wanes , it does so between doses as well , but each dose is assumed to return efficacy back to the initial level ., During the forecast period ( 2014–2033 ) , individuals in an age category ( 2 , 9 , or 16 year-olds ) are targeted for routine vaccination annually on Julian day 100 , one day after immune histories are transferred ., We also consider scenarios in which routine vaccination is supplemented with a one-time catch-up campaign , where vaccination occurs across many age groups ( from one year older than the routine age , up to age 30 ) in the first year of the forecast period ., When considering booster strategies for a waning vaccine , booster doses are provided to all previously vaccinated individuals , every two years ( irrespective of waning period ) from the final dose date ., Because the sizes of our age groups differ , we cannot simply hold the fraction of individuals vaccinated in an age group ( i . e . , coverage ) constant while assessing the outcomes for targeting different ages: at the population level , this would be assessing different vaccination rates and confound with other differences due to age ., Instead , we hold the number of doses constant across routine strategies , and across catch-up campaigns when used ., However , we base the number of doses on attaining 80% coverage in 9 year olds when targeting that age ( 30 , 100 vaccinees ) , and 60% coverage in the catch-up cohort associated with 9 year olds ( 448 , 500 vaccinees ) ., While we have represented some of the features of the Sanofi-Pasteur vaccine , our intent is to represent a generic , moderate-efficacy vaccine ., We assume that vaccine performance is affected by serostatus but not age of vaccinee per se , as that has not been specifically tested in the trials ., We do not address the potential complexities indicated by trial results in Southeast Asia , particularly any potential for disease enhancement 14 ., We assess vaccination strategies by contrasting the projected dengue burden with and without vaccine deployment over a 20 year forecast period ., Matched baseline and vaccination scenario runs are simulated for the forecast period , with these comparison runs sharing parameters produced by the fitting procedure ( see Section S5 in S1 Text ) as well as simulated history for the priming , vector control , and fitting periods ., We average across 1000 runs ( 100 parameter combinations times 10 samples each ) of the baseline and each scenario to get an expected number of cases each year ., In a time interval Δt , the total vaccine effectiveness ( Veff , Δt ) is 1 minus the proportion of the symptomatic cases in the vaccination scenario ( VSΔt ) relative to the number in a baseline with no intervention ( BΔt ) :, V eff , Δ t = 1 - VS Δ t B Δ t ( 1 ) We calculate this value for an annual ( Δt = 1 year , beginning 0 , 1 , 2 , … , 19 years after initiation of the intervention ) and cumulative ( Δt = 1 , 2 , … , 20 years , all from initiation of the intervention ) basis , across parameter combinations and replicates ., The model uses parameters from a wide range of sources ( see Section S5 in S1 Text ) ., Our synthetic human population was constructed using satellite imagery , microcensus , workplace and school data ( see Synthetic Human Population ) ., Seasonality in mosquito population and EIP is based on empirical temperature and precipitation time series ( see Section S3 in S1 Text ) ., When data were not available to inform or fit the model directly , as with certain vaccine performance parameters , we made assumptions that simplified the model implementation ., Epidemiological , entomological , and vaccine parameters were taken from the literature , or fit using Approximate Bayesian Computation ( see Table D in S1 Text ) ., We fit our model to reported case and seroprevalence data collected between 1979 and 2013 ( the fitting period , main text Fig 3 and Fig . P in S1 Text ) ., We retained the 100 best-performing parameter combinations ( out of 70 , 000–10 , 000 per set , 7 sets ) from the fitting procedure ., In our forecasts , we used each of those 100 parameter combinations 10 times , for a total of 1000 replicates ., Since transmission parameters vary seasonally , we estimate the basic reproduction number , R0 , by day of year ., For a vector-borne disease , R0 may be thought of as the number of additional human infections that are expected to result from a single infected human in a naïve population , after exactly one human-mosquito-human transmission cycle ., R0 does not take into account existing immunity in a population , but it nonetheless can provide some intuition about the seasonal timing and peak size of an epidemic ., We estimate R0 as follows: for each day of the year , we randomly infect an individual in an otherwise completely susceptible population , allow that person but no other people to infect mosquitoes , run the simulation forward until all infections clear , and count all human infections after the first ., We do this for the same 1000 samples used in the forecasting and average the number of secondary infections across the samples to compute R0 for that day ., Using the best parameter combinations from our fitting procedure , we simulated the historical and fitted period outbreaks to establish background immune profiles ., Our model generally predicts the size and timing of epidemics during most of the fitting period ( 1979–2008 ) , but not the large epidemics since 2009 ., We also forecast transmission from 2014 through 2033 ( the 20 year forecast period ) without any intervention to provide a baseline to compare interventions against ., Median results are reported here , and prediction intervals can be found in the supplement ( see Section S7 in S1 Text ) ., To generally characterize dengue transmission in the region , we calculate the seasonally-varying R0 for DENV1 ., We estimated a seasonal peak R0 of 5 . 2 , occurring in August ( Fig . A in S1 Text , panel C ) ., From late December through mid April , R0 is below 1 . 0 ., Dengue introductions to the population in the model can happen throughout the year , so stuttering transmission chains are still observed in those months ., We did not repeat this analysis for all serotypes , but the others would have generally lower R0 given our assumption that DENV1 has the highest risk of severe disease , and thus longer infectious periods than the other serotypes ., We evaluate the performance of the fitted model using seasonality and seroprevalence data that were not used in the fitting procedure ., Precipitation and temperature seasonality in the model ( see Section S3 in S1 Text ) drives changing mosquito populations and changing EIPs ( Fig . A in S1 Text ) ., These seasonal effects successfully reproduce the overall shape and timing of average weekly dengue incidence ( Fig 2 , average of years for which weekly data are available ) ., Seroprevalence , or the fraction of individuals with at least one past dengue infection , is a general indicator of whether an epidemic is possible and how large one might be if it occurs ., This relationship is more complicated for dengue given the temporary nature of cross-protection between serotypes and subsequent disease-enhancement , but still provides some insight ., To qualitatively assess our model fit , we compared age-stratified results from a recent serosurvey of Mérida 45 , 46 , the largest city in Yucatán , with simulated seroprevalence among Mérida residents in the synthetic population ( Fig 4 ) ., Our model results overlap with confidence intervals for the measurements , but are generally low for individuals below age 20 and high for those over 30 ., For scenarios that assumed vaccine-induced immunity did not wane over time ( a durable vaccine ) , annual effectiveness gradually increased for routine-only strategies ( Fig 5 ) and spiked early with catch-up campaigns followed by short increasing trend , then a gradual decline to roughly the same level as achieved with routine-only vaccination ., Routine-only vaccination started near 0% and increased to 65% annual effectiveness , while strategies with catch-up started near 65% and quickly increased to 75% , but after about 7 years , began to decrease to 65% by end of the forecast period ., Varying the target age for routine vaccination had a modest impact on annual effectiveness ( Table 2 ) , although strategies targeting older children generally out-performed those targeting younger children ., We expect a positive correlation between target age and effectiveness , based on the anticipated trend in seroprevalence with age ( confirmed for this population at the outset of the forecast period; see Fig 4 ) and our assumption that antibody-primed vaccinees benefit from enhanced vaccine efficacy ., However , annual effectiveness for all strategies appeared to be converging by the end of the forecast period ., We also considered vaccine-induced immunity that wanes linearly over time ( Fig . D in S1 Text ) , with three different half lives ., Under the 2 year half-life waning model , for example , vaccinees have immunity based on Table 1 immediately after each dose , and that efficacy declines linearly each day until it is 0 at 4 years post-vaccination ., Waning substantially reduced the long-term effectiveness of routine strategies that did not include additional booster vaccinations ( dashed lines , Fig 6A ) ., When vaccinees were re-vaccinated every 2 years , performance improved , reaching 50% effectiveness after 20 years ., Waning had a more dramatic effect on strategies with catch-up ( Fig 6B ) ., Catch-up campaigns with waning vaccines but no booster vaccination all resulted in negative annual effectiveness—performance worse than baseline—at some point within 20 years ., That outcome is the effect of delaying cases in the catch-up cohort: annual effectiveness initially looks good as cases are temporarily prevented relative to the baseline , but when the vaccine effect fades , vaccinees that have avoided natural immunizing infections soon experience infections that have already happened in the baseline , resulting in excess cases in later years ., However , cumulative effectiveness shows net case reduction–i . e . some of the cases are actually eliminated rather than just delayed ( see Section S7 in S1 Text ) ., Booster vaccination prevented negative annual effectiveness ( lighter solid lines ) , but overall performance was worse than the catch-up scenario with a durable vaccine ( dark solid line ) ., The rate at which immunity wanes was an important factor for strategies without booster vaccination , but not for those with it ., We fit an agent-based dengue transmission model to empirical data from Yucatán , Mexico , and then used this model to evaluate a range of vaccination scenarios ., For our effectiveness analysis , we used efficacy values based on phase III clinical trial results for Dengvaxia 11–13 ., We concluded that a Dengvaxia-like vaccine can be an effective tool for reducing the dengue burden , although a vaccine with waning efficacy would require a booster program ., We estimated a cumulative reduction in cases of 74% over 20 years for the most favorable scenario ( Table 2 ) , and scenarios with a durable vaccine converged near 65% reduction in annual case burden after 20 years ( Fig 5 ) ., Scenarios with waning vaccines required booster vaccination programs to maintain appreciable effectiveness; however , with boosting , they converge at around 50% annual effectiveness by the end of the forecast ( Fig 6 ) ., In general , vaccination strategies that include only routine vaccination at a particular age are much less effective in the first 10 years than those with a one-time catch-up ( Fig 5 ) ., However , as the initial catch-up cohort shrinks as a share of all vaccinees ( due to mortality and on-going routine vaccination ) , we expect the annual effectiveness of routine-only and catch-up strategies to converge ., For a durable vaccine , our model forecasts that effectiveness will converge around 65% after roughly 20 years ., The results for a waning vaccine are more complicated , particularly when there is a catch-up campaign ., To address concerns raised in a recently published analysis of long-term follow-up data from the phase III trials 14 , we considered vaccines that provide protective immunity that wanes with a half-life of 2 , 5 , or 10 years ., In that study , researchers found no significant reduction in dengue hospitalization risk for vaccinated versus control groups during the third year post-vaccination , suggesting that vaccine-induced protective immunity may begin to wane ., Without other adjustments to deployment strategies , we found that vaccination in these waning scenarios provides minimal long-term benefit ., Furthermore , if there was an initial catch-up campaign , some years have increased incidence relative to no vaccination , though there are still small cumulative benefits ( see Section S7 in S1 Text ) ., For these scenarios , the vaccine initially prevents large epidemics , leading to a decline in naturally acquired immunity compared to baseline scenarios ., When the relatively large cohort of catch-up vaccinees then collectively loses its vaccine-induced immunity over a short period of time , larger-than-baseline epidemics can result , which leads to years with expected negative annual effectiveness ., However , adding booster doses to the vaccination strategy can substantially offset waning , and results in annual effectiveness around 50% at the end of the forecast period ., As expected , vaccines with longer half-lives produce better effectiveness , but all waning vaccines had low long-term effectiveness without a booster program ., As a supplementary analysis , we also considered the effect of projected temperature increase associated with climate change on these results ( see Section S6 in S1 Text ) ., This sensitivity study suggests that increasing temperature would increase the projected dengue burden , but that estimates of annual vaccination effectiveness are robust to the increasing force of infection ., For other public health considerations , such as adherence to dose schedules and compliance with booster campaigns , we anticipate that changing these factors would have obvious directional effects ( e . g . lower coverage will lead to lower effectiveness ) , but there are not sufficient data at this time to make meaningful quantitative predictions ., While compliance rates to inform such analyses might reasonably be inferred from other vaccine programs , the most critical issue is unclear: the appropriate model of vaccine action ., Until there are appropriate data on Dengvaxia performance , attempts to quantify the nuanced effects of vaccine delivery are premature ., For all of our scenarios , we assumed that the vaccine efficacy for individuals who have not had a natural infection ( i . e . , antibody-naïve ) is half of that for those who have had one ( i . e . , antibody-primed; see Table 1 ) ., Thus , as the vaccine drives down natural infection rates , it will become less effective , lowering the long-term benefit ., Previous analyses have suggested that interventions ( either vector control 47 or vaccination 48 ) in a population that historically experienced high force of infection would initially look effective , but then have declining benefit ., We observed this effect for the scenarios with a catch-up cohort: the initially high effectiveness declines after about 10 years to what appears to be a new steady state that reflects both routine vaccination coverage and a reduced level of natural infections associated with reduced transmission ( Fig 5B , solid lines ) ., In addition to long-term outcomes , relative vaccine efficacy also influences the effectiveness of vaccination strategies based on the target age for routine vaccination: older vaccinees are more likely to be antibody-primed ., This results in higher effectiveness for these strategies , but the effect is modest in the modeled Yucatán population ., Therefore , other considerations such as distribution logistics might reasonably take precedence when choosing which age group to vaccinate ., In general , our fitting procedure reproduces several features of the observed data well ( Table F in S1 Text ) , but the model is not fit to and is not intended to replicate the exact historical time series ., Dengue epidemics in Yucatán are highly variable , undoubtedly influenced by factors we do not consider ( e . g . the circulating serotypes in adjacent regions , inter-annual environmental variation ) ., We predict the approximate timing of peaks in 1980 , 1984 , and 1997 due to the introduction of serotypes that had not recently circulated , but we do not predict the large epidemics observed near the end of the fitting period ( Fig 3 ) ., As a consequence , we under-predict seroprevalence in young people ( Fig 4 ) , and thus may be under-predicting short-term performance of the vaccine ., These recent large epidemics are unlikely to be driven by gradual trends , such as might be captured by improved data on natural history of dengue generally , mosquito ecology in the region , and demographic and economic trends in the region ., Large epidemics after quiet years are historically associated with the introduction of novel serotypes ., Thus , a more substantial change to the model , such as introducing a novel strain of DENV2 capable of re-infecting people with past DENV2 infections as suggested in 49 , may be necessary to replicate the end of the fitted period and the age distribution of seroprevalence ., Despite the model’s inability to reproduce the most recent large epidemics , we believe it is informative for forecasting vaccine performance for two reasons ., First , even though the vaccine has twice the efficacy for seropositive versus seronegative recipients , average efficacy is relatively insensitive to changes in seroprevalence: e . g . if 50% of vaccinees are seropositive , overall efficacy to DENV1 is 0 . 45 , while reducing seroprevalence by half to 25% only reduces average efficacy to 0 . 38 , a ∼15% reduction ., Second , the increased efficacy in seropositive vaccinees produces a stabilizing effect: if epidemics become large , the vaccine performs better thus driving epidemics smaller , while if epidemics have been small , overall efficacy decreases , permitting larger epidemics ., These offsetting effects make population-level effectiveness relatively robust ., Our ability to forecast vaccination impact is primarily limited by the current uncertainty regarding whether and how vaccine efficacy wanes over time and how vaccine efficacy is affected by prior infection ., Nevertheless , our model provides a useful perspective on how vaccine properties and strategic choices affect the relative size and severity of | Introduction, Methods, Results, Discussion | Dengue vaccines will soon provide a new tool for reducing dengue disease , but the effectiveness of widespread vaccination campaigns has not yet been determined ., We developed an agent-based dengue model representing movement of and transmission dynamics among people and mosquitoes in Yucatán , Mexico , and simulated various vaccine scenarios to evaluate effectiveness under those conditions ., This model includes detailed spatial representation of the Yucatán population , including the location and movement of 1 . 8 million people between 375 , 000 households and 100 , 000 workplaces and schools ., Where possible , we designed the model to use data sources with international coverage , to simplify re-parameterization for other regions ., The simulation and analysis integrate 35 years of mild and severe case data ( including dengue serotype when available ) , results of a seroprevalence survey , satellite imagery , and climatological , census , and economic data ., To fit model parameters that are not directly informed by available data , such as disease reporting rates and dengue transmission parameters , we developed a parameter estimation toolkit called AbcSmc , which we have made publicly available ., After fitting the simulation model to dengue case data , we forecasted transmission and assessed the relative effectiveness of several vaccination strategies over a 20 year period ., Vaccine efficacy is based on phase III trial results for the Sanofi-Pasteur vaccine , Dengvaxia ., We consider routine vaccination of 2 , 9 , or 16 year-olds , with and without a one-time catch-up campaign to age 30 ., Because the durability of Dengvaxia is not yet established , we consider hypothetical vaccines that confer either durable or waning immunity , and we evaluate the use of booster doses to counter waning ., We find that plausible vaccination scenarios with a durable vaccine reduce annual dengue incidence by as much as 80% within five years ., However , if vaccine efficacy wanes after administration , we find that there can be years with larger epidemics than would occur without any vaccination , and that vaccine booster doses are necessary to prevent this outcome . | Dengue is a mosquito-transmitted viral disease that is common throughout the tropics ., Despite a long history in humans and extensive efforts to control dengue transmission in many countries , the number , severity , and geographic range of reported cases is increasing ., Most control efforts have focused on controlling mosquito populations , but the main vector , Aedes aegypti , flourishes in human-disturbed and indoor environments ., Because the mosquitoes prefer to bite during the day when people are active and potentially moving around high-risk locations , fixed barriers like bed nets are not effective ., Several dengue vaccines are being actively developed and may become valuable tools in dengue control ., Using historical dengue data from Yucatán , Mexico , we fit a detailed simulation of human and mosquito populations to project future transmission , then used efficacy data from vaccine trials to evaluate the benefit of potential vaccination deployment strategies in the region ., For a durable vaccine , we find that population-level , annual vaccine effectiveness approaches 65% by the end of the 20-year forecast period ., For waning vaccines , however , effectiveness is greatly reduced–and sometimes negative–unless booster vaccinations are used . | invertebrates, medicine and health sciences, infectious disease epidemiology, immunology, geographical locations, animals, vaccines, preventive medicine, north america, infectious disease control, vaccination and immunization, insect vectors, public and occupational health, infectious diseases, booster doses, epidemiology, disease vectors, insects, arthropoda, people and places, mosquitoes, mexico, immunity, biology and life sciences, organisms | null |
journal.pntd.0003404 | 2,015 | Genome and Phylogenetic Analyses of Trypanosoma evansi Reveal Extensive Similarity to T. brucei and Multiple Independent Origins for Dyskinetoplasty | Trypanosomatid parasites Trypanosoma evansi and T . equiperdum are responsible for animal diseases with extensive pathological and economic impact and closely related to the T . brucei group 1 , 2 ., The latter includes three subspecies: the human parasite T . b ., rhodesiense , the zoonotic parasite T . b ., gambiense , and the animal parasite T . b ., brucei ., Together T . brucei , T . evansi , and T . equiperdum comprise the subgenus Trypanozoon ., The exact nature of the phylogenetic relationship between these three species has been the subject of ongoing debate , with some evidence suggesting that T . evansi and T . equiperdum are monophyletic and other evidence suggesting that they are polyphyletic and have emerged multiple times from T . b ., brucei 3–6 ., Trypanosomatids are a family within the protist group Kinetoplastida , the eponymous feature of which is a large and complex network of circular DNAs ( kinetoplast or kDNA ) inside their single mitochondrion ., Two key biological features distinguish T . evansi and T . equiperdum from the T . brucei group ., Firstly , their transmission is independent from the tsetse fly as obligatory vector ., T . evansi is predominantly transmitted by biting flies and causes surra in a wide variety of mammalian species ( the name of the disease varies with geographical area ) , while T . equiperdum causes a sexually transmitted disease called dourine in horses 1 , 3 , 7 ., The altered mode of transmission has enabled both parasites to escape from the sub-Saharan tsetse belt and become the pathogenic trypanosomes with the widest geographical distribution ., Secondly , all strains of T . evansi and T . equiperdum investigated so far are dyskinetoplastic , i . e . , lacking all ( akinetoplastic ) or critical parts of their kDNA 8 ., The loss of kDNA is thought to lock T . evansi and T . equiperdum in the bloodstream life cycle stage , presumably because the absence of kDNA-encoded components of the oxidative phosphorylation system prevents ATP generation in the tsetse midgut 9 ., Nonetheless , whether dyskinetoplasty preceded the switch to tsetse-independent transmission or vice versa is unresolved 8 , 10 , 11 ., The kDNA that comprises the mitochondrial genome in T . brucei consists of numerous concatenated circles of two kinds: maxicircles that encode genes primarily involved in oxidative phosphorylation and minicircles that encode guide RNAs ( gRNAs ) 12 ., The majority of maxicircle mRNAs undergo RNA editing to insert or delete uridine residues as specified by template gRNAs in a process catalyzed by multiprotein complexes called editosomes 13–15 ., One kDNA-encoded transcript that requires editing is the F1FO-ATPase subunit 6 , which is essential in both bloodstream and insect stage T . brucei 16–19 ., However , it has recently been shown that mutations found in the nuclear-encoded ATPase subunit γ of some T . evansi and T . equiperdum strains can compensate for the loss of kDNA , explaining their viability 20 ., In an effort to better understand the causes and consequences of tsetse-independent transmission and kDNA-independent viability , we sequenced the genome of T . evansi strain STIB805 ., This strain was isolated in 1985 from an infected water buffalo in the Jiangsu province of China , shown to completely lack kDNA ( i . e . to be akinetoplastic ) , and suggested to belong to a possibly clonal group of T . evansi with worldwide distribution 4 , 21 , which is why it was chosen for this study ., The comparative genome analysis between this strain and the T . b ., brucei TREU 927/4 strain reference genome 22 revealed extensive similarities ., While the sizes of the chromosomes differ between T . evansi and T . brucei , the gene content within their respective genomes are largely similar , as 92 . 7% of T . evansi CDS have an identifiable ortholog in T . brucei ., Analysis of T . evansi variant surface glycoprotein ( VSG ) sequences shows broad conservation of N-terminal sub-types , with extensive phylogenetic similarity and no evidence of any species-specific expansion of clades ., An analysis of T . evansi CDS corresponding to the identified T . brucei mitochondrial proteome revealed that virtually all are retained , despite the lack of requirement in an akinetoplastic trypanosome for respiratory complexes I-IV or any proteins involved in maintenance or expression of the mitochondrial genome ., Phylogenetic analyses with several genetic markers conclusively show that extant strains of T . evansi or T . equiperdum are not monophyletic and evolved on at least four independent occasions ., Together , the results presented here show few critical differences between T . evansi and T brucei , indicating that dyskinetoplasty and concomitant tsetse-independent transmission are significant phenotypic changes underpinned by relatively subtle genomic alterations ., The rearing of animals was regulated by Czech legislation ( Act No 246/1992 Coll . ) ., All housing , feeding and experimental procedures were conducted under protocol 90/2013 approved by Biology Centre , Czech Academy of Sciences and Central Commission for Animal Welfare of the Czech Republic ., Trypanosomes ( T . evansi strain STIB805 ) were purified from mice by DEAE ( DE52 ) cellulose 23 ., Total DNA was extracted as described elsewhere 24 ., Briefly , the cells were lysed using SDS , and incubated with proteinase K and RNase ., DNA was harvested after phenol extraction and ethanol precipitation ., Four runs of single-end 454 sequencing plus 2 runs of paired-end 454 sequencing were obtained using GS FLX ( + ) System following the manufacturers instruction ( Roche ) and generated 1 , 904 , 327 reads ( 225 , 826 paired end , 1 , 678 , 501 single end ) 25 ., Approximately 10 µg of genomic DNA was sheared by nebulization into desired fragments sizes ( ∼400 bp for single-end 454 , ∼3 kb for paired-end ) and adaptor oligos ligated to create the library for sequencing ., Additional sequence data was obtained by shearing genomic DNA to ∼200–300 bp fragments sizes for sequencing on an Illumina GAIIx producing 19 , 701 , 740 tags with an ordered read length of 76mers ., Illumina reads for T . b ., brucei strains TREU 927/4 and Lister 427 ( provided by the Wellcome Trust Sanger Institute , Hinxton , UK ) were downloaded from the European Nucleotide Archive ( accession nos . ERX009953 and ERX008998 ) ., Genome assemblies and identification of sequence polymorphisms ( SNPs and indels ) were carried out with CLC Genomics Workbench ( CLC bio ) ., Reads for T . evansi STIB805 , T . b ., brucei TREU 927/4 and T . b ., brucei Lister 427 were mapped against the T . brucei TREU 927/4 version 4 reference ( Tb927 ) using the following mapping parameters: global alignment , similarity fraction =\u200a0 . 9 , length fraction =\u200a0 . 5 , insertion cost =\u200a3 , deletion cost =\u200a3 , mismatch cost =\u200a2 ., De novo assemblies for each strain were created using either all reads or those binned during the reference-based assemblies , using the following parameters: automatic word size = yes , bubble size =\u200a50 , similarity fraction =\u200a0 . 9 , length fraction =\u200a0 . 5 , deletion cost =\u200a3 , insertion cost =\u200a3 , mismatch cost =\u200a2 ., De novo contigs were aligned to Tb927 using standalone NCBI BLAST version 2 . 2 . 25 and Artemis Comparison Tool release 12 . 0 26 ., For RPKM ( reads per kilo base per million ) analysis , aligned STIB805 and TREU 927/4 reads were assigned to sequential 1-kb bins along the length of the Tb927 reference ., For each chromosome , the log2 ratio of binned reads from the two read-sets was calculated for each bin , and normalized to a median log2 ratio of 0 by offsetting all values by the median log2 ratio ., SNPs and indels were called with CLC bio using the following parameters: minimum coverage =\u200a10 , maximum coverage =\u200a100 , minimum variant frequency =\u200a30% , minimum central quality =\u200a20 , minimum average quality =\u200a15 ., Coding sequence prediction of T . evansi genome was done using a combination of de novo gene prediction approach and reference based gene transfers ., The Rapid Annotation Transfer Tool ( RATT ) was used to transfer the gene boundaries and functional annotations from Tb927 onto T . evansi chromosomes ( target hereafter ) 27 ., To prevent the transfer of paralogs to the same target region , thus resulting in multiple overlapping/duplicate gene calls , the annotation transfer was performed pairwise between one chromosome from Tb927 and the corresponding chromosome from the target ., This resulted in the transfer of more than 96 . 4% ( 9427/9776 ) of genes from Tb927_v4 onto TevSTIB805ra ., Though we observed extensive conservation of synteny between Tb927 and target genomes , this approach would likely have missed genes in targets that were shuffled by chromosomal fission and fusion that occurred during evolution ., We performed another RATT transfer considering all Tb927 chromosomes and one target chromosome at a time ., We then parsed out genes that were predicted uniquely by this approach ( i . e . from regions where the initial RATT transfer failed to predict any genes ) ., Combining these two approaches allowed us to predict genes that were shuffled across the chromosomes while avoiding predictions that overlap each other ., We then used an in-house consensual de novo gene prediction suite called AutoMagi to predict protein coding genes from T . evansi 28 ., AutoMagi internally uses three gene prediction algorithms ( genescan , testcode , codonusage ) and predicts a consensus gene model out of individual gene predictions ., The codon usage table required by AutoMagi was generated by ‘cusp’ using the RATT output of T . evansi STIB805 assembly 29 ., We then used an in-house built prolog system to combine the de novo gene predictions and reference-based gene predictions ., The genes that are unique to RATT predictions were automatically included in the final set ., Genes that are unique to AutoMagi were compared with the NCBI non-redundant ( NR ) database ( accessed 23 February 2011 ) using BLASTp ( default criteria: EXPECT\u200a=\u200a−10 , WORD SIZE\u200a=\u200a−3 , MATRIX_NAME\u200a= BLOSUM62 , GAP COST\u200a= Existence:11 Extension:1 ) algorithm ., All of the BLASTp results were reviewed manually , and genes meeting the following two criteria were retained:, ( a ) an E-value of 5e-6 or lower to the matched NR sequence;, ( b ) coding strand identical to nearest neighbors on either side ( i . e . on same poly-cistronic unit ) ., This manual curation removed 165 putative CDS from making into the finalized TevSTIB8805ra ., The gene calls that overlapped between RATT and AutoMagi were subdivided into the following 4 groups ., 1 ) Identical: overlapping exactly , 2 ) AM_subsetof_RATT: AutoMagi prediction is entirely contained within RATT prediction , 3 ) RATT_subsetof_AM: RATT prediction is entirely contained within AutoMagi prediction 4 ) StaggeredOverlap: both predictions overlap in a staggered fashion ., In the first two cases ( Identical & AM_subsetof_RATT ) , AutoMagi gene calls were ignored and RATT models were retained ., In the third case ( RATT_subsetof_AM ) coordinates from AutoMagi were combined with annotation information from the RATT model ., Genes in the fourth category ( Staggered ) were subjected to a thorough manual review process ., The review process involved, ( i ) BLASTp search against NCBIs NR database to identify coordinates for queries and subject;, ( ii ) a ClustalW multiple sequence alignment of both candidate genes with their potential homolog ( s ) from T . brucei ., Manual review of BLAST and ClustalW was performed in each case to decide either to split/merge/choose one of the AutoMagi/RATT predictions ., These newly derived coordinates were then combined with the annotation information from the RATT model ., GeneIDs of the final set of protein coding genes were unified and ordered from left to right end of the chromosome ., Entire genome and all the predicted genes are publicly available at TriTrypDB ( http://tritrypdb . org/tritrypdb/ ) ., The fastq files containing the Illumina read data for T . evansi STIB805 are available at: http://www . ebi . ac . uk/ena/data/view/ERA000101 ., Bloodstream parasites at 2×108 cells/ml were purified using DEAE cellulose ( DE52 ) chromatography , and subsequently used to prepare chromosome blocks as previous described 30 ., DNA from T . b ., evansi and T . b ., brucei cells was embedded in low-melt agarose blocks ( final concentration of 5×107 cells/ml ) according to 31 , and was resolved using a 1% Megabase agarose ( Bio-Rad ) gel with 0 . 5X TBE buffer in the CHEF-DRIII system ( Bio-Rad ) ., S . cerevisiae DNA was used as a size marker ., Pulse field gel electrophoresis ( PFGE ) was run at 14°C under the following conditions: switch time A increased from 28 . 6 s to 228 s for 24 hrs , followed by switch time B , with increase from 28 . 6 s to 1 , 000 s for another 24 hrs ., The angle was set to 120° and voltage gradient to 3 V/cm ., The PFGE gel was stained with ethidium bromide after the run ., VSG sequences were extracted from the genome assembly using hidden markov models ( HMM ) and HMMER 3 . 0 32 ., HMMs were constructed for a- and b-type VSG respectively using multiple sequence alignments of T . brucei TREU 927/4 and T . b ., gambiense DAL972 sequences 33 ., All open reading frames>100 bp were marked up and the predicted amino acid sequences were searched for matches to either HMM using HMMER 3 . 0 ., Significant matches were checked manually to ensure that each VSG was complete and gene boundaries were correct ., Intact VSG were extracted and aligned approximately using Clustal X 34 ., The aligned sequences were combined with existing alignments of T . brucei TREU 927/4 a- and b-type VSG 35 and modified by eye ., C-terminal domains were trimmed ( due to recombination these present an inconsistent phylogenetic signal ) , resulting in a- and b-VSG alignments of 470 and 492 characters , respectively ., Neighbour-joining trees were estimated for each amino acid sequence alignment using PHYLIP v3 . 6 , with a JTT rate matrix and 100 non-parametric bootstrap replicates ., A frequency distribution of species-specific clade sizes , ( i . e . three T . evansi VSG clustered together with a T . brucei TREU 927/4 sequence as the sister lineage has a clade size of 3 ) was calculated to express the degree of intercalation of sequences from the two strains ., Putative VSG orthologs were extracted from the phylogenies ., In situations where single VSG from T . evansi STIB805 and T . brucei TREU 927/4 were sister taxa , supported by a bootstrap value>95 , these genes were interpreted as orthologs ., The rates of synonymous and non-synonymous nucleotide substitutions per site were estimated for these orthologous pairs ., The ratio of these rates ( ω ) was estimated using Ka/Ks Calculator v1 . 2 36 using GY and MS methods ., For comparison , this was repeated for 151 pairs of non-VSG orthologs chosen at random ., The dihydrolipoamide dehydrogenase ( LipDH ) CDS ( Tb927 . 11 . 16730 ) was PCR-amplified from total parasite DNA using primers 5′-ATA AAG CTT ATG TTC CGT CGC TGC-3′ ( forward ) and 5′-ATA AGA TCT TTA GAA GTT GAT TGT TTT GG-3′ ( reverse ) and Phusion polymerase ( New England Biolabs , NEB ) ., In cases where direct sequencing of the amplicon revealed heterozygosity , sequence information for individual alleles was obtained after cloning ., After removal of the primer sequences , LipDH CDSs were aligned using ClustalX 37 ., Phylogenies were reconstructed using the program MrBayes 38 via the Topali platform 39 , implementing the GTR substitution model and a discrete gamma rate distribution model with four rate categories ( to account for rate heterogeneity among sites ) as the most appropriate nucleotide substitution model ., Four independent MCMC chains of 1×106 generations were sampled every 100th generation ., A 50% majority rule consensus tree was derived after the first 25% of trees were discarded as burn-in ., The ATP synthase γ subunit sequence ( Tb927 . 10 . 180 ) was PCR-amplified from total parasite DNA with primers 5′-GCG GAA TTC GAA GCA GAT GAC ACC TAA-3′ ( forward ) and 5′-GCG GAA GAC CTT GCT GCG GAG CCA CTC T-3′ or 5′-GGC GAC ATT CAA CTT CAT-3′ ( reverse ) and Phusion polymerase ( NEB ) ., The sequence was determined by direct amplicon sequencing or , in cases of heterozygosity , after cloning ., A partial 812-bp sequence of cytochrome oxidase 1 ( COX1 ) was obtained by PCR and used for phylogenetic analysis as previously described 40 ., Briefly , we assessed phylogenetic relationships among T . equiperdum and T . brucei isolates using a haplotype network constructed using the statistical parsimony approach implemented in TCS v . 1 . 21 41 ., Subnetworks were created with 95% confidence limit and then unconnected subnetworks>10 mutations apart were connected by relaxing the confidence limit ., To verify that the haplotypes containing T . equiperdum isolates were on phylogenetically distinct branches , we estimated a phylogenetic tree using the Bayesian approach implemented in MrBayes v . 3 . 2 42 ., PartitionFinder v . 1 . 0 . 1 43 determined the Hasegawa , Kishino and Yano nucleotide substitution 44 with invariant sites ( HKY+I ) without partitioning by codon positions as the most appropriate model for the MrBayes analysis ., Microsatellite genotyping was carried out exactly as described previously 40 ., Briefly , isolates were typed for eight microsatellite markers 45 ., Principal component analysis ( PCA ) was performed in R using the package adegenet , as described 40 ., PFGE was performed to visualize the pattern of chromosomes in the akinetoplastic T . evansi STIB805 strain ., Multiple bands corresponding to megabase chromosomes ranging in size from ∼1 to ∼5 Mb could be visualized ( Fig . 1 ) ., The most noticeable differences in the chromosomal pattern in comparison to T . brucei are the five intermediate chromosome bands that range from ∼300 kb to ∼800 kb ., Additionally , two size groups of minichromosomes ( ∼100 kb and ∼200 kb ) were observed in T . evansi STIB805 ., Such a degree of variability is within the range observed among strains of T . b ., brucei 46 ., Genomic DNA isolated from T . evansi STIB805 was subjected to both 454 and Illumina sequencing , generating a combined total of 21 , 445 , 221 reads ., Using the T . brucei TREU 927/4 genome sequence 22 ( version 4; from here onwards abbreviated Tb927 for convenience ) as a scaffold , a reference-based assembly of these T . evansi reads was generated ., This assembly , called TevSTIB805ra , incorporates 17 , 856 , 165 reads and has an average coverage of 57 . 2 reads per nucleotide across the entire assembly and of 48 . 4 reads per nucleotide for the ‘core regions’ i . e . regions excluding telomeres , subtelomeres , and internal regions that consist of repetitive coding sequences such as VSGs , expression site-associated genes ( ESAGs ) or retroposon hot spot genes ( RHS ) ., An assembly approach based on the genome of a closely related reference strain allows the reliable identification of homologous gene pairs , of any differences that might exist between these sequences , and of genes that are missing or highly diverged in the genome of interest ., This approach has limited power in identifying structural differences between genomes and , by definition , cannot identify sequences that are present in the genome of interest but not in the reference genome ., For that reason we have supplemented the reference-based approach with the analysis of contigs that were assemble de novo ., As detailed below , this allowed the confirmation of Tb927 genes that are absent in in T . evansi STIB805 and the identification of candidate genes that might be present in the latter but absent in the former ., However , genome rearrangements that are not associated with differences in gene content will have been missed and the T . evansi STIB805 genome as published on TriTrypDB may indicate gene synteny where it does not exist ., Annotation of the TevSTIB805ra T . evansi genome was performed using a combination of RATT and AutoMagi , followed by manual curation ., RATT ( Rapid Annotation Transfer Tool; 27 ) identified likely orthologs in the T . brucei reference genome and transferred their functional annotations to the T . evansi genome , while AutoMagi predicted genes de novo using the consensus of three gene prediction algorithms ( genescan , testcode , codonusage ) 28 ., A total of 10 , 110 CDS were identified , with 9368 CDS annotated as T . brucei orthologs by RATT and subsequent manual inspection ( Table 1 , columns C , E and F; S1 Table ) , and 742 CDS uniquely predicted by AutoMagi ( Table 1 , column D; S2 Table ) ., Thus , 92 . 7% of the identified T . evansi CDS have an identified ortholog in T . brucei , while 7 . 3% were uniquely detected by de novo gene prediction ., Analysis of the 742 AutoMagi CDS by BLAST searching of the GenBank non-redundant database revealed at least one hit ( E-value ≤5e-6 ) for each of these CDS to a gene from a Trypanosoma species ( T . b . gambiense 568 CDS; T . b . brucei 168 CDS; T . equiperdum , T . vivax , or T . congolense 6 CDS ) ., The most common annotations among these BLAST hits were for ‘hypothetical unlikely’ ( 415 CDS ) or other hypothetical ( 150 CDS ) sequences ( S2 Table ) ., Orthologous sequences had not been annotated in the Tb927 reference , presumably due to more stringent criteria for gene calling ., Thus , most if not all of the de novo predicted genes in TevSTIB805ra are also present in T . brucei ., Of the 9368 T . evansi CDS identified as T . brucei orthologs , 8421 CDS were classified as non-repetitive genes; the remaining 947 CDS were VSG , ESAG , RHS , or duplicate sequences ., A comparison of the 8421 non-repetitive CDS between T . brucei and T . evansi revealed 7970 ( 94 . 6% ) had a nucleotide identity of 95% or greater , 320 ( 3 . 8% ) had a nucleotide identity between 70–95% , and 131 ( 1 . 6% ) had a nucleotide identity less than 70% ( Fig . 2; S1 Fig . ) ., After RATT annotation transfer , a total of 503 ( 5 . 1% ) Tb927 GeneIDs did not have identified orthologs in the TevSTIB805ra annotated genome ( Table 1 , column G; S2 Table ) ., The majority of these ( 406 ) represent repetitive genes ( e . g . VSG , ESAG , and RHS ) or ‘hypothetical unlikely’ genes , which were not analyzed further ., Five CDS correspond to predicted pseudogenes ., T . evansi STIB805 orthologs for 49 Tb927 CDS were shown to be wholly or partially missing in TevSTIB805 reads that mapped to these gene loci ( see below ) ., T . evansi homologs for the remaining 43 Tb927 GeneIDs were ‘missed’ by RATT and AutoMagi , but were subsequently identified by manual examination of the T . evansi sequence ., Thus , the majority of T . brucei CDS have extremely similar T . evansi orthologs , and very few T . brucei CDS were not found in T . evansi ., This result is consistent with a very close phylogenetic relationship between these parasites ., The initial RATT approach only identified T . evansi CDS with an annotated Tb927 homolog in the syntenic position ., To identify potential CDS in TevSTIB805ra that are syntenic to unannotated sequences in Tb927 , and are homologous to annotated CDS in other ( non-syntenic ) locations , RATT was performed a second time by comparing each TevSTIB805ra chromosome to the entire Tb927 genome ., This approach complemented de novo gene calling by AutoMAGI and identified 112 CDS in TevSTIB805ra ( Table 1 , columns E and F; S2 Table ) ., The majority of these CDS were annotated as hypothetical proteins ( 64 CDS ) or repetitive sequences such as VSG , ESAG , and RHS ( 39 CDS ) ., For almost all loci , manual inspection revealed that either, ( i ) a syntenic CDS existed in Tb927 that was unannotated or, ( ii ) a syntenic , annotated CDS did exist in Tb927 but the presence of a highly similar sequence elsewhere caused a RATT artefact ., An exception was TevSTIB805 . 6 . 770 , where the syntenic sequence in Tb927 was found to be disrupted by frame-shifts , but in T . b ., gambiense DAL972 to be annotated as Tbg972 . 6 . 420 ( hypothetical protein , unlikely ) Two approaches were used to find genes in the T . evansi genome that may not be present in the T . brucei genome ., Firstly , reads from T . evansi that did not match Tb927 were de novo assembled to make TevSTIB805dn and putative CDS called with AutoMagi ( Table 2 and S2 Table ) ., The 22 CDS with a homolog in Tb927 that was not a repetitive or hypothetical gene included multiple copies of pseudogenes for putative UDP-Gal/UDP-GlcNAc-dependent glycosyltransferase and lytic factor resistance-like protein , an adenosine transporter , calmodulin , a DNA topoisomerase , a leucine-rich repeat protein , ribulose-phosphate 3-epimerase , major surface protease B , and a galactokinase pseudogene ., 36 of the 202 CDS that did not get a hit against Tb927 had a BLASTx hit in the NCBI non-redundant database: 23 CDS matched hypothetical Trypanosoma spp ., and 9 appeared to be contaminants from other - mostly bacterial - organisms ., The remaining 4 BLAST hits were for T . vivax RNA-dependent DNA polymerase and T . evansi VSGs ( 2 hits ) and diminazene-resistance-associated protein ., Because Tb927 was created using a traditional sequencing approach and only covers the 11 large chromosomes 22 , sequences found in the de novo assembly of T . evansi deep sequencing reads might not be unique to this strain or species , but rather reflect a difference in sequencing methodology or stem from intermediate-sized or mini-chromosomes ., To address this , Illumina reads from sequencing T . brucei TREU 927/4 that also did not match Tb927 were de novo assembled into Tb927dn ., Comparison of Tb927dn to TevSTIB805dn showed that 45 . 6% of the binned T . evansi reads matched to Tb927dn ., The remaining T . evansi reads were then assembled into TevSTIB805dn_sub and putative CDS called with AutoMagi ( Table 2 and S2 Table ) ., Of the 84 BLASTp matches to non-repetitive genes , 71 are annotated as hypothetical proteins , 2 are lytic factor resistance-like proteins , 1 is a putative transporter , and 10 are annotated as putative ( or pseudogene ) UDP-Gal or UDP-GlcNAc-dependent glycosyltransferase ., Glycosyltransferases in T . brucei are frequently found in subtelomeric regions that are difficult to assemble due to repetitive sequences , and are known to be highly variable among trypanosome strains 33 , 47 ., Of the 117 TevSTIB805dn_sub CDS for which no match was identified in Tb927 , 18 had a BLASTx hit in the NCBI non-redundant database: 9 CDS matched hypothetical genes from various Trypanosoma species and 1 matched a diminazene resistance-associated protein that was previously suggested to convey diminazene aceturate ( Berenil ) resistance to certain strains of T . evansi 48 ., Thus , gene prediction found very few CDS in de novo T . evansi assemblies that are not present in T . brucei , and analysis of these CDS revealed differences reminiscent of those observed among strains of the same species ., When T . evansi STIB805 de novo contigs were filtered for minimum coverage of at least 5x , the number of genes without obvious homologs in Tb927 was reduced considerably ( Table 2 ) ., To test the potential absence of these genes in T . b ., brucei more rigorously , short read data sets for strains TREU 927/4 and Lister 427 were searched for matches to the 22 candidates from TevSTIB805dn and the 6 candidates from TevSTIB805dn_sub ., Only seven CDS candidates remained where the sequence was either entirely absent from both T . b ., brucei datasets , or did not contain an undisrupted ORF ( S2 Table , highlighted ) ., Whether these candidates are indeed functional CDS , and whether they are generally absent from T . brucei ssp ., and present in T . evansi , and therefore are of potentially diagnostic value , requires further investigation ., A non-redundant list of the ORFs that did not have a BLASTx hit in NCBI ( 200 ORFs in total ) is provided as S1 Data File ., A total of 49 CDS found in Tb927 were not identified in TevSTIB805ra ., These loci were analyzed individually by, ( i ) specifically searching for matching T . evansi STIB805 reads ( S2 Table ) ;, ( ii ) comparing reads per kilobase per million ( RPKM ) coverage plots of these regions for T . brucei and T . evansi;, ( iii ) aligning contigs of a full T . evansi STIB805 de novo assembly to the respective regions in Tb927 ., These analyses confirmed the absence or disruption of several of these CDS in T . evansi STIB805 ( including the iron/ascorbate oxidoreductase loci , the Tb927 . 9 . 7950/Tb927 . 9 . 7960 repeat , the Tb927 . 4 . 3200- . 3270 region , both Tb927 . 8 . 490 and Tb927 . 8 . 500 , and the Tb927 . 8 . 7300- . 7330 region ) , as illustrated in S2–S9 Fig ., , with the procyclin loci described in detail below ., These cases have in common that the loci in question show considerable variation among Trypanozoon strains and CDS appear to be absent in T . b ., brucei Lister 427 and/or T . b ., gambiense DAL972 as well 33 ., The differences observed in T . evansi STIB805 compared to Tb927 therefore are unlikely to be relevant for kDNA loss or tsetse-independent transmission ., In T . evansi STIB805 , the procyclin loci either lack or have disrupted versions of several CDS found in T . brucei ., GPEET and EP procyclins and associated genes are encoded in loci on chromosomes 6 and 10 , respectively , in various T . brucei strains 49 ., Procyclin proteins are GPI-anchored coat glycoproteins that are expressed exclusively in the procyclic insect form of T . brucei , and they have been hypothesized to be involved in protection against tsetse fly midgut hydrolases 50 ., Experiments in T . brucei have shown that knocking out all of the procyclin genes ( Null mutants of GPEET and EP3 on chromosome 6; EP1 and EP2 on chromosome 10 ) causes no growth defect in vitro and permits completion of the entire life cycle , but causes a selective disadvantage during co-infection with wild type cells in the tsetse fly midgut 51 ., These procyclin loci also contain procyclin-associated genes and a gene related to expression site associated gene 2 ( PAG3 and GRESAG2 on chromosome 6; PAG1 , 2 , 2* , 4 , and 5 on chromosome 10 ) 49 ., The functions of the PAG proteins and GRESAG2 are unknown; although transcripts of PAG1–3 have been shown to increase during differentiation to procyclic forms , published experiments using cell lines with all PAG genes knocked out reported no obvious abnormal phenotypes in vitro or in vivo 49 ., In multiple T . brucei strains , the chromosome 10 procyclin loci are heterozygous , with one chromosome containing EP1/EP2/PAG1/PAG5/PAG2*/PAG4 and the other chromosome containing EP1/EP2/PAG2/PAG4 ( PAG2 being a fusion of the 5′ part of PAG1 and the 3′ part of PAG2* ) 49 , 52–56 ., In T . evansi STIB805 , chromosome 10 appears to be homozygous , with only the EP1/EP2/PAG2/PAG4 locus present , and the associated absence of the segment containing the 3′ part of PAG1 , PAG5 and the 5′ part of PAG2* ( Fig . 3 ) ., The full STIB805 de novo assembly contained a single 14 . 9 kb contig corresponding to the EP1/EP2/PAG2/PAG4 locus ( S2 Fig . ) ., Also on chromosome 10 , the EP2 in T . evansi contains a stretch of 12 divergent amino acids in the domain N-terminal to the EP repeat; this region is highly conserved in T . brucei 57 , 58 ., Although the function of this domain is unknown , these 12 amino acids are found in all sequenced T . brucei genomes with very few variations ., The chromosome 6 procyclin locus contains a triplication of three genes ( EP3/PAG3/GRESAG2 ) in T . brucei TREU 927/4 , with GPEET present only in front of the last unit ( Fig . 3 ) ., Copy numbers of these genes may vary among T . brucei strains , as Southern analysis showed that these are single copy genes in the AnTat1 . 1 strain 49 ., In TevSTIB805ra , coverage of th | Introduction, Materials and Methods, Results, Discussion | Two key biological features distinguish Trypanosoma evansi from the T . brucei group: independence from the tsetse fly as obligatory vector , and independence from the need for functional mitochondrial DNA ( kinetoplast or kDNA ) ., In an effort to better understand the molecular causes and consequences of these differences , we sequenced the genome of an akinetoplastic T . evansi strain from China and compared it to the T . b ., brucei reference strain ., The annotated T . evansi genome shows extensive similarity to the reference , with 94 . 9% of the predicted T . b ., brucei coding sequences ( CDS ) having an ortholog in T . evansi , and 94 . 6% of the non-repetitive orthologs having a nucleotide identity of 95% or greater ., Interestingly , several procyclin-associated genes ( PAGs ) were disrupted or not found in this T . evansi strain , suggesting a selective loss of function in the absence of the insect life-cycle stage ., Surprisingly , orthologous sequences were found in T . evansi for all 978 nuclear CDS predicted to represent the mitochondrial proteome in T . brucei , although a small number of these may have lost functionality ., Consistent with previous results , the F1FO-ATP synthase γ subunit was found to have an A281 deletion , which is involved in generation of a mitochondrial membrane potential in the absence of kDNA ., Candidates for CDS that are absent from the reference genome were identified in supplementary de novo assemblies of T . evansi reads ., Phylogenetic analyses show that the sequenced strain belongs to a dominant group of clonal T . evansi strains with worldwide distribution that also includes isolates classified as T . equiperdum ., At least three other types of T . evansi or T . equiperdum have emerged independently ., Overall , the elucidation of the T . evansi genome sequence reveals extensive similarity of T . brucei and supports the contention that T . evansi should be classified as a subspecies of T . brucei . | The single-cell parasite Trypanosoma evansi is the disease-causing trypanosome with the widest geographical distribution ., The disease , called surra , has significant economic impact primarily due to infections of cattle , horses , and camels ., Morphologically the parasite is indistinguishable from bloodstream stage T . brucei , a parasite causing sleeping sickness in humans and the disease nagana in animals ., T . brucei , however , is strictly bound to sub-Saharan Africa where its obligate vector , the tsetse fly , resides ., The lack of a complete mitochondrial genome in T . evansi further distinguishes this parasite from T . brucei ., Important questions regarding the biology of T . evansi include how it escaped from Africa , whether this has happened more than once , and how exactly it is related to T . brucei ., To help answer these questions we have sequenced the T . evansi nuclear genome ., Our phylogenetic analysis demonstrates that T . evansi , and the closely related horse parasite T . equiperdum , evolved more than once from T . brucei ., We also demonstrate extensive similarity to T . brucei , including the maintenance of numerous genes that T . evansi no longer requires ., Therefore , despite the significant functional and pathological differences between T . evansi and T . brucei , our analysis supports the notion that T . evansi is not an independent species . | developmental biology, kinetoplastids, veterinary parasitology, cell biology, genome evolution, evolutionary biology, molecular biology, parasite evolution, life cycles, biology and life sciences, molecular evolution, microbiology, computational biology, molecular cell biology, protozoology, protozoan life cycles, parasitology, evolutionary genetics | null |
journal.pgen.1003090 | 2,012 | Genome-Wide Fine-Scale Recombination Rate Variation in Drosophila melanogaster | Recombination is a biological process of fundamental importance in population genetic inference ., The crossing-over of homologous chromosomes during meiosis results in the exchange of genetic material and the formation of new haplotypes ., Accurate estimates of the recombination rate in different regions of the genome help us to understand the molecular and evolutionary mechanisms of recombination , as well as a host of other important phenomena ., For example , recombination rate estimates are needed in assessing the impacts of natural selection 1 , 2 , admixture 3 , and disease associations 4 ., Recombination rates have been observed to exhibit a number of interesting heterogeneities: they are known to vary in magnitude and distribution between species ( e . g . , 5–7 ) , between populations within species 8 , 9 , and between individuals within populations 9–12 ., There is also substantial variation in different regions of the genome at different scales ., At the broad-scale , for example , recombination rates in humans are known to be correlated negatively with the distance from telomeres 13 , while at the fine-scale , recombination events cluster in narrow hotspots of 2 kb width 4 , 13 , 14 ., In humans , hotspots are typically defined as those with statistical support in favor of at least a five-fold increase of the recombination rate 13 over the background or surrounding region , and many hotspots suggest a ten- or even hundred-fold increase ., Such hotspots exhibit a powerful influence on the recombination landscape; 70–80% of recombination events in humans occur in 10% of the total sequence 8 ., Extensive fine-scale variation and recombination hotspots have also been found in other species , including chimpanzees 7 , Arabidopsis thaliana 15 and yeast 16 ., The picture in Drosophila is however less clear ., Broad-scale maps of recombination have been constructed for D . melanogaster by fitting a third-order polynomial to each chromosome arm 17 , 18 ., These give an overview of the distribution of recombination along each arm , quantifying for example earlier observations of declining recombination rates with proximity to the telomeres and centromeres ., Variation on finer scales has been inferred by studies of linkage disequilibrium ( LD ) and by breeding experiments ., Rapid and consistent decay in LD 19 leads to an absence of long haplotype blocks ., There is scant evidence for hotspots either at the intensity or prevalence of those found in humans ., Experimental studies of variation have produced local , fine-scale maps in D . melanogaster 20 , D . persimilis 21 , and D . pseudoobscura 22 , 23 , providing a resolution typically on the order of 100 kb in the regions analyzed ., These experimental results suggest that regions of fine-scale variation—including some mild “hotspots” 22—do exist in several Drosophila species ., For example , Singh et al . 20 study a 1 . 2 Mb region of the X chromosome in D . melanogaster , and find 3 . 5-fold variation in this region , though no hotspots by the criterion mentioned above ., These experimental approaches are cumbersome to recapitulate , however ., A number of crucial questions concerning Drosophila therefore remain unanswered ., It is not known to what extent this variation is further localized to finer scales , or how common such variation is across the genome ., Further , intra-specific differences in recombination rate have not been characterized ., However , the advent of ambitious projects ( e . g . , see the Drosophila Genetic Reference Panel 18 and the Drosophila Population Genomics Project 24 ) sequencing tens of D . melanogaster genomes each from different global populations raises the exciting prospect of addressing these and other questions ., The patterns of LD in a random sample of contemporary genome sequences taken from a population contain a great deal of information regarding historical recombination events , and from these we can infer recombination rates across the genome ., A number of sophisticated and computationally-intensive statistical approaches have been developed for inferring recombination rates from such data 14 , 25–27 and for testing for the presence of recombination hotspots 28 , 29 , and are ostensibly suitable for this task ., In particular , LDhat 14 , 25 , 30 is a useful software package which scales well to large datasets , and it has therefore been applied to estimating recombination rates in humans 4 , 8 , 13 , 14 , chimpanzees 7 , dogs 31 , yeast 16 , and microbes 32 , among others ., Estimating fine-scale recombination rates from recently published D . melanogaster genomes is , however , challenging for several reasons: First , these data exhibit a much higher density of single nucleotide polymorphisms ( SNPs ) than those of other species and of earlier technologies ., For example , the African data considered in this paper exhibits a mean SNP rate of about 1 SNP per 38 bp for a sample of size 22 , far higher than those of other recent sequencing projects ( e . g . , 8 ) ., This promises an unprecedented opportunity to localize recombination rate variation to very fine scales , but making full use of these data raises further challenges in computational and statistical efficiency ., Second , data generated from short-read sequencing technologies give rise to numerous missing alleles ., It would be highly advantageous to be able to make use of sites in which some alleles are missing without the exponential increase in LDhats running time that this entails ., Third , the background recombination parameter in D . melanogaster is known to be an order of magnitude higher than in humans ( the species for which LDhats prior distributions and parameters are typically calibrated ) and it is not clear how this will affect the accuracy of subsequent rate estimates ., Fourth , there is a growing consensus that a considerable fraction of the genome of some Drosophila species is influenced by adaptive substitutions 2 , 33 ., Recurrent selective sweeps combined with genetic hitchhiking affect patterns of variation across many kilobases of sequence and have the potential to invalidate inferences of recombination , even leading to the possibility of spurious signals of recombination hotspots 34 , 35 ., By contrast , the footprints of positive selection in recent human evolution are less widespread 1 ., The model underlying LDhat assumes a neutrally evolving population of constant size ., While LDhat is known to be robust to mis-specification of the demographic model 14 , its susceptibility to the effects of selection is less clear cut ., In this paper , we develop a new method , called LDhelmet , which addresses the above critical issues ., While it employs a reversible-jump Markov Chain Monte Carlo ( rjMCMC ) mechanism similar to that of LDhat , our method has a number of modifications that render key advantages ., Briefly , by utilizing recent theoretical advances in asymptotic sampling distributions 36–41 , we introduce several analytic improvements to the computation of likelihoods in the underlying population genetic model , which reduce Monte Carlo errors and simultaneously provide likelihoods for all relevant samples with an arbitrary number of missing alleles ., Our refinements further improve accuracy by allowing us to make full use of a quadra-allelic mutation model in which realistic mutation patterns between the four nucleotides A , C , G , T can be taken into account ., Additionally , we utilize information from the available genomes of outgroup species by using them to infer a distribution on the ancestral allele at each polymorphic site in D . melanogaster ., Taken together , our method enables us to compute fine-scale , genome-wide recombination rates with considerably improved accuracy and efficiency ., LDhelmet generally produces recombination maps that are less noisy than that of LDhats ., In particular , while LDhat can infer spurious hotspots under certain types of selection , we demonstrate that our approach is much more robust ., We apply our method to data taken from two D . melanogaster populations , one from North America and the other from Africa , and estimate fine-scale recombination maps for each population ., Then , through a wavelet analysis , we capture levels of variability and correlation of the two recombination maps , and provide a quantitative view of genome-wide inter-population comparison of recombination rates in D . melanogaster ., We also employ the wavelet analysis to examine the correlation between various genomic features , including recombination rates , diversity , divergence , GC content , gene content , and sequence quality ., At the fine-scale , we perform a conservative , systematic search for evidence of the existence of recombination hotspots and find a handful of putative hotspots each with at least a tenfold increase in intensity over the background rate ., Also , we compare our recombination rate estimates with existing experimental genetic maps ., Our software LDhelmet and the fine-scale recombination maps described in this paper are publicly available at http://sourceforge . net/projects/ldhelmet/ ., Using the procedure described in Materials and Methods , we were able to designate the ancestral allele in 1 , 755 , 040 of 2 , 475 , 674 high quality ( quality score ) SNPs in the RAL sample ( 70 . 9% ) , and 2 , 213 , 312 out of 3 , 134 , 295 high quality SNPs in the RG sample ( 70 . 6% ) ., These collections of polarized SNPs yielded the following estimates for the mutation transition matrix , with rows and columns ordered as A , C , G , T:These results imply that simple diallelic models are inadequate for the Drosophila populations ., As expected , we see a transition:transversion bias ., We also observe a higher overall mutation rate away from C and G nucleotides—this pattern persists even after excluding CpG sites from our analysis ( not shown ) ., Indeed , each of the four nucleotides exhibits its own characteristic mutation distribution ., There appears to be no significant difference between the transition matrices for the two populations ., This is partly explained by the shared history of the two populations: There were 2 , 990 , 025 sites for which:, ( i ) data were available in both populations ,, ( ii ) two alleles were observed in the combined sample , and, ( iii ) one of the two alleles was assignable as ancestral ., Of these , 925 , 569 ( 31 . 0% ) were polymorphic in both populations , 800 , 118 ( 26 . 8% ) were private to RAL , 1 , 262 , 109 ( 42 . 2% ) were private to RG , and 2 , 229 ( 0 . 1% ) were fixed differences ., For simplicity , in the analysis described in this paper , we used the same mutation transition matrix for all sites in the genome ., However , we note that our method can easily handle site-specific mutation transition matrices at no extra computational cost; see Materials and Methods: for details ., To assess the accuracy of estimated recombination maps , we carried out an extensive simulation study with various simple recombination patterns , first assuming selective neutrality ( the case with natural selection is discussed in the subsequent section ) ., The simulations assumed a finite-sites , quadra-allelic mutation model , with the mutation transition matrix shown above and the population-scaled mutation rate per bp ., We used these transition matrix and mutation rate in LDhelmets inference ., For LDhat , we used the corresponding effective mutation rate per bp ( see Estimation of mutation transition matrices ) ., Incidentally , we note that per bp is the estimated effective mutation rate for the autosomes of RAL lines 24 ., Figure 1 shows representative examples of LDhelmets and LDhats results ., As the figure illustrates , our method LDhelmet generally produces recombination maps that are less noisy than that of LDhats; in particular , LDhelmet produces spurious “spikes” less frequently than does LDhat ., To illustrate the impact of the spikes on the total genetic distance , the corresponding cumulative recombination maps comparing LDhelmet and LDhat are shown in Figure S1 ., Additional comparisons between LDhelmet and LDhat can be found in Table S1 , and SNP statistics of the datasets are listed in Table S2 ., In general , we observed that LDhelmet is able to identify the location of hotspots reliably ., Furthermore , in the scenario considered in the second row of Figure 1 , the width and height of the hotspot could be estimated very accurately; on average the total rate in the hotspot region could be estimated within 2 . 5% of the true value ., To test the performance of LDhelmet in a more realistic scenario , we simulated 1 Mb regions each with a substantial amount variation in recombination rate and with a high average rate representative of the interior of the D . melanogaster X chromosome ., To specify realistic levels of recombination rate variability in these regions , we took as the true recombination map a 1 Mb excerpt from our estimated map for the RAL sample ., To specify realistic absolute levels of recombination , we rescaled this map to match the mean ( per megabase ) recombination rates inferred for the X chromosomes of RAL and of RG ., In Figure 2 , LDhelmets estimated recombination maps for these two scenarios are illustrated in blue , while the true maps are shown in red ., These results demonstrate that , even when the average recombination rate is high , LDhelmet with our chosen block penalty in the rjMCMC is able to capture the pattern of fine-scale variation rather well ., However , we note that in the top plot of Figure 2 , in which case the true average rate is per kb , the estimated map tends to be slightly more variable than the true map ., In contrast , if the true average recombination rate is substantially higher , as in the bottom plot of Figure 2 with the true average rate of per kb but otherwise the same pattern of variation , the estimated map tends to be somewhat smoother than the true map ., Clearly , there is no single block penalty value that is universally optimal in all cases , but the value we have chosen seems to yield reasonable results for D . melanogaster ( see Materials and Methods for further details on the choice of block penalty ) ., It has been previously shown 34 , 35 , 42 that hitchhiking can induce seemingly similar patterns of linkage disequilibrium as that created by recombination hotspots , while McVean 43 has argued that the precise signatures of selective sweeps and hotspots actually differ ., To test the robustness of our method to natural selection , we simulated data under various scenarios with positive selection and recombination rate variation , and assessed the impact on our estimates of recombination rates ., We generated data using a range of values for the selection strength and fixation time ., See Simulation study on the impact of natural selection for details of the simulation setup ., The results of LDhelmet and LDhat for a few cases are shown in Figure 3; each plot shows the results for 25 simulated datasets illustrated in 25 different colors ., The corresponding cumulative recombination maps are shown in Figure S2 ., For both methods , the estimated recombination maps are in general noisier than that for the neutral case ( c . f . , Figure 1 ) , though LDhelmet is still more robust than LDhat ., As can be seen in Figure 3 , LDhat tends to produce false inference of elevated recombination rates near the selected site more frequently than does LDhelmet ., A more detailed comparison is provided in Table S3 and SNP statistics of the datasets are listed in Table S2 ., Overall , although strong positive selection causes more noise and fluctuations in our estimates , it does not seem to produce a strong bias to the extent that would consistently lead to false inference of recombination hotspots ., The noise in our estimates of the recombination rate in the presence of selection depends on several factors ., Specifically , we observed that the accuracy of our estimates decreases as the selection strength increases , whereas the accuracy improves as the distance between the selected site and the region of estimation increases ., Furthermore , the more recent the time of fixation , the noisier are the estimates ., In addition to the case of a single , recent selective sweep , we also assessed the impact of recurrent selective sweeps 44 , 45 on the estimation of recombination rates ., Assuming that beneficial mutations fixate randomly at a given rate , we simulated three different sets of datasets with a background recombination rate of per kb , as detailed in Simulation study on the impact of natural selection ., The degree to which recurrent sweeps reduce diversity in each model is summarized in Table S4 ., In model RS3 , which has infrequent but strong sweeps , the mean number of SNPs reduces by more than a factor of relative to the neutral model ., Such a drastic drop in diversity significantly reduces the ability to perform accurate statistical inference of recombination ., To infer the location of a recombination hotspot , for example , at least a few SNPs must be present in the hotspot and near its edges ., The results of recombination rate estimation under recurrent sweep models are summarized in Table 1 and Table 2 ., Compared to a single sweep , recurrent selective sweeps tend to decrease the accuracy of recombination rate estimates more noticeably ., Furthermore , infrequent but strong selective sweeps ( model RS3 ) have more severe impact on the accuracy than do frequent but weaker selective sweeps ( model RS1 ) ., As discussed above and can be seen in Table 2 , detecting recombination hotspots in model RS3 would pose a great challenge ., Overall , LDhelmet generally underestimates the recombination rate in the presence of selection , suggesting that it is unlikely to produce spurious hotspots because of selection ., We also tested our method on datasets simulated under a variety of demographic scenarios ., Specifically , the demographic models we considered are those proposed by Haddrill et al . 46 , and by Thornton & Andolfatto 47 , comprising two exponential growth models and two bottleneck models ., As in the neutral simulations , we assumed a finite-sites , quadra-allelic mutation model , with the mutation transition matrix and the mutation rate per bp ., See Simulation study on the impact of demographic history for details on the other parameters used in the simulations ., Table 3 and Table 4 show the results of recombination rate estimation in this simulation study ., Although the estimates are clearly less accurate compared to the case with constant population size , they are reasonably accurate in most cases ., Note that the overall trend is to underestimate the true rates , in some cases only slightly ., As in the case of recurrent selective sweeps , demography may decrease diversity , thus hindering statistical inference of recombination ., Table S4 includes the SNP statistics for the demography models we considered ., In model B2 , which involves a very recent bottleneck , a reduction in diversity by about a factor of was observed , partly explaining the particularly poor estimates of the recombination rate ., Table S5 shows that the average SNP density of the D . melanogaster data considered in this paper; note that the average SNP density of each chromosome is substantially greater than the SNP density observed in simulation model B2 ., The population-specific average recombination rate for each major chromosome arm is summarized in Table 5 , which shows that the average rate for the African ( RG ) population is higher than that for the North American ( RAL ) population ., This difference could be explained partially , but not entirely , by a difference in population size ., Note that the average recombination rate in the X chromosome appears to be higher than that in the autosomes , much more so in RG than in RAL ., Table 5 shows the ratio of the average recombination rate of RG to that of RAL for each chromosome arm ., Although the ratio is more or less consistent for the autosomes , the ratio for the X chromosome is significantly higher ., Hence , a difference in population size could explain the higher recombination rate estimates in RG for the autosomes , but it does not explain the significant increase in the recombination rate for the X chromosome of RG over RAL ., Furthermore , for RAL , that the observed average recombination rate of the X chromosome is higher than that of autosomes is unexpected given that an excess of LD is observed on the X chromosome of this population 18 , 24 ., In both populations , arm 3R has a notably reduced recombination rate compared to the other arms ., This reduction is more pronounced in RG than in RAL , which could be partly explained by the fact that , in African populations , arm 3R has the largest number of common inversions 48 ., To study the effect of sample size on the estimation of recombination rates , we subsampled a 2 Mb excerpt of chromosome arm 2L from each population over several repeated trials ., We performed the subsampling on an excerpt rather than the entire genome for computational reasons ., The averages of the estimates are shown in Table S6 ., Despite a slight increase in the estimate as sample size increases , the effect is small and appears to diminish with increasing sample size ., We also analyzed the whole-genome RAL dataset down-sampled to match the sample size ( i . e . , 22 ) of RG ., As Table 5 shows , the genome-wide average estimates produced using 22 genomes of RAL were only slightly lower than those produced using all 37 genomes ., Encouragingly , our estimate ( per kb ) of the recombination rate for the X chromosome of RG is similar to the previous estimates for other African populations obtained using a different method: Haddrill et al . 46 estimated , and per kb for the X recombination rate in three African populations ., To assess the effect of SNP density , we thinned the SNPs on chromosome arm 2L and chromosome X of the RG dataset to the corresponding SNP densities of RAL , and performed inference on the resulting data ., The results summarized in Table S7 show that although SNP density indeed influences the ability to estimate recombination rates , the impact is not nearly large enough to account for the difference between the observed recombination rates of RAL and RG on the X chromosome ., Finally , as there exist several inversions in D . melanogaster , we analyzed regions of inversion excluding individuals known to carry the inversion 49 ., The comparison of excluding individuals with inversions and the original analysis is shown in Table S8 ., Note that for each inversion , only a small number of individuals carry it ., We found that excluding the individuals with inversions did not significantly affect the recombination rate estimates ., LDhelmets fine-scale recombination maps for RAL and RG are illustrated in Figure 4; files containing the corresponding numerical values are publicly available ., To assess the accuracy of our recombination estimates obtained via statistical analysis of population genetic variation data , we compared them to genetic maps obtained experimentally ., Singh et al . 20 examined the fine-scale recombination rate variation over a 1 . 2 Mb region of the D . melanogaster X chromosome using a genetic mapping approach , by crossing an African line with a line presumably of North American origin ( a cross between two lines from Bloomington Drosophila Stock Center ) ., For their experiment , Singh et al . genotyped SNPs and identified two flanking genes , white and echinus , with visible phenotypes ., They found statistically significant heterogeneity in this region , with differences in rate up to -fold ., In Figure 5 , estimates from LDhelmet for both the RAL and RG samples are shown , along with the genetic map from 20 ., Both estimates from LDhelmet mostly fall within the confidence intervals of the empirical estimate , with the exception of the outermost intervals ., The three maps share the same overall shape , including the location of the highest peak ., We find -fold variation in the RG estimate , which is comparable to the -fold variation obtained by Singh et al . The high correlation among the three maps give us confidence that our estimates from the statistical analysis of population genetic data accurately represent the true underlying recombination map ., In a second study , we compared our chromosome-wide recombination estimates with a consensus genetic map for each chromosome arm based on data hosted at the FlyBase website ( http://www . flybase . org 50 ) ., To facilitate a comparison with this map , resolution of which is roughly 200 kb , we binned our data into the same cytogenetic subdivisions 24 and LOESS-smoothed the results , with a span of 15%; a correspondingly LOESS-smoothed version of the FlyBase data was kindly provided to us by C . H . Langley ., A comparison of the maps is shown in Figure 6; evidently , the three estimates show broad agreement , each capturing key features such as the spike in recombination near position 10 Mb on arm 2L , as well as a series of dramatic changes in recombination rate across chromosome X . When the recombination map for RAL is regressed on the FlyBase maps , the coefficient of determination , or proportion of variability explained by the simple linear regression model , is and for chromosome arms 2L , 2R , 3L , 3R , and X , respectively; the corresponding values for RG are , and ., These correlations are lower than those seen in a comparison of statistically- versus experimentally-derived maps in humans ( e . g . 13 ) , though in that case the experimental data from pedigrees were of higher quality ., As noted by Langley et al . 24 , data on which the FlyBase map is based is highly edited and based on heterogeneous experimental conditions with sometimes conflicting results ., As discussed in the sec:introduction , it is well known that in humans and many other eukaryotes recombination tends to cluster in recombination hotspots , regions of approximately 2 kb wide in which the rate of recombination may be one or more orders of magnitude higher than the background rate 4 , 12–14 ., However , it is an open question whether hotspots exist in the D . melanogaster genome , or to what extent recombination rates vary on a fine scale ., We first searched for the most extreme forms of recombination rate variation—namely , recombination hotspots ., Using a highly conservative approach described in Materials and Methods , we initially identified nineteen and five putative autosomal recombination hotspots from the RAL and RG data , respectively ., Of these , respectively six and four were also detected by the hotspot detection software sequenceLDhot 29 ., These ten hotspots , the width of which ranges between 0 . 5 kb and 6 . 8 kb , are listed in Table 6 ., All were found in genic regions , with all except one overlapping exons and one contained within an intron ., An example of a RAL hotspot is shown in Figure 7 , where we also show the RG recombination map ., The fine-scale recombination maps in this region for the two populations are clearly highly correlated , with both RAL and RG exhibiting a tenfold increase in recombination rate within almost identical 4 kb intervals , though only the hotspot of RAL was also found by sequenceLDhot ., We note that the power of sequenceLDhot may be further reduced by the apparent preference ( not shown ) for putative hotspots to reside in regions in which the “local” background rate is higher than that of the chromosome arm as a whole ., In light of these factors , it is likely that several more hotspots would have been found in one or both populations under a more relaxed definition , though it is clear that they are far scarcer , and less hot , than in humans ., It is apparent from both RAL and RG maps shown in Figure 4 that recombination rates vary on multiple scales , from the very fine to the very broad , as has been observed in a number of other species 7 , 13–16 ., It is clear , for example , that recombination rates tail off towards each end of the arm , with the reduction towards the telomere much more precipitous than the pericentromeric reduction ., The latter reduction is evident from as far as the start of heterochromatic sequence a few megabases from the centromere , in agreement with other broad-scale estimates of recombination 17 , 18 , although we do not find a complete absence of recombination here ., Figure 8 shows that the recombination rate in the X chromosome between positions 10 kb and 20 kb is noticeably higher than the rate in the subtelomeric region to the right ., This trend is much more pronounced in the North American X than in the African X , consistent with a previous study by Anderson et al . 51 ., The telomere-associated sequence ( TAS ) , located to the left of position kb , was not available in our data , but Anderson et al . provided evidence that the TAS region in the North American X exhibits even higher recombination rate than that in the subtelomeric region between positions 10 kb and 20 kb ., As shown in Figure 4 , the largest difference between the estimated recombination maps of the two populations is observed in the X chromosome ., First , the recombination map in the African X is generally much higher than that in the North American X . Second , there is noticeably less variation in the estimated African X recombination map ., As mentioned earlier in the discussion of our simulation study , when the average recombination rate is as high as that of the African X , the amount of variation in our estimated map tends to be somewhat lower than the true variation ., Hence , the observed reduction in variation could be partially attributed to our method being not sensitive enough in that range of very high rates ., More generally , it is also true that Fishers information for data on sequence variation is lower in regions of high recombination ( ) , which could create an inherent limitation in our ability to infer recombination rate changes here ., The use of wavelets enables us to compare how changes in the rate of recombination along the genome correlate with other genomic features ., For each population we computed pairwise correlations between the detail coefficients of the following features: diversity ( mean fraction of pairwise differences between each individual in the population , within sequenced nucleotides ) , divergence ( fraction of differences between the reference sequences of D . melanogaster and D . simulans ) , GC content , gene content ( fraction of sites annotated as exonic ) , and sequence quality ( Phred score ) , as well as the recombination rate , with each feature measured in 250 bp windows ( see Materials and Methods ) ., Results are shown in Figure 12 and Figure S12 , and follow a similar analysis performed by Spencer et al . 53 on human data ., From these results we can make a number of observations detailed below ., We have developed a new method , LDhelmet , which is able to provide accurate estimates of recombination rates using genomic data from D . melanogaster ., Although our focus has been on this species , the features of our method should offer improvements in the estimation of recombination in other species too ., For example , the desire to efficiently incorporate sites in which some alleles are missing is a common issue when data are generated by next-generation sequencing technologies ., We believe that our method will find many further applications in other datasets ., Using our method , we have performed a genome-wide comparison of fine-scale recombination rates between two populations of D . melanogaster , one from Raleigh , USA ( labeled RAL ) and the other from Gikongoro , Rwanda ( labeled RG ) ., While earlier studies have largely been confined to regions of moderate resolution , we find extensive fine-scale variation across all chromoso | Introduction, Results, Discussion, Materials and Methods | Estimating fine-scale recombination maps of Drosophila from population genomic data is a challenging problem , in particular because of the high background recombination rate ., In this paper , a new computational method is developed to address this challenge ., Through an extensive simulation study , it is demonstrated that the method allows more accurate inference , and exhibits greater robustness to the effects of natural selection and noise , compared to a well-used previous method developed for studying fine-scale recombination rate variation in the human genome ., As an application , a genome-wide analysis of genetic variation data is performed for two Drosophila melanogaster populations , one from North America ( Raleigh , USA ) and the other from Africa ( Gikongoro , Rwanda ) ., It is shown that fine-scale recombination rate variation is widespread throughout the D . melanogaster genome , across all chromosomes and in both populations ., At the fine-scale , a conservative , systematic search for evidence of recombination hotspots suggests the existence of a handful of putative hotspots each with at least a tenfold increase in intensity over the background rate ., A wavelet analysis is carried out to compare the estimated recombination maps in the two populations and to quantify the extent to which recombination rates are conserved ., In general , similarity is observed at very broad scales , but substantial differences are seen at fine scales ., The average recombination rate of the X chromosome appears to be higher than that of the autosomes in both populations , and this pattern is much more pronounced in the African population than the North American population ., The correlation between various genomic features—including recombination rates , diversity , divergence , GC content , gene content , and sequence quality—is examined using the wavelet analysis , and it is shown that the most notable difference between D . melanogaster and humans is in the correlation between recombination and diversity . | Recombination is a process by which chromosomes exchange genetic material during meiosis ., It is important in evolution because it provides offspring with new combinations of genes , and so estimating the rate of recombination is of fundamental importance in various population genomic inference problems ., In this paper , we develop a new statistical method to enable robust estimation of fine-scale recombination maps of Drosophila , a genus of common fruit flies , in which the background recombination rate is high and natural selection has been prevalent ., We apply our method to produce fine-scale recombination maps for a North American population and an African population of D . melanogaster ., For both populations , we find extensive fine-scale variation in recombination rate throughout the genome ., We provide a quantitative characterization of the similarities and differences between the recombination maps of the two populations; our study reveals high correlation at broad scales and low correlation at fine scales , as has been documented among human populations ., We also examine the correlation between various genomic features ., Furthermore , using a conservative approach , we find a handful of putative recombination “hotspot” regions with solid statistical support for a local elevation of at least 10 times the background recombination rate . | stochastic processes, mathematics, statistics, population genetics, biology, evolutionary biology, statistical methods, probability theory | null |
journal.pntd.0001503 | 2,012 | Diversification of Schistosoma japonicum in Mainland China Revealed by Mitochondrial DNA | Schistosomiasis is one of the most neglected tropical diseases , with six species in the Schistosoma still infecting more than 200 million people in the world 1–3 ., Schistosomiasis japonica is distributed in Indonesia , Philippines , and China ., In mainland China , this parasitic disease is the most severe zoonosis infecting about 360 , 000 people and about 1% buffalo and/or cattle in endemic regions , particularly in lake/marshland and hilly areas of Hubei , Hunan , Anhui , Jiangxi and Jiangsu provinces and mountainous areas of Sichuan and Yunnan provinces 4 ., Over the last 50 years , continuous efforts involving various measures , such as health education , snail control , community chemotherapy and environmental management have contributed significantly to the dramatic reduction in infection levels and epidemic areas of this parasitic disease in China , setting China as one of the most successful countries in control of schistosomiasis in the world 5–8 ., However , recently available data have suggested that schistosomiasis has re-emerged over the last decade , probably as a severe threat once again to human health especially in rural areas of mainland China 9 , 10 ., The drastic pathogenesis , the number of reservoir hosts involved in epidemiology and the vast endemic areas of schistosomiasis japonica have inevitably resulted in a less investigated status for S . japonicum in respect with its genetic diversity , host immune response etc . when compared with other schistosomes 6 , 11 , 12 ., The genus Oncomelania , which is the intermediate host of S . japonicum , was classified into different species and/or subspecies according to their morphology , biogeography and phylogeny 13 , 14 ., With the distinct diversity of snails in the genus Oncomelania which has been verified using various markers 14–16 , the diversity of the parasite S . japonicum is of great interest for research from a co-evolutionary point of view ., How diverse the digenean S . japonicum really is in such a large geographical range has not been well assessed especially in mainland China ., An accurate measure of its population genetic diversity is certainly needed to clarify our understanding on the epidemiology of schistosomiasis 17 , which may be also useful for implementing control measures , and for developing drugs or potential vaccines , as worms of different genetic backgrounds may respond differently to such treatments 18 , 19 ., In recent years , several molecular markers have been used to detect the variability of S . japonicum populations ., Gasser et al . 20 found the variability among 7 geographical isolates across mainland China using the random amplified polymorphism DNA ( RAPD ) technique and suggested a potential strain complex for S . japonicum ., Sorensen et al . 21 reported differences between S . japonicum populations from 6 localities in mainland China using NADH dehydrogenase subunit 1 ( ND1 ) gene , but could not detect variability conclusively at the intrapopulation level ., Bøgh et al . 22 did find 15 types of ND1 conformations and 23 types of cytochrome c oxidase subunit 1 ( CO1 ) conformations in 9 populations from 7 provinces across mainland China by single-strand conformational polymorphism ( SSCP ) ., These results did in fact suggest the significant polymorphism among S . japonicum in mainland China , but provided very limited information relating to the population genetic diversity of this species ., Upon the identification of polymorphic microsatellite loci , Shrivastava et al . 6 investigated the genetic variation of S . japonicum populations from 8 geographical locations in 7 endemic provinces across mainland China , and a high level of polymorphism was reported between and within populations ., They considered that populations of S . japonicum in mainland China could be separated mainly into the populations in Sichuan and Yunnan provinces as being in southwest ( SW ) China and those in low-lying lake regions along the middle and lower ( ML ) reaches of Yangtze River ., With three partial mitochondrial genes ( cox3 , nad4 and nad5 ) from 28 individual adult worms , Zhao et al . 23 reported recently that all parasites from SW China were grouped together , whereas those from the ML reaches of Yangtze River were not clustered together ., However , the reports by Shrivastava et al . 6 and Zhao et al . 23 both contained limited specimens from relatively few localities , which may not represent the geographical distribution of this schistosome , and thus not its exact population genetic diversity , in mainland China ., A comprehensive analysis is therefore needed using more molecular markers to examine more populations of S . japonicum from a wide range of its geographical distributions , especially in severely endemic areas along the ML reaches of Yangtze River in China ., In this study , mitochondrial DNA sequences including Cytb-ND4L-ND4 , 16S-12S and ND1 were examined for S . japonicum collected from localities in seven provinces of China , where schistosomiasis is geographically endemic ., The diversity in nucleotides and haplotypes was analyzed for different populations based on each of the three mitochondrial sequences and their combined sequences ., Phylogenetic tree and parsimony network were constructed for observed haplotypes , and the genetic distance was examined against the geographical distance in order to understand the genetic diversity in populations of S . japonicum in mainland China ., The procedures involving animals were carried out in accordance with the guidelines of the Association for Assessment and Accreditation of Laboratory Animal Care International ( AAALAC ) ., The animal study protocol was approved by the Institutional Animal Care and Use Committee of Wuhan University ., The intermediate host , Oncomelania hupensis , from 18 localities of 7 schistosomiasis endemic provinces in mainland China , including Hubei , Hunan , Anhui , Jiangxi , Jiangsu provinces in the ML reaches of Yangtze River , and Sichuan and Yunnan provinces which are in the higher reaches of the river in SW China , but separated from the ML reaches by mountain ranges ( Fig . 1 and Table 1 ) , were collected and transported to laboratory from October 2005 to October 2006 ., After one month captivity , snails were washed and exposed individually in water for 3 h in a vial under light at 25°C to stimulate the emergence of cercariae for identifying the S . japonicum infection ., Overall , snails from different localities had an infection rate ranging from 0 . 1% to 4% ., To generate adult worms , the best source of DNA , 10 field-collected infected snails from each locality , with the exception of Zongyang in Anhui province ( AHzy ) and Pengze in Jiangxi province ( JXpz ) where only three and four infected snails were obtained respectively , were exposed to light for 4 hours to stimulate the emergence of cercariae ., Five laboratory mice were infected percutaneously with 30 cercariae per mouse for each geographical locality ., 6 weeks following the infection , adult worms were retrieved by perfusion from mesenteric veins using 0 . 9% NaCl , and worms from each mouse infected with cercariae were pooled together , and washed extensively in saline before being preserved in 95% ethanol at 4°C ., The total genomic DNA was extracted individually from both male and female schistosomes using a standard sodium dodecyl sulfate-proteinase K procedure 24 ., Each worm was incubated and thawed in 200 µl extraction buffer containing 50 mM Tris-HCl , 50 mM EDTA , 100 mM NaCl , 1% SDS and 100 µg/ml proteinase K , at 56°C for 2 h with gentle mixing ., DNA in solution was extracted using standard phenol/chloroform purification , followed by 3 M sodium acetate ( pH 5 . 2 ) and ethanol precipitation ., Pellets of DNA were washed in 70% ethanol , air-dried , and resuspended in 10 µl TE ( pH 8 . 0 ) ., For each adult worm , three fragments , i . e . Cytb-ND4L-ND4 , ND1 and 16S-12S of the mitochondrial genome were sequenced ., For the Cytb-ND4L-ND4 fragment , the forward primer ND4F ( 5′- TTGGGGGTTGTCATGCGGAGTA -3′ ) and the reverse primer ND4R ( 5′- CAAATACCCAATAGCAACGGAACAC -3′ ) were used based on available GenBank sequence AF215860 ., For the ND1 fragment , the forward primer ND1F ( 5′- TAGAGGGTTTGTTGGTTGTTTTG -3′ ) and the reverse primer ND1R ( 5′- ACCATACTTTCATACTACTGCC -3′ ) were used based on available GenBank sequence AF215860 ., For the 16–12S fragment , the forward primer 16S-12SF ( 5′- GATTATTTCTAGTTCCCGAATGG -3′ ) and the reverse primer 16–12SR ( 5′- TGTAACGCACAACAACCTATACC -3′ ) were used based on available GenBank sequence AF215860 ., The PCR protocols were 94°C for 3 min followed by 30 cycles of 94°C for 30 s , 58°C ( for ND1 ) or 63°C ( for Cytb-ND4L-ND4 and 16S-12S ) for 30 s , and 72°C for 90 s and then a final elongation step at 72°C for 10 min ., The amplified products were purified on 1 . 0% agarose gel stained with ethidium bromide , using the DNA gel extraction kit ( Omega Bio-Tek ) ., The purified PCR products were sequenced using ABI PRISM BigDye Terminators v3 . 0 Cycle Sequencing ( ABI ) ., The DNA sequences were deposited in the GenBank database under accession numbers FJ851893–FJ852573 ., Sequences were aligned using ClustalX1 . 83 25 at default settings followed by manual correction in SEAVIEW 26 for each molecular marker ., DNAsp version 4 . 0 27 was used to define the haplotype ., The three parts , i . e . Cytb-ND4L-ND4 , ND1 and 16S-12S , of mitochondrial data were also combined and aligned into a new combined mitochondrial data set , with this combined sequence named as combined mtDNA ., Nucleotide divergences within and between populations were calculated in Arlequin3 . 11 28 and DNAsp ., Genetic variation within different populations was estimated by calculating nucleotide diversity ( π ) and haplotype diversity ( h ) values ., Selective neutrality was tested with Tajimas D 29 and Fus F test 30 ., The pairwise genetic difference was estimated for all populations by calculating Wrights F-statistics ( Fst ) based on gene flow ( Nm ) ., A Mantel g-test to compare the correlation between pairwise distance and geographical distance among localities was analyzed in Arlequin , with geographic distances ( km ) for the correlation analysis between geographical distance and genetic distance calculated using the great circle distance between localities ., The phylogenetic analysis for 96 haplotypes generated using combined mitochondrial DNA data was performed with Bayesian inference ( BI ) , which was carried out with MrBayes 3 . 1 31 under the best-fit substitution model ., Analyses were run for 1×106 generations with random starting tree , and four Markov chains ( with default heating values ) sampled every 100 generations ., Posterior probability values were estimated by generating a 50% majority rule consensus tree following the discard of first 3000 trees as part of a burn-in procedure ., The HKY+I+G model was determined as the best-fit model of sequence evolution by using the hierarchical likelihood ratio tests implemented in Modeltest 3 . 7 32 ., The phylogenetic tree was rooted using Schistosoma mansoni as outgroup ., The genetic structure was phylogenetically evaluated by constructing unrooted parsimony network of haplotypes for combined mtDNA data sets using TCS version 1 . 21 33 ., The primary sequence data were obtained by amplifying and sequencing three partial regions of the mitochondrial genome , i . e . Cytb-ND4L-ND4 with 793–794 bp , ND1 with 767 bp , and 16S-12S with 1463–1466 bp ., Measures of diversity of haplotypes and nucleotides within populations on the basis of the three mitochondrial regions are presented in Tables S1 , S2 and S3 , respectively ., The highest values for the diversity were all observed for populations in the ML reaches , and the lowest all in populations from the SW ( for details regarding each fragment , see Tables S1 , S2 and S3 ) ., The pairwise genetic distance among all 18 populations showed a high degree of variation , as revealed respectively from the three different mitochondrial regions ( for details , see Tables S4 , S5 and S6 ) ., A significant correlation was observed between geographical distance and genetic distance ( pairwise Fst ) for all 18 populations for Cytb-ND4L-ND4 ( R\u200a=\u200a0 . 642 , P<0 . 001 ) and 16S-12S ( R\u200a=\u200a0 . 746 , P<0 . 001 ) , respectively , which indicates that genetic distance increased with the increase in geographical distance ( Fig . 2a , b ) ., No significant correlation was detected when ND1 was used , with the correlation coefficient R\u200a=\u200a0 . 080 ( P>0 . 05 ) ( Fig . 2c ) ., However , among 15 populations in the ML reaches , the value of the correlation coefficient decreased to 0 . 119 ( P>0 . 05 ) and 0 . 061 ( P>0 . 05 ) for Cytb-ND4L-ND4 and 16S-12S , respectively ( Fig . 2d , e ) , implying that the genetic distance was not correlated with the geographical distance for populations in the ML reaches of Yangtze River ., Although some base substitutions were observed , selective neutrality of the observed nucleotide polymorphisms was suggested for S . japonicum , as indicated either by Tajimas D or Fus F test ( P>0 . 05 ) in each of the three regions ., As many studies have shown that longer genes contain generally more variable characters with proportionally more signals , and hence yield accurate phylogenetic estimates than shorter ones 34–36 , the combined mitochondrial data sets were then deduced from 169 specimens by aligning combined Cytb-ND4L-ND4 , ND1 and 16S-12S sequences ( combined mtDNA ) , which had a range of 3024 to 3027 bp , resulted in 3028 characters , including gaps , and 166 variable sites ( 113 parsimony informative sites ) ., A total of 96 mitochondrial haplotypes was observed ( Table 1 ) ., Measures of haplotype and nucleotide diversity based on combined mtDNA are presented in Table 2 ., The highest values in the diversity of haplotype and nucleotide were all observed for populations in the ML reaches , and the lowest were all in populations from the SW , which is consistent with the findings from the three separate mitochondrial DNA sequences ., 88 haplotypes were isolated from 143 specimens in five provinces along the ML reaches , with the mean haplotype and nucleotide diversity being 0 . 987±0 . 003 and 0 . 0036±0 . 0001 , respectively ., However , only 8 haplotypes were isolated from 26 specimens in the SW , with the haplotype and nucleotide diversity being 0 . 766±0 . 075 and 0 . 0017±0 . 0003 , respectively ., The Fst of all pairwise analyses varied from 0 . 482 to 0 . 870 between populations in the ML reaches and those in the SW ( Table 3 ) , showing highly significant difference ( P<0 . 001 ) ., Among the 3 populations in the SW , the Fst between SCxc and two Yunnan populations ( YNey and YNhq ) showed highly significant differences ( P<0 . 001 ) , whereas no significant difference was observed between YNey and YNhq ( P>0 . 05 ) ., Among the 15 populations in the ML reaches , the Fst varied from 0 . 014 to 0 . 807 ( Table 3 ) , with most of them being significantly different ( P<0 . 05 ) ., When all specimens were classified into two populations according to whether they were from above or below the three Gorges region , i . e . population in the ML reaches of the Yangtze River and population in Sichuan and Yunnan provinces of the SW China , the value of genetic distance ( Fst ) and the gene flow ( Nm ) between them was 0 . 381 ( P<0 . 001 ) and 0 . 410 , respectively ., Significant correlation was also observed between geographical distance and genetic distance ( pairwise Fst ) among all 18 populations for combined mtDNA ( R\u200a=\u200a0 . 670 , P<0 . 001 ) , indicating that genetic distance increased with the increase in geographical distance ( Fig . 2f ) ., Among 15 populations in the ML reaches , the value of the correlation coefficient decreased to 0 . 077 ( P>0 . 05 ) ( Fig . 2g ) , implying that the genetic distance was not correlated with the geographical distance for populations in the ML reaches of Yangtze River ., As shown in the Bayesian phylogenetic tree ( Fig . 3 ) , two clades can be clearly separated ., Clade A contains almost all haplotypes from all five provinces in the ML reaches of the Yangtze River ., Although various divergence and some subclades were observed within this clade , support probabilities for each clade were generally very low ., Haplotypes in the ML reaches were clustered in various subclades , and no obvious lineage was observed for haplotypes from different provinces along the ML reaches ., However , subclades A1 and A2 include most haplotypes from Hubei , Hunan , Anhui , and Jiangxi provinces , and subclade A6 includes haplotypes from Hubei , Hunan , Anhui , and Jiangsu provinces ., It is apparent that clade B can be separated into two distinct subclades , B1 and B2 , with clade B1 having a high support probability and containing only haplotypes from Sichuan and Yunnan provinces in SW China , and B2 containing three haplotypes from three provinces in the ML reaches ., Surprisingly , other trees ( NJ , ML , MP; not shown ) , although inconsistent in their respects , all had such two clades containing haplotypes from SW China , and three from the ML reaches , despite a relatively low level of support probabilities ., The network constructed by statistical parsimony from 96 haplotypes on the basis of combined mtDNA sequences showed some characters as observed in the phylogenetic tree ., The haplotype network was rather complicated , without any obvious lineages for those haplotypes from localities in the ML reaches ( Fig . 4 ) ., However , all haplotypes from SW ( from H26 to H33 ) were clustered together ( Fig . 4 ) , which corresponded exactly to clade B1 in Fig . 3 , and this clade contained no haplotypes from the ML reaches of Yangtze River , but was related with a few haplotypes from the ML reaches ( Fig . 4 ) , as also indicated in clade B2 which formed , together with B1 , into clade B ( Fig . 3 ) ., A relatively large network containing haplotypes ( from H71 to H93 ) from about 10 localities ( Fig . 4 ) showed some similarity with clade A1 in Fig . 3 , in composition of haplotypes ., It is , however , impossible to detect any other patterns of haplotype networks , and impossible to find other geographical relationships or characteristic lineages in other network branches , which is largely consistent with the complex structure of clade A in Fig . 3 ., The difference in genetic diversity of S . japonicum populations was demonstrated in samples collected from currently epidemic areas of schistosomiasis in mainland China , with the use of three mitochondrial fragments , Cytb-ND4L-ND4 , ND1 and 16S-12S , respectively , and the combined sequences of these three fragments ., The present study contains the mostly widespread and the largest number of S . japonicum populations in any attempts so far to examine the parasite genetic diversity in China ., Overall , populations of S . japonicum in mainland China showed a relatively large degree of variation in terms of nucleotide and haplotype diversity ., However , it is apparent that across the geographical distribution of schistosomiasis endemic areas in China , the genetic distance was correlated significantly with geographical distance when Cytb-ND4L-ND4 , 16S-12S , and combined mtDNA were used , although non-significance was observed for ND1 ., It is even more obvious that as revealed through analyses of nucleotide and haplotype diversity , populations in Hubei , Hunan , Anhui , Jiangxi and Jiangsu provinces , namely in the ML reaches of Yangtze River showed a much larger degree of genetic variation than those in Sichuan and Yunnan provinces of the SW China in the upper reaches of the river , and no haplotypes were shared between populations in the ML reaches and those in the SW ., Significant difference was also observed in genetic distance between populations in the ML reaches and populations in the SW , as revealed in pairwise analyses using individual and/or combined mitochondrial sequences ., Along the Yangtze River , are the endemic areas of schistosomiasis , and severe epidemic areas are mainly in the ML reaches 5 ., However , in the Three Gorges area that is from Yichang going upwards to Yibin ( Fig . 1 ) , human schistosomiasis has never been reported 10 ., It is quite obvious that the distribution of S . japonicum is geographically separated by the gorge area of the river ., This apparent geographical separation may account for the observed difference in no-shared haplotypes , and in the genetic distance for S . japonicum between areas in the ML reaches of Yangtze River and areas in the SW China ., When populations from the ML reaches and from the SW were further grouped separately , the Fst value ( 0 . 381 ) was greater than 0 . 25 , a value which was considered to be ‘very great’ by Wright 37 for genetic differentiation between populations ., It is therefore all indicated that a large level of genetic differentiation has evolutionarily occurred for S . japonicum , due to at least the geographical separation by the Three Gorges area and mountains ., Phylogenetic analyses and haplotype network may support this conclusion , as parasites from Sichuan and Yunnan provinces in the SW were all closely clustered in the phylogenetic tree and the haplotype network ., Using different molecular markers , other authors 6 , 23 have also , to some extent , detected the genetic difference between S . japonicum populations in the SW and those in the flood plain of the ML reaches of the Yangtze River ., Despite the finding that the mean nucleotide and haplotype diversity of populations in the SW were rather low when compared with the same parameters in the ML reaches , the genetic distance had some significant difference between the population from Sichuan , SCxc , and the two populations from Yunnan , YNey and YNhq , as revealed by Fst of pairwise analyses using ND1 , 16S-12S , and the combined mtDNA sequences , with the exception of Cytb-ND4L-ND4 ., Sichuan and Yunnan provinces are both distributed in Hengduan Mountains , and schistosomiasis was reported historically in various localities in these two provinces 38 ., As various mountain ranges and rivers , as well as intermountain basins , are the general features in Hengduan Mountains 39 , there must be some degree of geographical isolation in the distribution of S . japonicum in this region at a large geographical scale ., However , only three populations were included in the present study and efforts to obtain more parasite samples have been unsuccessful , although the intermediate host snails were collected in a much wider range ( unpublished data ) , due possibly to the continuous and extensive practices in either snail control or human chemotherapy in the two provinces ., Thus , whether there is an effect of geographical isolation on populations of S . japonicum in this mountainous area will likely remain unknown , and whether the observed low level of genetic variation in these populations resulted from a recent bottleneck effect as a consequence of intensive control practices may also remain to be answered ., Ecological habitats were thought to affect population genetic diversity of S . japonicum in mainland China 40 ., The mountainous habitats in Sichuan and Yunnan provinces may differ obviously from the habitats for the intermediate host in the ML reaches , in several aspects such as in hydrology , altitude and soil etc ., 41 , 42 , but the difference should mostly be attributed to the geographical separation , rather than a simple impact from habitat difference ., In the ML reaches of Yangtze River , it was impossible to clarify any patterns of haplotype clustering in relation to types of sample localities or to provinces , as haplotypes from a single locality were generally clustered in different clades ., It can thus be speculated that S . japonicum might have experienced frequent gene flows in most populations in this region ( Table 3 ) ., The localities for O . hupensis in the ML reaches have extensive physical connections through channels with the Yangtze River ., With frequent occurrence of floods in the Yangtze River basin , especially in its ML reaches , snails in these habitats can be dispersed and subsequently deposited widely in various localities , and this naturally occurred instance was , in a previous research , proposed to explain the high genetic diversity of O . hupensis in the ML reaches 16 ., It was further considered that this distinct genetic diversity in snail intermediate hosts may have strong implications in genetic diversity of schistosomes in mainland China 16 , as demonstrated clearly in the present study ., The accumulation of mixed sources of snails , especially infected snails can reconstitute the parasite population , leading to the existence of various haplotypes within a single population , and also to the limited degree of genetic distance between populations in the ML reaches as observed in the present study , which supports the speculation by Davis et al . 43 that floods may be the cause of the widespread mixing and dispersal of snails , leading to greater genetic diversity in O . hupensis populations along the Yangtze River plains compared with populations in SW China ., Surprisingly , the number of haplotypes , being 80 and 13 for the intermediate host snails in the ML reaches , and in Sichuan and Yunnan provinces 16 , matches roughly , if not coincidently , with the number of haplotypes , 88 and 8 , for S . japonicum in the ML reaches and in Sichuan and Yunnan provinces in this study , respectively ., The intermediate host snails and the schistosome in China exhibit a lesser degree of genetic diversity in the SW , but a relatively larger degree in the ML reaches of the Yangtze River , as reported in a previous study on the intermediate host snails 16 and in this study ., No shared haplotypes were observed either in the intermediate host snails or in the schistosomes between localities from the ML reaches and from the SW ., Zhao et al . 44 recently reported that the intermediate host snails O . hupensis robertsoni in Sichuan and the snail O . hupensis hupensis in the ML reaches had a 10 . 3% genetic distance , strongly indicating that the two subspecies may differ at the species level ., In a phylogenetic study on the Schistosomatidae , Lockyer et al . 45 considered that schistosomes in east Asia and their intermediate hosts in the Pomatiopsidae may be considered as the only co-evolutionary model between schistosomes and their intermediate host snails ., Davis et al . 46 also speculated , as snail population forms have diverged genetically , so must their associated schistosomes or else become regionally extinct ., However , it would be only possible to examine such relationship if the intermediate host snails and schistosomes are collected from a large geographical range in east Asia ., In a very small-scale area in Anhui province of China , Rudge et al . 40 detected strong genetic differentiation in S . japonicum between two types of habitats , lake/marshland region and hilly region , and suggested that contrasting host reservoirs may be associated with the genetic differentiation , with rodents and dogs being important infection reservoirs in hilly regions and bovines in lake/marshland regions ., On the other hand , they found little or no parasite genetic differentiation among host species within most villages; but in another study , Wang et al . 47 reported that schistosomes were separated into two clades representing the parasites from different definitive hosts ., It seems likely that S . japonicum has undergone genetic differentiation in a relatively small-scale area , as in a large geographical region reported in this study ., In the above two studies , miracidia from definitive hosts were examined with microsatellite markers ., In the present study , adult parasites were obtained through infecting mice with cercariae ., As definitive host-based genetic variation in S . japonicum has been noticed 40 , 47 , the selection pressure through definitive host may need to be further investigated ., Unexpectedly , three haplotypes representing some schistosomes from three localities , each in Hubei , Hunan , Anhui provinces , were actually clustered together within another clade containing all haplotypes from Sichuan and Yunnan provinces ., It is , however , at present impossible to explain this mixed cluster ., As the movement of people has been frequent in China 48 , the possible transmission through definitive host cannot be ruled out as a possible interpretation ., In conclusion , substantial genetic diversity was demonstrated in populations of S . japonicum in schistosomiasis endemic areas in mainland China ., Overall , a significant correlation was observed between the genetic distance and the geographical distance among the populations ., It is apparent that the populations from Sichuan and Yunnan provinces in SW China exhibited a relatively low level of genetic variation , and were genetically different from the populations in the ML reaches of the Yangtze River , which had a much more complicated genetic diversity ., Such obvious genetic diversity should be taken into consideration in guiding any strategic control programmes and/or vaccine development/trials in the future . | Introduction, Materials and Methods, Results, Discussion | Schistosoma japonicum still causes severe parasitic disease in mainland China , but mainly in areas along the Yangtze River ., However , the genetic diversity in populations of S . japonicum has not been well understood across its geographical distribution , and such data may provide insights into the epidemiology and possible control strategies for schistosomiasis ., In this study infected Oncomelania snails were collected from areas in the middle and lower ( ML ) reaches of the Yangtze River , including Hubei , Hunan , Anhui , Jiangxi and Jiangsu provinces , and in the upper reaches of the river , including Sichuan and Yunnan provinces in southwest ( SW ) China ., The adult parasites obtained from experimentally infected mice using isolated cercariae were sequenced individually for several fragments of mitochondrial regions , including Cytb-ND4L-ND4 , 16S-12S and ND1 ., Populations in the ML reaches exhibited a relatively high level of diversity in nucleotides and haplotypes , whereas a low level was observed for populations in the SW , using either each single fragment or the combined sequence of the three fragments ., Pairwise analyses of F-statistics ( Fst ) revealed a significant genetic difference between populations in the ML reaches and those in the SW , with limited gene flow and no shared haplotypes in between ., It is rather obvious that genetic diversity in the populations of S . japonicum was significantly correlated with the geographical distance , and the geographical separation/isolation was considered to be the major factor accounting for the observed difference between populations in the ML reaches and those in the SW in China ., S . japonicum in mainland China exhibits a high degree of genetic diversity , with a similar pattern of genetic diversity as observed in the intermediate host snails in the same region in China . | Despite the existing threat of schistosomiasis in some rural areas along the Yangtze River , the genetic diversity of Schistosoma japonicum has not been investigated across its wide geographical distribution in China , and such information may provide insight into the disease epidemiology and the development of its control measures ., In this study , the adult parasites , obtained through infecting mice with cercariae from snails of the genus Oncomelania collected from a wide range of localities in currently endemic areas of schistosomiasis in the middle and lower ( ML ) reaches of the Yangtze River , and in Sichuan and Yunnan provinces in the upper reaches of the river in southwest ( SW ) China , were sequenced individually for mitochondrial genes ., In general , a relatively high degree of genetic variation was observed in populations in the ML reaches in terms of nucleotide and haplotype diversity , but a low level was observed in populations in the SW ., The significant difference in genetic diversity as revealed by F-statistics , and the existence of no shared haplotypes , were observed between populations in the ML reaches and those in the SW , indicating the effect of geographical separation/isolation upon the schistosomes and probably the parasite-snail system in China . | zoology, ecology, genetics, biology, genomics, evolutionary biology, population biology, genetics and genomics | null |
journal.pntd.0002357 | 2,013 | Symptomatic Versus Inapparent Outcome in Repeat Dengue Virus Infections Is Influenced by the Time Interval between Infections and Study Year | Dengue is a major health problem globally , with more than 40% of the worlds population at risk and over a hundred countries affected by epidemics 1 ., In the past 50 years , the incidence of dengue has increased considerably , affecting tens of millions of people annually ., Dengue is caused by an enveloped , positive-sense RNA virus in the genus Flavivirus of the Flaviviridae family , which is transmitted by mosquitoes of the Aedes genus ., There are four serotypes of dengue virus ( DENV ) : DENV-1 , DENV-2 , DENV-3 and DENV-4 ., Infection with DENV can be subclinical ( inapparent infection ) or cause a variety of clinical manifestations ranging from undifferentiated illness and Dengue Fever ( DF ) to severe life-threatening syndromes Dengue Hemorrhagic Fever ( DHF ) and Dengue Shock Syndrome ( DSS ) 2 ., Very little is known about the determinants of inapparent versus symptomatic DENV infection outcome ., By definition , inapparent infections are not detected in routine surveillance and can only be captured in the context of prospective cohort or index cluster studies ., In a cohort study in Thailand , study year , total DENV infection incidence in the current and previous year , circulating DENV serotype and the number of circulating serotypes were identified as factors influencing inapparent versus symptomatic infection outcome 3 , 4 ., Analysis of infection outcome is further complicated by immune responses to multiple infections with different DENV serotypes , which can be either protective or pathogenic ., Early experimental studies in DENV-naïve healthy volunteers showed that infection with one DENV serotype confers immunity to that particular serotype for up to 18 months 5 ., In fact , this protection is thought to be life-long ., On the other hand , infection with one serotype only conferred short-term ( <2 months ) complete protection against heterologous infection with a different serotype 5 ., In Sabins studies , heterologous protection waned over a period of several months ., Heterologous protection after a short interval but not after a longer period of time was also observed in rhesus monkeys , depending on the serotype sequence 6 ., In contrast , secondary heterologous infection is well documented as the single most important risk factor for severe dengue 7–11 ., Epidemiological data from dengue epidemics in Cuba also suggest that longer time intervals between infections might increase disease severity 11 ., In 1977 , DENV-1 caused the first dengue epidemic in the country ., This was followed by two DENV-2 epidemics caused by similar strains in 1981 and 1997 , respectively 12 , 13 ., Interestingly , death rates were significantly higher in 1997 compared to 1981 12 ., Altogether , these observations highlight the intricate interplay between host immunity and repeat DENV infections and suggest that the time between two consecutive infections is an important factor in infection outcome ., Few studies have compared inapparent versus symptomatic outcome in primary and secondary DENV infections ., In one of the first prospective dengue cohort studies in Bangkok , Thailand 9 , and in a multinational index cluster study with four sites in South-East Asia and Latin America 14 , the inapparent-to-symptomatic ratio was similar in primary and secondary infections ., We also previously reported similar ratios in primary and secondary DENV infections in Managua , Nicaragua 15 ., However , an index cluster study conducted in Kamphaeng Phet , Thailand , found very few symptomatic dengue cases among primary infections when compared to secondary infections , albeit the overall number of infections reported in the study was limited 16 ., Even less is known about the impact of second , third or fourth DENV infections ( collectively referred to as “secondary infections” ) on inapparent versus symptomatic outcome ., In fact , few reports exist in the literature of third and fourth DENV infections 17 , 18 ., In a hospital-based retrospective study , third and fourth DENV infections were estimated to present a lower risk of hospital admission 19 ., However , once hospitalized , the risk of DHF/DSS in third and fourth DENV infections was not different from that in second DENV infections 19 ., In Nicaragua , the first dengue epidemic was reported in 1985 and caused by DENV-1 and DENV-2 20 ., Several DENV-1 , 2 and 4 outbreaks occurred in the early 1990s , followed by a large DENV-3 epidemic in 1994–5 21 ., Since then , all four serotypes circulate , but in contrast to hyperendemic areas , one serotype is dominant each season 22–24 ., The dengue season starts after the first rains , with most cases occurring from August to January 15 ., However , some cases are detected throughout the year ., In 2004 , we established the community-based , prospective Pediatric Dengue Cohort Study ( PDCS ) in Managua , Nicaragua 25 ., Here , we analyzed serological data from all cohort participants , as well as neutralizing antibody titers in a subset of children who had experienced repeat DENV infections , using 8 annual healthy blood sample collections ., We combined these results with data about dengue cases in the PDCS from 7 dengue seasons to investigate the determinants of inapparent versus symptomatic DENV infection outcome ., In particular , we evaluated the impact of factors that can only be analyzed in the context of long-term cohort studies such as infection number and the time interval between infections in children with documented repeat DENV infections ., This study was approved by the Institutional Review Boards of the Nicaraguan Ministry of Health and the University of California , Berkeley ., Parents or legal guardians of all subjects provided written informed consent , and subjects 6 years of age and older provided assent ., In August of 2004 , a community-based pediatric dengue cohort study was established in District II of the capital city of Managua , a low-to-middle income area with a population of approximately 62 , 500 25 ., This ongoing study is based at the municipal Health Center Sócrates Flores Vivas ( HCSFV ) , which is the principal source of primary health care for the districts population ., Initially , participants aged 2–9 years were recruited through house-to-house visits; over time , the age range of the study was extended to 2 to 14 years of age ., Each year , additional children were enrolled to maintain the age structure of the cohort 25 ., Participants were encouraged to seek medical care for all illnesses through study physicians and in particular , to present early in case of a febrile episode ., Cohort participants were followed closely for all illnesses , and study physicians classified participants into febrile illnesses that met the WHO dengue case definition ( category A ) 2 , undifferentiated fever ( category B ) , fever with an apparent focus other than dengue ( category C ) , and non-febrile episode ( category D ) ., Children who met WHO criteria for suspected dengue ( category A ) as well as those with undifferentiated fever ( category B ) were evaluated for acute DENV infection 15 , 25 ., The cohort was sized such that even in years of relatively low DENV transmission , a minimum number of symptomatic cases would be identified ., A suspected dengue case was considered a symptomatic DENV infection when, 1 ) DENV RNA was detected by reverse-transcriptase polymerase chain reaction ( RT-PCR ) 26 , 27 ,, 2 ) DENV was isolated 26 ,, 3 ) seroconversion was observed in paired acute and convalescent phase sera by IgM capture ELISA 26 , 28 , or, 4 ) seroconversion and/or a ≥4-fold increase in total DENV-specific antibody titer in paired acute and convalescent sera was observed by Inhibition ELISA 29 , 30 ., Inapparent DENV infections were identified through serological testing of paired annual blood draws from healthy subjects 15 , 25 ., Participants whose paired annual samples demonstrated seroconversion or a 4-fold or greater increase in total DENV-specific antibody titer by Inhibition ELISA , but who had not experienced a documented febrile episode associated with acute DENV infection , were considered to have experienced an inapparent DENV infection 15 , 25 ., To evaluate the effectiveness of capture of febrile cases , yearly participant surveys were conducted ( Table S1 ) ., Overall , surveys showed that only 1 . 9% of the participants reported having a fever and attending a different healthcare provider and 2 . 3% reported not attending any medical provider ., Both symptomatic and inapparent DENV infections were assigned a dengue season whose limits were defined by the healthy annual blood collection ., As a specific date cannot be assigned to inapparent DENV infections , since by definition the infection is inapparent and thus not reported to the study health center , the inapparent infection date was assumed to be October 1st , during the peak of the corresponding season ., For consistency , the same procedure was followed for symptomatic infections ., Raji-DC-SIGN cells ( kind gift from B . Doranz , Integral Molecular , Philadelphia , PA ) were used for all neutralization experiments ., Cells were grown at 37°C at 5% CO2 in RPMI medium supplemented with 10% ( v/v ) Fetal Bovine Serum ( FBS ) , 1% ( v/v ) penicillin-streptomycin , and 0 . 1 M HEPES ( RPMI complete medium ) ., DC-SIGN ( CD209 ) expression was quantified by flow cytometry using a monoclonal antibody ( PerCP-Cy5 . 5 Mouse Anti-Human CD209 , BD Biosciences ) , and cell lots were used only if >95% of the cells were positive for DC-SIGN ., DENV Reporter Viral Particles ( RVP; DENV-1 , Western Pacific 74; DENV-2 , S16803; DENV-3 , CH53489; DENV-4 , TVP360 ) containing a GFP reporter RNA were produced by Integral Molecular as previously described 31 , 32 ., RVPs were stored at −80°C , and for experiments , were thawed rapidly in a water bath and kept on ice before use ., For each RVP lot , the optimal working dilution was determined ., Briefly , RVPs were serially diluted 2-fold in RPMI complete medium adjusted to pH 8 . 0 with 5 M NaOH ., Infection was carried out in a 96-well plate by mixing , in each well , 50 µl of diluted RVPs with 40 , 000 Raji DC-SIGN cells in a total volume of 100 µl complete RPMI media ., The cells were then incubated at 37°C in 5% CO2 for 48 hours and subsequently fixed in 2% paraformaldehyde ., The percentage of infected , GFP-expressing cells was determined by flow cytometry ( Becton-Dickinson LSRII or Beckman Coulter Epics XL-MCL ) using FlowJo version 7 . 2 . 5 ( TreeStar Software , Ashland , OR ) ., The highest RVP dilution yielding an infection rate of 7 to 15% was used for subsequent neutralization assays 32 ., RVP neutralization assays were performed as previously described 32 ., Briefly , RVPs were prepared according to the previously determined working dilution in a final volume of 25 µl of RPMI pH 8 . 0 complete medium ., RVPs were then mixed with an equal volume of serum ( eight 3-fold serial dilutions in RPMI pH 8 . 0 complete medium starting at 1∶5 , tested in duplicate ) in 96-well plates and incubated on a shaker for 1 hour at room temperature ., Infections were carried out as described above ., The percentage of infected , GFP-positive cells for each serum concentration was plotted as percent infection versus log10 of the reciprocal serum dilution using Prism 5 . 0 ( GraphPad , La Jolla , CA ) ., A sigmoidal dose response curve with a variable slope was then generated to determine the 50% neutralization titer , or NT50 – the serum dilution at which a 50% reduction in infection was observed compared to the positive ( no-serum ) control ., Background GFP levels were subtracted from all measurements using a negative control ( no-RVP ) ., Neutralization curves using reference sera ( polyvalent anti-DENV-1+2+3+4 serum code 02/186 , National Institute for Biological Standards and Control , UK ) were performed with serial 2-fold dilutions of all RVP lots to ensure that viral particles were neutralized according to the law of mass action 32 , 33 , such that serial dilutions of RVPs yielded the same NT50 , thus ensuring that the antibodies in the serum were in excess ., Polyvalent serum was used in each neutralization assay to confirm neutralization against all 4 RVPs ( neutralization control ) ., The RVP assay was standardized both at UC Berkeley and in Nicaragua ., For each NT50 result , the absolute sum of squares ( ABSS ) and the coefficient of determination ( R2 ) of the non-linear regression were calculated ., If the ABSS was >0 . 2 and/or the R2 was <0 . 9 , the data were excluded and repeated ., An NT50 of <10 indicates a calculated NT50 value of <10 or the failure of the sera to neutralize at the lowest dilution by at least 50% ., NT50 titers were independently calculated by two analysts ., Thirty-nine participants who entered the cohort dengue-naïve and had experienced at least two DENV infections as determined by total antibody titer measurements ( ELISA ) were selected ., As with antibody titration by ELISA , we used annual healthy serum samples and determined the NT50 for all four DENV serotypes ., All participants in this subset had entered the cohort between 2004 and 2007 , and annual samples through 2011 were used , except for participants withdrawn from the study before then ., The following rules for interpretation of the longitudinal NT50 data were established and implemented ., For participants who had no evidence of a previous DENV infection ( i . e . , NT50 titers for all 4 DENV serotypes in all previous years were <10 ) , primary DENV infections were defined by seroconversion ( from NT50<10 to NT50≥10 ) to a specific serotype ., For participants with evidence of prior DENV infection , secondary DENV infections were defined by seroconversion ( from NT50<10 to NT50≥40 ) or a ≥4-fold increase in NT50 ( fold-change was calculated as post-infection NT50/pre-infection NT50 ) ., When several serotypes met the infection criteria during the same study year , the serotype with the highest NT50 fold-change was chosen ., If the fold-change for more than one serotype was similar ( ±15% ) , an infection was assigned to the year but the serotype was recorded as unknown ., If a symptomatic DENV infection with a given serotype was identified , no other infection with the same serotype was assigned throughout the years ., If an inapparent DENV infection was identified , no other inapparent infection with the same serotype was assigned in later years ., Interpretation of the DENV infection history of each participant over time was discussed by six authors to reach a consensus ., For determination of the proportion of symptomatic DENV infections among total DENV infections , we only included symptomatic infections identified in participants who completed the study year and for whom paired annual samples were available ( 404 out of 448 symptomatic DENV infections ) ., Statistical analyses were performed in STATA , version 12 ( StataCorp LP , College Station , TX ) ., The binomial test was used to assess the distribution of DENV infections by sex ., Chi-square and Fishers exact tests were used to compare categorical variables among two ( or more ) independent groups ., The Mann-Whitney U test was used to compare intervals between consecutive DENV infections ., A total of 5 , 541 children participated in the Pediatric Dengue Cohort Study from August 2004 to March 2011: 3 , 713 were enrolled at the onset of the study and 1 , 828 in subsequent years ., We identified DENV infections during this period , corresponding to 7 dengue seasons ., First , participants who met the WHO criteria for a suspected dengue case 2 and those with undifferentiated fever were evaluated for acute symptomatic DENV infection using molecular , virological , and serological diagnostic techniques ( see Methods ) ., Second , inapparent DENV infections were identified using total DENV-specific antibody titers measured by Inhibition ELISA 29 , 30 in healthy annual blood samples from 8 annual collections ( 2004–2011 ) ., The average number of annual samples contributed per participant was 5 . 3±2 . 1 ( Fig . S1A ) ., DENV infections were stratified by study year; each year was delimited by two consecutive annual blood sample collections and encompassed a dengue season ., Moreover , sequential first , second and third DENV infections were identified in participants who entered the study with no detectable anti-DENV antibodies ( “naïve” ) ., As relatively few third infections were detected , an additional category was created to study post-second DENV infections by including, 1 ) third infections in naïve participants , and, 2 ) second and third infections experienced by children who entered the study with anti-DENV antibodies ( “non-naïve” ) ., To identify first , second , third and post-second infections , participants who contributed two or more consecutive annual samples were included ( N\u200a=\u200a5 , 082 ) ., The average number of consecutive samples provided by these participants was 5 . 6±2 . 1 ( Fig . S1B ) ., The average time interval between consecutive samples was 343±41 days ( Fig . S1C ) ., Overall , we identified 448 symptomatic and 1 , 606 inapparent DENV infections ( Table 1 ) ., Both symptomatic and inapparent infections were equally distributed by gender ., However , repeat DENV infections tended to be more frequent in males ( chi-square test p\u200a=\u200a0 . 060 ) ( Table 1 ) ., We then analyzed the proportion of symptomatic DENV infections among all DENV infections ., For this analysis , only participants with symptomatic DENV infections who had completed the study year were included ( n\u200a=\u200a404 ) ., The proportion of symptomatic DENV infections among all DENV infections was similar in females ( 20 . 8% ) and males ( 19 . 4% , chi-square test p\u200a=\u200a0 . 447 ) ., The mean age of infection was significantly higher ( p<0 . 001 ) , by 1 . 2 years , in symptomatic infections when compared to inapparent DENV infections ( Table 1 ) ., We first examined the proportion of symptomatic infections among all DENV infections per study year ., This proportion showed substantial differences , ranging from 4 . 9% in 2006–07 to 39 . 1% in 2009–10 ( “All infections” bars , Fig . 1A–G ) ., Then , we analyzed the effect of infection number ( first , second , third and post-second ) on inapparent versus symptomatic DENV infection outcome ., For each study year , trend analyses were performed with first , second and post-second DENV infections , as the number of third infections was limited ., For all study years but one , the proportion of symptomatic DENV infections was similar in first , second , and post-second infections ( Fishers exact test , p>0 . 05 , Fig . 1A–G ) ., In 2008–09 , no symptomatic second infections and very few symptomatic post-second infections were identified when compared to symptomatic first infections ( Fishers exact test , p\u200a=\u200a0 . 003 ) ( Fig . 1E ) ., Overall , this analysis suggests that , in this study , inapparent versus symptomatic outcome is similar in first , second and post-second DENV infections ., For participants with repeat DENV infections , we then examined whether symptomatic versus inapparent outcome of a prior infection influences outcome of a subsequent infection ., To this end , the proportion of symptomatic infections was calculated given the outcome of the previous infection ., No significant difference was observed , as the proportion of symptomatic DENV infection was 24 . 9% when the previous infection was inapparent ( N\u200a=\u200a293 ) and 23 . 5% when the previous infection was symptomatic ( N\u200a=\u200a34 ) ( chi-square test p\u200a=\u200a0 . 859 ) ., We then evaluated the effect of the time interval between infections on repeat DENV infections ., The interval between two consecutive infections was defined as the number of seasons between the infections ., For instance , the interval between an infection in 2005–06 and another infection in 2008–09 is 3 years ., In total , 341 intervals between DENV infections were calculated ., The mean interval was 2 . 4 years ., Next , we stratified the intervals between infections with respect to the outcome of both the prior and the subsequent infection ., Four different infection sequences were thus defined: an inapparent DENV infection followed by another inapparent infection ( inapparent-to-inapparent ) or by a symptomatic infection ( inapparent-to-symptomatic ) , and a symptomatic DENV infection followed by an inapparent infection ( symptomatic-to-inapparent ) or another symptomatic infection ( symptomatic-to-symptomatic ) ., The mean interval was calculated for each of the four groups ( Fig . 2A ) ., Notably , the inapparent-to-inapparent infection mean interval was significantly shorter than the inapparent-to-symptomatic infection interval ( 2 . 2 versus 2 . 7 years , Mann-Whitney U test p\u200a=\u200a0 . 021 ) ; all other pairwise comparisons were not significant ., We further stratified the infection sequences by infection number ., Specifically , for participants who entered the cohort dengue-naïve , infection sequences were divided into “first-to-second” and “second-to-third” DENV infections ., In the “first-to-second” group , the inapparent-to-inapparent infection interval was again significantly shorter than the inapparent-to-symptomatic infection interval ( 1 . 8 versus 2 . 6 years , Mann-Whitney U test p\u200a=\u200a0 . 018 ) ( Fig . 2B ) ., The other pairwise comparisons were not significant ., The symptomatic-to-symptomatic infection sequences were not included in the analysis as no “second-to-third” such sequence was observed ., Interestingly , no difference was observed when comparing inapparent-to-inapparent and inapparent-to-symptomatic infection intervals for “second-to-third” infection sequences ( 2 . 7 versus 2 . 5 years , p\u200a=\u200a0 . 692 ) ., Moreover , the inapparent-to-inapparent infection interval was significantly shorter in “first-to-second” ( 1 . 8 years ) than in “second-to-third” infection sequences ( 2 . 7 years , Mann-Whitney U test p\u200a=\u200a0 . 005 ) ., However , this observation was limited by the small number of “second-to-third” infections sequences analyzed ( 11 inapparent-to-inapparent and 13 inapparent-to-symptomatic ) ., To extend this observation , we created a new group of infection sequences by adding to the “second-to-third” sequences those infections observed in participants who entered the cohort non-dengue-naïve ., This new group was termed “other infection sequences” as it includes all possible DENV infection sequences except the “first-to-second” infection group ., Notably , no difference was observed between the inapparent-to-inapparent and inapparent-to-symptomatic infection intervals within this group ( Fig . 2C ) ., Furthermore , when comparing the inapparent-to-inapparent infection interval between the “first-to-second” and the “other infection sequences” groups , the former was found to be significantly shorter ( 1 . 8 versus 2 . 7 years , Mann-Whitney U p<0 . 001 ) ( Fig . 2B–C ) ., The symptomatic-to-symptomatic infection sequences were not included in this analysis due to the small number of observations ( “first-to-second” N\u200a=\u200a5; “other infection sequences” N\u200a=\u200a5 ) ., Taken together , these show that the interval between two inapparent infections is significantly shorter than the inapparent-to-symptomatic infection interval , but only when considering the first and second DENV infections of a given participant ., We then undertook a longitudinal analysis of DENV serotype-specific neutralizing antibody titers in a subset of cohort participants ., The objective of this analysis was to examine the feasibility of reconstructing participants DENV immune history using a Reporter Viral Particle ( RVP ) flow cytometry-based DENV neutralization assay 32 and to substantiate the results obtained with Inhibition ELISA by measuring neutralizing antibodies instead of total anti-DENV antibodies ., This assay yields reproducible serotype-specific neutralization titers that are in agreement with Plaque Reduction Neutralization Test ( PRNT ) results 32 ., First , we examined the ability of the 50% neutralization titer ( NT50 ) changes between pre- and post-infection annual samples to detect symptomatic DENV infections and to identify the correct DENV serotype in a subset of 27 confirmed symptomatic infections with serotype information available from RT-PCR and/or virus isolation ., The pre- to post-infection fold-change in NT50 was calculated for each DENV serotype ., Using the highest NT50 fold-change as an indicator , 26 out of 27 DENV serotypes were correctly identified ( Fig . S2 ) ., In one additional case ( participant M , Fig . S2 ) , taking into account the participants immune history allowed for the identification of the infecting serotype ( DENV-3 ) ., In this case , the participant had experienced an inapparent infection with DENV-2 prior to the symptomatic episode ., The NT50 fold-change was highest for DENV-2 but , consistent with the interpretation rules we had established , the infecting serotype was recorded as DENV-3 , which had the second highest NT50 increase ., Second , we analyzed longitudinal data from 39 cohort participants to determine their DENV-specific immunological history by compiling symptomatic and inapparent DENV infections as detected in consecutive annual samples ( see Methods for specific rules ) ., Longitudinal NT50 titers for two participants are shown in Figure 3 ., Both participants displayed an NT50<10 against all 4 serotypes in their initial sample and were therefore considered dengue-naïve ., Participant A apparently experienced an inapparent DENV-2 infection in 2005–06 followed by an inapparent DENV-4 infection in 2006–07 ., Subsequently , NT50 titers did not display any major changes until 2010 , when titers for all four serotypes increased more than 4-fold ., However , the most likely infecting serotype was determined to be DENV-3 as the increase in NT50 against DENV-3 was the greatest , aside from DENV-2 , which had caused the first infection ., In fact , this participant experienced a symptomatic DENV-3 infection in 2009–10 as determined by RT-PCR and viral isolation using acute and convalescent samples from the period of illness ., Participant B experienced 3 inapparent DENV infections: DENV-1 in 2005–06 , DENV-2 in 2007–08 and DENV-3 in 2009–10 ., Overall , 75 inapparent DENV infections were detected among the 39 participants analyzed ( Table S2 ) ., For most infections ( 73/75 ) , the likely infecting serotype was identified ., For the remaining two , a comparable fold-change in NT50 titers was observed for two serotypes , making it difficult to assign an infecting serotype ., Finally , we compared DENV serotype circulation in each study year as determined by neutralization assay using annual samples to symptomatic DENV infections detected in the entire cohort by RT-PCR and/or virus isolation ., Serotype circulation was similar using both approaches , showing that the circulating serotype ( s ) cause both inapparent and symptomatic DENV infections and further validating the neutralization method ( Fig . S3 ) ., The only striking difference was DENV-4 circulation in 2006–07 , 2007–08 and 2009–10 , which only caused inapparent infections ., These data are consistent with limited PRNT data that we obtained as part of a study of DENV neutralizing antibodies in a random 10% of the cohort from 2004 to 2007 and in a subset of inapparent infections in different individuals each year from 2004 to 2008 , where inapparent DENV-4 infections were also identified in 2006–07 and 2007–08 ( M . J . Vargas , A . Balmaseda , E . Harris , unpublished results ) ., Using the same approach as for total antibody titers above , the intervals between consecutive DENV infections were determined in the subset of cohort participants examined using the neutralization assay ., The mean interval between two DENV infections was 2 . 41 years ( N\u200a=\u200a54 ) ., Despite the fact that the neutralization titer dataset contained approximately 6 times fewer consecutive infection sequences than the ELISA dataset from the entire cohort , the value obtained in the neutralization subset was similar to the mean interval determined using total antibody titer ( 2 . 35 years ) ., We then stratified the infection sequences by infection outcome and infection number ., Only inapparent-to-inapparent and inapparent-to-symptomatic infection sequences were compared , as the number of symptomatic-to-inapparent infections was small ( N\u200a=\u200a4 ) and no symptomatic-to-symptomatic infection sequences were observed ., When comparing all intervals , the inapparent-to-inapparent infection interval was significantly shorter than the inapparent-to-symptomatic infection interval ( Mann-Whitney U test p\u200a=\u200a0 . 025 ) ( Fig . 4A ) ., However , when we stratified by infection number , this difference was only observed in “first-to-second” subset ( Mann-Whitney U test p\u200a=\u200a0 . 003 , Fig . 4B ) and not when considering “second-to-third” infection sequences ( Fig . 4C ) ., These results corroborate the findings obtained with consecutive DENV infection interval using total antibody titers in the entire cohort ., In this study , we analyzed several determinants of inapparent versus symptomatic DENV infection , taking advantage of our long-term Pediatric Dengue Cohort Study in Managua , Nicaragua ., Data from 1 , 606 inapparent and 448 symptomatic DENV infections were collected over 7 years using annual total anti-DENV antibody titers as measured by Inhibition ELISA and “enhanced” passive surveillance of febrile cases , respectively ., Overall , symptomatic DENV infections were equally distributed by gender but more frequent in older children ., The proportion of symptomatic DENV infections among all DENV infections varied substantially across study years but was not significantly affected by infection number ( i . e . , first , second , third , or post-second infections ) ., In participants with documented repeat DENV infections , the outcome of a previous DENV infection did not influence the outcome of the subsequent infection; however , the time interval between two consecutive infections did ., In fact , the interval between two inapparent DENV infections was significantly shorter that the interval between an inapparent and a symptomatic infection ., However , this result was only observed when considering the first and second DENV infections of a given participant ., Moreover , this finding was confirmed using a flow cytometry-based neutralization assay to quantify serotype-specific anti-DENV neutralizing antibodies in a subset of cohort participants ., The proportion of symptomatic DENV infections among total infections was found to be similar in females and males , consistent with observations in other studies 3 , 14 ., However , age played a role in influencing symptomatic outcome , as symptomatic DENV infections tended to occur more frequently in older children ., Interestingly , this effect was not observed in the Kamphaeng Phet ( Thailand ) cohort 3 ., The most striking determinant of infection outcome was the study year ., We had previously reported large variations in the proportion of symptomatic DENV infections in the first four dengue seasons of the Pediatric Dengue Cohort Study ( 2004–05 to 2007–08 ) 15 ., Here , we extended this analysis through 2010–11 and found even more dramatic variations , from ∼5–6% in 2004–05 and 2006–07 to almost 40% in 2009–10 and 2010–11 ., Similar temporal variations have been reported in other studies in Peru 34 and Thailand 3 , 4 , 35 ., The factor ( s ) driving these differences in our Nicaraguan cohort are not completely known , although in 2007–08 a clade replacement within DENV-2 is thought to have contributed to the higher proportion of symptomatic infections 24 , and in 2009–10 the concurrent H1N1 influenza pandemic may have played a role 23 ., Overall , we did n | Introduction, Methods, Results, Discussion | Four dengue virus serotypes ( DENV1-4 ) circulate globally , causing more human illness than any other arthropod-borne virus ., Dengue can present as a range of clinical manifestations from undifferentiated fever to Dengue Fever to severe , life-threatening syndromes ., However , most DENV infections are inapparent ., Yet , little is known about determinants of inapparent versus symptomatic DENV infection outcome ., Here , we analyzed over 2 , 000 DENV infections from 2004 to 2011 in a prospective pediatric cohort study in Managua , Nicaragua ., Symptomatic cases were captured at the study health center , and paired healthy annual samples were examined on a yearly basis using serological methods to identify inapparent DENV infections ., Overall , inapparent and symptomatic DENV infections were equally distributed by sex ., The mean age of infection was 1 . 2 years higher for symptomatic DENV infections as compared to inapparent infections ., Although inapparent versus symptomatic outcome did not differ by infection number ( first , second or third/post-second DENV infections ) , substantial variation in the proportion of symptomatic DENV infections among all DENV infections was observed across study years ., In participants with repeat DENV infections , the time interval between a first inapparent DENV infection and a second inapparent infection was significantly shorter than the interval between a first inapparent and a second symptomatic infection ., This difference was not observed in subsequent infections ., This result was confirmed using two different serological techniques that measure total anti-DENV antibodies and serotype-specific neutralizing antibodies , respectively ., Taken together , these findings show that , in this study , age , study year and time interval between consecutive DENV infections influence inapparent versus symptomatic infection outcome , while sex and infection number had no significant effect ., Moreover , these results suggest that the window of cross-protection induced by a first infection with DENV against a second symptomatic infection is approximately 2 years ., These findings are important for modeling dengue epidemics and development of vaccines . | The four serotypes of the mosquito-borne dengue virus ( DENV ) infect an estimated 100 million humans annually , resulting in tens of millions of dengue cases and hundreds of thousands of cases of severe disease ., However , infection with DENV does not always lead to clinical signs , and a large proportion of DENV infections are inapparent ., Here , we studied the factors that influence whether a DENV infection is inapparent or symptomatic ., Data from over 2 , 000 DENV infections ( ∼1 , 600 inapparent and ∼400 symptomatic ) were collected during 7 years from an ongoing prospective cohort study of children in Managua , Nicaragua ., We show that whether a person is infected for the first , the second , or the third time with different DENV serotypes , the proportion of symptomatic infections is similar ., However , the proportion of symptomatic infection varied substantially across study years , and symptomatic infections tended to happen in older children when compared to inapparent infections ., We also show that if a second DENV infection happens within a period of ∼2 years after the first infection , the second infection is more likely to be inapparent ., However , if the time interval between first and second DENV infections is longer , this protection wanes and the infection is likely to be symptomatic ., These findings are important for the modeling of dengue epidemics and the development of new vaccines . | medicine, infectious diseases, infectious disease epidemiology, epidemiology, dengue fever, neglected tropical diseases | null |
journal.pntd.0001634 | 2,012 | Treatment for Schistosoma japonicum, Reduction of Intestinal Parasite Load, and Cognitive Test Score Improvements in School-Aged Children | Many children in developing countries perform below academically desired levels 1 ., Helminth infections are a pervasive part of childrens environments in these settings that may contribute to poor educational outcomes through reduced iron status , inflammation , decreased macro-nutrient nutritional status , and distracting symptoms such as abdominal pain 2 , 3 ., Some epidemiologic studies have linked these infections to low academic achievement in resource-limited settings 4–7 ., However , many of the studies did not control for important confounders or had methodological differences that made comparability of findings across studies difficult 8 ., All but two prior studies 9 , 10 examined associations between cognitive performance and single helminth species ., Recently , polyparasitism , that is , the concurrent multi-species helminth infection , has been associated with childhood anemia and self-reported morbidity 11–13 ., Its relationship to performance in cognitive tests deserves specific investigation 8 ., An earlier cross-sectional study by our group found that moderate or higher intensity infection with Trichuris trichiura , Ascaris lumbricoides , and Schistosoma japonicum were , respectively , associated with low scores on tests of verbal fluency , and the memory and learning subscales of the Wide Range Assessment of Memory and Learning ( WRAML ) tests in school-aged children 14 ., It is expected that treatment for parasitic helminth infections will confer a range of benefits to child health , including improvements in academic performance among heavily infected children 15 ., However , empirical support for this claim is lacking 8 ., Short follow-up periods for most randomized controlled trials , variability in prevalence and baseline intensities of helminth infections , and a background of high re-infection pressure could explain failure to consistently find treatment-associated score improvements ., The ambiguity in the literature justifies further exploration of this subject and motivates this longitudinal study to determine the relationship between cognitive testscore improvement and independent declines of schistosome and single soil-transmitted helminth ( STH ) infections , as well as the impact of concurrent declines of two or more STHs on changes in cognitive testscores ., Specifically , we provide associations between cognitive testscore improvement and:, ( i ) treatment-induced changes in S . japonicum intensity ,, ( ii ) non-treatment-related or natural declines in single STH infections , and, ( iii ) joint infection decline for ≥2 STH species ., We hypothesize that no or low level S . japonicum re-infection after praziquantel treatment , and clearance or intensity reductions for single and polyparasitic STH infections will predict improvements in cognitive testscores during follow-up among school-aged children living in a schistosome and STH co-endemic area of Leyte , The Philippines ., The parent study and the nested study reported here were approved by the Brown University , Lifespan , and Philippines Research Institute of Tropical Medicine Institutional Review Boards ., Participants aged ≥18 years provided written informed consent ., In addition , all parents/guardians provided written informed consent on behalf of child participants , whereas children aged ≥8 years provided assent ., All participants were S . japonicum infected and were treated with the anti-schistosomal drug praziquantel ( 60 mg/kg over 4 hours ) at enrolment as part of the parent study ., Only cognitive testing was conducted in a subset of 253 children , aged 7–19 years , as part of this nested observational study ., There was no baseline treatment for STH infections as large-scale helminth treatment campaigns were not available in The Philippines at the time this study was conducted ., However , at the end of the study , children with STH infection were treated with albendazole and those that became re-infected with S . japonicum were treated with praziquantel ., An approach that includes waiting to treat children infected with STH would not be taken today given more recent published findings regarding subtle morbidities related to STH infections ., This study was conducted in Macanip , a malaria-free rural rice farming village in Leyte , The Philippines , where S . japonicum and STH infections coexist with high prevalence ., This is a nested prospective cohort study conducted in a subset of S . japonicu- infected Filipinos aged 7–30 years who were enrolled in a study of immune correlates of resistance to S . japonicum reinfection 16 ., Eligibility criteria included: baseline S . japonicum infection , age 7–19 years at enrolment , provision of parental consent , and child assent for participation in this study ., Exclusion criteria included pregnancy or lactation , severe malnutrition ( weight-for-height z-score<−3 ) , severe anemia ( hemoglobin<7 g/dl ) , or the presence of a serious chronic disease determined by history , physical examination , or laboratory findings ., Four cognitive tests were administered , including the Philippine nonverbal intelligence test ( PNIT ) , verbal fluency ( VF ) , and two domains of the Wide Range Assessment of Learning and Memory ( WRAML ) , namely verbal memory and learning ., Tests were chosen based on their ability to capture a range of cognitive processes including fluid intelligence ( PNIT ) , learning ( WRAML ) , and memory ( VF and WRAML ) while being adaptable across different cultures ., The PNIT is an intelligence test that measures concept recognition and abstract thinking 17 ., VF test is thought to be a good measure of the central executive component of working memory ., The WRAML assesses a childs ability to learn and recall new information ., Specifically , the WRAML learning subtests evaluate a childs performance over trials on tasks using the free-recall paradigm , while the WRAML verbal memory subtests assess a childs memory capabilities on meaningful ( i . e . , stories ) and meaningless material ( i . e . , strings of random digits and letters ) 18 ., Each of the domains assessed by the WRAML consists of three age-standardized subtests that are added together to derive a total age- and gender-scaled score per domain ., Unlike the WRAML , neither the PNIT nor the VF are age standardized; therefore , these tests were adjusted for age variation using linear regression from which we calculated the error terms associated with each childs testscore ., We then modeled as the dependent variable the error terms associated with performance in PNIT and VF tests ., All tests were translated , adapted for cultural appropriateness , and pilot tested among Filipino children from other S . japonicum-endemic villages near the study area ., Testing was conducted in a designated room adjacent to the field laboratory with sufficient lighting and minimal external noises ., Ambient temperature within the classroom was approximately 27°C ., All children were provided a snack about 30 minutes prior to testing ., Joint inter-rater and test-retest reliability with a 6-week interval between tests were evaluated ., Cronbachs alpha coefficient was used to assess the degree of internal consistency between tests in the WRAML learning ( α\u200a=\u200a0 . 54 ) and WRAML verbal memory ( α\u200a=\u200a0 . 81 ) domains ., For all tests , higher scores correspond to better performance ., Details of each test and its psychometric properties have been previously reported 14 ., More details about the rationale for choosing specific tests and their respective properties are presented in Appendices S1 and S2 ., Cognitive assessments were made at months 0 , 6 , 12 , and 18 ., All infections were assessed at baseline and quarterly thereafter ., We have previously reported on cross-sectional associations between helminth infections and performance in the aforementioned tests 14 ., Here we determine associations between post-treatment testscores and:, ( i ) post-treatment re-infection with S . japonicum and, ( ii ) natural infection clearance/decline for STH infections ., Only cognitive assessments at 6 , 12 , and 18 months are included in the outcome matrix to preserve temporal sequence between infections and testscore changes ., The origin of this prospective analysis is the cohort-wide interval of least infection intensity for all species ( i . e . , months 1–3 ) ., STH and schistosome infections were assessed at months 0 , 3 , 6 , 9 , 12 , 15 , and 18 ., For S . japonicum only , an additional assessment ( one month post-treatment ) was done to evaluate treatment efficacy ., The number of eggs per gram ( EPG ) of stool was determined via duplicate examination of three stool samples by the Kato-Katz method for all species 19 ., EPGs were used to define none , low , moderate , or high intensity categories for each species using World Health Organization EPG thresholds 20 ., For each individual helminth species , except hookworm , a separate dichotomous baseline intensity indicator was defined as: uninfected/low vs . moderate/high infection to accommodate the intensity distribution in this cohort ., For hookworm infection only , baseline infection intensity was defined as none vs . any infection , since >40% of participants were hookworm-free at enrollment and those infected had predominantly low infections ., Children were initially grouped by the intensity of concurrent infection with hookworm , A . lumbricoides and T . trichiura as having:, ( i ) one or zero low;, ( ii ) two or three low;, ( iii ) one moderate/high STH;, ( iv ) two moderate/high; and, ( v ) three moderate/high intensity coinfections 11 ., These categories were further combined into one baseline polyparasitic STH indicator to distinguish children with ≥2 STH species at moderate/high intensity ( which may include zero or one low infection of the third STH species ) from those with at most one STH infection at moderate or higher intensity STH coinfection ( other STHs are either absent or present at low intensity only ) ., Given our treatment-reinfection design and study inclusion predicated on S . japonicum infection , the most dynamic infection changes occurred with respect to S . japonicum during follow-up; however , STH infection intensity also varied over time ., These non-treatment related changes in STH intensity may be due to one or more of the following factors:, ( i ) natural changes in STH infections within individuals over time ,, ( ii ) the limited sensitivity of some STH species to praziquantel 21 , 22 , and, ( iii ) lower diagnostic sensitivity for the Kato-Katz method especially when used for the simultaneous assessment of multiple STH species at low intensity in the same host 23 ., We defined three post-treatment infection intervals: 1≤t1<6 , 6≤t2≤12 and 12<t3≤18 months; to correspond with the three repeated cognitive assessments ., For each STH , t1 infection value ( I1 ) was the mean EPG at month three , whereas for S . japonicum I1 was the mean of EPGs at months one and three ., T2 infection ( I2 ) was the mean of EPGs at months six and nine , and t3 infection ( I3 ) was the mean of EPGs at months 12 , 15 , and 18 ., Within respective intervals , intra-individual infection change scores ( δit ) were defined by species as follows: t2: δi2\u200a=\u200aIi2 - Ii1; and t3: δi3\u200a=\u200aIi3 - Ii1 ., Hence , δit ranged from −∞ to +∞ and will be negative , zero , or positive for a given STH species if the childs infection was lower , equivalent to , or greater than their infection intensity at t1 ., For each species , separate δit values were defined and ultimately dichotomized into high vs . low categories as δit≥0 vs δit<0 ., For S . japonicum only , infection-free duration was defined as a four level categorical variable that is:, ( i ) 0 if not reinfected by month 18;, ( ii ) 1 if reinfected between months 12 and 18;, ( iii ) 2 if reinfected between months 6 and 12; and, ( iv ) 3 if never cured or S . japonicum positive in t1 , t2 , and t3 ( reference group ) ., Children reinfected by 6 , 12 , or 18 months were compared to those not reinfected by study end ., We determined the number of concurrent STH declines as the sum of individual STH intensity declines using the previously described dichotomous infection decline variable based on δit ., Possible values for polyparasitic STH declines were: 0\u200a=\u200ano decline/increase STH species , 1\u200a=\u200aany one STH , 2\u200a=\u200aany two STH to 3\u200a=\u200aall STH species intensity decline in a given interval ., Using these values , polyparasitic STH decline within intervals was defined as: concurrent intensity decline of ≥2 vs . ≤1 of 3 STH species ., We considered an extensive array of potential confounding factors ., Because exposure to helminth infection and cognitive testscores vary by age , sex , and socioeconomic status ( SES ) , these factors were considered non-time varying potential confounders ., SES measurements were based on baseline questionnaire data addressing four domains of social position; parental and child education , occupation , home/land ownership , and assets ., The method used to derive and validate this measure of SES has been described elsewhere 14 , 24 ., The derived summary SES variable is divided into four ordinal categories by the quartiles of its distribution ., Anemia and nutritional status at baseline were considered potential confounders and/or mediators of low testscores ., Anemia was defined on the basis of age- and sex-specific hemoglobin cutoffs recommended by the WHO 25 ., Hemoglobin measurement was based on complete hemograms determined on a Serono Baker 9000 hematology analyzer ( Serono Baker Diagnostics , Allentown , PA ) ., Nutritional status was assessed using weight-for-age z-scores ( WAZ ) calculated using the National Center for Health Statistics year 2000 reference values in EpiInfo software ( version 2000 , Atlanta , GA ) ., Normal and malnutrition status were defined by WAZ≥−2 and WAZ<−2 , respectively ., Multivariable random effects regression models were fitted separately to each cognitive test without adjusting for testscore at study enrollment ( month 0 ) given our observational study design 26 ., We assumed an unstructured covariance matrix to account for non-independence of repeated cognitive tests within individuals and accounted for clustering of observations within households by including a random intercept for household ., Empirical standard errors were used for all estimations to ensure that significance tests were robust against mis-specification of the covariance matrix ., In addition , we examined the relationship between test performance and S . japonicum-free duration in separate regression models ., Sample regression models for estimation of associations between testscores and S . japonicum infection decline and S . japonicum infection free duration are provided in Appendix S3 ., Finally , we examined the potential for modification in the association between infection change and testscore improvement by the following baseline factors: helminth infection intensity , underweight , and anemia ., For example , to examine whether the relationship between hookworm infection decline and testscore improvement was heterogenious by hookworm baseline infection intensity , we introduced a three-way multiplicative interaction consisting of the dichotmous indicator of hookworm infection decline , time , and baseline hookworm intensity in a multivariate models that in addition to other confounders also adjust for the baseline intensity of A . lumbricoides , T . trichiura and S . japonicum as well as each of the three dichotmous indicators of change in these infections from the interval of lowest infection ., We then examined the p-values associated with interaction terms and where P≤0 . 05 , results are presented by strata of baseline hookworm intensity ., The same approach was used to examine baseline underweight and baseline anemia as potential effect modifiers in separate multivariate regression models ., The prevalence of A . lumbricoides , T . trichiura and hookworm infections in this S . japonicum-infectected cohort at baseline were 79 . 9% , 95 . 6% , and 50 . 6% , respectively ., Of the 253 children , 97% were concurrently infected by S . japonicum and at least one STH species , approximately 36% were anemic and 60% were underweight relative to U . S . children of the same age and sex ( Table 1 ) ., The lowest intensity of S . japonicum infection ( mean\u200a=\u200a6 . 8 EPG ) occurred one month post-treatment at which 92% ( n\u200a=\u200a217 ) of the sample was infection-free ., However , re-infection was rapid and increased steadily until the 12th month of follow-up , at which point 70 . 8% of participants were infected with S . japonicum ., Only 25 ( 10 . 6% ) of the re-examined children were free of S . japonicum infection at 18 months ., Individual STH intensities also declined from enrollment with the lowest average infection for all STH species occurring at three months ., Infection intensity stabilized near this level throughout follow-up for hookworm and T . trichiura infections ., The cohort-wide , A . lumbricoides infection intensity by the 18th month was comparable to month zero despite the initial decline post-S ., japonicum treatment ( Figure 1 ) ., From multivariable models adjusted for sociodemographic characteristics and the intensity of coincident S . japonicum and STH species , declines in the intensity of T . trichiura , hookworm , and polyparasitic STH infections were independently associated with higher average scores on the learning and verbal memory domains of WRAML tests during follow-up ., Similarly , A . lumbricoides intensity decline was independently associated with higher scores in the learning sub-scale of WRAML ., The intensity of individual infections at enrollment were generally not associated with performance on any of the tests employed , except for moderate/high intensity polyparasitic STH infection , which was associated with lower scores on the PNIT ( Table 2 ) ., A decline vs . no change or an increase in S . japonicum intensity from the interval of least infection was not independently associated with improvements in any tests over the study period ( Table 2 ) ., We found no evidence that the relationship between S . japonicum infection decline and performance in respective tests differed within strata of S . japonicum intensity at enrollment ( data not shown ) ., However children who were S . japonicum free for ≥18 months or those who were S . japonicum infection free until 12 months post-treatment scored higher in all tests relative to rapidly re-infected or persistently infected children ., The strength of association was generally attenuated in multi-variable models that controlled for several sociodemographic characteristics and coincident STH and the baseline intensity of S . japonicum infection ., Nevertheless , never S . japonicum re-infected children and those S . japonicum infection-free for up to 12 months scored higher in the verbal memory sub-scale of WRAML and VF test , respectively ( Table 3 ) ., Anemia and underweight status at enrollment were not independently associated with performance in any tests ., However , among children with anemia at enrollment , S . japonicum decline was associated with higher scores on WRAML learning subscale ( mean\u200a=\u200a10 . 5 , 95% confidence interval ( CI ) : 4 . 8–16 . 3 ) ., There was no association between S . japonicum infection decline and performance in WRAML learning subscale among children without anemia at enrollment ( mean\u200a=\u200a−3 . 0 , 95% CI: −6 . 4–0 . 4 ) ., Given our observational study design , we cannot exclude residual confounding by unmeasured covariates as an alternative explanation for our findings ., By comparing children present at 18-months with those present at baseline on key factors , children scoring in the highest tertile of WRAML verbal memory at baseline and girls were over-represented among those lost to follow-up; however , there was no difference in average hemoglobin , SES , baseline STH intensity and average scores in WRAML learning , PNIT , and VF ., In addition , the Kato-Katz relative to other helminth diagnostic methods has been reported to be of lower sensitivity for detecting helminth eggs particularly for individuals with light infections 42 and those with concurrent multi-species infections 23 ., We expect that our duplicate assessment of three separate stool samples for each child would have improved the accuracy of helminth diagnosis in this study; however , we are unable to rule out the possible impact of limited sensitivity for lightly infected children ., To our knowledge , this is the first longitudinal study to investigate the independent effect of schistosome and individual STH infections as well as that of polyparasitic STH infection decline on learning domains of cognitive function , which may better reflect childrens ability to take advantage of limited educational opportunities ., The prospective study design , control for coincident helminth infections and numerous other confounders , and the explicit exploration of baseline infection , anemia and nutritional statuses as potential mediators of observed associations are additional strengths of this study ., We observed notable fluctuations in T . trichiura and A . lumbricoides intensity in this study even though only S . japonicum infection was treated at enrolment ., Praziquantel , however , has been shown to have some anti-hookworm activity 22 ., Unlike prior investigations of this question , our analytic strategy highlights the cognitive performance deficits associated with S . japonicum rapid reinfection following treatment as well as the cognitive benefits of natural declines in STH infections among school-aged children ., By modeling the relationship between helminth infections and cognitive testscores from the interval of least infection following S . japonicum treatment , we highlight the cognitive test performance advantage of sustained low level single and polyparasitic helminth infections that is derivable in the presence of systematic frequent deworming programs ., This relationship may be blunted or lost in an environment characterized by infrequent deworming and high helminth reinfection pressure ., Findings from this design and analytic strategy may be more generalizable to the actual implementation of deworming programs than randomized trials ., We conclude that declines in the burden of some helminth species and polyparasitic STH infections have beneficial long-term impacts on childrens cognitive performance ., Our results highlight the benefit of combined control for S . japonicum and STH infections; it further stresses the importance of sustained deworming for improving the learning , memory , and educational attainment of children in helminth-endemic settings ., The benefit of combined treatment for these infections notwithstanding , deworming is only a necessary first step in the implementation of a comprehensive integrated helminth control program , which must be tailored to a given endemic setting and include provision of clean water and improved sanitation to mitigate the fundamental causes of these infections and their associated adverse health effects among the most vulnerable populations 43 , 44 . | Introduction, Materials and Methods, Results, Discussion | To determine whether treatment of intestinal parasitic infections improves cognitive function in school-aged children , we examined changes in cognitive testscores over 18 months in relation to:, ( i ) treatment-related Schistosoma japonicum intensity decline ,, ( ii ) spontaneous reduction of single soil-transmitted helminth ( STH ) species , and, ( iii ) ≥2 STH infections among 253 S . japonicum-infected children ., Helminth infections were assessed at baseline and quarterly by the Kato-Katz method ., S . japonicum infection was treated at baseline using praziquantel ., An intensity-based indicator of lower vs . no change/higher infection was defined separately for each helminth species and joint intensity declines of ≥2 STH species ., In addition , S . japonicum infection-free duration was defined in four categories based on time of schistosome re-infection: >18 ( i . e . cured ) , >12 to ≤18 , 6 to ≤12 and ≤6 ( persistently infected ) months ., There was no baseline treatment for STHs but their intensity varied possibly due to spontaneous infection clearance/acquisition ., Four cognitive tests were administered at baseline , 6 , 12 , and 18 months following S . japonicum treatment: learning and memory domains of Wide Range Assessment of Memory and Learning ( WRAML ) , verbal fluency ( VF ) , and Philippine nonverbal intelligence test ( PNIT ) ., Linear regression models were used to relate changes in respective infections to test performance with adjustment for sociodemographic confounders and coincident helminth infections ., Children cured ( β\u200a=\u200a5 . 8; P\u200a=\u200a0 . 02 ) and those schistosome-free for >12 months ( β\u200a=\u200a1 . 5; P\u200a=\u200a0 . 03 ) scored higher in WRAML memory and VF tests compared to persistently infected children independent of STH infections ., A decline vs . no change/increase of any individual STH species ( β:11 . 5–14 . 5; all P<0 . 01 ) and the joint decline of ≥2 STH ( β\u200a=\u200a13 . 1; P\u200a=\u200a0 . 01 ) species were associated with higher scores in WRAML learning test independent of schistosome infection ., Hookworm and Trichuris trichiura declines were independently associated with improvements in WRAML memory scores as was the joint decline in ≥2 STH species ., Baseline coinfection by ≥2 STH species was associated with low PNIT scores ( β\u200a=\u200a−1 . 9; P\u200a=\u200a0 . 04 ) ., Children cured/S ., japonicum-free for >12 months post-treatment and those who experienced declines of ≥2 STH species scored higher in three of four cognitive tests ., Our result suggests that sustained deworming and simultaneous control for schistosome and STH infections could improve childrens ability to take advantage of educational opportunities in helminth-endemic regions . | Parasitic worm infections are associated with cognitive impairment and lower academic achievement for infected relative to uninfected children ., However , it is unclear whether curing or reducing worm infection intensity improves child cognitive function ., We examined the independent associations between:, ( i ) Schistosoma japonicum infection-free duration ,, ( ii ) declines in single helminth species , and, ( iii ) joint declines of ≥2 soil-transmitted helminth ( STH ) infections and improvements in four cognitive tests during18 months of follow-up ., Enrolled were schistosome-infected school-aged children among whom coinfection with STH was common ., All children were treated for schistosome infection only at enrolment with praziquantel ., Children cured or schistosome-free for >12 months scored higher in memory and verbal fluency tests compared to persistently infected children ., Likewise , declines of single and polyparasitic STH infections predicted higher scores in three of four tests ., We conclude that reducing the intensity of certain helminth species and the frequency of multi-species STH infections may have long-term benefits for affected childrens cognitive performance ., The rapidity of schistosome re-infection and the ubiquity of concurrent multi-species infection highlight the importance of sustained deworming for both schistosome and STH infections to enhance the learning and educational attainment of children in helminth-endemic settings . | medicine, public health and epidemiology, infectious disease epidemiology, parasitic diseases, helminth infection, hookworm infection, neglected tropical diseases, ascariasis, infectious diseases, soil-transmitted helminths, parasitic intestinal diseases, epidemiology, hookworm, gastrointestinal infections, schistosomiasis, pediatric epidemiology, trichuriasis | null |
journal.pgen.1000614 | 2,009 | Regulon-Specific Control of Transcription Elongation across the Yeast Genome | It is well known that RNA pol II accumulates under repressive conditions on some tightly regulated genes of higher eukaryotes , like human c-Myc 1 and Drosophila hsp70 2 ., RNA pol II pausing is in fact a frequent situation since a significant proportion of metazoan genes exhibits paused polymerases at promoter-proximal sites 3–8 ., Although pauses and arrests during transcription elongation seem to be also common phenomena further downstream 9 ., In yeast , almost 2500 repressed genes show poised RNA pol II in the stationary phase 10 but only a few , like CYC1 and those encoding NTP-biosynthetic enzymes , display an accumulation of RNA pol II at their 5′ region under repressive conditions in exponential growing cells 11 , 12 ., For NTP genes , transcription regulation works at the level of initiation through an attenuation mechanism 12 , 13 ., It is not clear whether the accumulation of RNA pol II at the 5′ end in the other cases responds to a pausing phenomenon ., In any case , RNA pol II pausing at promoter-proximal sites is not a frequent phenomenon in exponentially growing yeast 14 which has been proposed to reflect the different chromatin organization of the transcription start sites in yeast compared to metazoa 15 ., In the last 20 years , biochemical and genetic analyses have revealed a numerous set of factors playing auxiliary roles in RNA Polymerase II ( RNA pol II ) -dependent transcription elongation 16 ., The textbook view of transcriptional machinery is a uniform set of players that all genes require equally ., However , it is already well known that the diversity in core promoter elements throughout the genome reflects certain gene-specific roles of the general transcription factors involved in the pre-initiation complex ( PIC ) assembly ., For instance , yeast TATA box-containing genes are highly regulated and preferentially utilize SAGA rather than TFIID if compared to TATA-less promoters 17 ., According to such differences , a TBP regulatory network to explain gene-specific differences in the PIC assembly has been proposed 18 ., Similarly , several examples of gene-specific roles of elongation factors have been described ., Mutations affecting the integrity of the yeast THO complex , involved in transcription elongation and mRNP biogenesis , decrease the expression levels of long transcription units , but do not significantly influence the mRNA levels of the shorter ones driven by the same promoter 19–21 ., TFIIS , an elongation factor that is dispensable for the expression of most yeast genes , is absolutely required for the activation of IMD2 in response to NTP depletion 22 ., Mammalian splicing factor SC35 also plays a gene-specific role in transcription elongation since its depletion produces an accumulation of inactive RNA pol II on several , but not all , active transcription units 23 ., The transcription of the p53-dependent gene p21 does not require the phosphorylation of the carboxy-terminal domain of RNA pol II ( CTD ) in the serine residue situated at position 2 ( Ser2 ) ., This indicates that the requirement of P-TEFb for transcription elongation is also gene-specific 24 ., The chromatin factor FACT , involved in chromatin remodeling and reassembly during transcription elongation 25 , 26 , is also dispensable for the expression of p21 24 ., Likewise , the expression of the yeast CUP1 gene , which can be transcribed by a mutant version of RNA Pol II lacking the CTD 27 , is not affected by FACT depletion 28 ., Furthermore by comparing five genes under the control of the same promoter , we have previously shown that FACT is not equally required by all the genes during transcription , and that this differential requirement is related to the chromatin configuration of the transcribed region 28 ., In this work , we investigated the distribution of actively elongating and total RNA pol II by means of a new methodological approach that combines genomic run-on ( GRO ) and ChIP-on-chip ., We detected significant gene-specific differences in the proportion of active RNA pol II present in the transcribed regions ., The effect of FACT depletion was also differential for some gene functional categories such us those encoding mitochondrial proteins , or housekeeping genes encoding cytosolic ribosomal proteins and factors involved in ribosome biogenesis ., We found that the transcription elongation of ribosome-related genes responds to regulatory stimuli mediated by the protein kinase A pathway , and by the Rap1 transcription factor for those genes that encode structural ribosomal proteins ., We also found that an inactive form of RNA polymerase II , which is phosphorylated in the Ser5 residue of the CTD but is hypophosphorylated in Ser2 , accumulates along the full length of these genes , during standard growing conditions ., We measured the association of RNA pol II with yeast genes in exponentially grown cells in YPD by performing RNA pol II ChIP-on-chip experiments ( RPCC , Pelechano et al . , to be published elsewhere ) ., All normalized and processed genomic data are included in Table S1 ., We compared the Rpb1-binding data obtained by RPCC with the transcription rate ( TR ) data previously measured by GRO 29 ., We found that the ribosomal protein genes ( RP regulon ) were relatively enriched in Rpb1 ( using a Myc-tagged version of it , see Figure S1 ) ., Even clearer results were obtained when the RPCC experiments were performed with the 8WG16 antibody , which recognizes RNA pol II CTD ( Figure 1A ) ., Gene classes relating to cytosolic ribosome and translation presented significantly high ChIP/TR ratios ( Table 1 ) ., A prominent RNA pol II enrichment was also detected in RP genes with an antibody that recognizes the CTD repeats when these are phosphorylated in the serine residue situated at position 5 ( Ser5 ) ( Figure 1B and Table 1 ) ., All statistically significant GO categories found in all the genomic experiments are included in Table S2 ., We reason that the difference between the GRO and the RPCC data , reflected in the ChIP/TR ratios , could be due to the different degree of accumulation of non-actively elongating RNA pol II either in a step prior to initiation or arrested during elongation ( likely back-tracked ) ., Our data indicate that the inactive form of RNA pol II ( not producing a run-on signal ) which accumulated in RP genes was phosphorylated in the Ser5 residue of the CTD ., Therefore , we conclude that it should have passed the initiation step of transcription ., The detected imbalance between the amounts of RNA pol II bound to RP genes and their TR may either be an intrinsic feature of these genes or reflect the occurrence of a novel mechanism that regulates their expression ., In order to test these two possibilities , we calculated the ChIP/TR ratios in three different culture conditions:, i ) exponential growth in glucose medium ,, ii ) 2 h after transferring glucose-grown cells to galactose-containing medium ( non growing cells due to the metabolic shift ) , and, iii ) exponential growth in galactose medium ( 14 . 5 h in galactose ) ., Then we did a clustering analysis to group genes in accordance with their ChIP/TR patterns ., As shown in Figure 2A , two clusters were detected ( numbers 0 and 3 ) in which the ChIP/TR ratio clearly decreased during the shift from glucose to galactose ( 2, h ) , and continued to decrease when cells grew exponentially in galactose ( 14 . 5, h ) ., The difference between these two clusters was the kinetics of the ChIP/TR decrease , that is , more intense in the first step for cluster number 0 and deeper in the second step for cluster number 3 ( Figure 2A ) ., The genes belonging to the RP and RiBi regulons were significantly enriched in cluster 0 , although RiBi genes were also located in cluster 3 ( Figure 2A ) ., The RiBi regulon comprises all the genes encoding the non RP proteins involved in rDNA transcription , tRNA synthesis , ribosome biogenesis and translation ( see 30 for a more precise definition of RiBi ) ., The opposite scenario ( higher ChIP/TR ratios in galactose than in glucose ) was detected for clusters 6 and 8 , which were statistically enriched in mitochondria-related genes ., We detected a general genome-wide correlation between the ChIP/TR ratios of cells exponentially growing in glucose and those of cells exponentially growing in galactose , indicating that the lower ChIP/TR ratios shown by the RP and RiBi regulons in galactose and the higher ChIP/TR ratios displayed by mitochondria-related genes indeed reflect a specific regulatory phenomenon ( Figure 2B and Table 1 ) ., In the first step of the experiment ( 2, h ) , the TR and the amounts of RNA pol II binding to most genes , including the RP regulon , sharply decreased ( Figure 2C ) ., This reduction in the genome expression , particularly of the RP genes , is consistent with lack of growth after shifting the culture from glucose to galactose media ., Transcription increased in step two ( 14 . 5, h ) once cells recovered their exponential growth rate , as did the TR of the RP genes and the amount of RNA pol II bound to them ( Figure 2C ) ., During these successive down- and up-regulation steps however , the ChIP/TR ratios of RP and RiBi genes continuously decreased in relation to the genome average ( Figure 2D; Figure S2A , S2B ) ., This was mainly due to a decrease in the relative amount of bound RNA pol II ( Figure S2C ) rather than to a relative change in their TR ( Figure S2D ) ., Mitochondria-related genes also underwent a similar down- and up-regulation cycle in both their TR and the levels of bound RNA pol II ( Figure 2C ) , although the average ChIP/TR ratios increased in this case ( Figure 2D; Figure S2A , S2B ) ., This increase in the ChIP/TR ratio of mitochondria-related genes was also due to a change in the relative amount of bound RNA pol II ( Figure S2C ) rather than to a variation in the relative TR of respiration genes ( Figure S2D ) ., We conclude that the shift from glucose to galactose , and not the growth rate , was the stimulus responsible for the ChIP/TR regulation of RP , RiBi and mitochondria-related genes ., The transcriptional response of RP genes to glucose levels is mediated by the TOR and Ras-PKA pathways 31 ., To further confirm that glucose signaling regulates the ChIP/TR ratios of ribosomal related genes , we analyzed a Δtpk1 mutant ., Tpk1 is one of the three PKAs present in yeast and it is physically located on those genes which are highly transcribed 32 ., As shown in Figure 2E , the ChIP/TR ratio of RP and RiBi genes lowered , while the mitochondrial genes increased in the Δtpk1 when compared to an isogenic wild type grown in YPD ., The results of Δtpk1 resembled those of the wild type in galactose ( compare Figure 2B and 2E ) ., Tpk2 is a second PKA catalytic subunit which is physically located on the promoter regions of RP genes 32 ., The results of analyzing a Δtpk2 mutant showed similar patterns to those observed in Δtpk1 ., We did not observe any significant variation in the ChIP/TR ratios of RP genes between Δtpk1 and Δtpk2 ., We conclude that PKA mediates the signal which regulates the ChIP/TR ratios of RP , RiBi and mitochondrial genes in response to the glucose-galactose shift , but its participation in this regulation does not depend on a particular catalytic subunit ., The chromatin elongation factor FACT is formed in yeast by Spt16 , Pob3 and Nhp6 33 ., It has been shown that FACT is physically located on active yeast genes along the whole length of their transcription units 34 ., In order to test whether the accumulation of inactive RNA pol II on RP genes in glucose is extensive to transcription elongation factors , we measured the amount of Spt16 bound to yeast genes by performing ChIP and by hybridizing the same kind of arrays as we used for GRO and RPCC experiments ., As Figure 3A depicts , there is a general positive correlation between the amount of Spt16 bound to a gene and its TR ( see also Figure S3A ) , and similarly to RNA pol II , the RP genes show higher levels of Spt16 than those expected for their TR ( Figure 3A and Figure S3B ) ., In contrast , the RP genes showed no Spt16 enrichment in relation to the amount of RNA pol II bound to them ( Figure 3B ) ., Irrespectively of the TR , we found a close correlation between RNA Pol II and Spt16 levels of occupancy , as well as a constant Spt16/Rpb1 ratio ( Figure S3C ) , which suggests that the presence of RNA pol II on a gene , rather than its transcriptional activity , causes FACT recruitment ., By using a Tet-off::SPT16 strain , we have previously shown that there is some gene-specificity in the effect of Spt16 depletion on yeast transcription 28 ., In order to test whether the excess FACT present in RP genes was dispensable for their actual transcription rates in glucose , we analyzed the transcriptional effect of Spt16-depletion on a genome-wide scale ., Following the GRO procedure , we were able to calculate the effect of Spt16 depletion on TR of 5257 genes , which represents 91% of the genes present in the yeast genome ., We took these measurements at depletion times at which neither the growth rate nor the viability of the cells was affected ., Five hours after adding doxycycline the overall mRNA levels in the cell were not affected ( Figure S4A ) but the TR of most genes had decreased ( Figure 3C and Figure S4B ) , which is in agreement with the general positive roles played by FACT in transcription 35 , 36 ., By carrying out a gene-ontology analysis of the TR decrease , we detected functional classes of genes that were particularly sensitive or insensitive to Spt16 depletion ., We found that the RP and RiBi regulons were especially resistant to Spt16 depletion ( Figure 3C and Table 1 ) , whereas those genes related to the mitochondria were ranked as hypersensitive ( Table 1 ) ., Since ribosomal proteins genes are generally short and contain introns , we analyzed the influence of several structural gene features on sensitivity to Spt16 depletion ., No correlation with gene length , G+C content or intron presence was found ( Figure S5A , S5B , S5C ) ., Since RP genes are highly transcribed , we also checked the influence of TR itself on the response to Spt16 depletion ., We found a linear correlation between TR under depletion conditions and control conditions , thus ruling out that highly expressed genes were proportionally less sensitive to Spt16 depletion ( Figure S5D ) ., We performed additional GRO experiments with doxycycline-treated cells over a longer time to achieve a more severe depletion of Spt16 ., As shown in Figure 3C , the RP regulon showed a similar distribution of TR to the rest of the genome after 7 h of treatment with doxycycline ., At this depletion time almost no overrepresentation of gene ontology classes was observed ( Table 1 ) ., These results confirm that the slight accumulation of FACT on the RP and RiBi regulons , in relation to the levels of actively elongating RNA pol II present , makes these genes transiently resistant to Spt16 depletion ., Collectively , these results suggest that not only RNA pol II , but additional elements of the transcription elongation machinery , are enriched on the ribosome-related genes in glucose , if compared to their TR ., The exceeding signal of RPCC over GRO in glucose for RP genes suggests that an accumulation of non-transcribing RNA polymerases takes place ., The accumulation of inactive RNA pol II on ribosome-related genes is compatible with a post-recruitment mechanism of transcription regulation ., With paused metazoan genes , the intragenic distribution of RNA pol II is biased toward the 5′ end ., In order to know whether the RNA pol II enrichment of RP genes also involves a biased distribution of the enzyme , we analyzed in detail the distribution of RNA pol II on a representative RP gene ( RPS3 ) by ChIP ( Figure 4A–4D ) ., As expected , we found higher levels of total RNA pol II in glucose than in galactose ( Figure 4B ) , but we observed lower levels of Ser5- , and much lower levels of Ser2-phosphorylated RNA pol II in glucose on the 3′ end of the transcribed region ( Figure 4B ) ., These different intragenic distributions of RNA pol II are fully compatible with a lower elongation efficiency of RNA pol II in glucose in relation to galactose ., The representation of the data following the normalization procedure described by 37 supports this conclusion ( Figure 4C ) ., Likewise , the representation of the levels of phosphorylated RNA pol II , normalized by the total levels of the enzyme , reveals a clear difference between the two conditions ., Whereas phosphorylation in galactose followed the standard pattern , with a moderate decrease of Ser5-phosphorylation along the gene and a sharp increase of Ser2-phosphorylation towards the 3′ end , the increase of Ser2-phosphorylation in glucose along the gene was considerably less evident ( Figure 4D ) ., Very similar results were obtained with the detailed analysis of the gene encoding ribosomal protein L25 ( Figure S6A , S6B , S6C , S6D ) ., We also investigated the intragenic distribution of active RNA pol II by performing a detailed run-on analysis of the RPS3 gene ., In this case , we found similar patterns in both glucose and galactose with comparable levels on the 3′ and medium regions of the transcribed region , and with higher levels at the 3′ end of the gene in galactose than in glucose ( Figure 4E ) ., These results support the hypothesis that the accumulation of RNA pol II on RP genes in glucose took place during elongation and was due to a transcriptionally inactive form of RNA pol II that lacked normal levels of Ser2 phosphorylation ., In order to confirm the variation in the intragenic distribution of RNA pol II in RP genes from glucose to galactose , we repeated the RPCC experiments described before with a new type of DNA macroarrays containing 300 bp-long probes covering separately the 5′ and the 3′ ends of the transcribed regions of a set of randomly chosen genes ( Rodríguez-Gil et al , submitted ) ., We found that most of the RP genes present in this array presented a higher RPCC 5′/3′ ratio in glucose than in galactose ( Figure 4F and 4G ) ., This seems to be specific for RP genes since the RiBi genes represented in the array showed similar RPCC 5′/3′ ratios in the two media , as most non ribosomal genes did ( Figure 4F and 4G ) ., We also discovered that neither RP nor RiBi genes showed significantly higher GRO 5′/3′ ratios in glucose than in galactose , thus confirming that the enrichment of RNA pol II located toward the 5′ end of RP genes in glucose consisted of transcriptionally inactive molecules ( Figure S6E ) ., So far we have described a novel regulated phenomenon affecting the RP genes expression ., It is expected that the mechanism underlying it would be operated by the transcription factors that specifically regulate these genes ., A transcription factor playing a mayor role in RP genes transcription is Rap1 , a multifunctional protein that also acts as the main duplex DNA binding protein at telomeres , which not only contributes to silencing in both the subtelomeric regions and the mating type loci , but also activates the transcription of glycolytic genes ( reviewed by 38 , 39 ) ., Rap1 is essential for the RP expression as it organizes chromatin configuration at the RP genes promoters and allows the binding of Fhl1-Ifh1 , that is , the other two main transcription factors regulating the transcription of RP genes 40 ., An important domain of Rap1 is its silencing domain , which is involved in the subtelomeric recruitment of factors that regulate telomere length and gene silencing 41 ., Since mutants lacking the silencing domain of Rap1 are viable and do not show reduced levels of RP gene expression 42 , we decided to measure the influence of this mutation on the level of RNA pol II bound to RP genes and on their TR ., As shown in Figure 5A , RP were the most enriched genes in RNA pol II in both the wild-type and the rap1Δsil mutant ., However , and importantly , they were more transcribed in rap1Δsil than in the wild type ( Figure 5B ) ., Consequently , RP genes displayed a significantly low ChIP/TR ratio in the rap1Δsil mutant ( Figure 5C ) ., As expected , mitochondria-related genes were unaffected by rap1Δsil mutation ( Figure S7 ) ., The ChIP/TR ratios of RiBi genes , most of which are not directly regulated by Rap1 , did not undergo mayor change either ( Figure S7C ) ., In this case , they displayed slightly higher levels of both RNA pol II binding and transcriptional activity ( Figure S7A , S7B ) , which probably reflect their upregulation in response to the overexpression of RP genes caused by rap1Δsil ., As expected , rap1Δsil mutation also led to an increase in the TR of subtelomeric genes , but not in RNA pol II binding ( Figure S8 ) ., We also investigated the distribution of RNA pol II along RPS3 in the rap1Δsil mutant ., We found no clear difference in the intragenic distribution of bound RNA pol II when compared to the wild type ( Figure 5D ) ., However , we detected higher levels of active RNA pol II in the mutant measured by run-on , throughout the transcribed region ( Figure 5E ) ., Similar results were found for RPL25 ( Figure S9A , S9B ) ., We conclude that the silencing domain of Rap1 participates in the mechanism which controls the proportion of RNA Pol II that is effectively active on RP genes during transcription elongation ., In this work , we show that the transcriptionally active proportion of RNA pol II bound across the genome is gene-specific and can be regulated in response to physiological stimuli ., The presence of glucose causes an accumulation of inactive RNA pol II on RP genes ., FACT , a general chromatin factor that is recruited to transcribed genes , also presents an uneven distribution , similar to that shown by RNA pol II ., This indicates that not only RNA pol II accumulates on some genes , but other components of the transcriptional machinery that follow this enzyme during elongation also do ., Conversely , the presence of galactose , or more likely , the absence of glucose , leads to a decrease in the proportion of inactive RNA pol II on RP and RiBi regulons , and increases it on mitochondria-related genes ., Briefly , a set of at least 1000 genes ( more than 15% of the yeast genome ) coordinately changes the fraction of RNA pol II that is effectively active during their transcription ., Genome-wide analysis has shown that TOR and PKA pathways co-regulate several gene regulons in yeast , including RiBi , RP and respiration-related genes 31 ., Whereas TOR acts as an activator of all three regulons , the PKA pathway acts by repressing respiratory genes and by activating the RP and RiBi genes ., Here we show that the absence of either Tpk1 or Tpk2 , two of the yeast PKA variants , produces the same kind of changes in the ChIP/TR ratios on RP , RiBi and mitochondria-related genes as the changing growth of the wild type from glucose to galactose ., This indicates that the overabundance of inactive RNA polymerases is characteristic of some specific groups of genes , under particular growth conditions , and that it is regulated by the PKA pathway ., The fact that there is no difference between the lack of Tpk2 , a PKA subunit shown to be bound to RP genes promoters 32 and Tpk1 , a subunit bound to the body of most genes suggests that the effect is quantitative: a reduction in PKA activity caused either by the lack of the alternative subunits or the growth in galactose reduces the accumulation of inactive RNA pol II molecules on several kinds of yeast genes ., The importance of gene-specific regulation during elongation across metazoan genomes can be deduced from the occurrence of RNA pol II pausing , which is a frequent phenomenon mainly affecting tightly regulated genes 3–7 ., Our work indicates that the control of RNA pol II elongation is also a common regulatory mechanism in yeast ., Unlike metazoan genes however , whose paused RNA pol II concentrate at specific promoter-proximal sites , elongation-regulated yeast genes , at least RPS3 and RPL25 , accumulate inactive RNA pol II along the length of their bodies with only some bias toward their 5′ moiety ., This accumulation correlates with a decrease in Ser2-phosphorylated RNA pol II along these genes ., The experimental evidence described in this work reveals that an excess of RNA pol II , phosphorylated in Ser5 , accumulates on the yeast RP genes in glucose ., This situation is only compatible with a post-initiation form of RNA pol II ., However , the absence of a comparably high run-on signal indicates that this extra amount of RNA pol II , which accumulates in glucose media , should be arrested after backtracking ., A similar situation occurs in the Drosophila hsp70 gene upon depleting the TFIIS cleavage factor 43 ., Regulation of ribosome synthesis is a key element in controlling cell homeostasis , cell size and proliferation 30 ., A coordinated and balanced expression of all the ribosomal protein genes is also needed to ensure efficient ribosome assembly 44 , and to avoid the potential toxicity of free ribosomal proteins 45 ., Regulation at the transcription elongation level may provide a gear box-like mechanism which enables a fine-tuning of RP and RiBi transcription by rapidly adjusting the proportion of recruited machinery that is effectively active in response to the specific translational requirements of each physiological state ., According to a recently proposed model , a certain level of backtracking during elongation , in combination with a fast initiation step , provides a steadier mRNA population level than that which would be produced by an initiation model alone 46 ., Accordingly , the regulation of RP transcription elongation would allow the expression of balanced amounts of translational machinery components ., It would also contribute to avoid transcription bursts , which would be incompatible with the low transcriptional noise that characterizes yeast constitutive genes 47 and , more specifically , the RP expression 48 ., In addition , and as suggested for Drosophila genes 5 , regulating the transcription at the elongation level enables a continuously open promoter configuration , an essential situation for genes like RP which are being permanently transcribed ., A feedback regulation mechanism operating at the intron splicing level has been demonstrated for certain RP genes 49 ., Since exon definition takes place during transcription elongation , an attractive hypothesis would be the existence of coordination between transcription elongation and RNA splicing in RP genes 50 ., However , our data do not support such a hypothesis since RNA pol II enrichment , compared to TR , was detected in both intron-containing and intron-less RP genes ( data not shown ) ., RP and RiBi regulons show different RNA pol II- and FACT-ChIP/TR ratios , which suggest that their regulation mechanisms are not identical ., We have also detected this kind of regulation in the group of mitochondria-related genes , thus confirming the previously described control of CYC1 transcription after RNA pol II recruitment 11 , 51 ., In this case , the ChIP/TR ratios were reciprocal to those of RP genes ., If we consider this diversity , it is likely that at least one subset of the molecular elements regulating the proportion of active RNA pol II during elongation is gene-specific ., We provide evidence for the specific involvement of the silencing domain of Rap1 in the mechanism required to maintain significant levels of inactive RNA Pol II on RP genes ., We also show that the absence of either Tpk1 or Tpk2 produces the same phenotype on RP transcription ., These results indicate that the proportion of inactive RNA pol II on RP genes is controlled by the factors that specifically regulate the transcription of RP genes ., In such a scenario , Rap1 would regulate the transcription of RP genes at both the RNA pol II recruitment and transcription elongation levels ( Figure 6 ) ., Tethering experiments using lexA-Rap1 chimeras have shown that the Rap1 DNA-binding domain itself is required for the activating function of Rap1 in RP transcription 40 ., This observation , together with the ability of Rap1 to clear nucleosomes from the vicinity of its binding sites 52 , suggests that the positive contribution of Rap1 to RP transcription is exerted in a pre-initiation step ., This is likely to be done by arranging a chromatin configuration of the promoter to allow the hosting of other RP transactivators like Fhl1-Ifh1 40 and , eventually , the pre-initiation complex itself ., The persistent occupancy of Rap1 on RP promoters , even under transcriptionally inactive conditions ( stress ) , suggests that this factor may also play a repressive role 53 ., We provide evidence of a negative role of Rap1 on RP transcription elongation which is mediated by its silencing domain ., This domain , located in the C-terminal part of the protein , has been previously shown to be important for the downregulation of RP transcription in response to certain defects in the secretory pathway 54 , 55 ., Graham et al . 42 showed that it also affects the mRNA steady-state levels of RP genes ., They attributed this effect to the secondary post-transcriptional consequences of the rap1Δsil deletions ., Our results indicate that it is in fact a transcriptional effect since the silencing domain has a negative influence on the TR of both subtelomeric ( Figure S8 ) and RP genes ( Figure 5 ) without affecting RNA pol II recruitment ., In Drosophila cells , hundreds of genes show that RNA polymerase II molecules paused after initiation ( about 20–50 bp from the TSS ) , which has been argued to deal with the presence of an H2AZ-containing positioned nucleosome 56 ., In yeast , the advanced position of the first nucleosome , overlapping the TSS 15 , makes such a mechanism unlikely ., However and as we show herein , the regulation of the chromatin configuration by DNA-binding proteins like Rap1 may also have an effect on elongation ., The Rap1-dependent control that we propose for RP genes should not be the only one acting at the elongation level across the yeast genome as we have shown that at least two other functional groups of genes , RiBi and mitochondria-related genes , display a regulated variation in the proportion of active RNA polymerases ., This variation is controlled by PKA , but does not depend on the silencing domain of Rap1 ., It is tentative to hypothesize that the PKA pathway regulates a plethora of genes during the elongation step of transcription by using different chromatin-related factors ., The yeast strains used in this work are described in Table S3 ., Cells were grown in YPD ( yeast extract 1% , peptone 2% , glucose 2% ) with agitation at 28°C , at OD600\u200a=\u200a0 . 5 ., In the experiment done in the galactose medium , cells were harvested and changed to YPGal ( yeast extract 1% , peptone 2% , galactose 2% ) and grown for 2 h ( lag phase ) and 14 . 5 h ( exponential growth ) ., For the SPT16 shut-off experiments , 5 µg/ml doxycycline was added to exponentially growing SJY6 cells ( OD600\u200a=\u200a0 . 1 ) ., Since the experiments in this work were performed in rich media ( YPD ) , shorter times of incubation with doxycycline were required to reach the same level of Spt16 depletion described in 28 ., Control cells were harvested after 5 hours of mock treatment ., Genomic run-on ( GRO ) was done essentially as described in 29 ., See supplementary materials and methods in Text S1 ., To determine the intragenic distribution of elongating RNA pol II molecules , we used macroarrays containing 300 bp probes from the 5′ and 3′ ends of the coding regions of 377 yeast genes ., These 5′-3′ macroarrays were manufactured by printing PCR products onto a nylon Hybond N+ membrane , similarly to that described for whole genome ORF macroarrays 57 ., PCR products were obtained by using either yeast genomic DNA as a template and the primer pairs listed in Table S4 , or | Introduction, Results, Discussion, Materials and Methods | Transcription elongation by RNA polymerase II was often considered an invariant non-regulated process ., However , genome-wide studies have shown that transcriptional pausing during elongation is a frequent phenomenon in tightly-regulated metazoan genes ., Using a combination of ChIP-on-chip and genomic run-on approaches , we found that the proportion of transcriptionally active RNA polymerase II ( active versus total ) present throughout the yeast genome is characteristic of some functional gene classes , like those related to ribosomes and mitochondria ., This proportion also responds to regulatory stimuli mediated by protein kinase A and , in relation to cytosolic ribosomal-protein genes , it is mediated by the silencing domain of Rap1 ., We found that this inactive form of RNA polymerase II , which accumulates along the full length of ribosomal protein genes , is phosphorylated in the Ser5 residue of the CTD , but is hypophosphorylated in Ser2 ., Using the same experimental approach , we show that the in vivo–depletion of FACT , a chromatin-related elongation factor , also produces a regulon-specific effect on the expression of the yeast genome ., This work demonstrates that the regulation of transcription elongation is a widespread , gene class–dependent phenomenon that also affects housekeeping genes . | Transcription of DNA–encoded information into RNA is the first step in gene regulation ., RNA polymerases initiate transcription at the promoter region and elongate the transcripts traveling throughout the gene until reaching the termination sequences ., Classical models of transcriptional regulation were focused on the initiation step , but there is increasing evidence for gene regulation after initiation ., We have investigated the importance of elongation in gene regulation using the yeast Saccharomyces cerevisiae , one of the main experimental systems in modern biology ., By comparing the genomic distribution of RNA polymerase molecules with the actual transcriptional signal across the genome , we have detected that many genes are regulated at the elongation level ., We show that yeast cells use this step to modulate the expression of several groups of genes , which have to be simultaneously regulated in a very coordinated manner ., Genes encoding essential functions , like those related to protein synthesis and respiration , change their transcriptional activities in response to environmental stimuli , without changing in the same extension the amount of RNA polymerase that is physically associated to them ., We also show that this kind of regulation , in spite of taking place during the elongation step , can be mediated by promoter-binding transcription factors . | genetics and genomics/gene expression, genetics and genomics/functional genomics, computational biology/transcriptional regulation, molecular biology/transcription elongation, biochemistry/transcription and translation | null |
journal.pgen.1003264 | 2,013 | Polygenic Modeling with Bayesian Sparse Linear Mixed Models | Both linear mixed models ( LMMs ) and sparse regression models are widely used in genetics applications ., For example , LMMs are often used to control for population stratification , individual relatedness , or unmeasured confounding factors when performing association tests in genetic association studies 1–9 and gene expression studies 10–12 ., They have also been used in genetic association studies to jointly analyze groups of SNPs 13 , 14 ., Similarly , sparse regression models have been used in genome-wide association analyses 15–20 and in expression QTL analysis 21 ., Further , both LMMs and sparse regression models have been applied to , and garnered renewed interest in , polygenic modeling in association studies ., Here , by polygenic modeling we mean any attempt to relate phenotypic variation to many genetic variants simultaneously ( in contrast to single-SNP tests of association ) ., The particular polygenic modeling problems that we focus on here are estimating “chip heritability” , being the proportion of variance in phenotypes explained ( PVE ) by available genotypes 19 , 22–24 , and predicting phenotypes based on genotypes 25–29 ., Despite the considerable overlap in their applications , in the context of polygenic modeling , LMMs and sparse regression models are based on almost diametrically opposed assumptions ., Precisely , applications of LMMs to polygenic modeling ( e . g . 22 ) effectively assume that every genetic variant affects the phenotype , with effect sizes normally distributed , whereas sparse regression models , such as Bayesian variable selection regression models ( BVSR ) 18 , 19 , assume that a relatively small proportion of all variants affect the phenotype ., The relative performance of these two models for polygenic modeling applications would therefore be expected to vary depending on the true underlying genetic architecture of the phenotype ., However , in practice , one does not know the true genetic architecture , so it is unclear which of the two models to prefer ., Motivated by this observation , we consider a hybrid of these two models , which we refer to as the “Bayesian sparse linear mixed model” , or BSLMM ., This hybrid includes both the LMM and a sparse regression model , BVSR , as special cases , and is to some extent capable of learning the genetic architecture from the data , yielding good performance across a wide range of scenarios ., By being “adaptive” to the data in this way , our approach obviates the need to choose one model over the other , and attempts to combine the benefits of both ., The idea of a hybrid between LMM and sparse regression models is , in itself , not new ., Indeed , models like these have been used in breeding value prediction to assist genomic selection in animal and plant breeding programs 30–35 , gene selection in expression analysis while controlling for batch effects 36 , phenotype prediction of complex traits in model organisms and dairy cattle 37–40 , and more recently , mapping complex traits by jointly modeling all SNPs in structured populations 41 ., Compared with these previous papers , our work makes two key contributions ., First , we consider in detail the specification of appropriate prior distributions for the hyper-parameters of the model ., We particularly emphasize the benefits of estimating the hyper-parameters from the data , rather than fixing them to pre-specified values to achieve the adaptive behavior mentioned above ., Second , we provide a novel computational algorithm that exploits a recently described linear algebra trick for LMMs 8 , 9 ., The resulting algorithm avoids ad hoc approximations that are sometimes made when fitting these types of model ( e . g . 37 , 41 ) , and yields reliable results for datasets containing thousands of individuals and hundreds of thousands of markers ., ( Most previous applications of this kind of model involved much smaller datasets . ), Since BSLMM is a hybrid of two widely used models , it naturally has a wide range of potential uses ., Here we focus on its application to polygenic modeling for genome-wide association studies , specifically two applications of particular interest and importance: PVE estimation ( or “chip heritability” estimation ) and phenotype prediction ., Estimating the PVE from large-scale genotyped marker sets ( e . g . all the SNPs on a genome-wide genotyping chip ) has the potential to shed light on sources of “missing heritability” 42 and the underlying genetic architecture of common diseases 19 , 22–24 , 43 ., And accurate prediction of complex phenotypes from genotypes could ultimately impact many areas of genetics , including applications in animal breeding , medicine and forensics 27–29 , 37–40 ., Through simulations and applications to real data , we show that BSLMM successfully combines the advantages of both LMMs and sparse regression , is robust to a variety of settings in PVE estimation , and outperforms both models , and several related models , in phenotype prediction ., We begin by considering a simple linear model relating phenotypes to genotypes : ( 1 ) ( 2 ) Here is an -vector of phenotypes measured on individuals , is an matrix of genotypes measured on the same individuals at genetic markers , is a -vector of ( unknown ) genetic marker effects , is an -vector of 1 s , is a scalar representing the phenotype mean , and is an -vector of error terms that have variance ., denotes the -dimensional multivariate normal distribution ., Note that there are many ways in which measured genotypes can be encoded in the matrix ., We assume throughout this paper that the genotypes are coded as 0 , 1 or 2 copies of a reference allele at each marker , and that the columns of are centered but not standardized; see Text S1 ., A key issue is that , in typical current datasets ( e . g . GWAS ) , the number of markers is much larger than the number of individuals ., As a result , parameters of interest ( e . g . or PVE ) cannot be estimated accurately without making some kind of modeling assumptions ., Indeed , many existing approaches to polygenic modeling can be derived from ( 1 ) by making specific assumptions about the genetic effects ., For example , the LMM approach from 22 , which has recently become commonly used for PVE estimation ( e . g . 24 , 44–46 ) , is equivalent to the assumption that effect sizes are normally distributed , such that ( 3 ) Specifically , exact equivalence holds when the relatedness matrix in the LMM is computed from the genotypes as ( e . g . 47 ) ., 22 use a matrix in this form , with centered and standardized , and with a slight modification of the diagonal elements ., For brevity , in this paper we refer to the regression model that results from this assumption as the LMM ( note that this is relatively restrictive compared with the usual definition ) ; it is also commonly referred to as “ridge regression” in statistics 48 ., The estimated combined effects ( ) , or equivalently , the estimated random effects , obtained from this model are commonly referred to as Best Linear Unbiased Predictors ( BLUP ) 49 ., An alternative assumption , which has also been widely used in polygenic modeling applications 18 , 19 , 34 , and more generally in statistics for sparse high-dimensional regression with large numbers of covariates 50 , 51 , is that the effects come from a mixture of a normal distribution and a point mass at 0 , also known as a point-normal distribution: ( 4 ) where is the proportion of non-zero and denotes a point mass at zero ., This definition of follows the convention from statistics 19 , 50 , 51 , which is opposite to the convention in animal breeding 27 , 32–34 , 40 ., We refer to the resulting regression model as Bayesian Variable Selection Regression ( BVSR ) , because it is commonly used to select the relevant variables ( i . e . those with non-zero effect ) for phenotype prediction ., Although ( 4 ) formally includes ( 3 ) as a special case when , in practice ( 4 ) is often used together with an assumption that only a small proportion of the variables are likely to be relevant for phenotype prediction , say by specifying a prior distribution for that puts appreciable mass on small values ( e . g . 19 ) ., In this case , BVSR and LMM can be viewed as making almost diametrically opposed assumptions: the LMM assumes every variant has an effect , whereas BVSR assumes that a very small proportion of variants have an effect ., ( In practice , the estimate of under LMM is often smaller than the estimate of under BVSR , so we can interpret the LMM as assuming a large number of small effects , and BVSR as assuming a small number of larger effects . ), A more general assumption , which includes both the above as special cases , is that the effects come from a mixture of two normal distributions: ( 5 ) Setting yields the LMM ( 3 ) , and yields BVSR ( 4 ) ., we can interpret this model as assuming that all variants have at least a small effect , which are normally distributed with variance , and some proportion ( ) of variants have an additional effect , normally distributed with variance ., The earliest use of a mixture of two normal distributions for the regression coefficients that we are aware of is 52 , although in that paper various hyper-parameters were fixed , and so it did not include LMM and BVSR as special cases ., Of the three assumptions on the effect size distributions above , the last ( 5 ) is clearly the most flexible since it includes the others as special cases ., Obviously other assumptions are possible , some still more flexible than the mixture of two normals: for example , a mixture of three or more normals ., Indeed , many other assumptions have been proposed , including variants in which a normal distribution is replaced by a distribution ., These variants include the “Bayesian alphabet models” – so-called simply because they have been given names such as BayesA , BayesB , BayesC , etc ., – that have been proposed for polygenic modeling , particularly breeding value prediction in genomic selection studies ., Table 1 summarizes these models , and some other effect size distributions that have been proposed , together with relevant references ( see also 53 and the references there in ) ., Among these , the models most closely related to ours are BayesC 34 and BayesR 35 ., Specifically , BayesC without a random effect is BVSR , and with a random effect is BSLMM ( which we define below ) ., BayesR models effect sizes using a mixture of three normal components plus a point mass at zero , although the relative variance for each normal distribution is fixed ., Given the wide range of assumptions for effect size distributions that have been proposed , it is natural to wonder which are the most appropriate for general use ., However , answering this question is complicated by the fact that even given the effect size distribution there are a number of different ways that these models can be implemented in practice , both in terms of statistical issues , such as treatment of the hyper-parameters , and in terms of computational and algorithmic issues ., Both these types of issues can greatly affect practical performance ., For example , many approaches fix the hyper-parameters to specific values 27 , 32 , 33 , 40 which makes them less flexible 34 , 54 ., Thus , in this paper we focus on a particular effect size distribution ( 5 ) , which while not the most flexible among all that could be considered , is certainly more flexible than the one that has been most used in practice for estimating PVE ( LMM ) , and admits certain computational methods that could not be applied in all cases ., We consider in detail how to apply this model in practice , and the resulting advantages over LMM and BVSR ( although we also compare with some other existing approaches ) ., A key contribution of this paper is to provide new approaches to address two important practical issues: the statistical issue of how to deal with the unknown hyper-parameters , and the computational issue of how to fit the model ., Notably , with the computational tools we use here , fitting the model ( 5 ) becomes , for a typical dataset , less computationally intensive than fitting BVSR , as well as providing gains in performance ., With this background , we now turn to detailed description of the model , its prior specification and its computation algorithm ., In this paper we focus on the simple linear model ( 1 ) with mixture prior ( 5 ) on the effects ., However , the computational and statistical methods we use here also apply to a more general model , which we refer to as the Bayesian Sparse Linear Mixed Model ( BSLMM ) , and which includes the model ( 1 ) with ( 5 ) as a special case ., The BSLMM consists of a standard linear mixed model , with one random effect term , and with sparsity inducing priors on the regression coefficients: ( 6 ) ( 7 ) ( 8 ) ( 9 ) where is an -vector of random effects with known covariance matrix ., In referring to as the “random effects” we are following standard terminology from LMMs ., Standard terminology also refers to the coefficients as “fixed effects” , but this phrase has a range of potential meanings 55 and so we avoid it here ., Instead we use the term “sparse effects” for these parameters to emphasize the sparsity-inducing prior ., It is straightforward to show that when , BSLMM is equivalent to the simple linear model ( 1 ) with mixture prior ( 5 ) on the effects ., However , our discussion of prior specification , computational algorithms , and software , all apply for any ., When we say that ( 6 ) is equivalent to ( 1 ) with ( 5 ) , this equivalence refers to the implied probability model for given and the hyper-parameters ., However , and are not equivalent ( explaining our use of two different symbols ) : in ( 6 ) the random effect captures the combined small effects of all markers , whereas in ( 1 ) these small effects are included in ., Since both our applications focus on the relationship between and , and not on interpreting estimates of or , we do not concern ourselves any further with this issue , although it may need consideration in applications where individual estimated genetic effects are of more direct interest ( e . g . genetic association mapping ) ., A related issue is the interpretation of the random effect in BSLMM: from the way we have presented the material is most naturally interpreted as representing a polygenic component , specifically the combined effect of a large number of small effects across all measured markers ., However , if there are environmental factors that influence the phenotype and are correlated with genotype ( e . g . due to population structure ) , then these would certainly affect estimates of , and consequently also affect estimates of other quantities , including the PVE ., In addition , phenotype predictions from BSLMM will include a component due to unmeasured environmental factors that are correlated with measured genotypes ., These issues are , of course , not unique to BSLMM – indeed , they apply equally to the LMM; see 56 and the response from 57 for relevant discussion ., Finally , given the observation that a mixture of two normals is more flexible than a point-normal , it might seem natural to consider modifying ( 6 ) by making the assumption that comes from a mixture of two normal distributions rather than a point-normal ., However , if then this modification is simply equivalent to changing the values of ., The BSLMM involves ( hyper- ) parameters , , and ., Before considering prior specification for these parameters , we summarize their interpretations as follows: Appropriate values for these parameters will clearly vary for different datasets , so it seems desirable to estimate them from the data ., Here we accomplish this in a Bayesian framework by specifying prior distributions for the parameters , and using Markov chain Monte Carlo ( MCMC ) to obtain approximate samples from their posterior distribution given the observed data ., Although one could imagine instead using maximum likelihood to estimate the parameters , the Bayesian framework has several advantages here: for example , it allows for incorporation of external information ( e . g . that most genetic markers will , individually , have small effects ) , and it takes into account of uncertainty in parameter estimates when making other inferences ( e . g . phenotype prediction ) ., For the mean and the inverse of error variance , , we use the standard conjugate prior distributions: ( 10 ) ( 11 ) where and denote , respectively , shape and rate parameters of a Gamma distribution ., Specifically we consider the posterior that arises in the limits , and ., These limits correspond to improper priors , but the resulting posteriors are proper , and scale appropriately with shifting or scaling of the phenotype vector 58 ., In particular , these priors have the property that conclusions will be unaffected by changing the units of measurement of the phenotype , which seems desirable for a method intended for general application ., Prior specification for the remaining hyper-parameters is perhaps more important ., Our approach is to extend the prior distributions for BVSR described in 19 ., Following 19 we place a uniform prior on : ( 12 ) where is total number of markers being analyzed ., The upper and lower limit of this prior were chosen so that ( the expected proportion of markers with non-zero ) ranges from to ., A uniform prior on reflects the fact that uncertainty in in a typical GWAS will span orders of magnitude ., A common alternative ( see e . g . 18 , 34 ) is a uniform distribution on , but as noted in 19 this puts large weight on large numbers of markers having non-zero ( e . g . it would correspond to placing 50% prior probability to the event that more than half of the markers have non-zero , and correspond to placing 90% prior probability to the event that more than 10% of the markers have non-zero ) ., To specify priors for and , we exploit the following idea from 19: prior specification is easier if we first re-parameterize the model in terms of more interpretable quantities ., Specifically we extend ideas from 19 to re-parameterize the model in terms of the ( expected ) proportion of phenotypic variance explained by the sparse effects and by the random effects ., To this end , we define PVE ( the total proportion of variance in phenotype explained by the sparse effects and random effects terms together ) and PGE ( the proportion of genetic variance explained by the sparse effects terms ) as functions of , and : ( 13 ) ( 14 ) where the function V ( x ) is defined as ( 15 ) These definitions ensure that both PVE and PGE must lie in the interval ., PVE reflects how well one could predict phenotype from the available SNPs if one knew the optimal as well as the random effects ; together with PVE , PGE reflects how well one could predict phenotype using alone ., Since PVE and PGE are functions of , whose distributions depend on hyper-parameters , the prior distribution for PVE and PGE depends on the priors assigned to these hyper-parameters ., In brief , our aim is to choose priors for the two hyper-parameters and so that the induced priors on both PVE and PGE are roughly uniform on 0 and 1 ., ( Other distributions could be chosen if desired , but we consider this uniform distribution one reasonable default . ), However , because the relationship between the distribution of PVE , PGE and the hyper-parameters is not simple , we have to make some approximations ., Specifically , we introduce as approximations ( they are ratios of expectations rather than expectations of ratios ) to the expectations of PVE and PGE , respectively: ( 16 ) ( 17 ) where is the average variance of genotypes across markers , and is the mean of diagonal elements in ., In other words , and , where and are the th elements of matrices and , respectively ., See Text S1 for derivations ., Intuitively , the term captures the expected genetic variance contributed by the sparse effects term ( relative to the error variance ) , because is the expected number of causal markers , is the expected effect size variance of each causal marker ( relative to the error variance ) , and is the average variance of marker genotypes ., Similarly , the term captures the expected genetic variance contributed by the random effects term ( relative to the error variance ) , because is the expected variance of the random effects ( relative to the error variance ) when the relatedness matrix has unit diagonal elements , while properly scales it when not ., The parameter provides a natural bridge between the LMM and BVSR: when BSLMM becomes the LMM , and when BSLMM becomes BVSR ., In practice , when the data favors the LMM , the posterior distribution of would mass near 0 , and when the data favors BVSR , the posterior distribution of would mass near 1 ., In summary , the above re-parameterizes the model in terms of instead of ., Now , instead of specifying prior distributions for , we rather specify prior distributions for ., Specifically we use uniform prior distributions for : ( 18 ) ( 19 ) independent of one another and of ., Since and approximate PVE and PGE , these prior distributions should lead to reasonably uniform prior distributions for PVE and PGE , which we view as reasonable defaults for general use ., ( If one had specific information about PVE and PGE in a given application then this could be incorporated here . ), In contrast it seems much harder to directly specify reasonable default priors for ( although these priors on do of course imply priors for ; see Text S1 ) ., Note that we treat and as approximations to PVE and PGE only for prior specification; when estimating PVE and PGE from data we do so directly using their definitions ( 13 ) and ( 14 ) ( see below for details ) ., To facilitate computation , we use the widely-used approach from 52 of introducing a vector of binary indicators that indicates whether the corresponding coefficients are non-zero ., The point-normal priors for can then be written ( 20 ) ( 21 ) ( 22 ) where denotes the sub-vector of corresponding to the entries ; denotes the sub-vector of corresponding to the other entries , ; and denotes the number of non-zero entries in ., We use MCMC to obtain posterior samples of ( ) on the product space , which is given by ( 23 ) The marginal likelihood can be computed analytically integrating out ; see below for further details ., We use a Metropolis-Hastings algorithm to draw posterior samples from the above marginal distribution ., In particular , we use a rank based proposal distribution for 19 , which focus more of the computational time on examining SNPs with stronger marginal associations ., We use the resulting sample from the posterior distribution ( 23 ) to estimate PVE and PGE as follows ., For each sampled value of , we sample a corresponding value for from the conditional distribution ., We then use each sampled value of to compute a sampled value of PVE and PGE , using equations ( 13 ) and ( 14 ) ., We estimate the posterior mean and standard deviation of PVE , PGE , from these samples ., The novel part of our algorithm is a new efficient approach to evaluating the likelihood that considerably reduces the overall computational burden of the algorithm ., The direct naive approach to evaluating this likelihood involves a matrix inversion and a matrix determinant calculation that scale cubically with the number of individuals , and this cost is incurred every iteration as hyper parameter values change ., Consequently , this approach is impractical for typical association studies with large , and ad hoc approximations are commonly used to reduce the burden ., For example , both 37 and 41 fix to some pre-estimated value ., As we show later , this kind of approximation can reduce the accuracy of predicted phenotypes ., Here , we avoid such approximations by exploiting recently developed computational tricks for LMMs 8 , 9 ., The key idea is to perform a single eigen-decomposition and use the resulting unitary matrix ( consisting of all eigen vectors ) to transform both phenotypes and genotypes to make the transformed values follow independent normal distributions ., By extending these tricks to BSLMM we evaluate the necessary likelihoods much more efficiently ., Specifically , after a single operation at the start of the algorithm , the per iteration computational burden is linear in ( the same as BVSR ) , allowing large studies to be analyzed ., Full details of the sampling algorithm appear in Text S2 ., Software implementing our methods is included in the GEMMA software package , which is freely available at http://stephenslab . uchicago . edu/software . html ., Both the LMM and BVSR have been used to estimate the PVE 19 , 22 ., Since the LMM assumes that all SNPs have an effect , while BVSR assumes that only a small proportion of SNPs have an effect , we hypothesize that BVSR will perform better when the true underlying genetic structure is sparse and LMM will perform better when the true genetic structure is highly polygenic ., Further , because BSLMM includes both as special cases , we hypothesize that BSLMM will perform well in either scenario ., To test these hypotheses , we perform a simulation using real genotypes at about 300 , 000 SNPs in 3 , 925 Australian individuals 22 , and simulate phenotypes under two different scenarios ., In Scenario I we simulate a fixed number of causal SNPs ( with ) , with effect sizes coming from a standard normal distribution ., These simulations span a range of genetic architectures , from very sparse to highly polygenic ., In Scenario II we simulate two groups of causal SNPs , the first group containing a small number of SNPs of moderate effect ( or ) , plus a second larger group of SNPs of small effect representing a “polygenic component” ., This scenario might be considered more realistic , containing a mix of small and larger effects ., For both scenarios we added normally-distributed errors to phenotype to create datasets with true PVE\u200a=\u200a0 . 6 and 0 . 2 ( equation 13 ) ., We simulate 20 replicates in each case , and run the algorithms with all SNPs , including the simulated causal variants , so that the true PVE for typed markers is either 0 . 6 or 0 . 2 ( if we excluded the causal variants then the true PVE would be unknown ) ., Figure 1A and 1C , show the root of mean square error ( RMSE ) of the PVE estimates obtained by each method , and Figure 1B and 1D summarize the corresponding distributions of PVE estimates ., In agreement with our original hypotheses , BVSR performs best ( lowest RMSE ) when the true model is sparse ( e . g . Scenario I , or in Figure 1A , 1C ) ., However , it performs very poorly under all the other , more polygenic , models ., This is due to a strong downward bias in its PVE estimates ( Figure 1B , 1D ) ., Conversely , under the same scenarios , LMM is the least accurate method ., This is because the LMM estimates have much larger variance than the other methods under these scenarios ( Figure 1B , 1D ) , although , interestingly , LMM is approximately unbiased even in these settings where its modeling assumptions are badly wrong ., As hypothesized , BSLMM is robust across a wider range of settings than the other methods: its performance lies between LMM and BVSR when the true model is sparse , and provides similar accuracy to LMM when not ., Of course , in practice , one does not know in advance the correct genetic architecture ., This makes the stable performance of BSLMM across a range of settings very appealing ., Due to the poor performance of BVSR under highly polygenic models , we would not now recommend it for estimating PVE in general , despite its good performance when its assumptions are met ., We also compare the three methods on their ability to predict phenotypes from genotypes , using the same simulations ., To measure prediction performance , we use relative prediction gain ( RPG; see Text S1 ) ., In brief , RPG is a standardized version of mean square error: RPG\u200a=\u200a1 when accuracy is as good as possible given the simulation setup , and RPG\u200a=\u200a0 when accuracy is the same as simply predicting everyone to have the mean phenotype value ., RPG can be negative if accuracy is even worse than that ., Figure 2 compares RPG of different methods for simulations with PVE\u200a=\u200a0 . 6 ( results for PVE\u200a=\u200a0 . 2 are qualitatively similar , not shown ) ., Interestingly , for phenotype prediction , the relative performance of the methods differs from results for PVE estimation ., In particular , LMM performs poorly compared with the other two methods in all situations , except for Scenario I with , the Scenario that comes closest to matching the underlying assumptions of LMM ., As we expect , BSLMM performs similarly to BVSR in scenarios involving smaller numbers of causal SNPs ( up to in Scenario I ) , and outperforms it in more polygenic scenarios involving large numbers of SNPs of small effect ( e . g . Scenario II ) ., This is presumably due to the random effect in BSLMM that captures the polygenic component , or , equivalently , due to the mixture of two normal distributions in BSLMM that better captures the actual distribution of effect sizes ., The same qualitative patterns hold when we redo these simulation comparisons excluding the causal SNPs ( Figure S1 ) or use correlation instead of RPG to assess performance ( Figure S2 and Figure S3 ) ., For a potential explanation why LMM performs much less well for phenotype prediction than for PVE estimation , we note the difference between these two problems: to accurately estimate PVE it suffices to estimate the “average” effect size reliably , whereas accurate phenotype prediction requires accurate estimates of individual effect sizes ., In situations where the normal assumption on effect sizes is poor , LMM tends to considerably underestimate the number of large effects , and overestimate the number of small effects ., These factors can cancel one another out in PVE estimation , but both tend to reduce accuracy of phenotype prediction ., To obtain further insights into differences between LMM , BVSR and BSLMMM , we apply all three methods to estimate the PVE for five traits in two human GWAS datasets ., The first dataset contains height measurements of 3 , 925 Australian individuals with about 300 , 000 typed SNPs ., The second dataset contains four blood lipid measurements , including high-density lipoprotein ( HDL ) , low-density lipoprotein ( LDL ) , total cholesterol ( TC ) and triglycerides ( TG ) from 1 , 868 Caucasian individuals with about 550 , 000 SNPs ., The narrow sense heritability for height is estimated to be 0 . 8 from twin-studies 22 , 59 ., The narrow sense heritabilities for the lipid traits have been estimated , in isolated founder populations , to be 0 . 63 for HDL , 0 . 50 for LDL , 0 . 37 for TG in Hutterites 60 , and 0 . 49 for HDL , 0 . 42 for LDL , 0 . 42 for TC and 0 . 32 for TG in Sardinians 61 ., Table 2 shows PVE estimates from the three methods for the five traits ., PVE estimates from BVSR are consistently much smaller than those obtained by LMM and BSLMM , which are almost identical for two traits and similar for the others ., Estimates of PVE from both LMM and BSLMM explain over 50% of the narrow sense heritability of the five traits , suggesting that a sizable proportion of heritability of these traits can be explained , either directly or indirectly , by available SNPs ., These results , with LMM and BSLMM providing similar estimates of PVE , and estimates from BVSR being substantially lower , are consistent with simulation results for a trait with substantial polygenic component ., One feature of BSLMM , not possessed by the other two methods , is that it can be used to attempt to quantify the relative contribution of a polygenic component , through estimation of PGE , which is the proportion of total genetic variance exp | Introduction, Methods, Results, Discussion | Both linear mixed models ( LMMs ) and sparse regression models are widely used in genetics applications , including , recently , polygenic modeling in genome-wide association studies ., These two approaches make very different assumptions , so are expected to perform well in different situations ., However , in practice , for a given dataset one typically does not know which assumptions will be more accurate ., Motivated by this , we consider a hybrid of the two , which we refer to as a “Bayesian sparse linear mixed model” ( BSLMM ) that includes both these models as special cases ., We address several key computational and statistical issues that arise when applying BSLMM , including appropriate prior specification for the hyper-parameters and a novel Markov chain Monte Carlo algorithm for posterior inference ., We apply BSLMM and compare it with other methods for two polygenic modeling applications: estimating the proportion of variance in phenotypes explained ( PVE ) by available genotypes , and phenotype ( or breeding value ) prediction ., For PVE estimation , we demonstrate that BSLMM combines the advantages of both standard LMMs and sparse regression modeling ., For phenotype prediction it considerably outperforms either of the other two methods , as well as several other large-scale regression methods previously suggested for this problem ., Software implementing our method is freely available from http://stephenslab . uchicago . edu/software . html . | The goal of polygenic modeling is to better understand the relationship between genetic variation and variation in observed characteristics , including variation in quantitative traits ( e . g . cholesterol level in humans , milk production in cattle ) and disease susceptibility ., Improvements in polygenic modeling will help improve our understanding of this relationship and could ultimately lead to , for example , changes in clinical practice in humans or better breeding/mating strategies in agricultural programs ., Polygenic models present important challenges , both at the modeling/statistical level ( what modeling assumptions produce the best results ) and at the computational level ( how should these models be effectively fit to data ) ., We develop novel approaches to help tackle both these challenges , and we demonstrate the gains in accuracy that result in both simulated and real data examples . | genome-wide association studies, animal genetics, statistics, quantitative traits, mathematics, biostatistics, biology, heredity, genetic association studies, genetics, human genetics, genetics of disease, statistical methods, genetics and genomics, complex traits | null |
journal.pcbi.1003313 | 2,013 | Computational Protein Design Quantifies Structural Constraints on Amino Acid Covariation | Evolutionary selective pressures on protein structure and function have shaped the sequences of todays naturally occurring proteins 1–3 ., As a result of these pressures , sequences of natural proteins are close to optimal for their structures 4 ., Natural protein sequences therefore provide an excellent test for computational protein design methods , where the goal is to predict protein sequences that are optimal for a desired protein structure and function 5 ., It is often assumed that given a natural polypeptide backbone conformation , an accurate protein design algorithm should be able to predict sequences that are similar to the natural protein sequence ., This test is commonly referred to as native sequence recovery 4 and it has been used extensively to evaluate various protein design sampling methods and energy functions 6–8 ., Beyond simply recovering the native sequence , a further challenge in computational protein design is to predict the set of tolerated sequences that are compatible with a given protein fold and function 9–13 ., Predicting sequence tolerance is important for applications such as characterizing mutational robustness 14 , 15 , predicting the specificity of molecular interactions 16–20 , and designing libraries of proteins with altered functions 21 , 22 ., Recent methods developed for this goal involve generating an ensemble of backbone structures similar to the native structure and then designing low energy sequences for the different structures in the ensemble 9 , 16 , 19 , 23–25 ., These flexible backbone design methods can produce sequences that are highly divergent from the native sequence but may still fold into the desired structure , which makes simple native sequence recovery a poor indicator for the accuracy of these methods ., A more useful computational test of these approaches involves comparing designed sequences with a set of reference sequences , either naturally occurring or experimentally derived , that share the desired protein fold ., This comparison can be based on sequence profile similarity , which involves quantifying the difference between the frequencies of observing each amino acid at corresponding positions in the designed and reference sequences 16 , 17 , 19 ., While high similarity between designed and reference sequence profiles can be informative to gauge the accuracy of a protein design method , it does not guarantee that the method will predict sequences that fold into the desired structure ., This is because sequence profile comparisons evaluate amino acid positions independently from each other and therefore ignore the details of amino acid interactions that are critical for protein structure and function ., Naturally occurring protein structures are formed cooperatively and each amino acid can physically interact with multiple neighboring amino acids ., Evolutionary selective pressures have acted upon these interactions , resulting in the patterns of amino acid covariation that can be observed within todays naturally occurring protein families ., Accordingly , previous studies have used information theoretic methods to detect amino acid covariation in multiple sequence alignments of many different protein families 26–28 and have used contact prediction based on covariation to dramatically improve the accuracy of protein structure modeling 29 ., Despite the clear occurrence of amino acid covariation in natural protein sequences , the extent to which different selective pressures have shaped amino acid covariation in diverse protein families is unknown ., Additionally , it is difficult to dissect to what extent phylogenetic bias has influenced the observations of amino acid covariation ., Previous work has indicated that networks of covarying amino acids play a role in allosterically linking distant functional sites , suggesting that amino acid covariation is driven by protein functional constraints 30 , 31 ., However , other studies have shown in two test cases that computational protein design can recapitulate naturally occurring covariation in the cores of SH3 domains 4 , 13 , 32 and for two-component signaling systems 33 ., These results indicate that constraints imposed by protein structure have played a role in producing the covariation in the studied examples , but it has not yet been shown that these observations are general ., In this paper , we use computational protein design to measure the extent to which protein structure has shaped amino acid covariation in a diverse set of 40 protein domains ., Since computational protein design predicts sequences that are energetically optimal based on protein structure alone , we expect that pairs of amino acids that highly covary in both designed and natural sequences to have likely covaried to maintain protein structure ., We find significant overlap in the sets of highly covarying amino acid pairs between designed and natural sequences for all 40 domains examined , suggesting that maintenance of protein structure is a dominant selective pressure that constrains the evolution of amino acid interactions in proteins ., Our analysis furthermore quantifies the extent to which different types of interactions explain the observed covariation ., Finally , we demonstrate the utility of amino acid covariation recapitulation as a sensitive test for evaluating different protein design methods ., We find that flexible backbone design significantly improves covariation recapitulation relative to fixed backbone design and that recapitulation of amino acid covariation is exquisitely sensitive to both the magnitude and mechanism of backbone flexibility ., Taken together , these results provide fundamental insights into the physical nature of amino acid co-evolution and , more practically , provide a new benchmark that may help improve the accuracy of computational protein design methods ., To compare amino acid covariation in natural and predicted designed protein sequences , we selected 40 protein domains that were diverse with respect to their secondary structure composition and fold class ( Table 1 ) ., We then quantified natural amino acid covariation for each domain by creating a multiple sequence alignment for the domain , followed by computing covariation between every pair of columns in the multiple sequence alignment by using a mutual information based method 28 ( see Methods ) ., Pairs of amino acid positions with a covariation score that is two standard deviations above the mean or greater were considered to be highly covarying pairs ., We predicted designed protein sequences for each of the 40 domains using RosettaDesign 4 , 34 ., We first used the standard RosettaDesign fixed backbone protocol 34 , which takes a crystal structure as input and runs Monte Carlo simulated annealing , to predict 500 designed sequences for each domain structure ., We then quantified amino acid covariation in the designed sequences and compared it to natural amino acid covariation for each domain ., We calculated the similarity between designed and natural covariation based on the percent overlap of the highly covarying pairs in each set ( see Methods ) ., We found this overlap to be significant ( p<0 . 001 ) for all 40 domains ( Table S1 ) ., Given the observation that fixed backbone protein design can recapitulate a significant fraction of naturally covarying amino acid pairs , we next aimed to understand how incorporating backbone flexibility into the design protocol affects this recapitulation ., To accomplish this , we generated a conformational ensemble of 500 backbone structures for each domain using the “backrub” method 35 in Rosetta 36 , which iteratively applies local backbone perturbations throughout the protein structure combined with adjustments in side-chain conformations ., We then used RosettaDesign to predict a low energy sequence for each backbone structure in the ensemble , resulting in 500 designed sequences ., Figure 1 shows a flow chart of this approach applied to an SH3 domain ., To investigate the effect of the magnitude of backbone flexibility in the design protocol , we varied the temperature parameter in the Monte Carlo simulations used in the backrub protocol to generate conformational ensembles with different amounts of structural variation ( Figure 2A ) ., We designed sequences for each ensemble ( kT\u200a=\u200a0 . 3 , 0 . 6 , 0 . 9 , 1 . 2 , 1 . 8 , 2 . 4 ) and quantified similarity to natural covariation for each set of sequences ., We compared these results with sequences designed using the fixed backbone design protocol described above ( “Fixed” ) ., Figure 2B shows a significant increase in covariation similarity for the flexible backbone simulations relative to the fixed backbone simulation ., Moreover , the distributions of covariation similarity for the 40 domains show that there is an optimal degree of structural variation , as low-temperature and high-temperature simulations perform significantly worse than mid-temperature simulations ( Table S2 ) ., We observed this same trend when we repeated this analysis using a different method for quantifying covariation 37 ( Figure S1 ) , suggesting that our results are not dependent on the method used to quantify covariation ., To better understand the basis of this trend , we examined several other sequence and structural characteristics: sequence recovery , sequence profile similarity , sequence entropy and structural variation ( see Methods ) ., The resulting distributions for these characteristics are shown in Figure 2C ., Sequence entropy and sequence profile similarity showed similar trends to covariation similarity ( sequence entropy is most similar to natural sequences and profile similarity is highest at 0 . 9 kT ) , suggesting that backbone flexibility allows for sampling diverse sequences with native-like properties ., These trends are consistent with the observation that sequence recovery decreases with increasing amounts of backbone flexibility ., As diversity within a set of sequences increases , those sequences tend to become more dissimilar to any individual sequence , including the native sequence of the crystal structure used as input for design ., Structural variation in the 0 . 3 , 0 . 6 , 0 . 9 and 1 . 2 kT simulations is less than the structural variation among naturally occurring protein structures with these domains , which could be due to the fact that natural proteins use additional mechanisms of generating structural variation that are not being modeled , such as the insertion or deletion of amino acids in loop regions ., Taken together , these results suggest that a moderate degree of backbone flexibility allows for the accommodation of sequences that differ from the native sequence and yet are similar to naturally occurring sequences with respect to their sequence profiles , sequence entropies and patterns of amino acid covariation ., Next we examined whether or not these results were specific to the method used to generate the conformational ensembles for design ., We tested two other Monte Carlo based methods that iteratively perform perturbations to the backbone ., One method performed Kinematic Closure ( “KIC” ) , which involves randomizing phi/psi torsions in a local region of the backbone while keeping the rest of the backbone fixed , thus introducing a chain break , and then using inverse kinematics to solve for the torsions that will close the chain 38 ., The other method performs potentially non-local moves by perturbing the phi and psi torsions of residues by a random small angle ( “Small” ) 39 ., We ran both of these methods for the same number of trials and for the same values of kT as the backrub protocol ., The resulting distributions of covariation similarity show the same trend we observed previously with the backrub simulations , where mid-range temperature simulations result in an optimal degree of covariation similarity ( Figure S2 ) ., While the optimal simulation temperature parameter was comparable for each of the methods tested , the methods achieved a different optimum level of covariation similarity with the natural sequences ., We found that the two local move simulations ( KIC and Backrub ) outperformed the non-local move simulation ( Small ) ., To test if this observation holds true more generally , we tested two additional methods of generating conformational ensembles that make non-local moves ., These methods included FastRelax ( “Relax” ) , which consists of multiple rounds of side-chain repacking and all-atom minimization while increasing the weight of the repulsive term in the Lennard–Jones ( LJ ) potential from 2% to 100% of its default value , and AbInitioRelax ( “AbRelax” ) , which performs fragment-based ab initio structure prediction followed by FastRelax 40 ., As an additional control , we also designed sequences using a fixed backbone structure with an energy function that dampens the weight of the repulsive LJ term ( “Soft” ) ., The resulting covariation similarity distributions show that recapitulation of natural amino acid covariation is sensitive to the method used to generate conformational ensembles ( Figure 3A ) ., Both local move simulations ( KIC , Backrub ) achieved higher median covariation similarities than the non-local move simulations ( Small , AbRelax , Relax ) and the fixed backbone simulations ( Fixed , Soft ) ( see Table S3 for p-values ) ., We also evaluated each of these methods using the other metrics described above: native sequence recovery , sequence profile similarity , sequence entropy and structural variation ( Figure 3B ) ., Unexpectedly , the AbRelax method , which resulted in conformational ensembles with the greatest structural variation , achieved the highest sequence profile similarity with the natural sequences of any method tested ., A possible explanation for this behavior is that local interactions are preserved in AbRelax generated structures , but the overall topology of the protein is incorrect ., To test this hypothesis , we examined covariation similarity in the AbRelax sequences by splitting all covarying pairs into the following two sets: pairs separated by fewer than 10 residues in sequence ( “Near” ) and pairs separated by greater than 10 residues in sequence ( “Far” ) ., This analysis revealed that whereas AbRelax sequences have relatively high covariation similarity with natural sequences for pairs close in sequence , they have low covariation similarity for pairs that are distant in sequence ( Figure 3C ) ., In contrast , covariation similarity for “near” and “far” pairs were similar for simulations using backrub ensembles ., These results suggest that AbRelax can model local interactions within a secondary structural element or between adjacent secondary structures , but it does not correctly capture non-local interactions that are likely critical for achieving a cooperatively folded , stable tertiary structure ., This observation demonstrates the importance of using amino acid covariation to evaluate the accuracy of protein design methods , since it is possible to obtain deceptively high sequence profile similarity scores with highly divergent tertiary structures as long as local interactions are maintained ., Of all the flexible backbone design methods tested , Backrub , kT\u200a=\u200a0 . 9 resulted in sequences most similar to the natural sequences with respect to covariation similarity and sequence profile similarity ., Using the assumption that a method that gives higher similarity to natural sequences will better capture the mechanisms underlying covariation , we used Backrub , kT\u200a=\u200a0 . 9 as the representative flexible backbone sequences for the remainder of the study ., To understand how backbone flexibility influences the extent of covariation similarity between designed and natural sequences , we identified all pairs of amino acid positions that highly covaried in both the natural sequences and a set of flexible backbone sequences ( Backrub , kT\u200a=\u200a0 . 9 ) but did not highly covary in the fixed backbone sequences ., We then took all pairs of amino acids at these positions that were not sampled in the fixed backbone simulation and designed them onto the crystal structure backbone using fixed backbone design ., For each pair of these positions , we calculated mean interaction energies and compared these energies between fixed and flexible backbone design structures ( Figure 4A ) ., We calculated both one-body energies , which include the interaction of an amino acid residue with itself , and two-body energies , which include the interactions between two amino acid residues in the protein ( see Text S1 for description of the components of Rosetta one-body and two-body energies ) ., We found both the one-body and two-body energies of these pairs to be generally greater in the context of fixed backbones relative to flexible backbones ., Splitting the energies into their component terms revealed that the backbone-dependent Dunbrack rotamer energy ( fa_dun ) and Lennard-Jones repulsive ( fa_rep ) terms resulted in greater energy increases in the one-body and two-body energies , respectively , than any other term in the energy function ( Figure S3 ) ., These results suggest that amino acid pairs that covary in flexible backbone simulations but do not covary in fixed backbone simulations generally cannot be accommodated on fixed backbones without resulting in steric clashes or rotamers that are unfavorable for the given backbone ., Simply modifying the energy function by using a “soft” repulsive potential that reduces the energy of clashes does not increase sequence diversity or covariation similarity ( Figure 3B ) , suggesting that backbone movements are required to accommodate these amino acid interactions ., Figure 4B shows representative cases where some degree of backbone flexibility is required to correctly model the precise interaction details of specific amino acid pairings ., We have thus far compared amino acid covariation between natural and predicted designed sequences based on the extent of overlap between the sets of highly covarying pairs ., However , it is also important to consider the amino acid pair propensities at covarying positions to test whether the natural and designed covarying pairs utilize the same types of amino acid interactions ., To accomplish this , we calculated amino acid propensities at pairs of positions that covary in both the natural and designed sequences ( Figure 5A ) ., Over-represented amino acid pairs in both designed and natural sequences included those with opposite charges , hydrophobic pairs and hydrogen-bonding pairs ., Differences in the designed and natural amino acid pair propensities included the over-representation of cation-pi pairs in the natural sequences but not in the designed sequences ( such as W-R ) ., These differences highlight shortcomings of the energy function used for design , which does not currently account for cation-pi interactions ., To quantify the similarity between the natural and designed covarying pair propensities , we calculated the correlation coefficients between the natural and designed propensities for all sets of designed sequences ., We found these correlations to be dependent on both the magnitude and mechanism of backbone flexibility , as we previously observed with the overlap in covarying pairs ( Table S4 ) ., The comparison between natural and designed pair propensities for fixed backbone sequences and for a set of flexible backbone sequences ( Backrub , kT\u200a=\u200a0 . 9 ) are shown in Figure 5B , again supporting the conclusion that backbone flexibility improves recapitulation of amino acid covariation ., While similar pair propensities between natural and designed covarying pairs demonstrate that the same types of amino acid interactions occur in both natural and designed sequences , they do not show that the mechanisms underlying covariation are the same in both cases ., To investigate this , we first classified the mechanism of covariation for all pairs that covary in both designed and natural sequences and then quantified how often the same mechanism is used ., Figure 6A shows an illustration of three of the covariation mechanisms: size , hydrogen bonding and charge ., Classifying each of these mechanisms requires examining the transition from one amino acid pair to another ., For example , the transitions depicted in Figure 6A are IA–VV , AP–SS , RE–DR ., Covariation due to size involves a decrease in the size of one amino acid and an increase in the size of the other ( IA–VV ) ., Covariation due to hydrogen bonding involves a hydrogen bond that exists in one pair but not the other ( AP–SS ) ., Covariation due to charge involves a pair of amino acids with opposite charges that either swap sign ( RE–DR ) or become uncharged amino acids ., We also defined covariation mechanisms based on cation-pi interactions , pi-pi interactions , and other interactions not falling into any of the previous categories that we classify as hydrophobic , hydrophilic or mixed hydrophobic and hydrophilic ( see Methods for a detailed definition ) ., For each pair of positions that covaried in both the designed and natural sequences , we computed the ten most significant transitions between amino acid pairs at those positions and classified each transition based on the mechanism of covariation ., The resulting distributions of covariation mechanisms for the designed and natural pairs are shown in Figure 6B ., The designed and natural covariation mechanisms distributions share similar properties , including covariation due to charge being the most common mechanism , whereas cation-pi , pi-pi and other ( hydrophilic ) covariation mechanisms are more rare ., In both natural and designed distributions , hydrogen bonding and size covariation together account for approximately 30% of the total mechanisms ., However , a number of quantitative differences exist in the distributions , including charge occurring more frequently in the designed pairs , suggesting that the design method may be over-predicting charged interactions ., Additionally , in the natural pairs , size covariation is more common than hydrogen bonding covariation while the opposite is true in designed pairs ., The “other” categories are also more common in the natural pairs than in the designed pairs ., To better understand these differences , we split the pairs up based on the extent of their burial and compared the distributions of covariation mechanisms ( Figure S4 ) ., This analysis revealed that covariation mechanism is dependent on the extent of pair burial and that buried pairs have the most significant differences between natural and designed covariation mechanisms ., In natural buried pairs , the most common covariation mechanisms are size and other ( hydrophobic ) , whereas the most common mechanisms in designed buried pairs are hydrogen bonding and size ., This likely occurs due to insufficient penalization of buried polar groups during the design protocol , resulting in over-predicting polar amino acids at buried positions and therefore incorrect predictions of covariation mechanism ., To quantify how often the same covariation mechanism is used for specific pairs of positions in the designed and natural sequences , we calculated the percent of pairs sharing the same classification type in both the natural and designed sequences ( percent overlap ) for each type of covariation mechanism ( Figure 6C ) ., Covariation due to charge has the highest percent overlap between the designed and natural pairs , followed by hydrogen bonding , size , other ( hydrophobic ) and other ( mixed ) , which have roughly equal percent overlaps ., Covariation due to cation-pi and pi-pi interactions have relatively low percent overlaps between the designed and natural sequences , likely due to the fact that these types of interactions are not explicitly accounted for in the design energy function ., We repeated this analysis using fixed backbone design sequences and found a decrease in the percent overlaps for size and other ( hydrophobic ) interactions , indicating that backbone flexibility may aid in modeling these types of covariation mechanisms ( Figure S5 ) ., Taken together , this analysis provides insights into the mechanisms underlying amino acid covariation in naturally occurring proteins ., Overall , the analysis shows considerable agreement between naturally occurring and designed covariation mechanisms ., In some cases , it exposes pathologies in the design methods ( such as the over-representation of polar amino acids in cores under-representation of cation-pi and pi-pi interactions ) that can be addressed in future work using naturally occurring covariation as a reference point ., While computational protein design can model a significant fraction of naturally occurring covarying amino acid pairs , there remain pairs of amino acids that are highly covarying in the natural sequences but not in the designed sequences ( nature-specific pairs ) ., Moreover , there also exist pairs that highly covary in designed sequences but not in natural sequences ( design-specific pairs ) ., Figure 7A shows the classification of nature-specific , design-specific and overlap pairs for the SH3 domain ., To understand the basis for these differences , we first compared these sets of pairs based on their distances in three-dimensional structure ( Figure 7B ) ., We found the design-specific and overlap covarying pairs to be significantly closer in structure than the nature-specific pairs ., These results are consistent with the all-atom energy function used for generating the design sequences , which is most sensitive at short distances ., The long distances in the nature-specific pairs could result from a number of factors , including interactions that bridge monomers in an oligomeric complex 37 , interactions that exist in alternative conformations 37 , long-range correlations in protein dynamics or from phylogenetic bias in the natural sequences ., Another possibility is that in naturally occurring proteins , destabilizing substitutions ( that occur in functional sites ) co-vary with compensating stabilizing mutations in the protein that could be far away from the functional site ., In addition to analyzing design-specific and nature-specific pairs with respect to pair distance , we compared them based on extent of amino acid burial , the presence in interfaces or active sites , and amino acid pair propensity ., We observed a slight decrease in the percent of exposed pairs in the designed-specific pairs relative to the nature-specific pairs ( Figure S6 ) , which may be due to the difficulty of accurately modeling solvent exposed interactions in protein design ., We observed no difference in the design-specific and nature-specific pairs with respect to their presence in interfaces or active sites ( Figure S7 ) , suggesting that the constraints imposed by known functional sites are not responsible for the inability to model the nature-specific pairs ., We observed that the amino acid pair propensities of nature-specific and overlap pairs were different , while the amino acid pair propensities of design-specific pairs were highly correlated to those of the overlap pairs ( Figure 7C ) ., The latter observation indicates that the energetic interactions leading to design-specific and overlap pairs may be similar to each other ., A simple explanation may be that the design-specific pairs are equally compatible with the given protein structure , but may simply not have been sampled by nature ., Such design-specific pairs may provide opportunities for engineering proteins with novel amino acid interactions , such as re-designing the specificity of protein-protein interactions ., Our study tested the hypothesis that the structural constraints imposed by protein architecture are a major determinant of amino acid covariation in naturally occurring proteins ., If true , we reasoned that computational design methods that design sequences based on protein structure alone should be able to recapitulate amino acid covariation , provided that design predictions are sufficiently accurate ., Confirming these ideas , we found a significant overlap between amino acid covariation in natural and designed protein sequences across a set of 40 diverse protein domains ., These results quantify the influential role of the selective pressures for maintaining protein structure on shaping amino acid covariation ., Therefore , even though correlated changes are undoubtedly important to evolve new activities and regulatory mechanisms 30 , 31 the presence of covariation alone may not necessarily indicate a functional role ., Our study also illustrates how recapitulation of amino acid covariation serves as a stringent test for the ability of computational protein design methods to capture precise details of interactions between amino acids ., We demonstrate that modeling backbone flexibility significantly increases the similarity between natural and designed covariation , and that this similarity is exquisitely sensitive to the mechanism used to model backbone changes ., These findings indicate that protein backbone motions are required for allowing precise adjustments in amino acid interactions that enable covariation ., Moreover , simulations that perform local backbone movements ( Backrub and KIC ) result in sequences with more natural-like covariation than simulations that perform non-local backbone movements ( AbRelax , Relax , Small ) ., Proteins may have undergone local motions similar to Backrub and KIC moves to accommodate new mutations and amino acid interactions during evolution 24 , 35 , 36 , 41 ., Such motions could have provided proteins with a mechanism to allow subtle , incremental changes to their structures without adversely affecting protein structure or protein function ., While local motions may be a common mechanism for proteins to accommodate point mutations , larger structural adjustments may be necessary for dealing with insertions or deletions ., In this study , we found that a moderate degree of backbone flexibility best recapitulated natural amino acid covariation , however , the magnitude of structural variation produced by this degree of backbone flexibility was less than the structural variation among naturally occurring protein families ., This discrepancy is likely due to the assumption in the design method that the protein remains a fixed length ., This is not true in naturally occurring sequences; in fact , all 40 domains in our benchmark include loop regions that have varying lengths ., Mutations that change the length of a flexible loop could allow for secondary structure elements to re-orient themselves and slightly alter the tertiary structure ., The accumulation of mutations in loop regions can produce significant structural diversity that cannot be modeled using a protein design method that keeps the number of amino acids in a protein constant ., Future protein design methods , particularly those involving loop regions such as protein-protein interaction design or enzyme specificity design , could potentially benefit from incorporating moves that both change the conformation and length of the protein backbone ., In addition to observing significant similarity between the sets of natural and designed highly covarying amino acid pairs , we observed a high correlation in the amino acid propensities of these covarying pairs and showed that the structural mechanisms underlying covariation are similar for both natural and designed sequences ., Differences between natural and designed covarying pairs highlight areas for improvement in the energy function used for protein design ., For instance , cation-pi interactions , which are not explicitly accounted for in the energy function used in this study , have high propensities among naturally covarying pairs but not in designed covarying pairs ., Similarly , polar amino acid pairs are more frequent in the cores of designed proteins than in naturally occurring proteins ., Interestingly , we found differences in the pair propensities between natur | Introduction, Results, Discussion, Methods | Amino acid covariation , where the identities of amino acids at different sequence positions are correlated , is a hallmark of naturally occurring proteins ., This covariation can arise from multiple factors , including selective pressures for maintaining protein structure , requirements imposed by a specific function , or from phylogenetic sampling bias ., Here we employed flexible backbone computational protein design to quantify the extent to which protein structure has constrained amino acid covariation for 40 diverse protein domains ., We find significant similarities between the amino acid covariation in alignments of natural protein sequences and sequences optimized for their structures by computational protein design methods ., These results indicate that the structural constraints imposed by protein architecture play a dominant role in shaping amino acid covariation and that computational protein design methods can capture these effects ., We also find that the similarity between natural and designed covariation is sensitive to the magnitude and mechanism of backbone flexibility used in computational protein design ., Our results thus highlight the necessity of including backbone flexibility to correctly model precise details of correlated amino acid changes and give insights into the pressures underlying these correlations . | Proteins generally fold into specific three-dimensional structures to perform their cellular functions , and the presence of misfolded proteins is often deleterious for cellular and organismal fitness ., For these reasons , maintenance of protein structure is thought to be one of the major fitness pressures acting on proteins ., Consequently , the sequences of todays naturally occurring proteins contain signatures reflecting the constraints imposed by protein structure ., Here we test the ability of computational protein design methods to recapitulate and explain these signatures ., We focus on the physical basis of evolutionary pressures that act on interactions between amino acids in folded proteins , which are critical in determining protein structure and function ., Such pressures can be observed from the appearance of amino acid covariation , where the amino acids at certain positions in protein sequences are correlated with each other ., We find similar patterns of amino acid covariation in natural sequences and sequences optimized for their structures using computational protein design , demonstrating the importance of structural constraints in protein molecular evolution and providing insights into the structural mechanisms leading to covariation ., In addition , these results characterize the ability of computational methods to model the precise details of correlated amino acid changes , which is critical for engineering new proteins with useful functions beyond those seen in nature . | null | null |
journal.ppat.1003131 | 2,013 | ActA Promotes Listeria monocytogenes Aggregation, Intestinal Colonization and Carriage | Listeria monocytogenes ( Lm ) is a facultative intracellular Gram-positive bacterium and the agent of listeriosis , the deadliest foodborne infection in humans , with a mortality rate between 20 to 30% ., Listeriosis can manifest as gastroenteritis after ingestion of a high inoculum , as septicemia , meningitis and encephalitis primarily in immune-compromised individuals , and induce fetal-placental infection leading to in utero death , premature birth , abortion and neonatal infection ., Lm induces its internalization in non-professional phagocytes , such as epithelial cells , survives and multiplies in the cytosol of professional phagocytes and spreads from cell to cell ., These properties constitute crucial virulence determinants of Lm and their molecular mechanisms have been studied in detail ., InlA and InlB have been identified as critical surface proteins mediating Lm entry into epithelial cells 1 , 2 and crossing of the intestinal and placental barriers 3–6 ., Listeriolysin O ( LLO ) is a pore-forming toxin that mediates Lm escape from the internalization vacuole , and its access to the cytosol 7 ., It is a critical phenotypic marker for Lm identification and is the virulence factor that allows Lm survival in professional phagocytes 8 ., Once in the cytosol , Lm polymerizes actin to propel itself , forming protrusions at the host cell surface and spread from cell to cell ., ActA has been identified as the Lm factor necessary and sufficient on the bacterial side to polymerize actin and form comet tails 9 ., Recently , ActA has also been shown to allow Lm to escape autophagy 10 ., PrfA , a transcriptional activator that belongs to the cyclic AMP receptor protein family regulates most genes involved in Lm virulence , including inlA , inlB , hly ( which encodes LLO ) and actA 11–13 ., PrfA is expressed during Lm exponential growth and at the beginning of stationary phase 12 , above 30°C 11 ., This key regulator is selectively activated in vivo in the intestinal lumen , enabling Lm to switch on its virulence genes 14 ., PrfA is specific to the pathogenic species Lm and L . innocua ( Li ) , a non-pathogenic non-invasive Listeria species closely related to Lm , is devoid of PrfA and PrfA-regulated genes , including inlA , inlB , hly and actA 1 ., Because Lm is primarily regarded as a pathogen , its pathogenicity is the aspect of its biology that has been studied in the most detail ., Nevertheless , Lm can be shed asymptomatically , persist in human and animal feces and be released in the environment 15 , 16 ., Lm is a ubiquitous bacterium that also thrives in diverse external environments such as soil , water , decaying plants , and silage , exposing wild animals and cattle to multiple opportunities of ingestion and perpetuating Lm transmission 17 ., In the environment , bacteria can form biofilms , which favor their persistence 18 ., From a food-safety perspective and with the aim of limiting Lm transmission to humans , a lot of emphasis has been focused on reducing bacterial aggregation , biofilm formation and persistence of Lm on industrial surfaces and food 19 ., A number of factors , including the quorum-sensing-related proteins of the LuxS and Agr systems 20 , 21 , and stress responses factors 22–25 such as the transcriptional regulator SigB 26 , and PrfA 27 , 28 have been implicated in Lm biofilm formation ., Yet , neither Lm persistence nor the putative role of bacterial aggregation and biofilm formation has been investigated in the context of infection , Our study began with the serendipitous observation that Lm spontaneously sediments in test tubes whereas Li does not ., Fast sedimentation is usually triggered by tight interactions mediated by aggregation factors generally involved in biofilm formation 29 , 30 ., We show here that Lm rapid sedimentation results from PrfA-dependent aggregation ., Furthermore , we show that ActA is the PrfA-regulated factor promoting bacterial aggregation via direct ActA-ActA interaction ., Finally , we show that ActA-dependent bacterial aggregation leads to increased Lm persistence in the intestine , prolonged fecal shedding and thereby facilitates transmission ., This is a critical new function for ActA , which manifests extracellularly , and is independent of its role in actin-based motility ., Virulence factors may confer a selective advantage for pathogenic microbes , when they allow the colonization of otherwise sterile host tissues ., This newly observed property of ActA may also participate in the selective pressure on Lm to maintain ActA , as it favors bacterial dissemination ., When Lm ( EGD strain ) and Li cultures grown overnight in BHI , at 37°C with shaking , were switched to static conditions , Lm EGD sedimented within five hours whereas Li did not ( Figure 1A ) ., Microscopic examination of the pellet revealed bacterial aggregates ( Figure 1A ) and this phenotype was abolished when Lm was grown at 25°C ( data not shown ) ., Because Li lacks PrfA and PrfA-regulated genes , which are specific to Lm and regulated by temperature , we investigated whether prfA could be implicated in Lm aggregation ., An aggregation assay performed with EGD and an isogenic mutant ΔprfA showed that aggregation is prfA-dependent ( Figure 1B ) ., Similar results were observed for the other Lm reference strains , LO28 and EGDe , when cultivated in BHI ( Figures S1A–B ) or in DMEM ( Figures S1C–D ) , in which PrfA-regulated genes expression and the aggregation phenotype were increased 31 ., To confirm the role of prfA in Lm aggregation , we performed aggregation assay with clinical strains , randomly chosen from the collection of the French National Reference Center for Listeria and harboring a functional PrfA ( PrfA+ ) , and non-clinical isolates , naturally non-hemolytic and lacking phospholipase activity due to loss-of-function of PrfA ( PrfA− ) ( Table S1 ) ( our unpublished observations ) ., PrfA expression by both PrfA+ and PrfA− isolates was confirmed by immunoblot ( data not shown ) ., The mean aggregation in 24 h of the PrfA− isolates was significantly reduced ( p\u200a=\u200a0 . 001 ) when compared to the mean aggregation of PrfA+ ( Figure 1C ) , showing that the role of prfA in Lm aggregation is a general property of various Lm strains ., To determine how PrfA regulates Lm aggregation , we analyzed isogenic deletion mutants of the main PrfA-regulated virulence genes , i . e . inlA , inlB , hly and actA ., ΔinlA and ΔinlB EGD isogenic mutants displayed an ability to aggregate identical to that of WT EGD and the aggregation ability of Δhly mutant was marginally delayed compared to the WT ( Figures 1D–E ) ., In contrast , both ΔprfA and ΔactA mutants displayed very low aggregation , even after 24 h ( Figures 1D–E ) ., Consistent with these results , complementation of ΔprfA and ΔactA mutants either with prfA or actA fully restored WT aggregation ability ( Figure 1F ) ., Similar results were obtained with LO28 and EGDe strains ( Figures S1E–F ) ., Observation by scanning electron microscopy ( SEM ) of WT Lm sediment retrieved after five hours under static conditions showed dense bacterial aggregates , whereas no aggregate was detected with the ΔactA mutant ( Figure 1G ) ., Together , these results demonstrate that ActA is the PrfA-regulated gene product involved in the formation of Lm aggregates ., As bacterial aggregation is a key step of biofilm formation 18 , we investigated the contribution of ActA to Lm biofilm formation in vitro with EGD isogenic mutants ΔprfA , ΔinlA , ΔinlB , ΔactA and Δhly ., Whereas biofilm biomass of WT EGD could be homogenously and strongly stained by crystal violet on the surface of the wells , the ΔprfA mutant displayed a 70% reduction in biofilm biomass , which was only present in the center of the wells ( Figure 2A ) ., ΔinlA formed slightly but significantly more biofilm than WT , ΔinlB was equivalent to WT and Δhly formed slightly less biofilm as compared to WT ( Figure 2A ) ., In contrast , ΔactA displayed 55% biofilm reduction and was the only strain impaired in covering the bottom of wells as observed for ΔprfA ( Figure 2A ) ., This suggests that ActA is the major PrfA-regulated gene involved in biofilm formation ., To confirm the involvement of ActA in biofilm formation , we used continuous-flow microfermentors ., Whereas WT biofilm grew on both spatula and microfermentor walls , ΔactA exhibited a drastically reduced ability to form biofilm ( Figure 2B ) ., Comparisons of the biomass retrieved from biofilms formed on the spatula between the WT and the isogenic ΔactA mutant showed a 60-fold difference in optical density at 600 nm ( OD600 ) and a reduction of two orders of magnitude in CFUs ( Figure 2B ) ., To determine if other factors are required to trigger ActA-dependent biofilm formation , we expressed actA in Li , which only forms a very limited biofilm biomass in microtiter plate ., ActA expression in Li + actA was confirmed by immunoblot and immunofluorescence ( Figure 2C and data not shown ) ., Biofilm assay in microtiter-plate showed a significant increase of biomass following the expression of actA by Li ( Figure 2C ) , indicating that ActA is sufficient to promote biofilm formation in Li ., We next imaged EGD WT and ΔactA grown on static glass slide by confocal microscopy ., Whereas the ΔactA bacteria organized in a very thin and homogenous layer around 25 µm thick , the WT formed a deep mushroom-shaped and dense biofilm around 45 µm thick ( Figure 2D ) ., For an equivalent number of bacteria , there were one order of magnitude fewer WT clusters than with ΔactA , and the number of bacteria per cluster with WT bacteria was one order of magnitude higher than with ΔactA ( Figure 2E ) ., Taken together , these data show that bacteria expressing ActA aggregate into large clusters within biofilm structure thereby favoring biofilm formation , which is not the case for ΔactA ., ActA is a membrane-anchored protein exposed on the bacterial surface 9 ., Either direct or indirect ActA-ActA interaction may mediate bacterial aggregation and favor biofilm formation ., We observed that ActA-dependent aggregation occurs in PBS and H2O ( data not shown ) , suggesting that external factors are not required for Lm aggregation ., Moreover , when observed by SEM , bacterial aggregates did not exhibit visible matrix connecting bacteria to each other , suggesting that ActA-dependent aggregation occurs without any incorporation of matrix ( Figure 3A ) ., In order to determine whether aggregation is mediated by a direct ActA-ActA interaction , we performed aggregation assays by mixing EGD WT and/or ΔactA bacteria expressing green fluorescent protein ( GFP ) or not ., As expected , WT and WT GFP formed mixed aggregates ( Figure 3B–C ) ., In contrast , ΔactA and ΔactA GFP did not aggregate ( Figure 3B ) , and only constituted small and isolated mixed bacterial foci ( Figure 3C , two top rows ) ., In the case of mixed WT and ΔactA GFP bacteria , we observed an intermediate aggregation phenotype and aggregates contained almost exclusively WT bacteria , with some sparse ΔactA GFP bacteria trapped within the aggregative structure ( Figure 3B–C ) ., These results show that ΔactA bacteria are not able to aggregate with WT , and suggest that ActA-dependent aggregation requires a direct ActA-ActA interaction ., To study whether ActA is sufficient to promote Lm inter-bacterial interactions , aggregation assays were performed with ActA-expressing Li and Staphylococcus aureus strains ., We observed that ActA expression is sufficient to promote the aggregation of these two strains ( Figure 3D ) ., Finally , we performed an aggregation assay with latex beads coated with purified ActAHIS , InlBHIS or bovine serum albumin ( BSA ) 32 , 33 , 34 ., The coating of beads was assessed by immunofluorescence and a strong signal corresponding to either ActAHIS or InlBHIS coated on beads was detected ( Figure 3F ) ., The aggregation assays showed that ActAHIS-cotaed latex beads formed macroscopic aggregates within 15 minutes ( Figure 3E–F ) ., In contrast , latex beads coated with either BSA or purified InlBHIS did not , even after 24 hours ., Together , these data demonstrate that direct ActA-ActA interaction mediates aggregation ., ActA has a low isoelectric point ( pI of 4 . 95 ) , indicating that ActA-dependent aggregation at neutral pH , at which our experiments were performed , occurs when ActA is globally negatively charged ., We hypothesized that ActA charge could be important for aggregation and performed aggregation assays in a pH range of 1 to 9 ., Whereas overall bacterial aggregation within this pH range was roughly stable , ActA-mediated aggregation was maximal between pH 6 . 5 to pH 9 , a pH window within which no ActA-independent aggregation is detected ( Figure 3G ) ., To further investigate how ActA mediates Lm aggregation , we functionally mapped the ActA domains involved in bacterial aggregation ., The respective contribution of ActA domains in host actin polymerization have been previously determined ., These studies have shown that, ( i ) the NH2-terminal domain ( N region ) binds Arp2/3 complex , is involved in actin filament nucleation and is critical for actin polymerization ,, ( ii ) the central domain ( P region ) binds Ena/VASP , is not required for actin polymerization but contributes to the length of actin tails and the velocity of bacterial intracellular movement , and, ( iii ) the C-terminal or C region is dispensable for actin polymerization 32 , 35–43 ( Figure 4A ) ., When aggregation assays were performed with strains expressing ActA variants lacking the N , P or C region , or subdomains within the N region ( Figure S2B ) , we observed that only full-length ActA mediates full aggregation , suggesting that aggregation requires the native conformation of the full-length ActA protein ., We also observed that the consecutive 21–97 and 97–126 segments in N-region were only partially implicated in aggregation , allowing 31% and 36% of aggregation , respectively ., In contrast , the 126–231 segment of N-region appeared critical for aggregation ( Figure 4B ) ., Both mutants lacking P and C regions were also impaired in aggregation ., Because the C-terminal region of ActA , which is not involved in actin polymerization , is implicated in aggregation , we took advantage of this property to directly assess the contribution of ActA-dependent aggregation during infection , independent of the critical role of ActA in actin-based motility ., To this aim , we complemented EGD ΔactA mutant with a C-region-truncated actA ., We first confirmed that the EGD ΔactA + actAΔC ( ΔC+ ) strain was impaired in its abilities to either aggregate or form biofilm , as is the ΔactA mutant ( Figures 5A–B ) ., We also checked the ability of the ΔC+ mutant to polymerize actin in cultured cells ., We observed that ΔC+ bacteria formed actin comet tails as efficiently as WT and ΔactA + actA ( ActA+ ) ( Figure 5C ) ., Furthermore , ΔC+ intracellular bacteria were able to induce comet tails as WT and ActA+ ( Figure 5D ) and ΔC+ comet tails were of similar length than that of WT and ActA+ ( Figure 5E ) ., These results showed that ΔC+ mutant phenotype is similar to that of WT and ActA+ , as far as actin-based motility is concerned , but is impaired for biofilm formation and aggregation like ΔactA ., We next inoculated knock-in humanized E16P mEcad ( KI E16P ) mice , which are permissive to orally-acquired listeriosis , with either EGD ActA+ or ΔC+ strains , to investigate the role of ActA-dependent aggregation in vivo , independent of the critical role of ActA in actin-based motility 4 ., Four days after inoculation , no significant difference in CFU counts in the intestine and colon tissues , mesenteric lymph nodes , spleen and liver were detected ( Figure 6A ) ., This result shows that both ActA+ and ΔC+ are similarly invasive in vivo , and consequently that the ability of ActA to mediate Lm aggregation does not have an impact on Lm ability to infect tissues in the first four days of infection ., We next investigated if ActA-mediated aggregation occurs within the intestine , which pH is >6 . 5 , except in the stomach and proximal duodenum , and therefore optimal for ActA-mediated aggregation ., We first checked that ActA is expressed within the gut lumen ( Figure S2C ) ., We then performed a detailed imaging survey for bacteria within the whole small intestine , cecum and colon six hours after oral inoculation ., For both EGD ActA+ and ΔC+ strains , we observed rare isolated bacteria within the duodenal and ileal lumens , which fits with the rapid transit of Lm in the small intestine upon oral inoculation ( 44 , 45 and our unpublished observations ) ., Isolated intracellular bacteria were also found within the intestinal epithelium , and particularly in goblet cells , extruding cells and epithelial folds , which are the preferential sites for Lm entry within the intestine 3 , 6 , 46 ., Bacteria were also observed within the lamina propria of intestinal villi , confirming that both mutants are equally invasive ( Figure 6A–B ) ., Importantly , within the cecum lumen , ActA+ and ΔC+ strains exhibited distinct phenotypes as early as six hours post-inoculation: whereas ΔC+ bacteria remained mainly isolated , ActA+ bacteria formed small aggregates ., This distinctive phenotype was also observed within the colon lumen , in which ActA+ bacteria aggregates were detected , often trapped within mucus , whereas none was observed with ΔC+ ( Figure 6B ) ., Together , these results show that ActA-dependent aggregation is detectable in vivo in the cecum lumen as early as six hours post inoculation ., After four days of infection , ActA+ and ΔC+ Lm were eliminated from the small intestine lumen of infected mice ( data not shown ) ., In contrast , within the cecum lumen , we detected ActA+ bacteria forming aggregates , while ΔC+ bacteria remained essentially isolated in the lumen ( Figure 7A ) ., Indeed , the proportion of bacterial aggregates of more than three bacteria was four-fold higher in the cecum lumen of mice inoculated with ActA+ compared to ΔC+ bacteria ( p<10−6 ) ( Figure 7B ) ., Bacterial aggregates were also detected within stools of mice inoculated with ActA+ , whereas only rare and sparse bacteria were detected within stools of ΔC+-inoculated mice ( Figure 7C ) ., These results were confirmed using KI E16P mice inoculated with EGDe ActA+/ΔC+ ( Figure S3A ) ., Together , these results strongly suggest that the cecum is the site where Lm forms bacterial aggregates ., Having shown that Lm aggregates within the cecum and colon lumens , we investigated whether Lm intraluminal aggregation might favor its persistence in the gut and fecal shedding ., We inoculated KI E16P mice with EGD WT , ΔactA , ActA+ and ΔC+ bacteria and monitored Lm fecal carriage by enumerating daily bacterial CFUs in stools ., Within the first two days , we observed the elimination of the bulk of the inoculum 44 ., Fecal shedding of ΔactA and ΔC+ bacteria dropped steadily from day 1 and was no longer detectable after day 8 ( Figure 7D ) ., In sharp contrast , both WT and ActA+ bacteria showed increased fecal shedding between days 2 and 6 , followed by a gradual and slow decline to finally reach total clearance by day 17 ( Figure 7D ) ., Indeed , total fecal shedding of Lm from day 2 to clearance was three orders of magnitude higher and persisted for twice as long in mice inoculated with WT or ActA+ Lm relative to mice inoculated with ΔactA or ΔC+ ( Figure 7D–E ) ., Similar results were observed when KI E16P mice were inoculated with EGDe ActA+/ΔC+ ( Figures S3C–E ) , LO28 ActA+/ΔC+ ( data not shown ) , and the PrfA+/PrfA− isolates ( Figures S3B–D ) , which respectively express or not ActA ( Figure S2A ) ., These results show that even though ActA+ and ΔC+ bacteria invade mouse tissues at similar levels , their ability to colonize and persist the gut lumen strongly differs , illustrating that aggregating Lm display increased colonization and persistence in the gut than non-aggregating bacteria ., This indicates that ActA , independent of its well-established role in bacterial dissemination within tissues in the systemic phase of the infection , also plays a critical role in intestinal colonization and long-term carriage of Lm within the gut ., Lm is adapted to survive in various conditions , colonize diverse environments , notably as a biofilm ., It is also a facultative intracellular pathogen able to invade tissues and trigger a systemic infection in human and a wide range of animals ., These two complementary aspects of Lm biology have so far been considered separately ., We show here that , independently of its contribution to Lm actin-based motility that manifests intracellularly , ActA mediates Lm aggregation , colonization and persistence in the gut lumen , leading to its increased dissemination in the environment ., To our knowledge , this is the first time that a virulence factor is involved in microbial persistence and transmission , independently of its known role in pathogenesis , This new property of ActA that occurs when Lm is located outside of the host cell , may apply a positive selective pressure for the maintenance of its gene , during the extracellular phase of its life cycle ., While we were studying this novel and unexpected function of ActA , two different investigators reported on the implication of PrfA in biofilm formation 27 , 28 , a process involving bacterial aggregation 30 ., We show here that this process depends on ActA expression , which mediates inter-bacteria interactions and promotes biofilm formation ., We also observed minor modulation of biofilm formation by two others PrfA-regulated factor , LLO and InlA , which slightly promotes and reduces Lm biofilm formation , respectively ., Although the ΔprfA and ΔactA deletion mutants aggregation and biofilm phenotypes are indistinguishable , demonstrating that actA is the main PrfA-regulated gene accounting for the PrfA-dependence of Lm aggregation and biofilm formation , the contribution of InlA to Lm biofilm is in agreement with a previous study that showed that inlA mutations leading to InlA truncation slightly increase biofilm formation 47 ., Studies in reference strains such as LO28 and EGDe have shown that ActA is up-regulated by PrfA when Lm is within the cytosol , in which ActA mediates actin-based motility 48 ., ActA is also expressed in bacteria cultured in BHI liquid medium and within the gut lumen , although to a lower level than intracellularly ( Figure S2C and 14 ) ., Our initial observation of Lm aggregation was made in EGD , a reference strain that overexpresses ActA as a result of a gain-of-function mutation in prfA called prfA* ( our unpublished data ) ., We show here that in EGD , as well as in reference strains EGDe and LO28 , ActA expression in BHI is sufficient to promote bacterial aggregation in vitro ., This newly discovered property of ActA occurs at neutral pH and 37°C , the physiological environment of mammalian gut ., In contrast , no aggregation is observed when bacteria are grown at 25°C , when PrfA-regulated genes are off , suggesting that ActA-dependent aggregation may contribute to Lm persistence in warm-blooded hosts ( see below ) ., We demonstrate that Lm aggregation involves direct ActA-ActA interaction ., We consistently observed that ActA-dependent aggregation occurs in PBS and H2O , suggesting that ActA-dependent aggregation might implicate direct ActA-ActA interaction ., Consistent with this finding , SEM showed that Lm ActA-dependent aggregates do not contain detectable matrix or fiber-like material ., Previous studies have shown that Lm ActA-dependent actin based motility relies on ActA polar distribution 49 ., However , SEM on bacteria aggregates did not reveal any particular polar or lateral orientation in ActA-dependent bacterial interactions , which rather appeared to occur randomly ., This suggests that in contrast to ActA-dependent actin-based motility , the polar distribution of ActA is not critical for Lm aggregation ., We also show that the domain involved in ActA dimerization contributes to aggregation , indicating that ActA ability to dimerize might be implicated in the trans-dimerization of ActA molecules expressed by neighboring bacteria 50 ., Yet , as for ActA dimerization , this domain is not sufficient to mediate bacterial aggregation ., ActA is a particularly elongated molecule , largely made of random coils , which structure is responsible for many of its unique biochemical properties 51 ., Although its three-dimensional structure is unknown , our results show that aggregation requires all ActA structural domains , suggesting that the native conformation of the protein is critical for aggregation ., We have shown that ActA mediates Lm aggregation only above its pI , suggesting that ionic interactions between charged amino acids are essential in ActA-ActA interaction ., ActA contains a particularly large amount of charged amino acids , especially within the 126–231 domain that is critical for aggregation ., Because of its low pI ( 4 . 95 ) , ActA is strongly charged at neutral pH , with a mix of positively and negatively charged regions likely involved in ActA-ActA mediated aggregation ., However , ActA ortholog in L . ivanovii , IActA , which also mediates actin polymerization , does not mediate bacterial aggregation ( our unpublished data ) , despite an identical pI , 34% of sequence identity and 52% of sequence similarity with ActA 52 ., This suggests that ActA ability to mediate aggregation , although likely dependent on its charged residues , is a specific property of Lm ., The sequence variability of actA has been used for typing purposes , and several studies have reported a high degree of polymorphism within actA 53 ., Interestingly , a 105 bp deletion within actA region encoding the central proline rich repeat is frequently found in Lm 54 ., As this deletion does not affect ActA ability to polymerize actin 55 , we hypothesized that it may modify bacterial aggregation ., However , we detected no significant association between aggregation ability of strains harboring or not this deletion ( data not shown ) ., We showed that the N- , P- and C-domains of ActA are critical for bacterial aggregation ., Importantly , a mutant lacking C-region is still fully virulent ., We took advantage of this property of the ActA C-domain to study specifically the role of ActA-dependent aggregation in vivo , independently of ActA contribution to actin polymerization ., This led us to discover that the ability to form aggregate is associated to increased gut colonization and fecal shedding ., To our knowledge , our study is the first demonstrating the involvement of a virulence factor in gut colonization and transmission that is independent of the mechanism mediating virulence ., Indeed , although factors involved in gut colonization have been described for enteropathogenic bacteria such as Salmonella 56 , enteropathogenic and enterohaemorrhagic E . coli 57 , Citrobacter rodentium 58 and Campylobacter jejuni 59 , in all cases , these effects were directly linked to their enteropathogenicity ., We have shown that Lm , when able to aggregate in vitro , also forms aggregates in the cecum and colon lumen , and colonizes the gut far more efficiently and durably than when it does not form aggregates ., Lm is found in higher numbers in the cecum lumen than upstream in the small intestinal lumen 44 , 60 ., Furthermore , the gastric pH is highly acidic ( 1 to 2 . 5 ) , whereas the pH varies between 6 . 4 and 7 . 5 from the small intestine to the cecum and colon ., As ActA-mediated aggregation occurs between pH 6 . 5 to pH 9 , Lm is subjected to a pH permissive to ActA-dependent aggregation in the distal small intestine , cecum and colon lumens but not within the stomach or the proximal duodenum lumens , which luminal content is far too acidic for ActA-mediated aggregation to occur ., This hypothesis could not be verified as ingested bacteria were rapidly eliminated from small intestine lumen ., The cecum lumen is likely the best site for aggregates formation: not only its greater diameter than the small intestine results in decreased shear stress , but also the increased number of intraluminal bacteria 4 , 60 likely favors inter-bacterial contacts and hence aggregates formation ., Aggregates observed within the cecum and colon lumens appeared to be mainly trapped within mucus whereas isolated bacteria were not , suggesting that mucus may favor Lm aggregate formation and/or expansion in the gut ., ActA has been shown to be expressed before intestinal tissue invasion , within the intestinal lumen 14 but the significance of this somewhat premature expression remained unexplained so far , as the role of ActA was thought to be exclusively intracellular ., Here , we show that this extracellular expression of ActA allows intraluminal ActA-dependent aggregation , a property that correlates with increased gut colonization and fecal shedding ., The release of Lm aggregates , as opposed to isolated bacteria , may favor Lm survival in environment and its transmission to new hosts , including animals and humans 61 , 62 ., It should be noted however that Lm virulence and particularly its ability to cross the intestinal barrier and survive in host tissues also affects its ability to colonize the gut: ΔinlA or Δhly mutants for which virulence is attenuated in vivo also exhibit a reduced persistence in the intestine ( data not shown ) ., This suggests , as it has been recently proposed 44 , that bacteria are shed back from infected intestinal villi into the intestinal lumen ., Among virulent Listeria species , Lm is the most prevalent species harboring prfA 63 and Lm is also the most prevalent species infecting mammalian hosts 64 ., We demonstrate here that ActA favors long-term gut colonization and fecal shedding and that this advantage is Lm-specific ., How and under which selective pressure has Lm acquired and evolved prfA and PrfA-regulated genes is not known ., Virulence factors are thought to have been selected for as they allow pathogens to colonize otherwise sterile sites ., Yet , the fact that ActA mediates Lm aggregation and intestinal colonization may have also participated the selective pressure on Lm to maintain ActA , as it favors Lm release in the environment and access to new hosts ., Bacterial strains used in this study are listed in Table S1 ., Lm , Li and S . aureus bacteria were cultured in Brain Heart Infusion medium ( BHI , Difco ) or in Dulbeccos Modified Eagle Medium ( DMEM , Invitrogen ) , when specified ., E . coli was cultivated in Luria Broth medium ., Antibiotics were added when required at the following concentrations: erythromycin 5 µg/ml ( Li ) or 1 µg/ml ( S . aureus ) and chloramphenicol ( Cm ) 7 µg/ml ( Lm ) or 35 µg/ml ( E . coli ) ., EGD ΔprfA and EGDe ΔactA mutants were constructed as previously described 65 using primers listed in Table S2 ., Stable insertion of Cm resistance gene in PrfA+ isolates , of GFP in EGD ΔactA , as well as chromosomal complementation of EGD ΔactA , EGDe ΔactA and LO28 ΔactA with full-length actA ( ActA+ ) or actAΔC ( ΔC+ ) and EGD ΔprfA with prfA were realized as previously described 66 using plasmids pPL2 , pAD cGFP , pPL2-actA , pPL2-actAΔC and pPL2-prfA , respectively ., The pPL2-actA , pPL2-actAΔC and pPL2-prfA plasmids were constructed by PCR amplification from EGD chromosomal DNA of either full-length actA , actAΔC and full-length prfA using primers listed in Table S2 ., These PCR fragments were cloned into pPL2 plasmid 67 ., EGDe ΔprfA , LO28 Tn::prfA , S . aureus and S . aureus + actA were complemented after electroporation 68 of pMK4-prfA 12 , pAT18 2 or pAT18-actA 69 plasmids ., Aggregation assay was performed in BHI , or in PBS after culture in BHI , for the strains in EGD genetic background , Li strain and for mutants in LO28 or EGDe background , when specified ., Aggregation assay was realized in DMEM for the LO28 or EGDe backgrou | Introduction, Results, Discussion, Materials and Methods | Listeria monocytogenes ( Lm ) is a ubiquitous bacterium able to survive and thrive within the environment and readily colonizes a wide range of substrates , often as a biofilm ., It is also a facultative intracellular pathogen , which actively invades diverse hosts and induces listeriosis ., So far , these two complementary facets of Lm biology have been studied independently ., Here we demonstrate that the major Lm virulence determinant ActA , a PrfA-regulated gene product enabling actin polymerization and thereby promoting its intracellular motility and cell-to-cell spread , is critical for bacterial aggregation and biofilm formation ., We show that ActA mediates Lm aggregation via direct ActA-ActA interactions and that the ActA C-terminal region , which is not involved in actin polymerization , is essential for aggregation in vitro ., In mice permissive to orally-acquired listeriosis , ActA-mediated Lm aggregation is not observed in infected tissues but occurs in the gut lumen ., Strikingly , ActA-dependent aggregating bacteria exhibit an increased ability to persist within the cecum and colon lumen of mice , and are shed in the feces three order of magnitude more efficiently and for twice as long than bacteria unable to aggregate ., In conclusion , this study identifies a novel function for ActA and illustrates that in addition to contributing to its dissemination within the host , ActA plays a key role in Lm persistence within the host and in transmission from the host back to the environment . | Listeria monocytogenes ( Lm ) is a ubiquitous bacterium that survives and thrives within the environment , and a facultative intracellular pathogen that induces listeriosis ., So far , these two complementary facets of Lm biology have been studied independently ., Here we identify ActA , which is a major Lm virulence determinant mediating actin-based motility , as critical for bacterial aggregation and biofilm formation ., ActA promotes Lm aggregation via direct ActA-ActA interaction and ActA C-terminal region , which is not involved in actin polymerization , is essential for aggregation ., Whereas ActA-mediated Lm aggregation is not observed in infected tissues , it occurs in the gut lumen ., Strikingly , ActA-dependent aggregating bacteria exhibit an increased ability to persist within the gut lumen , and are shed in the feces three order of magnitude more and for twice as long than bacteria unable to aggregate ., This study identifies a novel function for ActA , which plays a key role in Lm persistence within the host and transmission . | bacteriology, medicine, microbiology, host-pathogen interaction, animal models, bacterial diseases, model organisms, gastroenterology and hepatology, microbial growth and development, bacterial pathogens, infectious diseases, bacterial and foodborne illness, microbial pathogens, biology, gastrointestinal infections, pathogenesis | null |
journal.pgen.1005310 | 2,015 | Functional Constraint Profiling of a Viral Protein Reveals Discordance of Evolutionary Conservation and Functionality | To comprehensively describe the functional roles of a given protein , which are often diverse for many viral proteins and include catalytic activity , intermolecular interactions , and/or cofactor binding , it is necessary to identify the individual functional residues that carry out the biochemical mechanism ., Sequence conservation analysis is a common strategy to search for functional residues and is facilitated by the availability of public protein sequence databases 1–3 ., The underlying logic is composed of two parts ., First , functional residues are essential ., Second , essential residues are conserved ., However , the reverse may not hold true − conserved residues are not necessary essential ., With the extensively studied influenza A virus , several groups have experimentally demonstrated that conserved residues need not be essential for viral replication 4–6 ., In addition , a residue shown to be essential for viral replication can also be the result of stability constraints , where the residue is essential for protein stability and expression levels , rather than due to functional constraints 7–10 ., Another caveat of sequence conservation analysis is the inefficacy for identifying species-specific functional residues ., This issue is often overlooked ., During natural evolution , continuous diversification and adaptation leads to the acquisition of new functions ., For example , NS1 from influenza B but not influenza A interacts with ISG15 11; NS1 from influenza A but not influenza B interacts with CPSF30 12 ., Furthermore , certain phosphorylation sites are not conserved across influenza A and B viruses 13 ., In fact , non-conserved functional residues have been demonstrated in various organisms 14–17 ., Consequently , when comparing the sequence identities of a set of diverse homologs , as is the case when comparing influenza types A , B , and C , species-specific functional residues may appear as non-conserved residues and be classified as non-functional ., As a result , development of a sequence conservation-independent approach is needed to provide an unbiased assessment for the functionality of individual residues and to permit a systematic interrogation of the relationship between functionality and evolutionary conservation ., The influenza A virus PA polymerase subunit consists of a ∼ 25 kD N-terminal domain and a ∼ 55 kD C-terminal domain 18 , 19 ., Structural information for both domains is available 20–23 ., PA forms a heterotrimer complex with two other influenza virus proteins , PB1 and PB2 ., Together , they function as an RNA-dependent RNA polymerase ., The three subunits perform distinct functions , which contribute to the replication and transcription of the viral RNA genome ., PB1 binds to the viral promoter and is the catalytic subunit for viral RNA synthesis 24 ., PB2 is essential for the transcription of viral RNA and can bind to the 5’ cap of host pre-mRNAs for “cap-snatching” 25–27 ., PA is required for both replication and transcription of the viral RNA and contains an endonuclease catalytic site for cleaving the capped RNA primer 28–31 ., It has also been reported that PA may be involved in other viral processes , such as viral assembly 32 , 33 , and may possess protease activity 34 , 35 ., Recently , several groups have proposed targeting the influenza PA polymerase subunit for antiviral drug development as it is an essential component for viral replication 36–42 ., In this study , we have developed a systematic approach that is independent of any prior knowledge in sequence conservation to identify functional residues at single amino acid resolution ., In this strategy , we coupled a high-throughput fitness profiling platform with an in silico mutant stability prediction ., We employed the influenza A virus PA polymerase subunit as the target protein , due to the availability of structural information and the extensive information available for natural sequence variants ., The fitness effects of amino acid substitutions were profiled across 94% of all PA protein residues using a novel “small library” approach ., Computational modeling predicted the stability effect of all individual substitutions , thus uncovering the structural constraints for individual residues ., By integrating the fitness and structural information , we identified known functional sites previously documented in the literature and provide additional insight into the structure-function relationship of the influenza PA protein ., We further examined the relationship between evolutionary conservation and functional constraints and show that functional residues are not necessarily conserved ., This study not only describes a novel functional annotation platform that provides insight into the relationship between functionality and sequence conservation , but also presents valuable information for drug development and future functional studies of the influenza A virus PA protein ., More importantly , this approach has the potential to be adapted for any protein where structural information is available ., High-throughput genetic approaches have been applied to the study of various proteins ( reviewed in 43 ) , which include several from influenza virus and HIV 44–49 ., Generally , a mutant library is monitored using deep sequencing , and the relative fitness of each mutation can be inferred by changes in the frequency of mutation occurrence throughout the fitness selection process ., Mutant library construction represents a key step in these high-throughput genetic approaches ., An ideal mutant library should contain only one point mutation per genome , which poses a challenge for high-throughput mutagenic strategies ., Existing approaches have used viral genomes that contain multiple mutations within the mutant library ., However , the short read length in current deep sequencing technologies disallows the examination of any possible linkage between distantly placed mutations within each genome ., Consequently , genetic interactions between mutations may exist during the selection process , but are not accounted for during the fitness calculation for individual point mutations ., To resolve this drawback in existing high-throughput genetic approaches , we have developed a “small library” strategy ( Fig 1A ) ., Each mutant library contains a mutated region that can be covered by a single sequencing read ., Here , we generated a 240 bp mutated amplicon by error-prone PCR , which is then cloned into a PCR-generated vector using type IIs restriction enzymes ( BsaI or BsmBI ) ., The resulting plasmid mutant library was constructed from ∼ 50 , 000 clones ., A total of nine different “small libraries” for influenza A/WSN/33 PA were constructed ., Together , these nine “small libraries” covered the entire PA gene ., Each viral mutant library was rescued by transfecting the plasmid mutant library with the other seven wild type ( WT ) plasmids of the influenza A/WSN/33 eight-plasmid reverse genetic system 50 ., A549 cells were then infected with the viral mutant library for 24 hours ., The plasmid mutant libraries ( DNA library ) , post-transfection viral mutant libraries ( transfection ) , and post-infection viral mutant libraries ( infection ) were subjected to deep sequencing ., In this study , we included a technical replicate for sequencing the DNA library , a biological replicate for transfection , and a biological replicate for infection to estimate the reproducibility of individual steps ( S1 Fig ) ., In addition , we also sequenced the WT PA plasmid as a control ., The amplicon sequencing library was prepared for the Illumina MiSeq 250 bp paired-end sequencing , using either DNA ( DNA library or WT plasmid ) or cDNA ( transfection or infection ) ( Fig 1B ) ., For each “small library” , the 240 bp mutated region was amplified by a primer pair that contained a BpmI restriction site ., A subsequent BpmI digestion excised the primer region from the PCR amplicon ., As a result , the entire 240 bp mutated region would be covered by both forward and reverse reads ( S2 Fig ) ., This enabled sequencing error correction by read-pairing ., We obtained a coverage of at least 20 , 000 ( range = 20 , 128 to 965 , 488 ) for each sequencing library ( S3 Fig ) ., The design of our high-throughput genetic platform enables us to examine the mutation in individual genomes ., On average , 44% ( range = 25% to 76% ) of viral genomes contain no mutation ( i . e . WT ) , 33% ( range = 20% to 36% ) of viral genomes contain a single mutation , and 23% ( range = 3% to 42% ) of viral genomes contain at least two or more mutations ( S4 Fig ) ., While a fraction of the genomes in the mutant library contain more than one mutation due to the nature of error-prone PCR , they were filtered out for downstream analysis ., Occurrence frequency for each point mutation was computed from genomes that contained only one mutation ., This allowed a precise fitness calculation for individual point mutations without complication by genetic interactions that may exist with additional mutations ., Individual point mutations exhibited an occurrence frequency of 0 . 04% ( range = 0% to 0 . 3% ) across all DNA libraries ., Whereas the mutation frequency obtained from sequencing the WT plasmid , which served as a control for sequencing error rates , was 0 . 005% ( range = 0% to 0 . 07% ) ( S5 Fig ) ., Comparison of the relative frequency of individual point mutations between replicates was performed to assess the reproducibility of our “small library” high-throughput genetic platform ( see Materials and Methods for the calculation of relative frequency ) ., A Pearson’s correlation of 0 . 95 was obtained for the technical replicate of DNA library , 0 . 76 for the biological replicate of transfection , and 0 . 96 for the biological replicate of infection ( Fig 2A ) ., The strong correlations between replicates validated the design of our high-throughput genetic platform ., Only those point mutations with an occurrence frequency of ≥ 0 . 03% in the DNA library were included in the downstream analysis , which covered 42% of all possible point mutations on the PA gene , to avoid fitness calculations being obscured by sequencing errors ., The relative fitness index ( RF index ) was used as a proxy to estimate the fitness effect for each point mutation ., The RF index of silent mutations ( mean = 0 . 98 ) was significantly higher than that of nonsense mutations ( mean = 0 . 09 ) ( P < 2e−16 , two-tailed Student’s t-test ) ., Furthermore , the RF index distributions of silent mutations versus nonsense mutations were well-separated ( Fig 2B ) , validating that fitness selection was taking place ., The fitness effects of substitutions were profiled across 94% of all amino acid residues in PA ., The fitness profiling data is shown in Fig 2C ., Next we aimed to identify amino acid residues that were functionally essential , but not structurally important ., Essential residues in viral replication can be systematically mapped by high-throughput fitness profiling experiments 46–48 , 51–53 ., However , fitness profiling only quantifies essentialness , but does not partition the structural versus functional role of individual residues ., Several studies have shown that mutating functional residues imposed minimum stability cost to the proteins in which they reside 54–58 , suggesting that functional residues can be pinpointed by identifying substitutions that are deleterious to the virus but not destabilizing to the protein ., Using Rosetta software we predicted the effect of individual substitutions on protein stability ., We used the parameters from row 16 of Table I in Kellogg et al . , which has been shown to give a correlation of 0 . 69 with experimental data and a stability-classification accuracy of 0 . 72 59 , 60 ., We were able to identify substitutions that had a low RF index , but did not destabilize the protein ( Fig 3A ) ., We hypothesized that these residues had large functional constraints with little structural effects to the protein upon substitution ., To identify the substitutions of interest , a cutoff was set at an RF index < 0 . 15 ( based on the separation point of silent mutations and nonsense mutations ) and a predicted ΔΔG < 0 ( not destabilizing ) ., A total of 32 substitutions ( 22 unique residues ) in the PA N-terminal domain and 110 substitutions ( 81 unique residues ) in the PA C-terminal domain satisfied these criteria ., A number of functional residues in the PA protein have been experimentally characterized in the literature ( S1 Table ) ., Out of 32 substitutions of interest in the PA N-terminal domain , eight were at residue positions that carried known biological functions ., This included five substitutions in the endonuclease active site ( E80V , E80G , E80K , E119V , K134 ) 20 , 21 , and six substitutions in the cRNA promoter binding site ( E166D , R170W , R170M , R170K , T173I , T173A ) 18 , 61 ., We also found multiple residues with known biological functions among the 110 substitutions of interest in the C-terminal domain ., This included a substitution at a residue required for endonuclease activity ( H510R ) 28 , a substitution at a residue required for small viral RNA ( svRNA ) binding ( R566W ) 62 , four substitutions at residues required for viral genome replication ( E410V , E524V , K539M , K539E ) 28 , and six substitutions at the PB1-binding site ( N412I , N412Y , Q670R , Q670L , F710I , F710Y ) 22 , 23 ., For all residues that carry a deleterious substitution ( RF index < 0 . 15 ) , residues identified as functional residues ( ΔΔG < 0 ) had a larger relative SASA ( solvent accessible surface area ) versus amino acid positions that were not ( P = 4 . 2e−9 , two-tailed Student’s t-test ) ( Fig 3B ) ., This indicates that the identified functional residues were mostly surface exposed , as expected if they mediate possible interactions with biomolecules ., In fact , ∼ 50% of the solvent exposed residues that carried a deleterious mutation ( relative SASA > 0 . 2 and RF index < 0 . 15 ) were identified as functional residues ( Fig 3C ) ., Since our mutagenesis technique was based on error-prone PCR , which results in a non-comprehensive sampling of all the possible amino acid substitutions at each site , there may be some functional substitutions that were not sampled in our study ., Nonetheless , these results demonstrate the feasibility of combining high-throughput fitness profiling with mutant stability prediction to identify functional sites at single amino acid resolution ., Since the PA C-terminal region’s structure-function relationship remains largely unclear , we aimed to identify functional residues in this region to provide additional insight into the role of PA during viral replication ., Ten previously uncharacterized substitutions with an RF index < 0 . 15 and a predicted ΔΔG < 0 were individually reconstructed and analyzed ., Their spatial locations were distributed throughout the PA C-terminal domain ( Fig 4A and S6 Fig ) ., The effect of these substitutions on the influenza polymerase activity was tested using an influenza A virus-inducible luciferase reporter assay 63 ( Fig 4B ) ., Three substitutions , K281I , K413M , and F681S , completely abolished the influenza polymerase activity ., This defect is unlikely to be a protein destabilizing effect since all ten mutants analyzed did not alter protein expression levels as compared to WT ( Fig 4C ) ., The fact that nine out of ten mutants had a decrease in polymerase activity as compared to WT further validated our high-throughput approach in identifying deleterious mutations ., Interestingly , we found six substitutions ( D426G , E427V , G429E , E430G , L470H , and R512W ) that retained > 10% of the WT influenza polymerase activity ( Fig 4B ) ., A rescue experiment was performed using the influenza A/WSN/33 eight-plasmid reverse genetic system 50 ., Unexpectedly , R512W , which had ∼ 60% of the WT polymerase activity , completely abolished the production of viral particles ( Fig 4D ) ., In addition , E430G , which had a polymerase activity comparable to WT , displayed a four-log drop in virus titer as compared to WT ., In contrast , although D426G and E427V displayed a polymerase activity that was only ∼ 10%-20% of WT , each could produce a much higher amount of infectious virus as compared to other substitutions in this set ( one-log to two-log higher titers as compared to E430G ) ., Our results suggest that the E430G and R512W substitutions each had a functional defect that is unrelated to the polymerase activity ., E430G and R512W were selected for further functional characterization because they exhibited the strongest polymerase activity among all the individually analyzed substitutions , despite their defect in producing infectious virus ., During a viral rescue experiment , there was an accumulation of viral copy number in the supernatant for WT , but not for the E430G and R512W viral mutants ( S7A Fig ) ., In contrast , both mutants displayed an accumulation of intracellular viral copy number similar to WT ( S7B Fig ) ., At 72 hours post-transfection , the HA titer of R512W and E430G was undetected , indicating viral particles were present at a very low amount , if present ( S7C Fig ) ., These results further confirm that E430G and R512W have a defect that is unrelated to polymerase activity ., When this study was initiated , PA was the only influenza polymerase subunit with structural information available ., The structural information for the other two influenza polymerase subunits , PB1 and PB2 , were largely unknown ., Nonetheless , after the completion of this study , the crystal structure of the complete influenza A virus polymerase complex bound to the viral RNA promoter has been published 64 , which provides an independent reference to validate and interpret our data ., Our functional profile identified a subset of PA residues that interact with PB1 ( S8A Fig ) , PB2 ( S8B Fig ) , and the viral RNA promoter ( S9 Fig ) ., Moreover , six out of the 10 validated functional residues participate in these interaction interfaces: − D426 , E427 , and F681 interacted with PB1; L470 interacted with PB2; K281 and R512 interacted with the viral RNA promoter ., Our data also identified functional residues that were not involved in polymerase complex formation or RNA binding activity ., For example , E430 did not interact with either PB1 , PB2 , or the viral RNA promoter ( S10 Fig ) ., This is consistent with our data that E430 is involved in a non-polymerase function ., In addition , a putative functional subdomain independent of the polymerase-interacting surface was identified in our functional profiling data ., This putative functional subdomain is composed of a series of charged or polar residues − D286 , N412 , K413 , R454 , D529 , K559 , and K635 ., Interestingly , this patch of functional residues was adjacent to residue 552 , which has been shown to be a host-specific determinant 65 ., This indicates a possible biological significance of the putative functional subdomain we identified ., Consistently , substitutions at positions D286 , N412 , K413 , R454 , D529 , and K635 were shown to abolish the polymerase activity in our validation experiment ( Fig 5B-5C ) , further confirming the functional importance of this subdomain in viral replication ., Overall , our profiling data is consistent with the polymerase complex-viral RNA promoter complex structural data , which provides an independent validation of our approach ., There are three types of influenza viruses , namely type A , B , and C . Phylogenetic analysis indicates that PA displays a high inter-type diversity ( evolutionary distance among viral strains within the same influenza type ) , while the intra-type diversity is limited ( evolutionary distance between viral strains of different influenza types ) ( S11 Fig ) ., The average inter-type amino-acid sequence identity is < 40% and that of intra-type is > 95% ., The huge divergence among different types of influenza viruses leads us to hypothesize that a significant number of functional residues are type-specific and are non-conserved across different influenza types ., Consequently , we aimed to interrogate the relationship between functional constraints , structural constraints and evolutionary conservation ., In this study , sequence conservation for each residue was computed using Shannon’s entropy 66 ., The higher the entropy , the less conserved a residue is ., Here , we divided all profiled residues into three groups:, 1 ) Functional residues , which had at least one substitution that displayed an RF index < 0 . 15 and a predicted ΔΔG < 0 . 2 ) Structural residues , which did not satisfy the condition of functional residues but had at least one substitution that displayed an RF index < 0 . 15 ., 3 ) “Other” residues , which contained all other profiled residues that were neither functional nor structural residues ( i . e . all profiled substitutions at “other” residues displayed an RF index ≥ 0 . 15 ) ., The entropy calculation was performed on a multiple sequence alignment of 3837 strains from different influenza types ( Fig 6A ) ., In general , functional residues were more conserved than structural residues ( P = 0 . 032 , Wilcoxon rank-sum test ) , and structural residues were more conserved than “other” residues ( P = 2 . 9e−9 , Wilcoxon rank-sum test ) ( S12 Fig ) ., From this analysis , 58% of functional residues , 43% of structural residues , and 26% of “other” residues were highly conserved ( entropy < 0 . 1 ) ., This indicates that a significant number of functional residues are not conserved across the different types of influenza virus ., We further computed a phylogenetic-based dN/dS analysis on each codon across the influenza A virus PA coding sequence with FUBAR 67 ( Fig 6B ) ., A mild , yet statistically significant , correlation was detected between dN/dS and RF index ( Spearman’s rank correlation = 0 . 38 , P < 2 . 2e−16 ) ( S13A Fig ) ., On average , functional residues and structural residues had a lower dN/dS as compared to “other” residues ( P = 7 . 2e−8 and P = 1 . 5e−8 respectively ) ( S13B Fig ) ., However , the difference of dN/dS between functional residues and structural residues was not significant ( P = 0 . 57 ) ., This result shows that dN/dS may not be a good indicator to distinguish functional residues from structural residues ., In addition , some functional residues exhibited a dN/dS that was well within the range of “other” residues , demonstrating that some functional residues could not be identified by dN/dS analysis alone ., The utility of dN/dS is largely determined by the phylogenetic depth of the sequences being analyzed ., In fact , it has been shown that when the genetic diversity is low , as is the case of PA protein sequences from type A influenza virus , dN/dS becomes less sensitive to purifying selection 68 , and may not be able to identify functional residues ., We next examined individual residues validated in this study ., Among the 13 validated functional residues , three ( K281 , K413 , and E430 ) had both entropy and dN/dS at the median level ( Fig 6C ) ., Moreover , these residues are not conserved across different influenza types ., These results confirm that functional residues may not be identified by phylogenetic-based analysis alone ., As expected , sequence conservation-based functional site prediction software was unable to predict these functional residues ., We tested three software approaches , firestar 69 and two classification schemes under FRpred 70 , 71 , namely FRcons and FRsubtype ., FRcons and FRsubtype were each able to identify only one of our validated functional residues ( D286 for FRcons and K413 for FRsubtype , respectively ) using a category cutoff of ≥ 8 ., Firestar was not able to identify any of our validated functional residues ., Furthermore , out of a set of 28 functional residues identified in the literature ( S1 Table ) , our approach identified 12 , whereas FRcons , FRsubtype , and firestar were only capable of identifying 4 , 2 and 5 functional residues , respectively ., This comparison demonstrates that our methodology can outperform phylogenetic approaches in identifying functional residues ., We aimed to further investigate the structural basis of type-specific functional residues ., The RNA binding function is required for viral replication and is conserved among type A and B influenza viruses ., In the validation above , substituting lysine K to isoleucine I at residue 281 completely abolished the polymerase activity ., This highlights the importance of the hydrogen bond formed between K281 and the RNA phosphate backbone in the influenza A virus ( Fig 7A and boxed in S14 Fig ) ., However , PA K281 is not conserved between type A and B influenza viruses ., All influenza B viruses carry an alanine A at residue 281 , which is unable to form a hydrogen bond with the RNA backbone ., The critical hydrogen bond mediated by K281 in influenza A virus is replaced by the main chain of G569 in the influenza B virus ( Fig 7B and boxed in S15 Fig ) ., In fact , structural analysis indicates that type A 64 and B 72 influenza viruses display different hydrogen bonding patterns between PA and the viral RNA promoter ( S14 Fig and S15 Fig ) ., Thus , conserved functions may not necessarily require conserved functional residues ., Together , these analyses show that while certain functional residues were completely conserved among different types of influenza viruses , a significant number of residues that mediate critical viral functions may not be conserved , and suggests that some residues may have acquired functionality in recent evolutionary history ., Traditionally , sequence conservation is the common approach for identifying functional residues ., In this study , we coupled two high-throughput techniques , experimental fitness profiling and in silico mutant stability prediction , to systematically identify functional residues in the influenza A virus PA protein ., This strategy provided a direct measure of essentialness and enabled the partitioning of functional constraints versus structural constraints at each residue position ., This approach is independent of any prior knowledge of sequence conservation ., Therefore , it is devoid of the caveats associated with sequence conservation analysis and possesses the power to identify species-specific functional residues ., A number of functional residues identified in this study , are not completely conserved across different types of influenza viruses , suggesting that even functional residues may not be conserved ., This disparity between conservation and function highlights the power of our approach to identify functional residues that may not be identified by traditional sequence conservation analysis alone ., We anticipate that this method can be further improved as the accuracy of mutant stability prediction methodology improves ., It has been shown that although most force fields exhibit a correct trend in ΔΔG prediction , the precision is still lacking as compared to experimental methods 73 ., For example , in this study , N412I decreases protein expression levels , despite being predicted as a stabilizing mutant ., In addition , it is known that most proteins are able to buffer a small destabilizing effect without becoming unfolded , and hence without attenuating the fitness 74 , 75 ., As a result , understanding the stability buffer margin will help to determine the optimal ΔΔG cutoff in our approach ., It is also known that many proteins have multiple conformations , which may further complicate the ΔΔG prediction ., Together , these caveats may explain the weak correlation between the predicted ΔΔG and RF index in this study ., To obtain a more accurate measurement of protein stability , high-throughput experimental analysis on protein stability may provide an alternative 76 , 77 ., All the advances stated above will improve the accuracy of our platform in identifying functional residues within a target protein ., During natural evolution , continuous accumulation of protein mutations drives speciation and divergence from the common ancestor ., The genomic plasticity of an evolving species permits the acquisition of new function through mutations 78 ., Evolution of a new function has been demonstrated in bacteriophage λ within an experimental timescale 79 , and a long-term evolution experiment on Escherichia coli80 ., Therefore , it is not surprising to see species-specific function even in recently separated species ., Based on the sequence comparison of hemagglutinin , it was estimated that type A and B influenza virus diverged from type C ∼ 8 , 000 years ago , whereas type A influenza virus diverged from type B ∼ 4 , 000 years ago 81 ., This length of time is sufficient for the influenza virus to develop a type-specific function as exemplified by type-specific virus-host interactions in NS1 11 , 12 ., Furthermore , conservation of protein function does not necessarily support that sequence conservation exists at the primary sequence level , which is evidenced by the differences between the nuclear localization signal of influenza A and B NP proteins 82 , 83 ., In fact , this study reveals that type-specific functional residues are prevalent in the influenza virus PA protein ., These results not only provide insight into how functional residues evolve through species diversification , but also highlight the caveats encountered when identifying functional sites from conservation-based approaches ., In the past decade , proteins from different medically important viruses , such as influenza , HIV , and HCV , have been crystallized 84–86 ., The approach described in this study systematically integrates the available structural information with mutation fitness information to examine the structure-function relationship of a viral protein of interest and to map functional subdomains ., Profiling datasets will facilitate functional characterization of the protein of interest , and will promote targeted drug discovery and rational drug design ., The emergence of drug resistant mutations is a major challenge for antiviral drug development ., Therefore , it is important to target functional subdomains that are less tolerable to substitution to increase the genetic barrier for developing drug resistant mutations ., Our profiling technique can help locate such functional subdomains that are suitable for drug development ., More importantly , our technique can potentially be adapted to study any protein , provided the relevant structural information is available ., The PA plasmid mutant libraries were created by performing error-prone PCR on the PA segment of the eight-plasmid reverse genetics system of influenza A/WSN/1933 ( H1N1 ) 50 ., To generate the mutated insert , we PCR-amplified regions of the PA gene from pHW2000-PA plasmid with error-prone polymerase Mutazyme II ( Stratagene , La Jolla , CA ) according to the manufacturer’s instructions ., The following primers were used:, Library 1 insert: 5’-CAG GTC TCA TCA AAA TGG AAG ATT TTG TGC GA-3’ and 5’-CAG GTC TCA ATA CTG TTT ATT ACT GTC CAG GC-3’, Library 2 insert: 5’-CAG GTC TCA TCG AGG GAA GAG ATC GCA CAA TA-3’ and 5’-CAG GTC TCA CTG GTT TTG ATC CTA GCC CTG CT-3’, Library 3 insert: 5’-CAG GTC TCA CCG ACT ACA CTC TCG ATG AAG AA-3’ and 5’-CAG GTC TCA TTT ACT TCT TTG GAC ATT TGA GA-3’, Library 4 insert: 5’-CAG GTC TCA ACG GCT ACA TTG AGG GCA AGC TT-3’ and 5’-CAG GTC TCA TAA TTT GGA TTT ATT CCC TTT TC-3’, Library 5 insert: 5’-CAG GTC TCA AAC CCA ATG TTG TTA AAC CAC AC-3’ and 5’-CAG GTC TCA GCC TTG TTG AAC TCA TTC TGA AT-3’, Library 6 insert: 5’-CAG GTC TCA AAT TGA GGT CGC TTG CAA GTT GG-3’ and 5’-CAG GTC TCA CCC TCC TTA GTT CTA CAC TTG CT-3’, Library 7 insert: 5’-CAG GTC TCA ATT TCC AAT TAA TTC CAA TGA TA-3’ and 5’-CAG GTC TCA TTA ATT TTT GAG GTT CCA TTT GT-3’, Library 8 insert: 5’-CAG GTC TCA GGC CTA TGT TCT TGT ATG TGA GG-3’ and 5’-CAG GTC TCA TGT G | Introduction, Results, Discussion, Materials and Methods | Viruses often encode proteins with multiple functions due to their compact genomes ., Existing approaches to identify functional residues largely rely on sequence conservation analysis ., Inferring functional residues from sequence conservation can produce false positives , in which the conserved residues are functionally silent , or false negatives , where functional residues are not identified since they are species-specific and therefore non-conserved ., Furthermore , the tedious process of constructing and analyzing individual mutations limits the number of residues that can be examined in a single study ., Here , we developed a systematic approach to identify the functional residues of a viral protein by coupling experimental fitness profiling with protein stability prediction using the influenza virus polymerase PA subunit as the target protein ., We identified a significant number of functional residues that were influenza type-specific and were evolutionarily non-conserved among different influenza types ., Our results indicate that type-specific functional residues are prevalent and may not otherwise be identified by sequence conservation analysis alone ., More importantly , this technique can be adapted to any viral ( and potentially non-viral ) protein where structural information is available . | The analysis of sequence conservation is a common approach to identify functional residues within a protein ., However , not all functional residues are conserved as natural evolution and species diversification permit continuous innovation of protein functionality through the retention of advantageous mutations ., Non-conserved functional residues , which are often species-specific , may not be identified by conventional analysis of sequence conservation despite being biologically important ., Here we described a novel approach to identify functional residues within a protein by coupling a high-throughput experimental fitness profiling approach with computational protein modeling ., Our methodology is independent of sequence conservation and is applicable to any protein where structural information is available ., In this study , we systematically mapped the functional residues on the influenza A PA protein and revealed that non-conserved functional residues are prevalent ., Our results not only have significant implication on how functionality evolves during natural evolution , but also highlight the caveats when applying conservation-based approaches to identify functional residues within a protein . | null | null |
journal.pcbi.1002500 | 2,012 | A Spatial Model of Mosquito Host-Seeking Behavior | We provide a review of the biological literature that influenced our modeling choices ., The scenario that we consider is illustrated in Fig . 1 , where hosts are distributed in groups , or patches , and they emit an odor ( e . g . ) that gets carried by and diffused in the wind in the same way that a puff of smoke dissipates in time ., Uniform , laminar wind extends host odor into a long , thin plume with sharp transverse gradients and shallow longitudinal gradients ., If the wind is turbulent , the odor plume is highly intermittent , but still retains relatively shallow average longitudinal gradients compared to the transverse gradients 13 ., We class mosquito host-seeking behavior as either plume finding , which is flight in search of an odor plume , or plume tracking , which is flight within the odor plume ( these terms are adopted from 12 ) ., In the model , hosts emit a single gaseous compound that attracts mosquitoes , is convected by the wind , and diffuses in the air ., For the purpose of this paper , we assume that the attractant is ., The distribution over time is modeled by a convection-diffusion partial differential equation ., The convection velocity of the wind is given by a vector , and the concentration of is described by the equation ( 1 ) where is time and are spatial coordinates on a square domain of length ., Throughout the paper , we refer to this square computational domain as the “ simulation region . ”, The constant diffusion coefficient reflects the rate at which diffuses in air in the absence of wind ., The term represents the odor cue emitted at a constant rate by the hosts in units of ppm/s:The wind velocity consists of two components:where is used to introduce drifts or relatively large features produced by the air and is a stochastic velocity vector used to approximate the effect of small-scale wind variations in the domain ., The direction of at each point in the domain is chosen uniformly from and the magnitude is chosen from a normal distribution centered around zero ., The large-scale velocity field is a constant speed flow from bottom to top ( see Fig . 2 ) in most of our simulations ., However , we sometimes use a meandering plume in place of the straight plume , which is given by the expression ( 2 ) where the frequency is ., It is easily checked that this flow is incompressible ., Equation ( 1 ) is numerically evolved using second-order centered differences to approximate the Laplacian and a first-order conservative upwind finite difference method for the convection term ( similar to page 636 of 30 ) ., The normal components of the concentration gradient are zero at the boundary ( Neumann conditions ) ., We integrated the equations with a forward Euler method and confirmed that our solution converged by verifying that significantly decreasing the spatial grid size and time step size had a minimal change in the solution ., In most of the simulations in this paper , we consider length scales and time frames consistent with the mosquitoes being in close proximity to the hosts ., The length of a side of the square domain is 10 m and the simulations cover time periods of 50–500 seconds ., The hosts are situated in the middle of the domain , 5 m from the top and bottom domain edges ( see Fig . 2 ) ., Initially , the domain is bare of and it takes about 45 seconds for the plume to reach the domain edge ., At that point , the mosquitoes are released into the domain , with starting positions dependent on their particular flight behavior ., For upwind plume-finding behavior , the mosquitoes are all released downwind of the hosts; for downwind and crosswind behaviors , the entry is along the upwind side of the domain ., Mosquitoes and hosts are modeled as discrete individuals , or agents ., Hosts are stationary , motivated by the interaction between nocturnal Cx ., quinquefasciatus mosquitoes and their roosting bird prey ., In this section we explain the mosquito navigation model , which differs during plume finding and plume tracking ., Assumptions about mosquito agent behavior include: We assess model performance by comparing different mosquito navigation strategies and by evaluating the sensitivity of the model to a subset of the simulation parameters ., We find that the gradient and sampling methods can exhibit comparable performance and that , in general , the parameter set in Table 1 is robust to small changes ., Intuitively , a crosswind flight strategy should result in a larger number of contacts than upwind or downwind behavior if the plume is straight ., This is because mosquitoes close to the plume are more likely to intercept it if their motion is primarily crosswind ., But for a meandering plume , it is not obvious if a crosswind flight strategy is more effective than an upwind or downwind strategy ., We simulated mosquito behavior in both straight and meandering plumes and recorded the proportion of mosquitoes that found a host and the average time that it took a mosquito to locate a host ., We estimated the effectiveness of each plume-finding behavior using these results and found that a crosswind strategy is superior to upwind and downwind strategies , but is less efficient ., The odor plumes were produced by a regular arrangement of nine hosts with a density of 1 host per 1 ( 0 . 09 ) located in a single small patch in the center of the square simulation domain ., The time series of random velocity fields superposed over the large-scale flow was the same for all simulations ., See Table 1 for parameter choices and Eq ., ( 2 ) for the formula for the meandering plume ., An example of the general form of the meandering plume is shown in Fig . 5 , alongside a straight plume for comparison ., The meandering plume covers more area and achieves a greater width than the straight plume ., The outermost contour in both plots is , the sensing threshold of the mosquito ., The other contours are equally divided between and the maximum concentration within each plume ., The meandering plume has a maximum concentration that is approximately three times higher than that of the straight plume ., Since the sources are the same in both simulations , the concentration difference is due solely to the differing velocity fields ., The results in Table 3 show the proportion of mosquitoes that found a host , , and the average time to locate a host , ., The column labeled gives the value of after only 35 s of host seeking ., We make the following observations from the table ., We explored the number of contacts that occur in two host groups of equal density but unequal number ., An equal per capita contact rate across both host groups would indicate that distributing the hosts into two separate patches has no effect ., However we found unequal per capita rates between groups , indicating that a mosquito is less likely to detect a large group of hosts than it is to detect the same number of hosts distributed in smaller groups or as individuals ., We also found unequal numbers of contacts between two unequally-sized groups , indicating that one group was easier to find than the other ., We performed a set of simulations in which we held the total number of hosts constant in the domain ( 10 birds ) , but split them between two groups in pairs of 9 and 1 , 8 and 2 , etc . , down to 5 and 5 hosts per group ., Fig . 2 shows an example of the 7-3 distribution ., All other parameters were the same as in the previous section for the straight plume ., The two straight plumes from the host groups were well separated from each other and from the left and right domain edges ., The exact positions of the hosts were allowed to vary from one simulation to the next , and this affected the plume shape and the internal distribution of over the plume ., To compare results between groups we plotted the ratio of the mosquito-host contacts in the smaller group divided by the mosquito-host contacts in the larger group ( S/L ) against the ratio of the number of hosts in the smaller group over the larger group ( see Fig . 6 ) ., Each point in the error bar plot shows the mean and standard deviation over 150 simulations for each S/L ratio ., The diagonal line in Fig . 6 denotes the case when the per capita contact rates are the same between groups ., A value of 1 on the represents the case when the number of contacts was the same between both groups ., When there are 5 hosts in both groups the number of contacts is the same , confirming that there is no left-right bias in the velocity field ., When the group sizes are unequal , the per capita contact rates are higher in the small group , but the total number of contacts is higher in the large group ( except for the nearly equal 6-4 distribution ) ., This occurs in the region between the lines and ., The crosswind strategy is closest to having equal numbers of contacts in both host groups , which corresponds to equal ease in locating either group ., In this section , we consider the effect of varying host density given a constant number of hosts ., Intuitively , the size of the region where the hosts are congregated will affect the number of mosquito-host contacts for two reasons ., First , a larger host area is more likely to be found by mosquitoes; second , the spatial arrangement of the hosts affects the size and shape of the odor plume they generate ., We performed a set of simulations in which 10 hosts were distributed in a single patch in the center of the domain ., Patch area was varied from 10–80 ( 1–7 . 4 ) with a corresponding host density varying from 1–8 per host ( 0 . 1–0 . 74 ) ., Our simulation region was 1076 ( 100 ) , so that the patch occupied less than 10% of the simulation region ., For each host density , we ran 150–450 simulations to average the effects of host position and individual mosquito choices , with more simulations performed for larger patch areas ., The proportion of mosquitoes that made contact in the group is shown by the solid markers in Fig . 7 , with the bars corresponding to one standard deviation ., All three plume-finding behaviors exhibit a slight upward trend in with increasing host patch area ., Crosswind plume finding resulted in the most contacts , while upwind and downwind plume finding exhibit nearly identical results ., In Fig . 7 , is plotted against the percentage , where is the side length of the square host patch divided by the square computational domain side length ., is a dimensionless ratio that relates a length scale important to the host distribution to one that is important to the mosquito distribution ., As increases , the hosts are spread over a larger area and the host density decreases ., We approximated the results by the line , where is the proportion of contacts made in the hypothetical case where the area of the host patch is zero ., depends on plume-finding behavior , wind velocity , number of hosts , and other simulation parameters ., We performed a least squares fit to find for each plume-finding behavior ( , , and ) and the lines are shown in Fig . 7 ., In the Discussion , we present a first attempt to link this linear approximation from our agent-based model to a contact rate that has been used in standard ( nonspatial ) epidemiology models ., Our results generally show that crosswind plume finding most reliably led mosquitoes to a blood meal source ., The effectiveness of the crosswind strategy over periods of minutes is compatible with the conclusions of the mark-release-recapture studies of Anopheles gambiae Giles in 34 , where the dispersion of recaptured mosquitoes was related primarily to the distribution of human settlements ( over time scales of days ) ., We find that this effectiveness was accompanied by a high cost in time , as also seen in 12 ., If the success of host location was restricted to occur within 35 s of mosquito release , then crosswind flight was the superior strategy only in a straight odor plume ., In a meandering plume under the time limit , up- and downwind searching were better than crosswind flight ., These results are similar to the conclusions drawn by Sabelis and Schippers 20 , who used a geometric argument to show that crosswind plume finding is less effective when wind direction is highly variable ., When hosts were arranged in two groups of different size , the larger group consistently attracted more mosquitoes regardless of the plume-finding strategy , although the difference was less pronounced for crosswind flight ( Fig . 6 ) ., At the same time , the smaller group experienced a higher per capita contact rate ., These results are consistent with Foppa et al . 7 , which reports on indoor experiments of Cx ., quinquefasciatus feeding on roosting chickens ., They found that the per capita feeding rate on a single chicken was about 4 . 27 times higher than that on a chicken in a group of nine ., We find similar mean per capita ratios ( 4 . 3 and 4 . 4 ) for the upwind and downwind plume-finding behaviors when ten hosts are split into a group of nine and a singleton ., The close match is surprising since our simulations included wind , whereas the experiments were conducted indoors ., However , the uncertainty in the simulations and experiments is high ., The results indicate that it is advantageous for birds to roost together in larger groups on this spatial scale , because on average they will receive fewer bites ., This phenomenon of attack abatement is well-known in the literature on predator-prey interactions 35 , and likely has two causes in the situation considered here ., First , the odor plume exuded by a group does not grow linearly with the number of hosts , because the hosts are clustered together ., This leads to fewer mosquitoes locating the group than would find the same number of well-spaced individuals ( an avoidance effect ) ., Secondly , mosquitoes only need a fixed amount of blood , and so they will not attack additional individuals even if they are available ., This is a dilution effect , also seen in the reduction of groups at risk for malaria resulting from urbanization 36 ., The resulting contact heterogeneity arising from attack abatement could have important implications for transmission dynamics ., Standard models of mosquito-borne transmission assume that the mosquito contact rate on one host is inversely proportional to the numbers of hosts ( reviewed , e . g . by 37 ) ., This is likely true when hosts are so abundant that a mosquito will always be able to locate one ., Our simulations indicate that the probability that a mosquito will locate a host is largely determined by the shape of the odor plume ( see Table 3 ) ., However , the shape of an odor plume is difficult to predict , since it depends on local wind velocity and turbulence due to landscape features ., We therefore propose an approximation that does not depend on plume shape , and is instead based on , which can be viewed as a patchiness parameter ( Fig . 7 ) ., This allows us to model contact rates when hosts are variably distributed over patches within a larger domain more realistically even if exact features of the relevant odor plumes are unknown ., We derive a patchiness-driven contact rate model by starting with the contact rate for malaria transmission from Chitnis et al . 38 which applies specifically in the case of homogeneous mixing in a fixed area without considering host-seeking mechanisms ., They propose a vector-host contact rate ofwhere is the number of contacts each mosquito wants per unit time , is the maximum number of contacts a host can receive per unit time , is the total number of vectors and is the total number of hosts ., This implies that is the approximate contact rate when the mosquito population is small since every mosquito in a small population is expected to locate a host when the populations are homogeneously mixed ., We seek to modify this expression to account for host patch area , as depicted in Fig . 1 , and mosquito behavior ., When the hosts are limited to a small patch of area within a larger region , we make two modifications to the formula above ., First , not every mosquito in a small population can be expected to find a host , since not all of the mosquitoes will come into contact with the odor plume ., For this reason , we replace the contact rate with , where represents the ratio of the diameter of the patch where the hosts are located to the diameter of the simulation region ., The second modification comes from the fact that even when the patch size becomes small ( and ) , the number of contacts must be nonzero since even a tiny area produces a plume that mosquitoes can follow to a host ., We want our modified contact rate function to be consistent with the original function in 38 when , which is the case where the host patch is equal to the entire simulation region ., Based on these observations we propose the modified contact function ( 8 ) where represents the proportion of mosquitoes that find a host as the patch of area containing the hosts becomes infinitesimally small ., We note that when the host patches are small ( i . e . small ) , Eq ., ( 8 ) can be linearized to read , where the factor in brackets , , is interpreted as the proportion of mosquitoes that find a host for a given patch-to-region length ratio ., This proportion can be computed from the simulations , as shown by the lines in Fig . 7 ., Therefore , reasonable assumptions about the patchy distribution of hosts in the domain allow us to devise a quantitatively reasonable estimate of the host-finding probability of mosquitoes for the space and time scales considered here ., The length and time scales in the simulations presented here were motivated by the indoor experiments in 7 and a desire to quantify mosquito success near the end of a host-seeking flight ., Our conclusions must be validated for spatial domains or time periods that are significantly larger ., Furthermore , there are other probability distributions that can be used for mosquito navigational choices ( see e . g . 12 ) and a variety of initial mosquito spatial arrangements that may affect our conclusions ., Such parameter choices are within the capabilities of the model , and to demonstrate we briefly present a simulation of a growing odor plume in a large domain ., Mosquitoes engaging in all three types of plume finding ( 2000 mosquitoes each ) were initially uniformly distributed over a disk of radius 100 m centered on four clusters of ten roosting birds each ., The wind meandered according to an incompressible flow appropriate for the large domain size and the crosswind mosquitoes flew in the same direction for a number of decisions chosen from a Lévy distribution having a power law in the tail of 39 ., The Lévy distribution was centered at 0 . 7 s , the mean value of in Table 1 , and only positive values were sampled ., The mosquitoes were placed in the domain after 150 seconds when the odor plume was approximately 35 meters long ., Fig . 8 shows the state of the odor plume after 250 s and 500 s , along with the positions of all nearby mosquitoes ., The simulation was run for 1500 s , at which point the odor plume was about 180 m in length and 4 . 2% , 6 . 4% , and 26 . 6% of the upwind , downwind , and crosswind mosquitoes had located a host respectively ., The success rate was lower compared to the smaller domain , particularly for the upwind and downwind mosquitoes that rapidly left the plume behind ( see Fig . 8 , right ) ., The crosswind mosquitoes moved downwind at the same rate the plume did , and therefore had many more opportunities to intercept it ., There were still a substantial number of crosswind mosquitoes interacting with the plume when the simulation ceased ., We hope to expand our careful analysis of the smaller domain to a larger domain in the future ., In larger domains , it would be interesting to include the effect of mosquito breeding sites and how their locations affect the biting rate of hosts near them in comparison to hosts living farther away ., Significant differences in the biting rates in spatial relation to breeding sites have been reported for malaria in 40 ., There are other potential model extensions that are equally interesting ., Disease transmission via mosquito bite , host movement , infection , and demographic processes in both vertebrate hosts and mosquitoes and the dependence of these processes on biotic and abiotic factors could be integrated with the existing model for explicit small-scale modeling of disease spread ., This modeling framework is capable of accommodating many further levels of complexity , such as gusting wind , moving hosts , multiple host types , odor-baited traps , variable breathing rate , compound odors , repellents , etc ., The challenge will be to identify the components that most strongly affect the behavior of the model system and the underlying reality on which it is based ., For example , in the simulations presented here the spread of the host odor is dominated by turbulent convection and diffusion plays a very minor role ., It is therefore reasonable to hypothesize that the particular odor cue used by the mosquitoes will have little effect , and that substituting lactic acid for ( for example ) will not impact the resultant contact rates ., See 13 for a discussion of turbulent mixing versus molecular diffusion in the context of chemotaxis ., Additional factors could be included in the model in order to capture subtle differences between mosquito species , lighting effects , and other elements not included in the current work ., Hosts tend to attract mosquitoes in unequal ways 41 ., Differential attractiveness , the emission of different levels of by different hosts as well as multiple odor cues can also be introduced into the model to study situations like those presented in 42 , where mosquitoes of many species finding humans distributed in huts , or in 43 , where mosquitoes were collected inside village houses in Tanzania ., These studies report an approximate direct relationship between the number of inhabitants per house and the number of mosquitoes collected ., Our model could be used to study which factors influence more strongly this relationship ., From the point of view of controlling the vector population , the model presented here may offer some insights into how the spatial distribution of mosquito traps may affect the overall control ., This can be accomplished by replacing the hosts in the model with odor-baited mosquito traps and adjusting appropriate parameters ., This was addressed in a non-spatial model by Okumu et al . 10 , where homogeneous mixing of hosts was assumed ., They discuss the importance of space in the vector-host contact process and indicate that the rate at which an individual host is discovered by an individual vector depends on the distance between hosts and vectors as well as on the size of the odor plumes generated by the hosts ., Further , spatial characteristics such as the topography and wind direction are known to be influential in the rates at which individual hosts are found ., Our model can explicitly include spatial features to compare strategies of where to place mosquito traps relative to blood-source hosts ., According to the studies in 44 the distances at which various species of mosquitoes responded to baits by initiating orientation toward the them was 30 meters or less ., Therefore , the small model length scales presented here are appropriate for such simulations . | Introduction, Model, Results, Discussion | Mosquito host-seeking behavior and heterogeneity in host distribution are important factors in predicting the transmission dynamics of mosquito-borne infections such as dengue fever , malaria , chikungunya , and West Nile virus ., We develop and analyze a new mathematical model to describe the effect of spatial heterogeneity on the contact rate between mosquito vectors and hosts ., The model includes odor plumes generated by spatially distributed hosts , wind velocity , and mosquito behavior based on both the prevailing wind and the odor plume ., On a spatial scale of meters and a time scale of minutes , we compare the effectiveness of different plume-finding and plume-tracking strategies that mosquitoes could use to locate a host ., The results show that two different models of chemotaxis are capable of producing comparable results given appropriate parameter choices and that host finding is optimized by a strategy of flying across the wind until the odor plume is intercepted ., We also assess the impact of changing the level of host aggregation on mosquito host-finding success near the end of the host-seeking flight ., When clusters of hosts are more tightly associated on smaller patches , the odor plume is narrower and the biting rate per host is decreased ., For two host groups of unequal number but equal spatial density , the biting rate per host is lower in the group with more individuals , indicative of an attack abatement effect of host aggregation ., We discuss how this approach could assist parameter choices in compartmental models that do not explicitly model the spatial arrangement of individuals and how the model could address larger spatial scales and other probability models for mosquito behavior , such as Lévy distributions . | Mosquito-borne diseases can spread when a mosquito bites a vertebrate host to obtain a blood meal for egg-laying ., The first step in the transmission process consists of the mosquitoes seeking and finding a host ., Mosquitoes use the wind direction and odors , such as carbon dioxide , emitted by the hosts in order to locate a host to bite ., We present a spatial computational model of the host-seeking process in a region where hosts are heterogeneously distributed in clusters ., The model is used to analyze the success in finding hosts once the mosquitoes are close to the host ., We show that the number of mosquito-host contacts increases as hosts become more widely spaced within their clusters; that mosquito flight perpendicular to the wind leads to greater success in locating a host; and that the number of bites per host decreases when hosts aggregate into larger clusters . | mathematical computing, mathematics, population modeling, biology, computational biology, infectious disease modeling | null |
journal.pgen.1007172 | 2,018 | Novel genetic polymorphisms associated with severe malaria and under selective pressure in North-eastern Tanzania | Sub-Saharan Africa bears a disproportionately high share of the global Plasmodium falciparum malaria burden , with 90% of the estimated 212 million annual cases and 92% of 429 , 000 annual deaths , mostly in children under five years of age 1 ., Whilst the majority of cases of Plasmodium falciparum infection are asymptomatic or cause only mild to moderate clinical symptoms , a subset of affected individuals present with severe manifestations such as severe malarial anaemia and cerebral malaria ., Risk factors for severe malaria and its various clinical subtypes are poorly understood , although host and parasite genotype , age and immune status have all been established as playing a significant role in individual host susceptibility 2 ., Plasmodium falciparum has also exerted significant selection pressure upon the human genome , as evidenced by the geographical concurrence of malaria parasite prevalence with sickle cell trait ( HbAS ) and other haemoglobinopathies , such as the thalassemias and glucose-6-phosphate dehydrogenase ( G6PD ) deficiency ., Recent studies , set in a region of high malaria transmission in north-eastern Tanzania , estimated that host genetic factors account for approximately 22% of the total variation in severe malaria risk 3 , consistent with previous findings in a Kenyan family-based study 2 ., Less than half of this variation can be explained by erythrocyte-associated polymorphisms 4 , including HbS ( sickle cell trait ) , alpha-thalassaemia , ABO blood group 5 and G6PD deficiency 4 ., Novel polymorphisms in or around USP38 , FREM3 3 , glycophorins gypA/B/E 6 , 7 , DDC 8 , MARVELD3 and ATP2B4 9 account for additional variation but , in sum , are less protective than heterozygous carriage of HbS 3 ., Moreover , the effects of some of these loci are subtype- , location- , or population-specific 3 , 6 , 7 , 9 , reinforcing the need for targeted genome-wide association studies ( GWAS ) in different African populations ., Utilising such an approach with robust malaria phenotypes in parallel with whole genome sequencing of study populations is crucial to unravelling host genetic factors that could lead to a greater understanding of protective immunity and development of new tools for disease prevention ., To identify novel loci associated with severe malaria in north-eastern Tanzania , we applied genome-wide association and haplotype-based selection methods to a case-control dataset with extensive phenotypic data for 914 participants and 15 . 2 million SNPs ., In addition to the expected HbS ( sickle cell ) association , our analyses reveal multiple novel loci under association or selection ., Association analysis highlighted significant SNPs clusters within IL-23R , IL-12RB2 , LINC00944 , and KLHL3 whilst lone SNP associations were also present within TREML4 and ZNF536 ., Further , we reveal loci under recent positive selection including GCLC and loci within the Major Histocompatibility Complex ( MHC ) ., These analyses were supported by whole genome sequencing of an independent dataset consisting of 247 Tanzanian individuals within parent-child trios , which was used to confirm the allele frequencies of putative associations and determine if there are any linked common structural variants in chromosome regions encoding important polymorphisms ., All severe malaria cases ( n = 449 ) and controls ( n = 465 ) came from the Tanga region of North-Eastern Tanzania ., Severe malaria cases presented with varying combinations of hyperlactataemia ( 57 . 0% ) , severe malarial anaemia ( 49 . 2% ) , respiratory distress ( 27 . 6% ) and cerebral malaria ( 26 . 7% ) ( Table 1 ) ., Compared to controls , malaria cases were younger ( t test P<2 . 2x10-16 ) and marginally more likely to be male ( Chi squared P = 0 . 012 ) ( Table 1 ) ., DNA from all samples ( n = 914 ) was genotyped on the Illumina Omni 2 . 5 million SNP chip , and imputed against the 1000 Genomes reference panel ( Phase 3 ) 10 and a Tanzanian parent-child trio panel ( see below ) , using Beagle 4 . 1 11 , leading to 15 . 2 high quality SNPs ., These markers were complemented by 180 SNPs within malaria candidate genes , including HBB ( encoding HbS ) 3 , 4 , 5 on the same cases and controls ., DNA from a validation cohort of 78 healthy parent and child trios and 13 independent individuals ( “Trios dataset” , n = 247 ) were whole genome sequenced using Illumina HiSeq2500 technology ., For the GWAS samples , a principal component analysis ( PCA ) using all genome-wide SNPs revealed a low degree of stratification by ethnicity and case-control status ( S1 Fig ) and potential cryptic relatedness due to familial clustering ., A similar analysis revealed that GWAS and Trio sample clusters overlap , and there is some separation from the other 1000 Genome African populations , including Yoruba ( Nigeria ) and Luhya ( Kenya ) ( S1 Fig ) ., GWAS analysis was undertaken with EMMAX mixed model regression 12 , controlling for age as a fixed effect and relatedness ( represented in a kinship matrix ) as a random effect to account for the cryptic population clustering ., Separate models of association were fitted for each SNP ( additive , heterozygous , dominant , recessive ) , with their respective genomic inflation factors all being close to one ( see S1 Fig for the heterozygous results ) , consistent with reliable adjustment for stratification ., A total of 53 SNPs ( in 16 genomic regions ) were identified with a significance level below our threshold ( P<1x10-6 ) ( Fig 1 , Table 2 , S1 Table ) ., Relaxing the stringency would lead to 258 SNPs with a p-value below 1x10-5 and 2 , 322 below a threshold of 1x10-4 ., As expected , the most significant association was with the sickle cell locus , rs334 ( P = 8 . 59x10-13 , heterozygous odds ratio = 0 . 07 ) ( Table 2 ) ., Controlling for HbS status through a complementary conditional GWAS demonstrated our top associations as robust against linkage with rs334 ( Table 2 , S1 Table ) ., Novel associations of note also include SNPs within the KLHL3-MYOT region ( 13 SNPs , Min P = 5 . 85x10-7 , Additive OR = 0 . 590 ) , the IL23R-IL12RB2 region ( 7 SNPs , Min P = 7 . 98x10-7 , Recessive OR = 0 . 479 ) , FAM155A ( 6 SNPs , Min P = 6 . 24x10-7 , Additive OR = 0 . 207 ) , and CSMD1 ( 5 SNPs , Min P = 7 . 98x10-7 , Additive OR = 4 . 795 ) ., ( Fig 2 ) ., Three significant SNPs are also found within both LINC00943/4 and lincRNA AF146191 . 4–004 ., Lone SNP associations are present within proximity of TREML4 ( P = 1 . 21x10-7 , Heterozygous OR = 4 . 087 ) , zinc finger-containing ZNF536 ( P = 8 . 69x10-7 , Recessive OR = 0 . 507 ) , C4orf17 ( P = 3 . 75x10-7 , Recessive OR = 0 . 289 ) , and near LINC00670 ( P = 2 . 15x10-7 , Additive OR = 3 . 867 ) ., And finally , three intergenic regions display clusters of significance , most notably a region within chromosome 5 ( 43 , 892 , 232–43 , 964 , 366bp; Min P = 2 . 17x10-7 , Heterozygous OR = 0 . 354 ) , as well as regions within chromosome 7 and 11 ., As expected , allele frequencies of the putative polymorphisms within the Trios dataset are generally equivalent to frequencies in our case and control groups , whilst there were some differences from the 1000 Genomes populations , including within the HBB locus ( Table 2 ) ., Using the Trios dataset , we sought to identify structural variants that could confound the association analysis or be putative hits ., We identified no structural variants within HBB , IL-12RBR2 or LINC00943/4 , one deletion ( 2 , 904bp ) within IL23R , and 152 deletions within KLHL3 ( 63 distinct variants , all singletons except for one 1 , 325bp deletion in 91 individuals ) ( S2 Table ) ., None of the common variants are in linkage disequilibrium with the putative GWAS hits , and eight putative regions had structural variants in the Tanzanian trios , but were absent in the 1000 Genomes populations ( S2 Table ) ., Subtype specific association analyses were undertaken for those SNPs found to be significantly associated with severe malaria in the primary GWAS ( Table 2 ) ., The majority of significant associations are with the hyperlactataemia subtype , a phenotype that includes 57 . 0% of cases , with variants within FAM155A , and the HBB and KLHL3/MYOT regions exhibiting associations exceeding our 1x10-6 significance threshold ., In contrast , variants within IL-23R , IL-12RB2 , CSMD1 , ZNF536 and TREML4 appear to be most significantly associated with severe malarial anaemia , who comprised 49 . 2% of cases ., Candidate SNPs identified in previous studies , with the same individuals , were included to provide appropriate context for novel findings ., ABO blood group , USP38 , FREM3 and alpha-thalassemia have previously been putatively associated with severe malaria in a Tanzanian population 3 , 5 , but these associations are no longer statistically significant ( P>10−4 ) at a more stringent GWAS significance threshold ( S3 Table ) ., We also performed targeted imputation of HLA haplotypes within the MHC , finding the most significant SNP to be rs1264362 , which demonstrated a marginal association with hyperlactatemia ( additive model P = 2 . 33x10-5 ) ., For the analysis of structural variation within candidate regions in the Trios dataset , we identified 28 distinct deletions within FREM3 , of which all but one are present in only one individual , and six distinct deletions in GYPB , for which copy number variation has previously been identified 6 ., Nine distinct variants were identified in ABO , including six duplications , one deletion , one insertion and one inversion ., All such ABO variants are present in single individuals , though 18 individuals have a 23bp insertion ., In contrast to a diversity of structural variation present within HLA and the wider MHC region , minor frequency variants were identified in ATP2B4 ( 25 deletions across 25 samples ) , MARVELD3 ( five deletions across five samples ) , HBA2 ( 3 deletions across three samples ) , and HBA1 ( one sample with one deletion ) ., No structural variants were found in HBB or USP38 ( S2 Table ) ., We imputed structural variants within the wider region of human glycophorin genes ( gypA , gypB , gypE ) on chromosome four , using 55 distinct large polymorphisms identified in 59 individuals within our Trios dataset ( S2 Table ) ., The glycophorin region is structurally highly diverse , and specific individual variants are of low frequency ( mean frequency: Case Control dataset = 0 . 098 , Trios dataset = 0 . 022 ) , consistent with observations in other African populations 7 ., Whilst these large variants could be potentially protective against severe malaria , we identified no significant associations looking at each individually ( P ≥ 0 . 301 ) ., Grouping these variants into forms based on genomic location and function may enhance signals within this region , but could also introduce experimenter bias ., Further , there exists a multitude of potential variant combinations analysis of which , without specific hypotheses , could risk so-called ‘P hacking’ ., A full and in-depth analysis of this region is required but beyond the scope of this study ., Two approaches were applied to identify regions under recent positive selection within the Tanzanian GWAS population as a whole ( Integrated Haplotype Score , iHS ) 13 , or between the cases and controls ( Cross-Population Extended Haplotype Homozygosity , XP-EHH ) 14 ., A common genome-wide absolute score threshold of 4 ( equivalent to P = 6 . 3x10-5 ) was established for both approaches ., At this threshold iHS identified 244 loci , 116 ( 47 . 5% ) within chromosome 6 ., Ninety-four of these significant signals are within the MHC , with three loci within HLA-DOA having an absolute score greater than 6 ., Other MHC genes with significant signals include the immunophilin FKBP5 ( 35732-37931bp , 9 SNPs , iHS: 4 . 00–4 . 84 ) , SAMD3 ( 12490-13053bp , 6 SNPs , iHS: 4 . 00–4 . 68 ) and the exocytosis regulator RIMS1 ( 72805811-72828559bp , 3 SNPs , iHS: 4 . 17–4 . 62 ) ( S4 Table ) ., Most notably , two regions within chromosome 17 ( 3496105-3689132bp , 6 SNPs , 8 genes including integrin ITGAE ) and chromosome 19 ( 38743962-38900106bp , 14 SNPs , 10 genes including two transmembrane channels ) represent regions with a high density of selection signals , akin to those within the MHC ., Further signals of note include the transcription factor ZFHX3 ( ATBF1 ) ( chr . 16 , 16326-73133bp , 3 SNPs , iHS: 4 . 27–4 . 91 ) , ABHD5 ( chr . 3 , 43794949bp , 1 SNP , iHS: 4 . 75 ) , DUSP19 & NUP35 ( chr . 2 , 99180–18528 , 3 SNPs , iHS: 4 . 30–4 . 66 ) , surface tyrosine-kinase receptor ERBB4 ( chr . 2: iHS: 4 . 31–4 . 54 ) , transcription-associated RORC ( chr . 1 , 151792842-151817543bp , 2 SNPs , iHS: 4 . 31–4 . 66 ) ., No structural variation was identified in ABHD5 or DUSP19 , whilst variants were present but rare for the remaining iHS hits ( S2 Table ) ., In total , two deletions were identified in RORC , three in ZFHX3 , NUP35 , and ITGAE , one of which is an 86bp deletion found in seven individuals , and 14 deletions and a 31bp insertion in RIMS1 ., Particularly variable are ERBB4 and FKBP5 for which we identify 49 and 75 distinct variants respectively ., ERBB4 consists of 44 deletions , four insertions , and one inversion , whilst FKBP5 consists of 70 deletions , three duplications , one insertion , and one inversion ., The between case-control XP-EHH approach identifies 10 significant SNPs across six genetic regions ( S5 Table ) ., Relative selection for the control population lies within three regions , including POM121L12 ( XP-EHH: 4 . 33 ) , SYNJ2BP , ADAM21 , ADAM20 ( XP-EHH: 4 . 12 to 5 . 57 ) and ERG ( XP-EHH: 4 . 04 ) , whilst three regions are under relative selection in the case population , including MCUR1 ( XP-EHH: -4 . 26 ) , GCLC ( XP-EHH: -4 . 69 ) and the MHC ( XP-EHH: -4 . 02 ) ( S5 Table ) ., We identify no structural variants within POM121L12 , ADAM21 , or ADAM20 , but a singleton 75bp deletion in SYNJ2BP , two deletions within MCUR1 , three deletions in GCLC , and 20 distinct deletions within ERG , of which 8 individuals share a 106bp deletion and 6 share a 325bp deletion ( S2 Table ) ., As expected , the most significant SNP association is the heterozygous protective rs334 effect ( P = 2 . 61x10-13 ) , with thirty-nine further SNPs within HBB also being significantly associated with resistance to severe disease ., SNPs in other candidate genes , including FREM3 , GYPA , GYPB and USP38 3 , did not exceed a significance threshold of 1x10-6 , and their p-values were different ( greater ) to those previously published because of the use of the more conservative EMMAX mixed model regression 12 ., Marginal evidence for a role of HLA association with severe malaria was also identified , and is broadly consistent with previous work in a West African population that demonstrated that carriers of HLA Class I Bw53 and HLA class II DRB1*1302-DQB1*0501 were protected against severe malaria 15 ., Note that our targeted imputation of HLA utilised a Caucasian reference panel and may therefore overlook further true associations within the HLA locus ., Further , we identified signals of positive selection within the MHC region , this being consistent with malaria as a driver of MHC polymorphism in the human population 16 , 17 ., Of the novel SNP associations identified here , two of the top candidates are located between the interleukin receptors IL-23R and IL-12RB2 , a region that has been identified in GWAS of other inflammatory and immune-linked diseases 18 ., IL-12 and IL-23 are related pro-inflammatory cytokines that share both the p40 subunit and the IL-12Rβ1 receptor subunit ., IL-12 signals through a receptor comprising IL-12Rβ1 and IL-12Rβ2 and is a potent inducer of IFN-γ which mediates both clearance of infection and immunopathology in infections with Plasmodium parasites ., IL-23 signalling ( through its receptor , comprising IL-12Rβ1 and IL-23R ) promotes transcription of RORC which encodes RORγ , a transcription factor involved in generation of IL-17 ., RORC was found to be under recent positive selection in our analysis , further supporting the importance of the pathway ., Decreased IL-12 levels have been associated with progression from uncomplicated malaria to severe disease , specifically an increased risk of severe malarial anaemia in children 19 , 20 ., Variants in IL-12B have been linked to P . falciparum parasite density and associated with protection against cerebral malaria in children whilst , variants in the related IL-12A and IL-12RB1 loci have been associated with protection against severe malarial anaemia among children in western Kenya 19 ., Conversely , the IL-23/IL-17 immune pathway has been implicated in the development of inflammatory reactions in children that develop severe malarial anaemia 21 , in multi-organ dysfunction and acute renal failure in adult P . falciparum cases from India 22 and with the risk of cerebral malaria in Africa 23 ., IL-23R haplotypes have also been associated with increased susceptibility to severe malarial anaemia in Kenya 24 ., Three significantly associated SNPs are present within LINC00944 , with one being 80bp from a known CTCF binding site 25 ., Structurally , although the LINC00943/4 region is a known deletion site 25 , we identified no such deletions within the region in our ‘Trios’ dataset ., Broader functionality of this long intergenic non-coding RNA is unclear , given limited experimental characterisation , making it difficult to determine a role for these SNP variants ., A strong association peak was also identified within KLHL3 , kelch-like protein 3 , being a region known to contain an enhancer and various deletions 25 ., Correspondingly , we identify 152 such deletions within our Trios reference panel , of which 62 distinct variants are present in only one individual and one 1 , 325bp deletion is present in 91 individuals ., This frequent deletion is located within an open chromatin-containing region between 137 , 022 , 562 and 137 , 023 , 887bp ., Mutations of KLHL3 have previously been linked with hypertension and metabolic acidosis 26 suggesting that these novel SNP associations and deletions may prime individuals to have a greater risk for severe malarial acidosis ( hyperlactataemia ) ., A number of the most significantly associated SNPs are present as lone , or paired , associations rather than “stacks” ., This includes SNPs within or very near to TREML4 and ZNF536 ., Whilst this may demonstrate false positive outliers , the existence of these SNPs and their minor frequencies are confirmed in our Trios reference panel ., The broad picture of whole population iHS selection is unsurprising , with the MHC region demonstrating the most striking evidence for recent selective sweeps ., Our results are also consistent with a number of previously identified iHS signals , such as those for loci containing the alcohol dehydrogenase ADH7 , cadherin PCDH15 , synaptotagmin SYT1 , the nociception receptor TRPV1 , and the transmembrane protein SPINT2 27 ., It should also be emphasised that our iHS signals reflect selection within our case-control dataset and therefore oversample , relative to a general Tanzanian population , for those signals associated with susceptibility to severe malaria ., Recent differential selection between the case and control groups , as determined by XP-EHH , identified very few significant signals ., There is likely to be limited differential selection between subsets of a closely related population , despite malaria infection being a strong selector ., We identified the MHC , GCLC , MCUR1 , POM121L12 and the SYNJ2BP-ADAM21-ADAM20 region ., The strongest of these signals covers ADAM20 and ADAM21 , both members of a larger family of disintegrins and metalloproteinases that are believed to be exclusively expressed in the testis 28; this association might simply reflect differences in the gender ratio between the cases and controls , for which XP-EHH does not control ., Selection for this region is more likely driven by a variant of SYNJ2BP , a Synaptojanin-2 binding protein with potential roles in membrane trafficking and signalling 29 ., Our previous work has demonstrated that novel associations with potentially significant roles in malaria susceptibility remain to be uncovered 3 , and here we show that an integrated approach that identifies signals of association , selection and structural variation can empower such studies ., However , with only 914 individuals in this study , sample size is a notable limitation for interpretation ., Initial approaches to account for this were pursued through robust contextualisation of novel variants within the secondary ‘Trios’ dataset , and the wider 1000 Genomes project ., More generally , it remains vital that further validation , through larger scale studies , be undertaken to better characterise the SNP and structural variants uncovered ., This is particularly true for structural variation such as within KLHL3 , which may impact gene expression and would therefore benefit from incorporation of transcriptomic data ., Distributions of human genetic variants with putative roles in P . falciparum malaria susceptibility are diverse ., The HbS sickle cell polymorphism is present across most regions of sub-Saharan Africa but is known to have arisen multiple times leading to a number of distinct haplotypic backgrounds 30 ., Similarly , other variants , such as G6PD polymorphism and glycophorin structural variants vary both in frequency across populations and in their direction of association , leading in some cases to allelic heterogeneity that may be subtype specific ., Many protective variants identified within our study , such as IL-23R and KLHL3 , were found at similar frequencies within the ‘Trios’ dataset but differed from the global 1000 Genomes panel , and may therefore represent examples of Tanzanian- or regional-specific associations ., Such variants are informative to our understanding of human-parasite interactions , yet risk being overlooked in inadequately designed studies ., Ultimately , human GWAS in parallel with whole genome sequencing of host and parasites in large study populations across Africa will be crucial to unravelling host genetic and parasite interactions that could lead to novel malaria control measures such as vaccines ., All DNA samples were collected and genotyped following signed and informed written consent from a parent or guardian ., Ethics approval for all procedures was obtained from both LSHTM ( #2087 ) and the Tanzanian National Institute of Medical Research ( NIMR/HQ/R . 8a/Vol . IX/392 ) ., All participants were from the Tanga region of North-Eastern Tanzania , as described previously 3 ., Briefly , severe malaria cases ( n = 449 ) were recruited in the Teule district hospital and surrounding villages in Muheza district , Tanga region , Tanzania between June 2006 and May 2007 ., The controls ( n = 465 ) were recruited , matched on ward of residence , ethnicity and age ( given in months ) , during August 2008 from individuals without a recorded history of severe malaria 3 ., Four severe malaria subtypes were identified within case individuals including hyperlactatemia ( Blood lactate > 5 mmol/L , n = 256 ) , severe malarial anaemia ( Hemocue Hb < 5g/dL , n = 221 ) , respiratory distress ( n = 124 ) and cerebral malaria ( Blantyre coma score <5 , n = 120 ) ( Table 1 ) ., Parasite infection was initially assessed by rapid diagnostic test ( HRP-2 –Parascreen Pan/Pf ) and confirmed by double read Geimsa-stained thick blood films ., A further 247 anonymously sampled individuals , consisting of 78 healthy parent and child trios ( 156 parents , 78 children , 13 singletons; 80 ( 32 . 4% ) Chagga , 77 ( 31 . 2% ) Pare , 90 ( 36 . 4% ) Wasambaa ) , were collected between 2007 and 2008 ., These individuals are those that had no current illness or no history of malaria ., The samples were collected from highland , medium and lowland villages near the Kilimanjaro , Pare and West Usambara mountains in the Tanga region of Tanzania ., This is a region that experiences low to medium to high levels of malaria transmission ., This dataset was used to confirm allele frequencies and identify candidate region structural variation within the general Tanzanian population , as well as to impute variants onto the case-control set ., DNA was extracted from processed blood samples , as described previously 3 , 5 ., The DNA was genotyped on the Illumina Omni 2 . 5 million SNP chip and SNP genotypes called by the MalariaGEN Resource Centre at the Sanger Institute and the Wellcome Trust Centre for Human Genetics , using previously described methods 6 , 7 ., These data were complemented by Iplex genotyping assays that included 180 single nucleotide polymorphisms ( SNP ) across 50 loci on the same individuals 3 ., 107 additional candidate SNPs , including the HbS SNP rs334 , were included from previous candidate genotyping of the same case-control individuals; their collection having been described previously 3 ., DNA for the individuals in the Trio dataset ( n = 247 ) was sequenced using Illumina HiSeq2500 technology at the Sanger Institute , and aligned to the GRCh37 build of the human genome 7 ., The minimum genome-wide coverage across the samples was 22-fold ., SNPs were called from the alignments using the standard samtools-bcftools pipeline 31 ., This process led to 2 , 788 , 671 high quality SNPs with quality scores of at least 30 ( 1 error per 1000bp ) and perfect trio-consistent genotype calls ., Haplotypes were phased from genotypes using SHAPEIT ( www . shapeit . fr; default settings ) ., Structural variants , including duplications , deletions , insertions and inversions , were identified within the secondary ‘Trios’ dataset for candidate regions using DELLY version v0 . 7 . 3 32 ., This software was applied using default settings , and its use in pipelines has been shown to reliably uncover structural variants from the 1000 Genomes Project , and validation experiments of randomly selected deletion loci show a high specificity 32 ., Structural variants greater than 100 , 000 basepairs in length were removed to conservatively exclude false positives ., To increase genome-wide SNP resolution , our initial case-control dataset was imputed using a combined reference panel of the Phase 3 1000 Genomes project 10 and children within the trio dataset , using Beagle 4 . 1 11 ., This allowed for the inclusion of 13 . 5 million additional high quality SNPs , to a total of 15 . 2 million SNPs ., A total of 621 , 019 SNPs were removed from the pre-imputation dataset due to evidence of:, ( i ) deviations in genotypic frequencies from Hardy-Weinberg equilibrium ( HWE ) as assessed using a chi-square test ( >0 . 0001 ) ;, ( ii ) high genotype call missingness ( >10% ) ; or, ( iii ) low minor allele frequency ( <0 . 01 ) ., 51 individuals were removed due to:, ( i ) genotypic missingness ( >0 . 1 ) ;, ( ii ) abnormal PCA clustering or, ( iii ) missing malaria phenotype data ., 849 , 134 strand flips were identified with snpflip , with these being corrected pre-imputation with Plink v1 . 07 ., Raw hybridisation plots were manually verified for all top non-imputed GWAS associations , excluding rs334 for which the data was unavailable ., Linkage disequilibrium between SNPs in close genomic distance was calculated using Plink v1 . 07 33 ., Targeted imputation was performed for HLA haplotypes within the major histocompatibility complex using 9 , 785 high quality SNPs within the region; for this we utilised SNP2HLA software ( version 1 . 0 . 3 ) and the default Caucasian reference panel 34 ., Association tests for this targeted analysis were performed through the pipeline described above ., Similarly , 1 , 202 structural variants ( 698 deletions , 311 duplications , 19 insertions , 174 inversions ) within chromosome four were imputed into our primary ‘case-control’ dataset using IMPUTE2 with default parameters , akin to standard SNP imputation ., This approach allowed us to perform association analysis on those structural variants using EMMAX mixed model regression 12 ., Trio parental SNP data was also used to provide additional context for our case-control SNPs within the wider Tanzanian population , as seen in S1 Table ., Case-Control association analysis of SNPs was undertaken with EMMAX mixed model regression 12 , controlling for age as a fixed effect and relatedness ( represented by a kinship matrix ) as a random effect ( to reduce associations relating to familial clustering ) ., Several genotypic models were implemented separately , including additive , heterozygous , dominant and recessive ., Minimum P values from each model were utilised for top hit identification ., Odds ratios were estimated with Plink v1 . 07 33 ., Our complementary conditional GWAS shared the pipeline for the main GWAS , but with HbS status added as an additional covariate ., To evaluate the statistical potential of our GWAS study , we performed a retrospective power calculation ( using http://zzz . bwh . harvard . edu/gpc/cc2 . html ) ., A study of 460 cases and 460 controls can detect odds ratios of at least 2 for a high risk allele minor allele frequency of 5% with a statistical power of 85% ( and type I error of 10−6 ) ., A significance threshold of 10−6 was established using a permutation approach 35 ., In particular , both the case-control status of the chromosomes were randomly permuted 10 , 000 times ., From each of the 10 , 000 random experiments , we determined the maximum chi-square statistics ( across the four genotypic tests ) over all SNPs genotyped ., We ordered these statistics and then calculated the 95 percentile ., This was the estimate of the 0 . 05 significance level for the experiment performed , assuming inference is taken with respect to maximum chi-square statistic observed over all genotyped SNPs , and accounts for the linkage disequilibrium between SNPs and correlation between the results from applying the 4 genotypic tests ., Whole population Integrated Haplotype Scores ( iHS ) 13 and case-control Cross-Population Extended Haplotype Homozygosity ( XP-EHH ) 14 were calculated and normalised over the whole genome using selscan and norm 36 ., Core SNPs with a minor allele frequency below 0 . 01 were excluded from this analysis ., In this context , high iHS values indicate a whole population selection signal whilst positive XP-EHH values indicate relative selection within the control population and negative XP-EHH values indicate relative selection within the case population ., We looked for structural variants in regions with SNP-based signals of positive selection , as it possible that selection may actually be driven by structural variants ( see 37 for an example ) . | Introduction, Results, Discussion, Methods | Significant selection pressure has been exerted on the genomes of human populations exposed to Plasmodium falciparum infection , resulting in the acquisition of mechanisms of resistance against severe malarial disease ., Many host genetic factors , including sickle cell trait , have been associated with reduced risk of developing severe malaria , but do not account for all of the observed phenotypic variation ., Identification of novel inherited risk factors relies upon high-resolution genome-wide association studies ( GWAS ) ., We present findings of a GWAS of severe malaria performed in a Tanzanian population ( n = 914 , 15 . 2 million SNPs ) ., Beyond the expected association with the sickle cell HbS variant , we identify protective associations within two interleukin receptors ( IL-23R and IL-12RBR2 ) and the kelch-like protein KLHL3 ( all P<10−6 ) , as well as near significant effects for Major Histocompatibility Complex ( MHC ) haplotypes ., Complementary analyses , based on detecting extended haplotype homozygosity , identified SYNJ2BP , GCLC and MHC as potential loci under recent positive selection ., Through whole genome sequencing of an independent Tanzanian cohort ( parent-child trios n = 247 ) , we confirm the allele frequencies of common polymorphisms underlying associations and selection , as well as the presence of multiple structural variants that could be in linkage with these SNPs ., Imputation of structural variants in a region encompassing the glycophorin genes on chromosome 4 , led to the characterisation of more than 50 rare variants , and individually no strong evidence of associations with severe malaria in our primary dataset ( P>0 . 3 ) ., Our approach demonstrates the potential of a joint genotyping-sequencing strategy to identify as-yet unknown susceptibility loci in an African population with well-characterised malaria phenotypes ., The regions encompassing these loci are potential targets for the design of much needed interventions for preventing or treating malarial disease . | Malaria , caused by Plasmodium falciparum parasites , is a major cause of mortality and morbidity in endemic countries of sub-Saharan Africa , including Tanzania ., Some gene mutations in the human genome , including sickle cell trait , have been associated with reduced risk of developing severe malaria , and have increased in frequency through natural selection over generations ., However , new genetic mutations remain to be discovered , and recent advances in human genome research technologies such as genome-wide association studies ( GWAS ) and fine-scale molecular genotyping tools , are facilitating their identification ., Here , we present findings of a GWAS of severe malaria performed in a well characterised Tanzanian population ( n = 914 ) ., We confirm the expected association with the sickle cell trait , but also identify new gene targets in immunological pathways , some under natural selection ., Our approach demonstrates the potential of using GWAS to identify as-yet unknown susceptibility genes in endemic populations with well-characterised malaria phenotypes ., The genetic mutations are likely to form potential targets for the design of much needed interventions for preventing or treating malarial disease . | genome-wide association studies, medicine and health sciences, immunology, tropical diseases, parasitic diseases, parasitic protozoans, anemia, genetic mapping, clinical medicine, protozoans, genome analysis, molecular genetics, major histocompatibility complex, malarial parasites, cerebral malaria, molecular biology, hematology, haplotypes, eukaryota, clinical immunology, heredity, genetics, biology and life sciences, malaria, genomics, computational biology, organisms, human genetics | null |
journal.pgen.1000033 | 2,008 | New Insights into Human Nondisjunction of Chromosome 21 in Oocytes | The overwhelming majority of trisomy 21 , or Down syndrome , is caused by the failure of chromosomes to separate properly during meiosis , also known as chromosome nondisjunction ., As nondisjunction is the leading cause of pregnancy loss , mental retardation and birth defects , it is imperative that we understand the biology underlying this phenomenon ., Characteristics of chromosome 21 nondisjunction are typical of many of the other human autosomes ., That is , the overwhelming majority are due to errors during oogenesis: at least 90% of cases of chromosome 21 nondisjunction are due to maternal meiotic errors 1 , 2 ., In addition , among these maternal errors , the majority occur during meiosis I ( MI ) 3 , 4 ., It has been well established that increased maternal age , the most significant risk factor for nondisjunction , is associated specifically with errors occurring during oogenesis ., Interestingly , for chromosome 21 nondisjunction , advanced maternal age is associated with both maternal MI and meiosis II ( MII ) errors 5 ., The timing of meiosis in the human female suggests risk factors that may be involved in chromosome nondisjunction ., Meiosis is initiated at about 11–12 weeks of gestation and , after pairing , synapsis and recombination , arrests in prophase I until just prior to ovulation ., At that time , the oocyte completes MI and progresses to metaphase II where it remains until it is fertilized and the meiotic process is completed ., Thus , homologous chromosomes are arrested in prophase I for 10 to 50 years ., In contrast , spermatogenesis in the human male begins at puberty and cells entering meiosis move from one stage to the other with no delay ., This extended state of arrest in oocyte formation is hypothesized to be associated with the increased prevalence of maternal nondisjunction ., Chiasmata function to stabilize paired homologous chromosomes ( tetrads ) during MI along with sister chromatid and centromere cohesion ., They also help to properly orient homologous chromosomes on the meiotic spindle 5 ., A proportion of nondisjunction is associated with failure of homologues to pair or to recombine , leading to an increased risk for homologue malsegregation during MI 6–9 ., In our previous work 10 , it was estimated that 45% of maternal MI cases of trisomy 21 did not have an exchange along chromosome 21 ., We also found that the location of the exchange was associated with nondisjunction: a single exchange near the telomere of 21q increased the risk of maternal MI nondisjunction and the presence of an exchange near the centromere increased the risk for so called MII nondisjunction ., This association of a MI event ( i . e . , recombination ) with a MII error in chromosome segregation led us to suggest that MII nondisjoining errors are initiated during MI ., To represent this finding , we will refer to MII errors in quotes ., Most recently , we have explored the relationship between maternal age and recombination to gain further insight into potential mechanisms of abnormal chromosome segregation 11 ., We compared the frequency and the location of exchanges along 21q between women ( or “oocytes” ) of various maternal ages who had an infant with Down syndrome due to a maternal MI error ., While there was no significant association between maternal age and the overall frequency of exchange , the placement of meiotic exchange differed significantly by maternal age ., In particular , single telomeric recombinant events were present in the highest proportion among the youngest age group ( 80% ) , while the proportions in the oldest group of women with nondisjoined chromosomes 21 and in women with normally disjoining meiotic events were almost equal ( 14% and 10% , respectively ) ., We speculated that for young women then , the most frequent risk factor for MI nondisjunction is the presence of a telomeric exchange ., As a woman ages , her meiotic machinery is exposed to an accumulation of age-related insults , becoming less efficient/more error-prone ., The susceptible telomeric exchange pattern still increases susceptibility to nondisjunction , but now even homologous chromosomes with optimally placed exchanges are at risk ., Over time , the proportion of nondisjunction due to normal exchange configurations increases as age-dependent risk factors exert their influence ., As a result , the most prevalent exchange profile of nondisjoined oocytes shifts from susceptible to non-susceptible patterns with increasing age of the oocyte ., As mentioned above , our studies also identified an association between the presence of a meiotic exchange within the pericentromeric region of 21q and “MII” nondisjunction 10 , but further studies were not possible due to limited sample size ., We have now increased our sample size and , for the first time , have been able to investigate the relationship of exchange patterns stratified by maternal age for maternal “MII” cases of trisomy 21 ., This increase in sample size has also allowed us to refine our analysis of recombination in maternal MI cases by maternal age ., These analyses have provided further insight into the complex pathways leading to nondisjunction among oocytes ., Among normal disjoining maternal meiotic events , exchanges most often occur in the center of 21q 11 ., This observation suggests that the presence of a single medially placed exchange is important for normal segregation of homologous chromosomes 21 ., This pattern is in striking contrast to the chromosomes 21 that have undergone maternal MI or “MII” nondisjunction , where either no exchange occurs or single exchanges occur at the very ends of 21q 8 , 10 ., In order to better understand the factors that play a role in these recombination-related disjoined events , we have examined both the number and location of recombination along nondisjoined chromosomes 21 stratified by maternal age ., In these analyses , maternal age served as a proxy for the age of the oocyte ., First , among normally disjoining chromosomes 21 in oocytes , there was no obvious association between maternal age and the frequency of exchange or the location of exchange along chromosome 21 ., We did not expect to observe a maternal age association , as our comparison group , taken from the CEPH families , was relatively small compared to Kong et al . 9 , the only study that has noted such an association ., In that study , it took over 14 , 000 maternal meiotic events in order to identify that the frequency of exchanges increased with maternal age: an additional two recombinants genome-wide were estimated over a 25 year age span ., Thus , the magnitude of the observed association is not on the same scale as that observed for nondisjoined meiotic events ., Irrespective , we still must be cautious with our results and emphasize that the sample sizes of meiotic events , particularly those in the older age groups were small ( Table S1 ) and thus limited our ability to detect maternal age associations with recombination ., Whereas there was no obvious maternal age association with recombination patterns among normally disjoining chromosomes 21 , there was a significant one among maternal MI and “MII” errors ., One set of observations provides evidence for specific recombination patterns being the proximal cause of nondisjunction , while the others suggest an interaction between specific recombination patterns and maternal age-related risk factors ., Figure 1 provides an overall summary of our findings related to the spatial distribution of exchanges for MI and “MII” nondisjunction events ( using the data from Table 2 ) ., In Figure 2 , we interpret these findings , as well as those associated with the frequency of exchanges ( Table, 1 ) within the context of the overall rate of trisomy 21 among women of the three age groups ( see Materials and Methods for calculations ) ., In this figure , the overall rate of trisomy 21 among births by maternal age group is represented by the height of each bar and is estimated from Hecht and Hook 14 ., Within each bar , the proportion of those rates that are estimated to have a specific origin and recombination pattern is denoted by color ., Here , we have focused on meiosis occurring in the aging oocyte ., Several meiotic proteins that function to promote proper chromosome segregation have been shown to degrade with increasing age 15 , 16 ., This degradation is assumed to lead to increased frequency of nondisjunction; thus , more maternal-age related risk factors for nondisjunction exist among older women compared to younger women ., In the analyses presented here , we have compared the pre-disposing recombination patterns among the oocytes with nondisjoined events by maternal age ( Figure 1 ) ., Our expectation is that some recombination patterns will lead to susceptibility irrespective of other maternal age factors and these will predominate the youngest age group , or that group with no other risk factors ., We found that single telomeric exchanges follow this pattern ( Figure 2 , “MI: E1 int 6” ) , as reported previously 11 ., This type of error represents less than 8% of each maternal age group ., This same risk factor has been established in model organisms as well 17–19 ., Most likely , susceptibility is related to the minimal amount of the sister chromatid cohesion complex remaining distal to the exchange event 20 ., Specifically , when the exchange is too far from the kinetechore , this could prevent the biorientation of homologues on the meiotic spindle 18 , 21–23 ., Alternatively , the integrity of the chiasma may be compromised when a minimal amount of cohesin remains to hold homologues together ., Thus , bivalents may act as a pair of functional univalents during MI , as has been observed in human oocytes 24 , 25 ., The results related to lack of exchange are intriguing , although difficult to interpret at this time ., We did find that the proportion of E0s was the highest among the youngest group compared with the other two age groups , indicating a maternal-age independent mechanism ., However , the proportions did not decrease linearly with age ( Table 1 ) ., Conservatively , we can state that E0s lead to susceptibility irrespective of the age of the oocyte ., However , the non-significant increase in E0 in the older age group causes us to speculate further ., As noted in Figure 2 ( “MI: E0” ) , the lack of a linear decrease by age group suggests that a greater proportion of older oocytes at risk for trisomy 21 will have E0 tetrads compared with the other two age groups ., Perhaps these results provide preliminary evidence for a secondary mechanism that is age-dependent ., In model systems , there are known mutations that lead to increased nondisjunction of E0s ., For example , Drosophila with mutations in the gene nod ( no distributive disjunction ) , show increased nondisjunction of non-exchange chromosomes 26 ., This observation was the first to suggest a mechanism that functions to ensure the proper segregation of non-exchange homologues ., Studies in yeast also provide evidence for such a mechanism 27 ., Interestingly , proteins in humans that may have a similar function to those that play a role in the proper segregation of non-exchange homologues in yeast have been shown to be down regulated with increasing ovarian age 15 , 16 ., Thus , the age-dependent down-regulation of these essential proteins , or others , may lead to the decreased ability to properly segregate non-exchange chromosomes in aging oocytes ., However , this is only speculation at this point ., More data are needed to determine significance of our preliminary finding ., Interestingly , the analysis of the normally disjoining meiotic events from the CEPH data indicates a large proportion of E0s , 20% ., These data are based on genotyping a high density of chromosome 21-specific SNPs among 152 maternal meiotic events 28 ., Other studies have used the CEPH families and have obtained similar frequencies of observed recombinants and estimates of E0 frequencies 29 , 30 ., These data suggest a higher frequency of E0s compared with other studies that have used techniques that examine tetrads more directly , such as chiasma counts or MLH1 counts ., For example , Tease et al . 31 identified three E0 chromosome 21 bivalents out of a total of 86 counted ., However , all 86 oocytes analyzed came from only one ovary ., As variation in recombination rates among women is well established 28 , 32 , we need to be careful in drawing conclusions about the difference in estimates of E0 using MLH1 counts versus linkage studies ., Nevertheless , future studies are required to determine if the frequency of E0s is significantly different from zero for chromosomes 21 in oocytes ( e . g . , using MLH1 counts ) and in transmissions to births ( e . g . , linkage studies ) , each representing a different time point in oocyte development ., These studies will complement those among nondisjoined events to determine if a distributive pairing system similar to those in model systems exists in humans ., The other established susceptibility pattern that is associated with an increased risk for “MII” nondisjunction is the presence of a single exchange within the most proximal 5 . 2 Mb of 21q ., When we compared such events among age groups , we observed an enrichment of pericentromeric exchanges in the oldest age group of “MII” nondisjoined chromosomes 21 as summarized in Figure 1 ., This leads to a greater proportion of trisomy 21 cases among older women being related to pericentromeric exchanges ( Figure 2 , “MII: E1 int 1” ) ., This pattern can be explained in two different ways:, 1 ) a pericentromeric exchange sets up a suboptimal confirmation that exacerbates the effect of maternal age-related risk factors or, 2 ) a pericentromeric exchange protects the bivalent from maternal-age related risk factors allowing the proper segregation of homologues , but not sister chromatids ., An example of the former would be that a pericentromeric exchange compromises proteins involved in centromeric cohesion , exacerbating the normal degradation of this important complex with age ., Shugoshin , a protein important in protecting centromere cohesin during MI , would be an obvious target ., For example , in yeast cells that were shugoshin deficient , Marston et al . 33showed that homologous chromosomes segregated to opposite poles in MI , but sister chromatids prematurely separated prior to anaphase II and segregated randomly , sometimes leading to MII nondisjunction ., Interestingly , BubR1 , the protein required for the localization of shugoshin to the centromere , has been shown to have decreased expression with increasing maternal age in the human female 15 , 16 ., Perhaps the presence of a pericentromeric exchange exacerbates the degradation of this complex ., Alternatively , a pericentromeric exchange may protect the bivalent from maternal-age related risk factors ., The effect of degradation of centromere or sister chromatid cohesin complexes or of spindle proteins with age of the oocyte may lead to premature sister chromatid separation ., Perhaps a pericentromeric exchange helps to stabilize the compromised tetrad through MI ., This would lead to an enrichment of MII errors among the older oocytes ., Although there is no specific model system that points to this mechanism , findings can be interpreted with this mechanism in mind ., For example , the effects of a hypomorph of bubR1 were examined in female meiosis in Drosophila 34 ., In mutant females , most chiasmate X chromosome failed to segregate properly at MII , most likely due to premature sister chromatid separation in late MI anaphase or MII ., Interestingly , a subtle but repeatable increase in pericentromeric exchanges was identified along such chromosomes ., Lastly , we examined the hypothesis that the number of exchanges may be protective against maternal age-related risk factors ., This was first suggested by Robinson et al . 35 , who found that among maternal MI chromosome 15 nondisjunction errors , the age of the mother was significantly increased among cases with multiple recombinants compared with those having zero or only one observed recombinant ., From this , the authors suggested that cases with multiple recombinants might be more resistant to nondisjunction because of increased stability of the tetrad over time ., Similarly , an analysis of maternal nondisjunction of the X chromosome showed that the mean maternal age of cases with recombination was significantly older than that of cases with no recombination 36 ., This same pattern was observed for trisomy 18 , although the difference was not statistically significant 37 ., For chromosome 21 MI errors , we do not see this pattern ., Among the young , middle and older age groups , the observed data infer 40% , 23% and 33% of tetrads have multiple exchanges among our young , middle and old groups respectively ( Table 1 ) ., Among chromosome 21 “MII” errors , we observe a very different pattern: 78% , 49% and 44% of tetrads have multiple exchanges , respectively ., This pattern is opposite of that expected if multiple exchanges were protective ., Again , we need to be cautious in our interpretation for the following reason ., We have assumed that “MII” cases with no recombination are due to post-zygotic , mitotic events ., As shown in Figure 2 , these appear to be age-independent events ., However , some proportion may be true MII errors with no recombination and we do not have a method to distinguish these alternatives ., We have not discussed our observations related to the placement of multiple recombinants along the nondisjoined chromosomes 21 and the potential effects of altered interference ., This is due to the obvious fact that chromosome 21 is small , leading to only a few meiotic events on which we could derive exchange patterns ., There were approximately 20 meiotic events in each age category of MI and MII errors ., Thus , this type of investigation awaits a larger sample size , or , perhaps , should be based on larger nondisjoined chromosomes ( e . g . , chromosome 15 or the X chromosome ) ., The importance of understanding the causes of nondisjunction and the maternal age effect cannot be over-stated ., Many women are electing to delay childbearing until their mid-thirties or later , the time at which nondisjunction rates dramatically increase ., Irrespective of the exact mechanisms of nondisjunction , our findings indicate that nondisjunction is a complex trait and that there are different risk factors that play a role in age-independent and dependent nondisjunction ., The study design for identification of such environmental and genetic risk factors can be guided by our findings ., Clearly , examination of nondisjunction events stratified by maternal age , type of error and recombination pattern should increase the power to identify important factors that play a role in chromosome mal-segregation ., Families with an infant with full trisomy 21 were recruited through a multisite study of risk factors associated with chromosome nondisjunction 2 , 8 , 10 ., Parents and the infant donated a biological sample ( either blood or buccal ) from which DNA was extracted ., All recruitment sites obtained the necessary Institutional Review Board approvals from their institutions ., Only families in which DNA was available from both parents and the child with trisomy 21 were included in the present analysis ., A subset of families in the current analysis with maternal MI errors were also included in a previous study 6 ., Samples were genotyped for a minimum of 21 short tandem repeat ( STR ) markers specific to the long arm of chromosome 21 ( Figure 3 ) ., The most centromeric STR was D21S369 and the most telomeric was D21S1446 ., Our analysis of the number and location of recombination was restricted to 21q ., The long arm of chromosome 21 was divided into six relatively equal physical intervals with interval 1 comprising the most centromeric region of 21q and interval 6 comprising the most telomeric region ( Figure 3 ) ., The presence of a recombinant event was identified by changes in the status of adjacent informative markers from “reduced” to “nonreduced” ( or vice versa ) ., In most cases , the location of recombination was scored as belonging to one of six distinct intervals along 21q ., When one of the six intervals was uninformative , but markers defining the two flanking intervals were informative , we included the family ., Those with two or more adjacent uninformative intervals were excluded from our analysis ., In some instances , the recombinant event could not be located to one specific interval , but instead to one of two adjacent intervals ( e . g . , interval 1 or interval 2 ) ., The location of such events was treated as occurring at the midpoint of the two intervals ( e . g . , represented as interval 1 . 5 ) in most of our analyses ( see Statistical Analysis below ) ., Our final analysis included a total of 615 maternal MI cases and 253 maternal “MII” cases of trisomy 21 ., In order to determine the location of recombination along 21q in women who exhibited normal segregation of chromosome 21 , the transmission of maternal grandparental SNP genotypes to the maternal offspring was analyzed ., A maternal recombinant event was noted when the sharing of SNPs identical by descent switched from one maternal grandparent to the other ., Our final analysis included 152 informative maternal meioses . | Introduction, Discussion, Materials and Methods | Nondisjunction of chromosome 21 is the leading cause of Down syndrome ., Two risk factors for maternal nondisjunction of chromosome 21 are increased maternal age and altered recombination ., In order to provide further insight on mechanisms underlying nondisjunction , we examined the association between these two well established risk factors for chromosome 21 nondisjunction ., In our approach , short tandem repeat markers along chromosome 21 were genotyped in DNA collected from individuals with free trisomy 21 and their parents ., This information was used to determine the origin of the nondisjunction error and the maternal recombination profile ., We analyzed 615 maternal meiosis I and 253 maternal meiosis II cases stratified by maternal age ., The examination of meiosis II errors , the first of its type , suggests that the presence of a single exchange within the pericentromeric region of 21q interacts with maternal age-related risk factors ., This observation could be explained in two general ways:, 1 ) a pericentromeric exchange initiates or exacerbates the susceptibility to maternal age risk factors or, 2 ) a pericentromeric exchange protects the bivalent against age-related risk factors allowing proper segregation of homologues at meiosis I , but not segregation of sisters at meiosis II ., In contrast , analysis of maternal meiosis I errors indicates that a single telomeric exchange imposes the same risk for nondisjunction , irrespective of the age of the oocyte ., Our results emphasize the fact that human nondisjunction is a multifactorial trait that must be dissected into its component parts to identify specific associated risk factors . | Nondisjunction occurs when chromosomes fail to segregate during meiosis; when this happens , gametes with an abnormal number of chromosomes are produced ., The clinical significance is high: nondisjunction is the leading cause of pregnancy loss and birth defects ., We have studied trisomy 21 using DNA from individuals with Down syndrome and their parents to identify mechanisms underlying nondisjunction ., The results from these studies show that altered patterns of recombination , e . g . , no exchange , a single telomeric exchange and a single pericentromeric exchange , were associated with nondisjunction of chromosome 21 within the oocyte ., In this report , we stratified maternal cases of trisomy 21 by the type of nondisjunction error ( meiosis I or meiosis II ) and by maternal age ( ages <29 , 29–34 and >34 years ) and examined both the number and location of recombination by age group ., Our results suggest that the risk imposed by the absence of exchange or by a single telomeric exchange is the same , irrespective of the age of the oocyte ., In contrast , the risk imposed by a single pericentromeric exchange increases with increasing maternal age ., These findings , put into the context of proteins involved in the meiotic process , have enabled us to further understand mechanisms underlying nondisjunction . | molecular biology/recombination, genetics and genomics/chromosome biology | null |
journal.pbio.1002023 | 2,014 | Coevolution Drives the Emergence of Complex Traits and Promotes Evolvability | Life emerged on Earth some 4 billion years ago and has evolved increasingly complex traits , including intricate biochemical pathways , elaborate developmental networks , and powerful neural architectures 1 , 2 ., However , the processes responsible for promoting this complexity remain poorly understood 1–9 ., Is adaptation by natural selection largely responsible for this complexity and , if so , what is the nature of that selection ?, Or is this apparent trend an artifact that reflects the initial conditions and lower bounds to complexity ?, Given the limitations of historical data for answering these questions , experimental evolution offers an alternative approach to explore these issues and test specific hypotheses ., However , the emergence of complexity in nature is a slow process , one not readily replicated in the laboratory 8 , 10; and without an objective way to measure the complexity of organismal traits 11 , 12 , rhetorical arguments may obscure and delay empirical research on this fundamental problem ., Fortunately , computational approaches have advanced beyond traditional numerical simulations , and it is now possible to test evolutionary hypotheses by running experiments with computer programs that self-replicate , mutate , compete , and evolve 13 ., In one study , Lenski and colleagues 14 used the Avida 15 system to examine the role of selection for intermediate steps along many evolutionary paths to a particularly complex trait , the EQUALS ( EQU ) logic function ., Because Avida is computational , the authors could readily observe changes over thousands of generations; moreover , the complexity of traits could be objectively quantified as the number of building blocks ( in this case , NAND instructions ) required for their execution ., By allowing initially identical populations to evolve in different environments , Lenski and colleagues demonstrated that the most complex traits emerged only when simpler functions were also selectively favored , which promoted the accumulation of the necessary building blocks 14 ., Here we use this system to ask whether coevolution—specifically , parasite-host interactions—can drive complexity to higher levels than would otherwise be achieved ., Several authors , including Dawkins and Krebs 7 and Vermeij 16 , have proposed that coevolutionary “arms races” lead to increased complexity as adaptations and counter-adaptations favor more and more extreme traits 6 ., Indeed , we show that host-parasite coevolution produced substantially more complex host traits than did evolution in the absence of parasites ., Moreover , we show that this complexity arose in the evolving computer programs , in part , by an unexpected process: selection for increased evolvability , which was achieved by genetic mechanisms reminiscent of so-called “contingency loci” that are found in many pathogenic bacteria 17 ., In Avida , both host and parasite organisms are self-replicating programs that must expend CPU cycles to execute instructions in their genomes 18 ., The genetic instruction set includes basic arithmetic and input/output operations as well as operations that allow storage and manipulation of binary numbers in temporary memory via a set of stacks ., Coordinated execution of appropriate sets of instructions allows organisms to obtain resources ( in the case of hosts ) or infect hosts ( in the case of parasites ) and copy their genomes instruction-by-instruction to produce offspring ., The copying process occasionally introduces mutations including point mutations , insertions , and deletions that may affect the progenys phenotype ., As in nature , most mutations are deleterious or neutral , but occasional beneficial mutations improve an organisms ability to acquire resources , infect hosts or resist parasites , or reproduce ., These benefits may enable genotypes to increase in frequency as they displace less fit conspecifics because of their faster acquisition and more efficient use of CPU cycles ., Thus , populations of digital organisms , like their counterparts in nature , typically evolve to better fit their environments 13 ., Figure 1 shows a schematic overview of the relationships between hosts , functions , resources , and parasites in our experiments ., Hosts obtain the resources necessary for their reproduction by performing one or more logic functions , but those functions also make the host vulnerable to infection by a parasite that can perform the same function ., Thus , an infection can occur only if a particular host and parasite share at least one function , although the specific genetic encoding that a host and parasite employ to perform that function rarely , if ever , correspond at the sequence level ., After a successful infection , the parasite acquires 80% of the infected hosts CPU cycles , which the parasite uses to execute and copy its own genome , while imposing a severe cost on the host ., As a consequence , coevolution occurs when hosts and parasites acquire and lose functions ., The experimental configuration allowed for nine different logic functions , which require varying numbers of NAND instructions to be executed with the proper inputs used for each; NAND is the only logic function available in the genetic instruction set ., Although there are many potential measures of functional complexity , the Avida logic environment provides an intuitive metric , as follows ., The minimum number of NAND operations required for each functions performance is known and provides a simple , objective measure of the complexity of that function 14 ., The most complex function , EQU , requires five NAND operations , and the shortest program that can perform EQU requires nearly 20 precisely interacting instructions , although there are many longer programs that also encode EQU 14 ., In the absence of parasites , a previous study found that 23 of 50 populations evolved the ability to perform EQU when the other eight functions were rewarded with additional CPU cycles that increased with their complexity ( i . e . , minimum required NANDs ) , thus allowing essential building blocks to accumulate in the evolving genomes 14 ., Here , we test whether host-parasite coevolution can drive increased complexity without explicitly rewarding building blocks ., To that end , we ran similar experiments except with coevolving parasites in one-half of the replicates and without the progressive reward structure used in the previous work ., Figure 2 shows that coevolution with parasites drove host populations to evolve more complex functions in order to obtain the resources necessary for their replication , without any greater reward for performing the more difficult functions ., Host complexity increased in both the presence ( red ) and absence ( blue ) of parasites , but it did so much faster and reached much higher levels in the coevolution treatment ( p≪0 . 001 , Mann-Whitney U\u200a=\u200a2 , 304 ) ., The effect of parasites on the rise of complexity is exemplified by EQU , the most complex function; the ability to perform EQU evolved in 17/50 host populations that coevolved with parasites , but in none that evolved without parasites ( p≪0 . 001 , Fishers exact test ) ., In a third treatment , parasites were removed at the mid-point of the runs , and the cured host populations ( green ) evolved substantially reduced complexity relative to the coevolution treatment ( p≪0 . 001 , Mann-Whitney U\u200a=\u200a543 . 5 ) , although the cured hosts retained greater complexity than those that never saw the parasites ( p\u200a=\u200a0 . 002 , Mann-Whitney U\u200a=\u200a1 , 703 ) ., The increased complexity relative to the ancestor observed in the absence of parasites ( p≪0 . 001 , Wilcoxon signed-rank W\u200a=\u200a1 , 275 ) accords with a simple model that couples a random walk in complexity with a selective constraint that limits functional degradation; Gould dubbed this model the “drunkards walk , ” alluding to how a patron leaving a pub eventually stumbles to the curb because the pub itself limits backward movement 5 ., In our experiments , all populations started from the same ancestral program that could perform only the simplest function , NOT , and hence they were the least complex programs able to obtain resources and reproduce ., Any less complex genotypes generated by mutation could not reproduce and were thus eliminated ., More complex organisms also arose by mutation; although they obtained no additional resources for performing more complex functions ( and , in fact , might replicate more slowly ) , they nonetheless could reproduce and thereby persist ., Over time , this asymmetrical constraint allowed complexity to increase , albeit slowly and to a limited extent ., This explanation of complexity evolving as a “drunkards walk” does not imply that evolution as a whole operates as a random walk; instead , it only implies that complexity might follow such a pattern ., The coevolutionary process clearly produced greater functional complexity in the hosts ., In broad outline , this effect occurs because parasites constantly select for new host phenotypes and thereby cause host populations to explore adaptive landscapes more broadly than hosts that are evolving alone 19 ., However , it is not obvious why the effect was so large and continued for so long ., Understanding the initial increase in complexity is seemingly straightforward—hosts must evolve some function other than NOT to avoid infection yet still reproduce , and all except one of the other functions have higher complexity than NOT ., But this explanation alone cannot explain even the initial step , because the first new function to arise by mutation was , in the vast majority of cases , the one other function , NAND , that also requires executing only a single NAND instruction ., In fact , the average complexity of the first new function was only 1 . 10 ( 1 . 01–1 . 19 95% confidence interval ) , and the maximum was only 2 in any case ., What then might account for the large and sustained rise in complexity ?, One plausible explanation is an escalatory arms race that gives rise to progressively more extreme and complex adaptations 7 , 19 , 20 ., For example , coevolution between cheetahs and gazelles may have favored ever-increasing speed , which was achieved by evolving more complex musculoskeletal systems ., In many systems , however , coevolution does not occur along a single axis , but instead involves many traits 21 and can lead to fluctuating frequency-dependent selection instead of an arms race 22 ., For example , such frequency-dependent fluctuations appear to dominate the interactions between Daphnia magna and its parasite Pasteuria ramosa , as determined by reviving eggs and spores from various sediment depths representing different historical states of the interaction 23 ., Escalating arms races and negative frequency-dependent cycling , in general , are the two main outcomes of host-parasite coevolution ., Escalation could lead to an increase in complexity if , for example , more complex tasks provided hosts with resistance to any less complex parasites ., However , there is no such task “dominance” in Avida ., Instead , a particular parasite can infect a particular host provided they share at least one function ., Given that requirement , there is no inherent reason that escalation must occur 24 , 25—for example , the host and parasite populations could cycle repeatedly between two states—and so we can reject the arms-race hypothesis as a sufficient explanation for the emergence of more complex traits in hosts that coevolved with parasites ., Nonetheless , it is important to note that frequency-dependence and escalation are not mutually exclusive processes ., How could negative frequency-dependence drive a sustained increase in host complexity rather than producing simple cycles ?, One possible explanation is that parasites maintain a “memory” of previously encountered host states ., If so , then hosts can escape infection only by evolving in a previously unexplored direction—in the Avida system , by evolving an entirely new and therefore usually more complex function to acquire resources , rather than recycling one that was previously discarded after it was targeted by the parasite ., The simplest way to achieve such memory is if a parasite population evolves generalist phenotypes that can infect multiple hosts , including types no longer common in the community ., Indeed , the coexistence of multiple host types maintained by negative frequency-dependent selection would favor parasites with broad host-ranges ., To examine whether this population-genetic memory existed , we quantified the average number of functions that parasites could perform ., Consistent with the memory hypothesis , parasites evolved to become generalists that often performed four or five functions and thereby could infect several different host types ( Figure 3 ) ., By contrast , we expect the hosts to evolve primarily as specialists because an individual needs to perform only one function to obtain resources , and performing multiple functions makes it vulnerable to a broader range of parasites ., Indeed , most hosts performed only a single function ( Figure 3 ) , although that function became much more complex over time ( Figure 2 ) ., To verify that the parasites population-genetic memory drove the evolution of host complexity , we performed another set of coevolution experiments using a “challenge” design ., This design is analogous to a microbiological approach in which bacteria are challenged with phage , a single resistant mutant is isolated , the phage are then challenged with the resistant host , a single host-range mutant is isolated that can overcome the resistance , and the cycle is repeated 26 ., Using this design , diversity is lost because only individual mutants are retained at each step , and the advantage to the parasite of retaining a broad host-range ( i . e . , memory of prior hosts ) is reduced or eliminated ., Therefore , if the parasites population-genetic memory drove the evolution of host complexity in the original coevolution treatment ( Figure 2 ) , then we expect hosts to achieve reduced complexity under the challenge regime ., Indeed , the resulting host complexity was much lower in the challenge treatment than with coevolution ( p≪0 . 001 , Mann-Whitney U\u200a=\u200a2 , 373 ) ; in fact , the challenge treatment was indistinguishable from the populations that had evolved without parasites ( p\u200a=\u200a0 . 43 , Mann-Whitney U\u200a=\u200a1 , 298 ) ., We can form an intuitive understanding of the parasites population-genetic memory and its effects on the evolution of complexity using the imagery of an adaptive landscape ., Consider the case where increasing complexity is disadvantageous because performing more complex functions requires more resources than performing simpler tasks ., In the absence of parasites , hosts will evolve the simplest viable functions ( Figure 4A ) ., However , when this host is targeted by parasites , the landscape is deformed , creating a new peak at a slightly more complex task ( Figure 4B ) ., As coevolution continues , additional hosts and parasites will evolve and a diverse set may be maintained through negative frequency-dependent selection ., This community further depresses the landscape , thus moving the peak toward even higher levels of complexity ( Figure 4C and 4D ) ., To evaluate whether our experiments supported this intuitive model , we measured the proportion of parasites unable to infect hosts performing each one of the nine logic functions on its own ., That proportion represents a critical fitness component of the host because it reflects the hosts ability to resist infections by the parasites present in its environment ., Figure 4E–4H shows the empirical relationship between average host fitness ( i . e . , resistance ) and the complexity of the task performed over evolutionary time ., In support of our population-genetic memory hypothesis , the fitness peak shifted strikingly toward higher levels of complexity as coevolution progressed ., Thus , the diversity of parasites—with their individually and collectively broad host-ranges—sustained a memory of previously evolved host phenotypes and generated an adaptive landscape for the host that favored increasingly complex tasks ., To test whether the fitness landscape shaped by a coevolved population of parasites was sufficient to drive the evolution of complexity observed in our original coevolution treatment , we performed a new treatment in which the parasite population began with genotypes “frozen” at the frequency they occurred within each original replicate at 250 , 000 updates ( the halfway point , when the majority of host complexity and parasite diversity had evolved ) , but further evolution of the parasite was precluded ., To maintain constant frequencies of the parasite genotypes , each newly reproduced parasite was assigned a random genotype from the 250 , 000-update set ., After 500 , 000 updates in this complex-but-static environment of frozen parasite frequencies , hosts evolved significantly higher complexity than in the treatment without parasites ( p\u200a=\u200a0 . 003 , Mann-Whitney U\u200a=\u200a1 , 684 ) ., However , the hosts confronted with the complex-but-static parasite populations did not reach as high a level of complexity as when the parasites coevolved ( p≪0 . 001 , Mann-Whitney U\u200a=\u200a1 , 946 ) ( Figure 5 ) ., This disparity may indicate an effect of fluctuating environments , such that dynamic parasite environments favor increased host complexity more than complex-but-static parasite environments ., To test this hypothesis , we then allowed hosts to evolve in environments where we “replayed” the changing parasite genotype frequencies over time from the coevolution treatment , but where these parasite genotypes did not respond to the host evolution that was occurring within any particular replicate ., Again , the host populations that evolved in this replay treatment achieved significantly greater complexity than those that evolved without the parasites ( p\u200a=\u200a0 . 034 , Mann-Whitney U\u200a=\u200a1 , 529 ) , but the hosts in the replay treatment still did not reach as high levels of complexity as the coevolved hosts ( p≪0 . 001 , Mann-Whitney U\u200a=\u200a1 , 908 ) ( Figure 5 ) ., Thus , coevolved parasites—whether constant ( frozen ) or varying over time ( replayed ) —favored the evolution of hosts with more complex functions than hosts that evolved without parasites at all ., Nonetheless , the hosts under these treatments failed to evolve the highest level of complexity , which they achieved with coevolving parasites ., Coevolution involves reciprocal changes in which the host population influences how the parasite population responds , both ecologically and evolutionarily , and vice versa ., Although the parasite population was diverse in both the frozen and replayed treatments , and while it varied in time in the latter treatment , the evolution of the parasite population was decoupled from the evolutionary changes that occurred in the host population ., Taken together , these experiments thus indicate that the special push-and-pull of coevolution played a major role in the evolution of host complexity ., They also imply a more dynamic view of population-genetic memory , one in which negative frequency-dependence constantly tunes the parasite population in response to host evolution ., Without coevolutionary reciprocity , the interactions between host and parasite populations are dissonant and population-genetic memory is ineffective ., We started the previous experiments with hosts and parasites capable of performing only the simplest logic functions in order to understand how coevolution might drive the emergence of complex functions from simpler ones ., However , we can also ask whether coevolution would sustain the greater functional complexity if the experiments began with hosts and parasites that could perform the most complex function , EQU ., Indeed , coevolution maintains much higher complexity than evolution alone from this alternative starting point ( Figure 6 ) ., In these runs , host complexity initially declined rapidly when the parasites were present because the hosts readily escaped by performing new , simpler , tasks ., However , as the hosts exhausted the simple tasks that were easily evolved , their complexity leveled off at a higher value than without parasites ( p≪0 . 001 , Mann-Whitney U\u200a=\u200a2 , 305 ) ( Figure 6 ) ., Although the average complexity across populations never dropped below ∼3 , genotypes within the host populations explored the simplest functions ( Figure S1 ) ., In all but one population , the frequency of hosts that performed only the simplest tasks transiently exceeded 10% ., The apparent equilibrium levels of host complexity with and without coevolving parasites were evidently the same whether the experiments began at low or high complexity ( compare Figures 2 and 6 ) ., Coevolution with parasites also had profound effects on the phylogenetic structure of host populations and on the phenotypic evolvability of host genomes ., With respect to phylogenies , the frequency-dependent nature of host-parasite interactions promotes not only greater diversity at any given moment but also deeper branches that reflect the preservation of diversity through time ., In Avida , we can track genealogies precisely and thus construct exact phylogenetic trees , avoiding uncertainty about historical states and branch lengths ., Figure 7 shows representative trees for host populations that evolved in the presence and absence of parasites , and they differ strikingly in their coalescence profiles ., To formalize this difference , we calculated the time since the most recent common ancestor ( MRCA ) for all 50 host populations in the coevolution and evolution-without-parasites treatments ( Figure 8 ) ., The MRCA in coevolved host populations usually arose soon after the experiment began ( median 3% of the total elapsed time ) , whereas the MRCA in the absence of parasites typically dated to well after the midpoint ( median 70% ) , and this difference is highly significant ( p≪0 . 001 , Mann-Whitney U\u200a=\u200a1 , 742 ) ., Thus , coevolution not only affects the outcome of adaptation , but also fundamentally changes how those outcomes are reached ., Coevolution was similarly found to increase the rate of adaptation when embedded in multispecies networks of mutualists 27 ., Although the systems and form of interactions are different , their similar results suggest the important role reciprocity plays in evolving systems ., Previous research using Avida showed that different treatments could drive populations into qualitatively different regions of the fitness landscape; specifically , populations that experienced higher mutation rates evolved onto lower but flatter regions of genotypic space than populations that evolved at lower mutation rates , a phenomenon dubbed “survival of the flattest” 28 ., Here we examine whether coevolution with parasites produced host genomes that were more evolvable with respect to escaping infections ., To that end , we mapped phenotypic changes onto every possible one-step point mutation for the most common host genotype from all evolved and coevolved populations at the end of the experiment ., Several types of phenotypic changes are possible including the gain of a function , the loss of a function , or switching which function is performed without changing the total number of functions performed ., Mutations in the last category are of particular interest because , in the presence of parasites , the ability to switch functions without requiring intermediate steps ( adding a new function before losing the old one ) could be adaptive ., That is , more evolvable hosts would be able to change phenotypes faster and could thereby escape coevolving parasites more readily ., While selection does not directly favor hosts with more evolvable genotypes , they are more likely to produce surviving lineages when coevolving with parasites; thus , second-order selection could drive the evolution of evolvability ., In strong support of this hypothesis , function-switching mutations were >10-fold more common in hosts that evolved with parasites than in hosts that evolved without parasites ( p≪0 . 001 , Mann-Whitney U\u200a=\u200a2 , 338 ) ( Figure 9 ) ., To evaluate whether this effect might somehow merely reflect the more complex tasks typically performed by coevolved hosts , we analyzed pairs of genotypes from the coevolved and evolved host populations that perform identical sets of tasks ., The coevolved hosts were still significantly more evolvable than their paired evolved host ( P≪0 . 001 , Wilcoxon signed-rank W\u200a=\u200a112 , 616 . 5 ) ( Figure 10 ) , although the frequency of task-switching mutations tended to be lower in both treatments after this pairing procedure ., Thus , coevolution drove host populations to occupy more evolvable regions of the adaptive landscape ., Taken together , our experiments show that parasites pushed hosts to levels of functional complexity that were well beyond what they achieved by random walks ( Figure 2 ) ., This complexity resulted from population-level processes 29–31 , in which frequency-dependent interactions sustained generalist parasites ( Figure 3 ) that were supported by phenotypically and phylogenetically diverse hosts ( Figures 7 and 8 ) ., If population-level effects were eliminated , as in the challenge experiments , then host complexity remained low ., Moreover , if the coevolutionary feedback between hosts and parasites was broken by freezing or replaying parasite genotypes , then hosts did not evolve such complex tasks as when parasite populations could respond to the changing host population ( Figures 4 and 5 ) ., Although the form of interactions between the hosts , their resources , and parasites in our study system ( Figure 1 ) strongly constrained host evolution ( e . g . , hosts performing multiple functions were more broadly susceptible to parasites and rarely observed ) , hosts nevertheless overcame these limitations by becoming more evolvable ( Figure 9 ) ., In particular , host genomes evolved such that a much larger proportion of mutations caused a switch from one resource-acquisition function to another , thereby allowing hosts to escape , in a single step , parasites that targeted the first function ., These results—from an unusual but highly tractable system—add to growing evidence from experiments and theory that coevolutionary processes promote biological diversity , new functions , and evolvability 16 , 17 , 20–25 , 29–35 ., All experiments were performed using the Avida 2 . 13 . 0 software , which is available without cost ( http://avida . devosoft . org/ ) ., Configuration files with the parameter settings used and data files have been deposited into the Dryad Repository: http://dx . doi . org/10 . 5061/dryad . 485qq 36 ., Host and parasite populations lived in a well-mixed chemostat-like environment , with a single type of resource entering at a constant rate ., Hosts obtained resources required for replication by performing any of nine distinct one- and two-input logic functions , provided there were resources available in the environment ., A parasite could infect a host if they performed at least one function in common , and an infecting parasite then acquired 80% of its hosts energy ( CPU cycles ) 37 ., The ancestral hosts and parasites could perform only NOT , one of the two simplest functions ., We initially monitored evolution under two main treatments , each with 50-fold replication: host organisms evolved alone in one treatment , and they coevolved with parasites in the other ., Each replicate started with a different numerical seed , and the resulting sequence of pseudo-random numbers influenced mutations , parasite-host encounters , and other probabilistic events ., The parasites went extinct in 12 coevolution runs; except where otherwise noted , we included those runs in our analyses ., In a third treatment , the parasites were experimentally removed halfway through each run , with the first half being identical to a run in the coevolution treatment ( i . e . , using the same initial seed ) ., All runs lasted for 500 , 000 updates; an update is an absolute time unit in Avida equal to the execution , on average , of 30 instructions per individual host organism ., Generation times for the ancestral host and parasite genotypes were 63 and 23 updates , respectively , although generation times changed as genomes evolved ., Each host population began with one individual; the carrying capacity was 14 , 400 in the absence of parasites ., In the coevolution treatment , 400 parasites were introduced after 2 , 000 updates; only a single parasite could infect an individual host ., Mutation rates were 0 . 25 and 0 . 5 per genome replication for the ancestral host and parasite , respectively , of which 90% were point mutations and 5% each were insertions or deletions of single instructions ., Per-site mutation rates were constant , so total genomic rates varied with changes in genome length ., Mutations occurred at random with respect to genome position ., To eliminate all population-level interactions in both species , we screened individual hosts and parasites for defenses and counter-defenses , rather than using evolving populations ., Starting from the same ancestral host , we generated thousands of individuals using the same mutation regime as in the evolution experiments , and we randomly chose a single host mutant that was resistant to the ancestral parasite ., We then repeated this process for parasites , again using the same mutation regime as in the evolution experiments , and we isolated a host-range mutant able to infect that resistant host mutant ., We continued the pairwise challenges using the derived host and parasite genotypes for 50 rounds ., A challenge experiment was stopped if we failed to isolate a relevant mutant after screening 500 , 000 individuals ., In the comparisons with the evolution and coevolution treatments , we used 56 challenge experiments ( out of 100 started ) that achieved the full 50 rounds of reciprocal defenses and counter-defenses ., However , the truncated runs appeared to be indistinguishable from those that went the full duration ., In these experiments , we allowed host populations to evolve with either “frozen” or “replayed” parasites ., During the original coevolution experiments , we saved each replicates entire set of host and parasite genotypes every 1 , 000 updates ., We modified the Avida source code such that this record of genotypes can be loaded into an on-going run at any point by adding an option to override the normal replication process with one that samples from a genotype list ., When organisms reproduce , instead of inheriting their parents genome , the offspring is assigned a random genotype from the list ., This procedure can be implemented for hosts , parasites , or both; however , in the freeze and replay experiments presented here , we manipulated only the parasite populations using this new procedure ., In both treatments , we injected 1 , 500 parasites into the host population after 2 , 000 updates; this number was increased relative to the coevolution treatment to ensure that the frozen and replayed parasite populations , which were sometimes poorly adapted to the ancestral host , did not go extinct ., In the freeze treatment , each host population confronted a parasite population that was complex and diverse , but constant in its genotypic frequencies over an entire run ( except for the fluctuations associated with births and deaths of the parasites ) ., The composition of each parasite population was based on the list of parasites taken at the mid-point ( i . e . , 250 , 000 updates ) of one of the coevolution treatment runs ., Thus , the genetic composition of the parasite population was frozen throughout the run , although the total number of parasites co | Introduction, Results and Discussion, Materials and Methods | The evolution of complex organismal traits is obvious as a historical fact , but the underlying causes—including the role of natural selection—are contested ., Gould argued that a random walk from a necessarily simple beginning would produce the appearance of increasing complexity over time ., Others contend that selection , including coevolutionary arms races , can systematically push organisms toward more complex traits ., Methodological challenges have largely precluded experimental tests of these hypotheses ., Using the Avida platform for digital evolution , we show that coevolution of hosts and parasites greatly increases organismal complexity relative to that otherwise achieved ., As parasites evolve to counter the rise of resistant hosts , parasite populations retain a genetic record of past coevolutionary states ., As a consequence , hosts differentially escape by performing progressively more complex functions ., We show that coevolutions unique feedback between host and parasite frequencies is a key process in the evolution of complexity ., Strikingly , the hosts evolve genomes that are also more phenotypically evolvable , similar to the phenomenon of contingency loci observed in bacterial pathogens ., Because coevolution is ubiquitous in nature , our results support a general model whereby antagonistic interactions and natural selection together favor both increased complexity and evolvability . | Over billions of years , life has evolved into the extraordinarily diverse and complex organisms that populate the Earth today ., Although evolution often proceeds toward increasing complexity , more complex traits do not necessarily make organisms more fit ., So when and why is greater complexity favored ?, One hypothesis is that antagonistic coevolution between hosts and parasites can drive the evolution of more complex traits by promoting arms races with increased defenses and counter-defenses ., Here , by using populations of self-replicating host computer programs and parasitic programs , which steal processing power from their hosts , we demonstrated that coevolution promotes complexity and dissected how it does so ., Instead of simple escalation , we found that a diversity of coevolving lineages must arise for coevolution to drive complex traits ., Surprisingly , coevolution had a second effect; it promoted the evolution of more evolvable hosts ., As a consequence , mutations in the evolved host genomes that confer resistance to parasites occur at high rates , which help the coevolved hosts outrun their parasites ., Our experiments with an artificial system demonstrate how the naturally ubiquitous process of coevolution can promote complexity and favor evolvability . | evolutionary ecology, ecology and environmental sciences, coevolution, ecology, evolutionary modeling, biology and life sciences, computational biology, evolutionary biology, evolutionary emergence, evolutionary processes, evolutionary theory | Experiments using a digital host-parasite model system show that coevolution can drive the emergence of complex traits and more evolvable genomes. Homepage Title: Parasitism Drives the Evolution of Complexity |
journal.pcbi.1004857 | 2,016 | Integrating Antimicrobial Therapy with Host Immunity to Fight Drug-Resistant Infections: Classical vs. Adaptive Treatment | Overcoming antimicrobial resistance is currently considered an international medical priority 1 , 2 ., The evolution of drug resistance affects our ability to treat new infections as well as carry out hospital procedures that rely on the prophylactic use of antibiotics such as surgeries and organ transplants ., Despite extensive research , antimicrobial alternatives to antibiotics , are not yet a practical solution over current therapies ( reviewed in 3 ) ., It is thus critical to evaluate different treatment strategies in order to understand how the various parameters involved in the prescription of antibiotics can influence the selection and spread of drug resistance ., Optimization of antibiotic treatments to increase the effective life span of drugs , while reducing both the probability of resistance evolution and the adverse effects of treatments , is a key component of hospital antimicrobial stewardship programs 4 , as well as a research priority in evolutionary epidemiology 5–7 ., The problem of preventing the emergence of resistance is augmented with the problem of resistance management once it is already present in a population 8 ., Often , by the time bacterial infections cause symptoms and treatment is initiated , the within-host bacterial load is large enough to harbour mutants that are resistant to the treating antibiotic 9 ., More importantly , in hospital settings , resistant bacteria can already be acquired upon infection , requiring specialized therapeutic regimes 10–12 ., Classical wisdom in drug-resistance management recommends that treatments should be as aggressive as possible , using the highest possible dose to ensure that the pathogen load is eliminated , and to prevent de novo evolution of resistance mutations 13 ., These aggressive therapies have recently been questioned on the basis that the stronger the treatment applied , the stronger the selection favouring resistant pathogens , in particular in infections harbouring pre-existent resistance ., This conventional protocol of hitting hard and hitting fast might be relevant for highly mutable pathogens such as HIV , but in cases where resistant strains are more likely to be acquired in the community such as in TB 8 the advantages of aggressive therapies are less obvious 14 ., Alternative strategies could include more moderate treatments , or adaptive regimens where doses and treatment durations closely follow patient health 14–16 ., Current empirical and theoretical evidence has examples to support both therapeutic strategies , as well as for a mixed compromise such as high dose and short treatments ( reviewed by Kouyos et al . 15 ) ., For instance , experimental studies using rodent malaria parasites in laboratory mice have shown that less aggressive chemotherapeutic regimens substantially reduce the probability of onward transmission of resistance without significant changes in host pathology 16 ., In contrast , varying concentrations of vancomycin in vitro17 and in vivo using a rabbit model 18 has confirmed the advantage of high dose aggressive treatment in controlling the resistant populations of Staphylococcus aureus ., This multitude of results indicates that the problem of devising general practices for treatment is far from settled ., Conceptual frameworks can help compare aggressive and moderate chemotherapy 15 , but quantitative systematic analyses are also needed ., The current challenge is to identify among the diverse potential treatment regimes , those that minimize selection for drug-resistance while not compromising patient health 14 ., A general principle advocated to guide rational development of patient treatment guidelines is to impose no more selection than is absolutely necessary ., For this , it is important to understand when rules like ‘hit hard and hit early’ should apply 13 , and when more moderate treatment regimes would be more effective ., Mathematical models play instrumental role in this endeavour ., When focused on population level dynamics they can evaluate and guide antibiotic use regimens for hospitals 11 , 19 , 20 or wider communities 21 , generally in endemic , but also in epidemic scenarios ( e . g . antiviral usage 22 , 23 ) ., When modelling pathogen dynamics within host , mathematical approaches can outline the mechanisms of interaction and feedbacks among pathogen types , and quantify how this basic ecology is modulated by one drug 24 or multiple drugs 25 ., An important , but often overlooked factor in the process of infection clearance and resistance management is host immunity ., A strong immune response can substantially reduce the need for long treatments , as evidenced by some acute infections tending towards shorter drug treatments in hosts with intact immunity 26–28 ., The interplay between host immunity and antimicrobial drugs has recently been incorporated into mathematical models of infection 29–31 ., Previous work 29 has shown that the presence of an immune response can narrow down the mutant-selection window ( MSW ) , defined as the range of drug concentrations for which the drug is strong enough to remove the sensitive population 32 , but insufficient to remove the partially resistant pathogen population ., Along similar lines , Ankomah and Levin , 31 , using an explicit resource-based model for the interaction between pathogen and host immunity , have investigated infection scenarios , separating the effects of pathogen-dependent and pathogen-independent immune responses ., Yet , a quantitative understanding of host immunity as a player in optimal treatment of resistant infections remains under-developed ., A series of studies have recently addressed the role of timing of antimicrobial use at the population level 22 , 23 ., By considering the indirect and direct effects of antimicrobial use , models have found that optimal timing for treatment at the population level is well into the course of an epidemic , where the indirect effects of delays usually result from minimizing the degree of overshoot , i . e . minimizing the number of cases beyond the number that would be needed to reach the epidemic threshold ., There are parallels between transmission processes at the population level and pathogen growth dynamics at the within-host level , where timing effects of antimicrobial therapy have also been shown to be important 33 ., In this article , we combine these two important concepts to study antimicrobial treatment of drug-resistant infections:, i ) we zoom further into host immunity processes , and, ii ) we analyze explicitly the role of treatment timing on the success or failure of antibiotic therapies ., We consider a dynamic mathematical model that describes the interaction between the host’s immune system , pathogen density , and antimicrobial treatment in mixed infections of drug-sensitive and pre-existing drug-resistant pathogen strains ., By analysing a diverse range of therapeutic scenarios , and especially focusing on treatment timing , we uncover critical consequences for infection dynamics and selection of resistance , before , during and after treatment ., We also compare in depth through a mechanistic approach , classical and adaptive treatment protocols , applied to the same infection ., To facilitate insight into the driving factors of treatment efficacy , we simplify many aspects of host-pathogen interaction , focusing on key features ., We examine their interplay with treatment parameters , and their final impact on infection outcomes , such as total immunopathology , time to clearance , pathogen burden , and overall resistance ., Our framework formalizes and broadens up the question of what it means for a treatment to be optimal and how such optimality can be achieved in practice ., The within-host model is designed to investigate the interplay between antibiotic treatment regimes and host immune response in acute drug-resistant infections ., Our formulation is based on a previous within-host model of infection dynamics 33 , but here we consider two pathogen phenotypes: those sensitive to the drug , Bs , and those partly resistant Br ., These are distinguished by their intrinsic growth rates ( r0 and r1 ) and killing rates by antibiotic ( δ0 , δ1 ) ., We consider c = r0 − r1 ≥ 0 to be the fitness cost of resistance 34 and a = δ1/δ0 , ( 0 ≤ a ≤ 1 ) to represent the fitness benefit of resistance , i . e . the factor by which antibiotic killing rate is reduced in the resistant sub-population ., The action of host immunity , is considered explicitly , in terms of naive antigen-specific precursor cells N , effector cells E , and memory cells M . We thus implicitly consider those infections that may have escaped the first barrier of innate immunity in the host 35 ., The pathogen-dependent immune dynamics represents a typical CD8+ T-cell mediated immune response 36 , 37 , but also describes broadly key features of CD4+ cell responses 38 , 39 ., These are major players against intracellular bacterial pathogens , such as Listeria monocytogenes40 , and Legionella pneumophila41 , 42 , but have also been implicated in Haemophilus influenzae39 , 43 and protective responses against pneumococcal bacteria 44 ., In the interest of generality , we keep the detail of immune responses to a minimal level ., Thus , the model is inevitably a simplification of the complex interaction between host immunity , bacteria , and antibiotics 45 ., However , the underlying assumptions do capture crucial aspects of the expected immune responses in acute infections ., These include induction , activation , proliferation , decay and memory formation , typically studied in greater empirical detail in virus-host interactions 46 , 47 ., Several mathematical aspects of our formulation feature in other theoretical models of infection 30 , 31 , 48 ., Within-host dynamics for a mixed infection with a drug-sensitive ( Bs ) and pre-existing partially resistant ( Br ) strain are described with the following set of ordinary differential equations:, d B s d t = r 0 B s - d B s I - δ 0 B s η ( t ) A m ( 1 ) d B r d t = r 1 B r - d B r I - δ 1 B r η ( t ) A m ( 2 ) d N d t = - σ N B k + B ( 3 ) dEdt= ( 2σN+σE ) Bk+B−hE ( 1−Bk+B ) ( 4 ) dMdt=fEh ( 1−Bk+B ) , ( 5 ), where B ( t ) = Bs ( t ) + Br ( t ) is the total pathogen load at time t , and I ( t ) = N ( t ) + E ( t ) + M ( t ) is the total number of immune cells activated to clear the pathogen ., Naive precursor cells ( N ) are stimulated to divide and differentiate into effector cells ( E ) in response to increasing pathogen density ., Effector cells proliferate further upon antigen stimulation at rate σ as long as pathogen is still in circulation ., As bacteria are cleared , the majority of effector cells undergo apoptosis at rate h per cell , except for a fraction f that differentiate into memory cells ( M ) that persist indefinitely ., All three types of immune cells act to kill pathogen , but effector cells represent the dominant arm of the host immune defense , in particular in primary infection , which we focus on ., An important model assumption is that the killing rate d by lymphocytes is equal for both pathogen sub-types , regardless of their antimicrobial susceptibility ., Another important assumption regards the immunity stimulation function ., For immune stimulation by antigen , a monotonically increasing saturating function of pathogen density ( Hill function with coefficient 1 ) is assumed , where the parameter k represents the half-saturation constant for stimulation of lymphocytes to divide and differentiate ., We will hereafter refer to this parameter as the host immunity threshold ., To reflect the discrete nature of the pathogen , we assume an extinction threshold , when pathogen density of either sub-population falls below a critical level Bext ., Since the model is primarily designed to describe acute infection , we do not include a limiting resource for pathogen growth 24 , assuming main control via host immune responses ., A detailed description of model parameters is given in Table 1 ., Although our simulations are based on a limited set of parameter values , likely to apply to a range of acute infections , the theoretical analysis that we provide alongside simulations enables extrapolation of our results to settings and numerical values departing from the ones considered here ., As in another recent study 31 , the exact parameter values used for simulations do not reflect any particular antibiotic-species combination ., To model antimicrobial treatment we use an indicator function η ( t ) , which represents the rate of antimicrobial uptake per unit of time ., The dose of the drug deployed is denoted by Am ., The case when treatment onset ( τ1 ) and duration ( τ2 ) are fixed from the start corresponds to a classical treatment ., The case when drug uptake depends on bacterial density within host corresponds to an adaptive regime ., For classical treatment , the rate of administration of antimicrobials is:, η ( t ) = 1 if τ 1 ≤ t ≤ τ 1 + τ 2 0 if t < τ 1 or t > τ 1 + τ 2 ., ( 6 ) In the adaptive regime , treatment onset and duration are influenced by the bacterial dynamics in the infected host ., Previous authors have considered tight coupling between adherence to drug and bacterial load 29 ., In this study , similar to the study by Ankomah and Levin 31 , we only consider the simplest form of adaptive treatment that uses a threshold for total pathogen load , Ω , above which the patient takes the drug , and below which the patient does not ., The rate η ( t ) of antibiotic administration per unit of time , becomes a direct function of pathogen load B and the threshold Ω:, η ( t ) = 1 , if B ( t ) ≥ Ω 0 , if B ( t ) < Ω ., ( 7 ) Thus , over a given treatment window , the net average amount of drug taken by the host per unit of time in the adaptive case , may be less than the actual administered dose ., This alternative model of antimicrobial delivery could mirror a ‘take when feeling bad , stop when feeling good’ approach , requiring necessarily reliable translation between symptoms and pathogen load ., Focusing on the net effect of the antibiotic on the bacterial population , which has been shown to be relatively insensitive to changes in the frequency of administration of the drug 31 , we neglect the explicit pharmaco-dynamics of the antibiotic ., Thus we model only the average rate of antibiotic-mediated pathogen killing ( represented by the product δ0 Am and δ1 Am , respectively for Bs and Br ) , which simplifies analysis ., Both in the classical and adaptive regime , we explore treatment onset at various times over infection , departing from previous studies that typically link treatment initiation to a fixed pathogen load or the peak bacterial density 30 , 31 ., Our formulation of treatment delay is inspired by two recent studies 23 , 33 ., Its generality enables a deeper understanding of the trade-off induced by antibiotic treatment between reduction in host pathology and immunization , in the new context of resistant infections ., Assuming an extinction threshold when pathogen density of each subpopulation within host reaches Bext , we can compute text , the extinction time , or clearance time ., Because we simulate treated and untreated infections only up to a finite time horizon T , usually set to 30 days , infection duration is thus defined as:, D = m i n ( T , t e x t ) ( 8 ) The total resistance burden over the entire infection is calculated as, R t o t = ∫ 0 D B r ( t ) d t ., ( 9 ) The total pathogen burden over infection is B t o t = ∫ 0 D B ( t ) d t , and final host immune memory is M ( D ) ., We also track the resulting immunopathology 33 , which roughly reflects the cumulative damage to host health due to pathogen killing by cells of the immune response and associated inflammation 57 ., For the total immunopathology accumulated up to time t , H ( t ) , following 33 , we define: H ( t ) = ∫ 0 t d B ( s ) I ( s ) d s ., As the pathogen population grows and host immunity builds up , the cumulative immunopathology due to immune-mediated killing and inflammation also grows following the infection dynamics ., Upon pathogen clearance ( or at the end of a simulation ) the immunopathology accumulated over infection reaches, H t o t = ∫ 0 D d B ( s ) I ( s ) d s ., ( 10 ), We perform a systematic analysis of these infection outcomes , varying treatment regimes ( classical and adaptive ) and model parameters , such as the fitness cost and benefit of resistance , and host immunity characteristics ., We use as reference for comparison summary measures from infections in which no treatment is used ., All simulations are performed in Matlab® R2011a ., In the absence of treatment , the infection follows a typical acute dynamics ( Fig 1A ) ., Sensitive bacteria grow initially quasi-exponentially , while immune responses are not yet active ., Resistant bacteria also increase from their initially low numbers , but relatively more slowly , depending on their fitness cost c = r0 − r1 ., Resistant bacteria reach their peak around the same time as the sensitive sub-population , but at a lower density ., As sufficient immunity gradually builds up during the bacterial growth phase , bacterial clearance is initiated , primarily through the action of effector cells ., Following pathogen decline , effector cells also decline , with a fraction of them differentiating into persistent immune memory cells ., High levels of acquired immune memory will act as pre-existing immunity in a secondary infection with the same pathogen and lead to rapid clearance ., Mathematical analysis of the model in the absence of antibiotic confirms that stability of the infection-free state requires N* + M* > max ( r0 , r1 ) /d ( see S1 Text , part I ) ., Below we derive analytical expressions , to understand how characteristics of the pathogen and of the host , represented by different parameters , interact to determine outcomes of infection ., This serves as a starting point to then explore how perturbations like treatment , or variation in parameter values can affect these baseline dynamics ., Focusing on the ‘expansion phase’ of immune dynamics , as in 48 , 58 , we can simplify the rates of change in total bacterial density and host immunity by the following sub-system:, d B d t ≈ r 0 B - d B I ( 11 ) d I d t ≈ σ I B k + B ( 12 ), where I = N + E + M and B = Bs + Br ., By assuming negligible fitness cost of resistance ( r0 ≈ r1 ) , the equations above give somewhat an upper bound on total bacterial growth ., Biologically , the relative magnitudes of various parameters satisfy: B0 ≪ k , dI0 ≪ r0 , where B0 = B ( 0 ) is the initial pathogen density , and I0 = I ( 0 ) reflects the precursor frequency ,, i . e ., the number of initial immune cells specific to the pathogen at the time of infection ., Dividing the above equations , and integrating , we obtain:, log ( B+kB0+k ) =r0σlog ( II0 ) −dσ ( I−I0 ) ( 13 ) This equation gives the relationship between the number of immune cells and parasite density at any given time during the bacterial growth phase ., Thus , it allows us to calculate the level of immunity as a function of current pathogen load , and viceversa ., Under this approximation , the peak pathogen load , in the absence of treatment , occurs when a critical level of host immunity has been reached , namely when, I ≈ I c r i t = r 0 d ., ( 14 ), The peak pathogen density at the end of the growth phase in acute infection can be obtained from combining Eqs 13 and 14:, Bmax≈ ( B0+k ) ( r0deI0 ) r0/σ ( 15 ) The time it takes for the pathogen load to reach its peak can be approximated considering two phases of growth:, i ) the time it takes the pathogen load to reach k , required for half-maximal immune stimulation , and, ii ) the time it takes the immune response subsequently to grow from its initial level to the critical level Icrit ., This dynamic decomposition focusing on k is an analytically convenient choice , yielding:, tpeak=tk+tk→peak≈1r0log ( kB0 ) +1σlog ( IcritI0 ) ( 16 ) Such expressions , taken together , convey how host immunity characteristics , e . g . initial immunity I ( 0 ) 55 , or immune cell recruitment rate σ , affect different infection outcomes ., These may vary with host age 59 , or other aspects of immune competence ., One can also notice above the importance of the host immunity threshold , k , and maximal pathogen growth rate , r0 , which may vary too across host-pathogen systems ., In the absence of the drug , the difference in growth rate between resistant and sensitive bacteria does not significantly affect the dynamics of immune build-up , peak bacterial load , or the cumulative immunopathology over infection ., The cost of resistance ( c = r0 − r1 ) only changes the relative frequency of resistance in the total pathogen load ., This is because immunity gets equal stimulation from both bacterial types and kills them at the same rate ., After the immune ‘expansion phase’ , which leads to pathogen clearance , the ‘contraction’ and ‘memory’ phases of the immune response follow , provided that h > 0 , described in detail by Eqs 3–5 of the full model ., Summing those three equations , one can see that total immune response in the system keeps increasing whenever B > k h ( 1 − f ) E σ ( N + E ) ., This means some immune stimulation still continues during pathogen decline , thus the peak immune response reached over infection typically exceeds the critical value Icrit required for triggering clearance ., Notice that setting h = 0 in the full model would mimic a situation of non-waning immunity ( at least non-waning in the time-scale of interest ) , quantitatively captured by the simple system of Eqs 11 and 12 ., Most of the analysis above could thus be useful to understand also such a scenario , where for instance , the final level of immunity Ifinal accumulated after infection could be calculated from Eq 13 , by solving it for B = 0 ., In all these theoretical scenarios , infection in principle resolves through action of host immunity , but depending on the severity of parameter values , the total damage to the host can be overwhelming , such that administration of drugs is required ., By severity here we mean the clinical relevance or manifestation of Bmax in the absence of treatment ( Eq 15 ) , e . g . how close this peak density would be to a pathogenesis or lethal threshold for the host 58 ., This naturally depends on pathogen growth rate and host immune competence ., For example , slowly-growing pathogens might never trigger symptoms in their host ( thus may never need antibiotic treatment ) , and eventually will be cleared by the immune system without causing high levels of pathology ., Next , we analyze the full model with treatment , where antibiotics interact with host immunity ., In treated infections , the presence of a drug-resistant pathogen sub-population becomes relevant in either regime of drug delivery ( see Fig 1B and 1C ) ., The effect of the antibiotic can be encapsulated as a reduction in the intrinsic per capita net growth rate of the two bacterial types during the treatment phase ( τ1 ≤ t ≤ τ1 + τ2 ) ., The antibiotic reduces pathogen load and immunopathology , relieving the burden on host immunity ( Fig 1B ) ., However , its timing , dose , and duration can produce a diverse range of outcomes , as shown in Fig 2 ., With very aggressive treatments , resistant bacteria are not selected , and infections get cleared rapidly ., In other cases , treatment cessation may result in a second infection peak , or even multiple peaks of bacteria , which may be equal to or even higher than pre-treatment levels , and consist of sensitive or resistant organisms ., Treatment consequences vary especially depending on the phase of the infection in which treatment begins , where the growth potential of both strains is modulated by host immunity ., To understand critical treatment parameters , we must consider the respective growth rates of bacterial subpopulations within host at the time τ1 when treatment is applied ., In the presence of an immune response , the doses needed to halt growth of either subpopulation are decreasing functions of the immunity level I ( τ1 ) upon treatment onset:, A m ′ ( I ) = r 0 - d I δ 0 and A m ′′ ( I ) = r 1 - d I a δ 0 , ( 17 ), for the sensitive and resistant strains respectively ., In the absence of any immunity , the antibiotic doses that inhibit growth of sensitive and resistant bacteria are given by the maximum values:, A m * = r 0 δ 0 and A m * * = r 1 a δ 0 , ( 18 ), where A m * ≤ A m * * , if the cost and benefit of resistance balance in such a way that r0 ≤ r1/a ( the scenario we consider here ) ., In general , depending on the level of immunity , thus on the delay for treatment initiation , either bacterial type can decline , persist or grow during treatment , subject to how the actual dose that is deployed , Am , sits in this critical range ( Fig 3 ) ., As a consequence , immunity can also decline , persist or grow while antibiotics are applied ., If during treatment , the net change in dynamics results in an excessive decline of pathogen-dependent immunity , there is a window of possibility for pathogen relapse after treatment cessation , in case complete clearance has not been achieved with the drug ., The sub-population surviving at an advantage at the end of treatment , may be the one to dominate the relapse , provided that such advantage in total numbers is greater than its relative fitness cost in the absence of treatment ( Fig 2A , τ1 = 2 , Am = 4 ) ., When such recrudescence is caused by resistant bacteria ( e . g . for τ1 = 2 , Am = 10 in Fig 2A , or Am = 30 in Fig 2B ) , the lower the fitness cost of resistance is , the faster the new peak will be reached after therapy stops ., Clearly , the amount of drug interference with normal immune build-up during treatment depends on its dose and duration ., Thus , under pathogen density-dependent immunity , if bacteria persist or grow slightly during treatment it may not be so bad , given that such growth helps stimulate more immunity , and reduces the risk of relapse at the end of treatment ., By the same argument , removal of antigen stimulus too rapidly during treatment may have adverse effects , because surviving pathogens at the end of therapy could re-grow if immune responses in the meantime have declined to subcritical levels ( assuming waning immunity , h > 0 ) ., Selecting a moderate regime to balance between these scenarios is a challenge ., While finding an optimal intermediate regime , involving some degree of immune control , is far from trivial , the extreme therapeutic option that does not require immunity at all , is much easier to analyze ., Such antibiotic treatment is bound to be of an aggressive type ., Consider the total bacterial load at treatment onset B ( τ1 ) ., The scenario of drug-only-mediated clearance can be represented as an exponential decay of both bacterial subpopulations during treatment ., Notice that resistant bacteria are killed at lowest rate by the drug , so by approximating the total population decline at that lower rate , we explore the worst case scenario for the host ., Resistant bacteria are also more likely to suffer a fitness cost ( r1 ≤ r0 ) , thus by approximating total population growth at its highest possible rate , r0 , we are also considering a worst case scenario for the host ., In this way , by being conservative in bacterial growth and decline during treatment , we obtain a sufficient criterion for ultimate clearance during classical treatment with dose Am and duration τ2 as B ( τ 1 ) e ( r 0 − a δ 0 A m ) τ 2 ≤ B e x t , which is equivalent to requiring:, Am≥1aδ0 r0−1τ2log ( BextB ( τ1 ) ) ( 19 ) Thus , if the dose and duration of classical treatment , in combination satisfy the above inequality , relative to the pathogen density at treatment onset B ( τ1 ) , and pathogen extinction threshold Bext , infection clearance by the end of treatment is guaranteed , without relying on host immunity ., As the above expression shows , the earlier treatment begins , thus the lower B ( τ1 ) , the easier it is to meet the criterion with smaller doses and shorter treatment duration ( Fig 4 ) ., Generally , the dose Am and duration τ2 , can be traded-off against one another , and still satisfy the clearance criterion for different pathogen loads at classical treatment onset ., The caveat is to know whether these effects are possible with antibiotic doses below the toxic threshold for the patient ., Notice , that the criterion in Eq 19 , does not depend on the cost of resistance , and also does not exclude that clearance may be achieved with lower doses , because the additional pathogen killing by immunity , accumulated up to and during treatment , is not accounted for by this formula ., Here , we explore the interplay between antibiotic treatment and host immunity in the full dose range through numerical simulations of the complete model ., In several ‘theoretical experiments’ ( Fig 2 ) , we vary treatment onset τ1 , between 2 and 5 days post-infection , and consider treatment duration between 3 and 15 days , realistic for bacterial infections 56 ., Such duration may correspond to the prescribed therapy by a doctor , or may reflect the actual adherence by the patient ., Similarly , the delay can reflect the time over a typical infection course when a patient seeks treatment , and this may fluctuate from person to person ., Varying the antimicrobial dose , we observe that doses of the drug below A m * ( Eq 18 ) administered somewhat later over infection can be efficient in reducing the bacterial burden without promoting selection for resistance , because they yield the preponderate role in eliminating bacteria to the immune system ., As soon as doses go above r 0 − r 1 δ ( 1 − a ) , the fitness differential between sensitive and drug-resistant bacteria is reversed ( e . g . Fig 2A: Am = 4 , τ1 = 2 ) ., Small doses , just above A m * , start to interfere with immune build-up , but this interference decreases when treatment onset is delayed ( moving along the delay axis in Fig 2A and 2B ) ., Higher intermediate doses of the drug , between A m * and A m * * , instead , promote more selection of resistant bacteria during and after treatment , and ultimately infection clearance is achieved by the delayed action of the immune system ( e . g . Fig 2A: Am = 10 ) ., Yet , also here , optimal intermediate delays for initiating treatment , can help reduce host immunopathology and selection of resistance ( S1 Fig ) ., In contrast , higher doses of antimicrobial drug , beyond A m * * , are able to induce immediately the decline of both sensitive and resistant populations ( e . g . Fig 2A: Am = 20 , and Fig 2B Am = 40 ) , but at the risk of a resistant relapse if they are not high enough , or applied sufficiently long ( Eq 19 ) ., At the extreme case of very aggressive treatment , the host experiences minimal immunopathology from infection , but also does not accumulate any immune memory ., As a result of interference by the drug , at certain intermediate doses , relapses in pathogen load can be maintained indefinitely ., These arise when immunity at the end of treatment consists approximately only of effector cells I ≈ E , and coincides with r0/d , while pathogen load coincides with B, ( t ) = hk/σ ( see S1 Text , part I ) ., Since these values are sufficient to yield dE/dt = 0 and dB/dt = 0 , a persistence quasi-steady state is observed with oscillatory dynamics , as reported also in the model by 33 ., Such oscillations typically arise in predator-prey systems , making it hard for the immune response to clear the pathogen in the short term ., The total pathogen density may consist of sensitive or resistant bacteria , if the dose has been low or sufficiently high respectively ., Given enough time however , conversion of effectors into memory cells will gradually build up enough persistent immunity to enable final clearance ., When fixing the delay τ1 for treatment onset , we find many dose-duration combinations that s | Introduction, Methods, Results, Discussion | Antimicrobial resistance of infectious agents is a growing problem worldwide ., To prevent the continuing selection and spread of drug resistance , rational design of antibiotic treatment is needed , and the question of aggressive vs . moderate therapies is currently heatedly debated ., Host immunity is an important , but often-overlooked factor in the clearance of drug-resistant infections ., In this work , we compare aggressive and moderate antibiotic treatment , accounting for host immunity effects ., We use mathematical modelling of within-host infection dynamics to study the interplay between pathogen-dependent host immune responses and antibiotic treatment ., We compare classical ( fixed dose and duration ) and adaptive ( coupled to pathogen load ) treatment regimes , exploring systematically infection outcomes such as time to clearance , immunopathology , host immunization , and selection of resistant bacteria ., Our analysis and simulations uncover effective treatment strategies that promote synergy between the host immune system and the antimicrobial drug in clearing infection ., Both in classical and adaptive treatment , we quantify how treatment timing and the strength of the immune response determine the success of moderate therapies ., We explain key parameters and dimensions , where an adaptive regime differs from classical treatment , bringing new insight into the ongoing debate of resistance management ., Emphasizing the sensitivity of treatment outcomes to the balance between external antibiotic intervention and endogenous natural defenses , our study calls for more empirical attention to host immunity processes . | The evolution and spread of antimicrobial resistance is a major global problem , and a cause of substantial human mortality ., As the discovery of new antibiotics does not follow the rate at which new resistances develop , a more judicial use of available drugs is needed ., Here we develop a mathematical model of within-host infection dynamics that combines the effects of pathogen clearance by the host immune system and by the antibiotics ., Computer simulations and mathematical analysis are used to evaluate treatment protocols in order to identify those that can restore patient health and limit the overall pathogen burden and selection of resistance ., We focus our study on infections with pre-existing resistance , and explore two main treatment strategies: the classical treatment , characterized by fixed drug dose and treatment duration , and the adaptive treatment that closely follows infection outcomes and patient symptoms ., Our results highlight treatment strategies that promote synergy between host immunity and the antimicrobial drug ., This can be achieved by moderate treatments that combine appropriate timing , reduced drug dosage , and short treatment durations ., Our model is developed for bacterial infections but our framework and findings may apply to other biological scenarios featuring drug resistance . | antimicrobials, medicine and health sciences, pathology and laboratory medicine, pathogens, drugs, immunology, microbiology, antibiotic resistance, pharmaceutics, antibiotics, pharmacology, antimicrobial resistance, pathogenesis, immune response, immunity, host-pathogen interactions, microbial control, biology and life sciences, drug therapy | null |
journal.pcbi.1003884 | 2,014 | Reconstruction of the Gene Regulatory Network Involved in the Sonic Hedgehog Pathway with a Potential Role in Early Development of the Mouse Brain | Pattern formation in early animal development is controlled by signal transduction cascades , in which transcription factors ( TFs ) play crucially important roles as downstream effectors ., The signal transduction cascades together with the gene regulatory networks they activate determine the temporal and spatial expression of a wide range of genes for the specification of regions and differentiation of cells 1 ., Sonic hedgehog ( Shh ) is a classical signal molecule required for pattern formation in many aspects of animal development , not least in neural development ., In the central nervous system ( CNS ) , depending on the graded Shh concentration along the dorsal-ventral axis in the mouse ventral neural tube , particular TFs are activated in different regions , resulting in specification of these regions 2–5 ., The Shh signaling pathway is itself activated when Shh binds to its receptor Ptch1 , which , without the ligand , inhibits the cell membrane protein Smo ., Shh binding removes the inhibition on Smo and triggers the activation of three GLI family TFs , Gli1 , Gli2 and Gli3 , which further activate or inhibit specific TFs to determine regional cell fate ., Identifying those downstream TFs and how they work is a central task in the elucidation of early CNS development ., Several recent studies using high-throughput in situ hybridization ( ISH ) have provided a rich harvest of information on spatio-temporal gene expression in early mouse development ., The data are available in databases such as GenePaint 6 , Eurexpress 7 and Allen Brain Atlas ( ABA ) 8 ., GenePaint and Eurexpress have focused on whole mouse embryos at the E14 . 5 stage and covering almost the entire set of known mouse genes ., In contrast , ABA ( http://developingmouse . brain-map . org ) recently offered manually annotated ISH data for the developing mouse brain from three developmental stages: E11 . 5 , E13 . 5 and E15 . 5 ., It includes information about expression intensity , density and pattern for more than 2000 genes , many of which are TFs and key genes in early brain development ., Because ISH data contain high-resolution spatial information on gene expression , they are invaluable for in-depth study of gene regulation in pattern formation during early development ., For example , Visel et al . showed that it is possible to identify the probable targets of Pax6 , a key TF in early mouse brain , by the clustering of co-expressed genes using E14 . 5 ISH data 9 ., However , co-expression of genes does not guarantee that they are directly co-regulated by the same TF ., Furthermore , developmental genes in animals are often regulated by a combination of TFs acting through cis-regulatory modules 10 ., Therefore , high-throughput ISH data has to be integrated with direct gene regulatory data such as genome-wide ChIP-seq data to delineate the specific regulatory mechanisms underlying particular developmental processes ., Such an approach would be very useful in elucidating the genetic networks involved in early brain development ., Gata3 and Foxa2 are two key TFs implicated in early animal development , including early brain development ., Gata3 is a member of the GATA family , consisting of Gata1-6 , among which only Gata2 and Gata3 have been reported to be expressed in the CNS 11 ., Mice homozygous for a Gata3 null mutation were found to have serious malformations of the embryonic brain , revealing its essentiality for that stage 12 ., Furthermore , continuous expression of Gata3 in the brain from early embryo to adulthood suggests that it is important for the maintenance of brain functions beyond early development 13 ., Recent genome-wide studies of Gata3 have mainly focused on the molecular mechanisms underlying its critical roles in T cells 14 and breast cancer 15 ., In breast cancer , Gata3 has been shown to function as a “pioneer factor” to help open up condensed chromatin and recruit other TFs ., However , no genome-wide study on Gata3 in the CNS has been conducted so far ., Foxa family TFs including Foxa1 , Foxa2 and Foxa3 are involved in development , organogenesis 16 and metabolism 17 ., Similarly to Gata3 , there is increasing evidence that they also play crucial roles as pioneer factors 18 ., Unlike Gata3 , however , the role of Foxa2 in brain development has been better studied ., Notochord-secreting Shh required for patterning of the neural tube fails to form when Foxa2 is mutated , hindering the entire subsequent developmental process 19 ., Genome-wide ChIP-seq analysis of Foxa2 targets in midbrain dopaminergic neuron ( mDA ) progenitors revealed that Foxa2 directly regulates key genes in the Shh signaling pathway and that Foxa2 promotes gene expression in the floor plate while repressing the genes normally expressed and required in the ventro-lateral region of midbrain 20 ., In this study , we investigated gene regulation in the Shh signal transduction cascade in early developing mouse brain , focusing on Gata3 and Foxa2 as two putative key TFs ., We have found that they demarcate two mutually exclusive domains in the early mouse brain coinciding with two domains defined by the reciprocal expression patterns of Shh and its receptor Ptch1 ., These will be designated as the Shh+Ptch1− and Shh−Ptch1+ domains ., To understand the molecular functions of Gata3 in the early mouse brain , we used PC12 cell line established from rat adrenal medulla pheochromocytoma to mimic the Gata3-expressed domain in early mouse brain ., We performed Gata3 ChIP-seq in PC12 cells and RNA-seq experiments in Gata3 siRNA knockdown cells ., We found that Gata3 target genes that are down-regulated by Gata3 siRNA knockdown were enriched in the Gata3-expressed domain ., By contrast , Foxa2 target genes were enriched in both Foxa2- and Gata3-expressed domains ., These results suggested that the fates of the two domains were controlled by distinct regulatory mechanisms directed by Gata3 and Foxa2 ., The interaction between these two domains was transmitted via the Shh signaling pathway ., In addition , we identified , amongst Gata3 target genes , Slit2 and Slit3 , which are involved in axon guidance , as well as Slc18a1 , Th and Qdpr , which function in neurotransmitter synthesis and release ., From these findings and ChIP-seq data , we were able to reconstruct a gene regulatory network for the genes in Shh+Ptch1− and Shh−Ptch1+ domains ., Our study expands current knowledge of the Shh pathway and sheds new light on the gene regulatory mechanisms controlling cell fates in the early mouse brain ., We used ISH data of developing mouse brain at E11 . 5 stage from the ABA database ., The data consists of more than 2000 genes manually annotated by experts for the ISH image series ., Gene expression properties were characterized by utilizing three metrics: intensity ( Undetected , Low , Medium and High ) , density ( Undetected , Low , Medium and High ) and pattern ( Undetected , Full , Regional and Gradient ) ., In our study , to convert the textual annotation to numerical data , we used intensity as the metric and treated “Undetected” as 0 and “Low , Medium and High” as 1 for our downstream analysis ., Based on this gene expression data of E11 . 5 mouse brain , we observed that the expression patterns of Shh and its receptor Ptch1 were obviously stratified along the ventral-dorsal axis ., From this observation , we defined the Shh+Ptch1− domain as containing 20 brain sub-regions in the anatomical map provided by ABA , and the Shh−Ptch1+ domain , which we found contains 30 brain sub-regions ( Figure S1 ) ., Next , Fishers exact test was used to identify genes expressed exclusively in the Shh+Ptch1− domain ( P value <0 . 0001 , odds ratio >1 , expressed in more than 10 Shh+Ptch1− sub-regions ) and those in the Shh−Ptch1+ domain ( P value <0 . 0001 , odds ratio <1 , expressed in more than 15 Shh−Ptch1+ sub-regions ) ., These two sets were accordingly defined , respectively , as Shh+Ptch1−-pattern genes and Shh−Ptch1+-pattern genes ., A heatmap containing these two types of genes was generated by the R program ( http://www . r-project . org ) and is shown in supplementary material Figure S2 ., The gene annotations and repeat-masked genome sequences for six mammalian species including human , marmoset , mouse , rat , cow , pig were downloaded from ENSEMBL ( version 62 ) ., Promoter sequences defined as the region upstream 1000 bp to downstream 200 bp from transcriptional start site ( TSS ) were extracted using Perl Script from each species ., For each mouse gene , we obtained their orthologous gene information in the other five mammalian species using ENSEMBL homologs data ( version 62 ) ., Promoter analysis was performed based on Pscan program 21 , by which we can obtain the enriched transcription factor ( TF ) binding motifs in each set of promoters of orthologous genes ., The relationships between TF binding motifs and TFs were obtained from TRANSFAC 22 ., In this study , TF motif-target relationships were determined by selecting TF motifs with the criteria that enrichment P value less than 0 . 005 and the rank is at least top 20 ., Motif enrichment in the promoter sequences of Shh+Ptch1−- and Shh−Ptch1+-pattern genes were performed using Pscan solely on mouse genes ., Enriched TF groups were selected with P values less than 0 . 005 ., The functional analysis of gene sets based on gene ontology ( GO ) resources was performed using GOToolBox program ( http://genome . crg . es/GOToolBox ) with “Mouse Genome Informatics ( MGI ) ” and “Rat Genome Database ( RGD ) ” as the corresponding annotations respectively for distinct sets of genes ., Results are supplied in Table S1 ., The statistical significance of enrichment between gene group of interest and background gene group was calculated by applying the one-sided Fishers exact test ., PC12 cells were plated on a Poly-L-lysine-coated dishes ( Corning ) and maintained in DMEM/F12 ( Invitrogen ) with 5% FBS ( Biochrom ) , 5% horse serum ( Gibco ) and 1% penicillin/streptomycin at 37°C in 5% CO2 ., ChIP assays were carried out using materials from PC12 cells and performed as described previously 23 ., Briefly , cells were cross-linked with formaldehyde and sonicated to generate chromatin fragments size-enriched to between 200–600 bp ., Antibody against GATA3 ( 558686 , BD Pharmingen™ ) was used ., Chromatin from 20 million cells was used for each ChIP experiment , which yielded approximately 10 ng of DNA ., As input , 2% of sonicated chromatin was treated with proteinase K at 50°C for 2 hr and purified using the QIAquick PCR Purification Kit ( Qiagen Cat # 28106 ) ., Both input DNA and ChIP DNA fragments were blunt-ended , ligated to the Illumina adaptors , and sequenced with the Illumina Hiseq 2000 ., Sequencing reads of ChIP-seq were mapped to the rat genome ( Baylor 3 . 4/rn4 ) using Bowtie ( version 1 . 0 . 0 ) 24 , with the setting that sequence alignments can have no more than 3 mismatches ., Then MACS ( Model-based Analysis of ChIP-seq; version 1 . 4 . 2 ) 25 was used to identify Gata3 binding regions and peak summits which were further annotated by using CEAS 26 ., Two tools in Cistrome were deployed to calculate the correlation coefficient for our biological replicates and the PhastCons scores 27 ., De novo motif analysis was performed using MEME-ChIP version 4 . 9 . 0 28 after masking query sequences using RepeatMasker ( http://www . repeatmasker . org/ ) ., A gene was defined to be the target gene containing a binding site if this site is located between 10 kb upstream of transcription start site ( TSS ) and 3 kb downstream of transcription end site ( TES ) of this gene with the exception that the binding site on Th was found when we extended its promoter region to17 , 491 bp upstream of TSS ., The MACS output file about binding sites , together with the associated target genes , is provided in Table S4 ., We used two different custom-made siRNAs against Gata3 ., siGATA3-1 ( sense 5′-GUACUACAAACUCCACAAUTT-3′ and antisense 5′-AUUGUGGAGUUU GUAGUACTT-3′ ) , siGATA3-2 ( sense 5′-CCGUAAGAUGUCUAGCAAATT-3′ and antisense 5′-UUUGCUAGACAUCUUACGGTT3′ ) and negative control ( sense 5′-UUCUCCGAACGUGUCACGUTT-3′ and antisense 5′-ACGUGAC ACGUUCGGAGAATT ) were obtained from GenePharma ( Shanghai ) ., All siRNA experiments were conducted at a final concentration of 50 nM ., Transfections were conducted using Lipofectamine RNAiMAX ( Invitrogen ) ., Total RNA was isolated from cells using Trizol ( Invitrogen ) ., Purified mRNA was used to prepare the cDNA library as per the manufacturers instructions ., The short cDNA fragments were ligated to the Illumina sequencing adaptors and sequenced with the Illumina Hiseq 2000 ., Total RNA was isolated from cells to synthesize cDNA with SuperScript II Reverse Transcriptase ( Invitrogen ) ., qRT-PCR amplification mixtures ( 20 µl ) contained 3 µl water , 1 µM forward and reverse primer , 10 µl LightCycler 480 DNA SYBR Green I Master Mix buffer and 5 µl template cDNA ., All reactions were run on LightCycler 480 ( Roche ) ., All sequencing reads of RNA-seq were mapped to the rat genome using TopHat with default settings ( http://tophat . cbcb . umd . edu/; version 2 . 0 . 7 ) 29 ., The output data were analyzed by Cuffdiff to identify differentially expressed genes 30 ., The results were filtered by the criteria: “status”\u200a=\u200aOK and “P value”<0 . 05 ., Our ChIP-seq and RNA-seq data were submitted to ArrayExpress database with accession number: E-MTAB-2008 ., The original CEL files of GSE42565 from the Shh stimulation experiment 31 and GSE15942 performed in PC12 cells 32 were downloaded from Gene Expression Omnibus ( GEO ) ., The method “RMA” from R package “affy” was used to normalize the raw data ., For GSE42565 , students t-test was used to identify differentially expressed genes responding to the Shh stimulation ., 1677 genes were selected as the downstream genes of Shh on the basis that they had a P value <0 . 05 ., In this study , the expression patterns of 2074 genes , manually annotated based on ISH images for 78 regions in E11 . 5 mouse brain , were downloaded from Allen Brain Atlas ., We converted the textual annotation of ISH data to binary gene expression data of 0 and 1 ( Materials and Methods ) ., We observed that the genes coding for key signaling molecules , such as Fgf8 expressed in 9 regions , Shh in 31regions , Notch2 in 24 regions , Bmp1 in 12 regions and Bmp4 in 3 regions as well as critical developmental genes including En1 in 25 regions , En2 in 12 regions , Hes3 in 3 regions and Otx2 in 34 regions , showed restricted expression patterns at the E11 . 5 stage ., In particular , we found that the gene expression patterns of Shh and its receptor Ptch1 were clearly segregated along the ventral-dorsal axis , especially in the regions from midbrain to hindbrain ., Shh was highly expressed in the ventral brain region while Ptch1 was expressed just above the Shh-expressed region ., Shh occupied the entire floor plate while Ptch1 occupied the whole alar plate and most of the basal plate ( Figure 1A and 1B ) ., We used the expression patterns of Shh and Ptch1 to define two adjacent non-overlapping brain domains: a Shh+Ptch1− domain where Shh is expressed but Ptch1 is not expressed and a Shh−Ptch1+ domain , defined by the reciprocal pattern , where Shh is not expressed but Ptch1 is expressed ( Figure S1 ) ., Next , to identify the factors controlling the specification of these two domains , we searched for the genes specifically expressed in the Shh+Ptch1− and Shh−Ptch1+ domains respectively ., We identified 45 Shh+Ptch1−-pattern genes and 337 Shh−Ptch1+-pattern genes using Fishers exact test ( P<0 . 0001 ) ( Figure S2 ) ., We then cross-compared these two groups of genes with the genes specifically expressed in midbrain floor plate ( FP ) and ventral-lateral region ( VL ) of neural tissues obtained from an independent microarray study 33 ., We found that the Shh+Ptch1−-pattern genes were significantly enriched among the FP genes while the Shh−Ptch1+-pattern genes were enriched among the VL genes ( Figure S3A and S3B ) ., The consistency between the two datasets supported our method of defining the Shh+Ptch1−-pattern and Shh−Ptch1+-pattern genes based on ISH data ., Gene ontology ( GO ) enrichment analysis revealed that biological processes of system development , anatomical structure development and regulation of transcription were significantly enriched in both Shh+Ptch1−-pattern and Shh−Ptch1+-pattern genes , which highlighted their importance in early brain development ( Table S1 , Figure S3C and S3D ) ., To discover the potential transcriptional regulators for Shh+Ptch1−- and Shh−Ptch1+-pattern genes , we conducted promoter analysis for these two groups of genes ., Motif enrichment analysis showed that known TF binding motifs for GATA and GLI family TFs were significantly enriched in the promoters of Shh−Ptch1+-pattern genes ( p\u200a=\u200a5 . 95e–05 for GATA motif , p\u200a=\u200a6 . 82e–06 for GLI motif ) but not in Shh+Ptch1−-pattern genes ( Table S2 ) , indicating the importance of these two families of TFs in controlling the specification of the Shh−Ptch1+ domain ., Interestingly , GLI family TFs Gli1 and Gli2 , the downstream transducers of the Shh signaling pathway , are expressed in the Shh−Ptch1+ domain but not in the Shh+Ptch1− domain ., Furthermore , our promoter analysis predicted that both Gli1 and Gli2 directly target the GATA family member TF Gata3 , which is known to be a pioneer factor and strictly expressed in the Shh−Ptch1+ domain ., Therefore , it is likely that Shh secreted in the Shh+Ptch1− domain diffuses to the neighboring Shh−Ptch1+domain to exert its influence via the transcriptional activation of Gata3 ., In other words , Gata3 may determine the specification of the Shh−Ptch1+ domain via the Shh signaling pathway ., In contrast , in the Shh+Ptch1− domain , among all of the eight Shh+Ptch1−-pattern TFs annotated by ABA , Foxa1 and Foxa2 have been shown to function as master regulators to specify the identity of ventral midbrain progenitor cells by regulating Shh signaling 34 ., As shown in a published Foxa2 ChIP-seq dataset 20 , Twenty four out of the total of 45 identified Shh+Ptch1−-pattern genes , including Shh , were targeted by Foxa2 ., There is evidence that Shh is activated by Foxa2 while its downstream effectors , Ptch1 , Gli1 , Gli2 and Gli3 are all repressed by Foxa2 20 , 34 ., This would explain the absence of Ptch1 , Gli1 , Gli2 and Gli3 expression in the Shh+Ptch1− domain and suggests that Foxa2 plays a key role in determining the fate of the ventral Shh+Ptch1− domain ., These observations led us to propose that the Shh signaling pathway affects the pattern formation of the Shh+Ptch1− and the Shh−Ptch1+ domains in E11 . 5 mouse brain along the ventral-dorsal axis via the mutually exclusive expression of Foxa2 and Gata3 ., In total , we found 8 and 147 TF genes in the Shh+Ptch1−-pattern and Shh−Ptch1+-pattern , respectively ., Amongst the Shh−Ptch1+-pattern TFs , critical developmental genes such as Pax6 , Pax3 , Lhx1 , Irx3 , Isl , Ascl1 and Gata3 were found ., For these two groups of TFs , we were able to predict their regulatory targets within the two domain patterns by promoter analysis ., Some known regulatory relationships , such as Foxa1 targeting Foxa2 and Gli1/2 targeting Ptch1 , were correctly recapitulated by our promoter analysis 16 ., Five out of eight predicted targets of Foxa2 including Nfib , Aff3 , Foxa2 , Foxq1 and Nfia were supported by the Foxa2 ChIP-seq data 20 ., Notably , Foxa2 ChIP-seq data showed that Foxa2 targetsGata3 and our promoter analysis predicted that Gata3 targets Foxa2 ., Together with the non-overlapping expression patterns of Foxa2 and Gata3 ( Figure 1C and 1D ) , it seems to suggest a potential mutual inhibitory relationship between Foxa2 and Gata3 ., The role of Foxa2 in regulating the expression of Shh and other genes expressed in the Shh+Ptch1− domain has been previously characterized 20 ., Here , however , we investigated the functional role of Gata3 in mediating the Shh signaling pathway in the Shh−Ptch1+domain ., To this end , we sought a proper cell line that can mimic the gene expression pattern of this domain ., We therefore analyzed the expression of Shh−Ptch1+-pattern genes in published microarray data available in Gene Expression Omnibus ( GEO ) for neuron-like cell lines , including PC12 , neuro2a and N1E cells ., Shh−Ptch1+-pattern genes , when compared to other genes annotated by ABA , were only found to be significantly enriched among the highly expressed genes of the PC12 cells ( Fishers exact test , P value =\u200a0 . 00007 ) but not in neuro2a and N1E cells ., In particular , according to the microarray data as well as our Real-time PCR assay , Gata3 has high expression in PC12 cells while Foxa2 is not expressed ( Table S3 ) ., PC12 cells are able to synthesize noradrenaline 35 and have the properties of neurons in that their exposure to Neuron Growth Factor ( NGF ) causes them to stop dividing and begin to grow neurites similar to those of sympathetic neurons ., This neuron-like character makes this cell line a versatile model system for researches in neurobiology and neurochemistry 35 ., Therefore , we selected PC12 cells to perform ChIP-seq 36 analysis for Gata3 and specifically to identify its target genes ., Our ChIP-seq experiments included two biological replicates for ChIP and input materials respectively ., The high correlation ( Pearsons r\u200a=\u200a0 . 97 ) between the two ChIP replicates suggested that our ChIP experiments were highly reproducible ., After mapping all sequencing reads to the rat genome ( rn4 ) , we used the MACS program for peak calling , which yielded 1296 peaks with a default P value cutoff ( Table S4 ) ., De novo motif analysis of these binding regions by MEME-ChIP revealed a significantly enriched Gata3 motif ( Figure 2A ) ., The elevated average phastcon scores around the center of Gata3 binding sites suggested that Gata3 binding sites were more conserved compared with the neighboring regions , an indication of functional binding sites ( Figure 2B ) ., We used the CEAS program to examine the distribution of Gata3 binding sites across the genome ., We found that Gata3 binding sites were significantly enriched in the promoter regions with respect to the whole genome , i . e . 4 . 1% of ChIP regions fell within 1000 bp , 7 . 2% within 3000 bp and 13 . 6% within 10000 bp upstream of the transcription start site ( TSS ) of different genes ., Furthermore , 32 . 7% of Gata3 binding sites were located in the gene bodies compared to 26 . 3% in the genome background ., Among the binding sites in the gene bodies , 30 . 6% were within introns , 0 . 2% within the 3′UTRs and 0 . 7% within the 5′UTRs ( Figure 2C ) ., Using the gene annotation data of the rat rn4 genome downloaded from UCSC , we obtained 683 Gata3 target genes in PC12 cells ( Table S4 ) ., GO analysis showed that these Gata3 targets were involved in biological processes such as nervous system development , cell differentiation , and cell maturation ( Figure 2D ) ., While Foxa2 targets were significantly enriched in genes in both the Shh+Ptch1−- and Shh−Ptch1+-patterns , Gata3 targets identified in our study were only enriched in Shh−Ptch1+-pattern genes but not in Shh+Ptch1−-pattern genes , indicating that Gata3 mainly influences the Shh−Ptch1+domain ( Figure 3A and 3B ) ., The genes of eight Shh−Ptch1+-pattern TFs , including Abl1 , Cebpe , Gata2 , Isl2 , Myt1l , Nfib , Pou2f2 and Sox12 , were targeted by Gata3 , as shown by our ChIP-seq experiment ., Among Gata3 targets identified by ChIP-seq , we found two known regulators of the Shh signaling pathway , Sufu and Gsk3b ( Figure 4A ) ., Previous studies have shown that Sufu negatively regulates Shh signaling by direct interaction with Gli1 protein 37 and that Sufu is involved in Gli3 phosphorylation mediated by Gsk3b to induce the repression of Shh downstream genes 38 ., To further investigate the involvement of Gata3 in the Shh signaling pathway , we systematically searched for Shh downstream genes by analyzing a published microarray dataset ( GSE42565 ) on Shh stimulation in in vitro neural progenitors 31 ., We found 74 Gata3 ChIP-seq target genes among the downstream genes of Shh ., Among them , 45 genes including Slit2 and Slit3 were up-regulated by Shh stimulation ( Figure 4B ) and 29 genes were down-regulated by Shh ( Table S5 ) ., Sixteen out of the 45 Gata3 target genes up-regulated by Shh were annotated by ABA in E11 . 5 ISH data ., Eight of them were Shh−Ptch1+-pattern genes including Cotl1 , Foxn3 , Klhl29 , Limk1 , Mapt , Myt1l , Nfasc and Scg3 ( Figure 4C ) ., Nfasc is well-known as a cell adhesion molecule important for cell-cell communication and neurite outgrowth ., Nfasc also influences cell differentiation and maintenance in the brain but the signaling pathways upstream of Nfasc in the nervous system are unclear 39 ., Two of the genes in this set , Mapt and Limk1 , are known to be essential for brain development ., A Mapt mutation is associated with neurodegenerative disorders such as Alzheimer disease 40 , while the brain-specific Limk1 is implicated in axonal elongation 41 ., Our study indicated that the specification of these genes in the Shh−Ptch1+ domain is likely due to Gata3 regulation in the Shh signaling pathway ., Gata3 is known to control the synthesis of noradrenaline and serotonin 42 ., The ISH data for Th , Ddc , and Dbh , which are involved in noradrenaline synthesis , and Tph2 , which is involved in serotonin synthesis , support the idea that the Shh−Ptch1+domain includes brain regions that eventually develop into noradrenergic and serotonergic neurons ., Previous work has shown that a mutation in Gata3 reduced Th expression 43 ., In our ChIP-seq experiments , we found Gata3 binding sites located in the potential promoter region of Th ( Figure 4D ) ., Furthermore , we found that Gata3 regulated two other neurotransmitter-associated genes , Qdpr and Slc18a1 ( Figure 4D ) ., Qdpr is an enzyme involved in biosynthesis of tetrahydrobiopterin biosynthesis , which functions as a coenzyme in the reaction converting tyrosine to L-DOPA catalyzed by Th ., L-DOPA can further lead to the formation of neurotransmitters including dopamine , noradrenaline , and adrenaline . Slc18a1 is a vesicular transporter that transports neurotransmitters including dopamine , noradrenaline , adrenaline and serotonin into synaptic vesicles and which thus plays an important role in neurotransmitter release ., Functional disruption of Slc18a1 leads to neuropsychiatric diseases resulting from disorders of the corresponding neurotransmitter systems 44 ., Our discovery that Gata3 targets the promoters of Qdpr and Slc18a1 further supports Gata3s important role in neurotransmitter synthesis and release ., To further uncover the functional roles of Gata3 , we applied siRNAs to knockdown Gata3 in PC12 cells ., RNA-seq was performed in Gata3 siRNA-knockdown PC12 cells ., Comparing our RNA-seq data with a published microarray data in wild-type PC12 cells ( GSE accession: GSE15942 ) showed that they are highly correlated ( Spearmans Rho =\u200a0 . 77 , P value <2 . 2e–16 ) ., The gene expression values of two independent Gata3-knockdown samples with knockdown efficiencies of 50% and 51% respectively were also highly correlated ( Pearsons r\u200a=\u200a0 . 997; P value <2 . 2e–16 ) ., We integrated these results and obtained 1 , 121 differentially expressed genes compared to wild-type PC12 cells ., Among them , 731 that were down-regulated by Gata3-knockdown , including Gata3 itself , were enriched in Shh−Ptch1+-pattern genes ., By contrast , 390 up-regulated genes were not enriched in either the Shh−Ptch1+ or Shh+Ptch1− patterns ( Table S6 , Figure 3C and 3D ) ., This result supports our hypothesis that Gata3 preferentially up-regulates Shh−Ptch1+-pattern genes ., We then integrated the results of Gata3 ChIP-seq and Gata3 knockdown RNA-seq ., Seventy seven ( 77 ) differentially expressed genes in RNA-seq assays were directly targeted by Gata3 ., The RNA-seq analysis revealed that Slc18a1 was up-regulated after Gata3 knockdown ., Notably , the expression of two Gata3-targeted genes from ChIP-seq , Slit2 and Slit3 ( Figure 4B ) , together with Robo1 , were all down-regulated after Gata3-knockdown ., Other identified genes were further validated by our Real-time PCR analysis ( Table S3 ) ., SLIT/ROBO , functioning as a ligand/receptor signaling system , is involved in axon guidance and neuronal migration in the CNS ., Its special function in regulating axons to project across the midline has attracted a lot of attention 45 ., Furthermore , recent studies have suggested that the Slit2/Robo1 signaling might be enlisted for treating glioma because it can inhibit glioma cell migration 46 ., Earlier microarray analysis of Shh-induced expression also suggested that Slit2/3 were downstream genes of the Shh pathway ., Altogether , our results demonstrate that the SLIT/ROBO system was activated by Shh through the direct regulation of Gata3 in the Shh−Ptch1+ domain ., We next reconstructed a gene regulatory network downstream of the Shh signaling pathway in early mouse brain ., We downloaded all suitable ChIP-seq data from Gene Expression Omnibus ( GEO ) database or published papers 47–50 for Shh+Ptch1−-pattern TFs , including Foxa2 , Foxp1 , Phf19 , and Shh−Ptch1+-pattern TFs including Gli1 , Gata2 , Pbx1 , Sox11 and Ctnnb1 ( Table S7 ) ., Except for Foxp1 whose target genes were directly obtained from the original paper , as the raw data were not available , we downloaded the ChIP-seq data for all other TFs and annotated the target genes using the same procedure as our own Gata3 ChIP-seq data analysis ., In this network , only Shh+Ptch1−- and Shh−Ptch1+-pattern genes , as identified from our ISH data analysis , were included as potential target genes of the TFs ., The regulatory relationships between TFs and target genes were based on the result of ChIP-seq data ., Considering that many TFs can have both positive and negative regulatory functions , the TFs in one domain may target genes in the other domain as well ., The complete gene regulatory network is illustrated in Figure S4A ., As shown in this network , Gata3 was targeted by TFs Foxa2 and Phf19 ., Since the expression of Gata3 is mutually exclusive with Foxa2 and Phf19 , we propose that Gata3 is negatively regulated by Foxa2 and Phfl9 ., Similarly to Foxa2 , Phf19 is expressed only in the floor plate of the entire hindbrain at the E11 . 5 stage ., Studies showed that Phf19 , a subunit of the polycomb repressor complex 2 ( PRC2 ) , has essential functions in cellular differentiation and embryonic development , in binding to H3K36me3 and being associated with H3k36me3 histone demethylase NO66 , thereby mediating transcriptional silencing 49 , 51 ., We also found that Gli2 , Ptch1 and Foxa2 are all targeted by Ctnnb1 while Foxa2 targets Ctnnb1 ., Ctnnb1 encodes β-catenin , the signal transducer for the Wnt signaling pathway that is involved in early brain development 52 ., Our analysis suggests that Gli2 , Ptch1 and Foxa2 are downstream of this signaling pathway , reflecting its crosstalk with the Shh signaling pathway 53 ., Using the MCODE program , we identified two gene regulatory modules in our network ( Figure S4B and S4C ) ., In the first module , Dnmt3a , Nfasc and Mytl1 are co-regulated by both Gata3 and Pbx1 ( Figure S4B ) ., Pbx1 is expressed throughout the entire alar plate and basal plate in the E11 . 5 mouse brain ., It has been reported that embryos died at day E15/16 when Pbx1 was deleted , with developmental defects in multiple organs 54 ., In the second module , both Phf19 and Foxa2 target a total of 12 known genes that are enriched in the Shh−Ptch1+ domain , including Ptch1 , while Foxa2 is under the regulation of Phf19 and Foxa2 itself ( Figure S4C ) ., Furthermore , Pax3 in the Shh−Ptch1+ domain is targeted by Foxa2 in the Shh+Ptch1− domain , consistent with | Introduction, Materials and Methods, Results, Discussion | The Sonic hedgehog ( Shh ) signaling pathway is crucial for pattern formation in early central nervous system development ., By systematically analyzing high-throughput in situ hybridization data of E11 . 5 mouse brain , we found that Shh and its receptor Ptch1 define two adjacent mutually exclusive gene expression domains: Shh+Ptch1− and Shh−Ptch1+ ., These two domains are associated respectively with Foxa2 and Gata3 , two transcription factors that play key roles in specifying them ., Gata3 ChIP-seq experiments and RNA-seq assays on Gata3-knockdown cells revealed that Gata3 up-regulates the genes that are enriched in the Shh−Ptch1+ domain ., Important Gata3 targets include Slit2 and Slit3 , which are involved in the process of axon guidance , as well as Slc18a1 , Th and Qdpr , which are associated with neurotransmitter synthesis and release ., By contrast , Foxa2 both up-regulates the genes expressed in the Shh+Ptch1− domain and down-regulates the genes characteristic of the Shh−Ptch1+ domain ., From these and other data , we were able to reconstruct a gene regulatory network governing both domains ., Our work provides the first genome-wide characterization of the gene regulatory network involved in the Shh pathway that underlies pattern formation in the early mouse brain . | Recent large-scale projects of high-throughput in situ hybridization ( ISH ) have generated a wealth of spatiotemporal information on gene expression patterns in the early mouse brain ., We have developed a computational approach that combines publicly available high-throughput ISH data with our own experimental data to investigate gene regulation , involving signal molecules and transcription factors ( TFs ) , during early brain development ., The analysis indicates that two key TFs , Foxa2 and Gata3 , play antagonizing roles in the formation of two mutually exclusive domains established by the Sonic hedgehog signaling pathway in the developing mouse brain ., Further ChIP-seq and RNA-seq experiments support this hypothesis and have identified novel target genes of Gata3 , including the axon guidance regulators Slit2 and Slit3 as well as three neurotransmitter-associated genes , Slc18a1 , Th and Qdpr ., The findings have allowed us to reconstruct the gene regulatory network brought into play by the Sonic hedgehog pathway that mediates early mouse brain development . | developmental biology, biology and life sciences, gene regulatory networks, computational biology | null |
journal.pntd.0003901 | 2,015 | Interaction of Mean Temperature and Daily Fluctuation Influences Dengue Incidence in Dhaka, Bangladesh | Dengue virus ( DENV ) 1 transmission occurs in more than 100 countries; however , the burden of dengue is not evenly distributed ., Approximately half of the global population at risk of acquiring dengue infection resides in the South-East Asia Region of the World Health Organization 2 , a region characterized by strong seasonal weather variation and heavy monsoon rainfall ., This reflects the influence of local weather , particularly temperature and rainfall , on the transmission of DENV by Aedes mosquitoes ., Higher temperature , for example , shortens mosquito development time 3 , increases the frequency of blood feeding presumably by decreasing body size 4 , 5 , and reduces the extrinsic incubation period of DENV within mosquitoes 6 ., However , transmission of DENV is influenced not only by average temperature , but also by diurnal temperature range ( DTR , the difference between daily maximum and minimum temperature ) ., Temperature-dependent empirical and mathematical experiments show that temperature fluctuation influences vectorial capacity of Aedes aegypti , the principal mosquito vector of DENV , via biting rate , DENV transmission probability , extrinsic incubation period , and vector mortality rate 7–10 ., At high mean temperatures , vectorial capacity increases with narrow daily temperature variation 7–9 ., At low mean temperatures , the relationship between DTR and vectorial capacity is reversed 7–10 ., Temperatures above 30°C reduce survival of adult Ae ., aegypti 11 as does either very low or very high rainfall 12 ., The positive relationship between rainfall and dengue incidence has been observed in several locations 13–15 ., Seemingly paradoxical is the observation that the incidence of dengue increases in the dry season in some locations 16 ., Large scale climatic events , such as the Southern Oscillation , resulting from the interplay of large scale ocean and atmospheric circulation processes in the equatorial Pacific Ocean have been identified as a remote driver of inter-annual weather variability across the globe ., The warm and cold phases of the Southern Oscillation , El Niño and La Niña , respectively , are known to influence local temperature and rainfall and hence year-to-year variations in dengue incidence 13 , 17 , 18 ., Socio-demographic and economic factors also influence dengue incidence ., While the population at risk of dengue is likely to rise with increasing population , economic development would be expected to reduce risk 19 ., Bangladesh , a member country of the World Health Organization South-East Asia Region experienced its first epidemic of dengue fever in 2000 after more than three decades of sporadic dengue 20 ., Dengue is highly seasonal in Bangladesh with increased incidence during the monsoon ., From 2000 to 2009 , cases have been reported from 29 of the 64 Bangladeshi districts , with around 91 . 0% from the capital , Dhaka 21 ., Since 2010 very few cases have been notified from districts other than Dhaka 21 presumably because of a change in reporting criteria requiring confirmatory laboratory diagnosis ., Studies of dengue in Bangladesh before ours have not considered daily temperature variation 22 , 23 ., We present an analysis of the influence of daily temperature variation on the transmission of dengue adjusted for rainfall and population density , using a monthly dengue case time-series over 10 years from Dhaka ., We also considered anomalies in sea surface temperature ( SSTA ) , an index for El Niño-Southern Oscillation ( ENSO ) that is associated with extreme weather in Bangladesh and has not been included in other studies ., Analyses such as ours are critical for understanding the associations between weather , population , and dengue incidence and will allow the development of a reliable dengue early warning system ., The study was approved by The Australian National University Human Research Ethics Committee ., The national surveillance data of dengue fever cases was anonymized ., Dhaka district , comprising Dhaka Metropolitan area ( DMA ) and adjacent sub-districts , is a 1 , 464 km2 area near the center of Bangladesh ., Of the 64 districts this is the most densely populated , currently with 8 , 229 people per square kilometer ., Over the years 2001 to 2011 , there was a 41 . 0% increase in the population density of Dhaka 24 ., More than 37 . 0% of the population of DMA live in slums with a population density of 220 , 246 people per square kilometer 25 ., Slums have no access to piped water and temporary containers like drums and earthen jars are commonly used to store water in which Ae ., aegypti lays eggs 26 ., Inadequate supplies of piped water and an absence of proper waste management in most locations of Dhaka result in abundant indoor and outdoor mosquito breeding sites ., Both Ae ., aegypti and Aedes albopictus , the latter a secondary vector of dengue , were observed in Dhaka during the 2000 epidemic 27 ., Unscreened doors and windows permit mosquito entry to dwellings ., Dhaka has a hot and humid tropical climate , with an average temperature of approximately 25°C , which nearly always permits mosquito development and DENV transmission ., Rainfall is highly seasonal , with the wettest period ( June to September ) occurring during the warmest months ., About 80 . 0% of the annual rainfall of 2 , 000 mm falls during the monsoon ., Rainfall in Bangladesh is partly influenced by the Southern Oscillation with El Niño years usually associated with less than average monsoon rainfall while the opposite has been observed in La Niña years ., However , the influence of the Southern Oscillation on monsoon rainfall is not linear and is inconsistent , as observed in the moderate El Niño years causing flooding while some La Niña events during the monsoon preceded by El Niño are associated with reduced monsoon rainfall in Bangladesh 28 , 29 ., Monthly dengue cases for Dhaka district , from January 2000 to December 2009 , were obtained from the Directorate General of Health Services ., This time period was chosen to avoid the influence of the change in reporting practice started in 2010 ., The daily maximum , minimum , and mean temperatures ( °C ) , relative humidity ( % ) , and rainfall ( mm ) data for Dhaka were collected from the Bangladesh Meteorological Department ., A single missing value for maximum temperature was replaced by linear interpolation ., Diurnal temperature range was derived as the difference between maximum and minimum daily temperature ., Monthly means of these climatic variables were calculated from the daily records ., A monthly time series of SSTA over the Niño 3 . 4 region was obtained from the United States National Oceanic and Atmospheric Administration Climate Prediction Center ( http://www . cpc . ncep . noaa . gov/data/indices/ersst3b . nino . mth . 81-10 . ascii ) ., The Niño 3 . 4 index was used because of its correlation with Indian Ocean region monsoon rainfall ., An increase ( decrease ) of >0 . 5°C ( <-0 . 5°C ) in three-month moving average of SSTA is referred to as an El Niño ( a La Niña ) event ., Population estimates were extracted from the 1991 , 2001 , and 2011 census data ( there was no census taken between these years ) of the Bangladesh Bureau of Statistics ., Linear interpolation was used to calculate the monthly population for each of the years between 2000 and 2009 ., The population density ( people/km2 ) for Dhaka was estimated by dividing the district population size by the area ( km2 ) ., To examine temporal patterns over the study period , monthly dengue cases and climatic averages were plotted over the 10-year period ., To display seasonal patterns , monthly averages of mean temperature , DTR , and rainfall , and monthly numbers of total dengue cases over the 10 years were aggregated and plotted by month ., Overall correlation between dengue cases and climatic variables ( mean monthly temperature , mean monthly DTR , mean monthly relative humidity , mean monthly rainfall , and monthly SSTA ) were examined using Spearmans rank correlation test ., To avoid multicolinearity arising from correlated variables , the final set of candidate variables was restricted to those with pair-wise correlations of ≤0 . 8 ., Cross-correlation functions of dengue cases with each of the climatic variables were then estimated to investigate their lagged effects on dengue incidence ( p≤0 . 05 ) ., Time lags were included to account for the influence of climatic variables on the development , maturation , and survival of the vector ( Aedes mosquitoes ) as well as the extrinsic incubation period of DENV in the vector and the intrinsic incubation period in the human host ., Lags of up to three months were considered for all weather variables , with SSTA also considered at a lag of four months ., The counts of dengue cases were then fitted by a generalized linear model ( GLM ) with negative binomial distribution to allow for overdispersion in dengue counts ., The population of Dhaka was added as an offset to the model on a logarithmic scale to adjust for population size ., Population density was also included in the model to account for the potential influences of associated socio-demographic changes on dengue transmission in Dhaka ., An indicator variable for outbreak months was added to prevent occasional extreme values from distorting the analyses ., A month with the number of dengue cases exceeding the 10-year mean plus two standard deviations was defined as an outbreak month ., To account for the long term trend in dengue incidence over time , an indicator variable for year was incorporated in the model ., An autoregressive term at order 1 was also included to allow for autocorrelation in monthly numbers of dengue cases ., To determine whether seasonal variation had any influence on dengue incidence , a categorical variable for winter ( December–February ) , pre-monsoon ( March–May ) , monsoon ( June–September ) , and post-monsoon ( October–November ) was also considered ., The analyses were performed using STATA 13 . 1 ( StataCorp . , Texas , USA ) and figures were drawn using RStudio ( R development Core Team , 2015 ) ., Inter-annual and seasonal variations for dengue and weather over the period 2000–2009 are presented ( Figs 1 and 2 ) ., The number of dengue cases during winter is low and starts to increase from June ( Fig 2 ) with the advent of the monsoon with considerable annual variation ( Fig 1 ) ., The peak comes one month after the initial rainfall peak in July and starts declining afterwards ., Temperature reaches its peak in April and plateaus until October when it drops ( Fig 2 ) ., Because of the high correlation with mean temperature and DTR , relative humidity was excluded at the initial stage of model formulation ., Consideration of both temperature and rainfall was , however , expected to minimize the potential confounding effect of relative humidity on dengue incidence ., The categorical variable for season was also subsequently removed because it did not improve model fit ., Therefore , the model finally fitted is as follows:, yt~NegBin ( μt , θ ), log ( μt ) =α+∑j=03β1jTjt+∑j=03β2jDTRjt+∑j=03β3j ( Tjt×DTRjt ) +∑j=03β4jRjt+∑j=04β5jSSTAjt+β6Popdent+outbreak+year+yt−1+log ( Population ) +εt, ( 1 ), where yt is the dengue count in Dhaka in month t ( t = 1 , … , 120 ) ; μt is the corresponding mean dengue count; T , DTR , R , and SSTA are the mean monthly temperature ( °C ) , mean monthly diurnal temperature range ( °C ) , mean monthly rainfall ( mm ) , and monthly sea surface temperature anomaly respectively; ( T×DTR ) represents the interaction between mean monthly temperature and mean monthly DTR; j = 0 , … , 4 represent the time lag periods in months; outbreak is the categorical variable for outbreak months; year represents time trend; yt-1 is the dengue count of previous month; and εt is the error term ., Table 1 shows estimates of the significant covariates from model ( 1 ) ., Mean temperature , DTR , and the interaction between these two variables are all significant predictors of dengue incidence at a lag of one month ., However , the opposing directions of main and interaction effects indicate a negative synergy between mean temperature and DTR ., Therefore , dengue incidence increases with higher temperature and lower DTR or lower temperature and higher DTR in the previous month but decreases when both are either high or low ., Rainfall at lag one and two months was found to be positively associated with dengue incidence , suggesting that increased incidence of dengue in a given month is associated with higher rainfall during the previous two months ., The negative effect of SSTA on dengue incidence at lag zero month indicates that the incidence goes up with increasing negative values of the SSTA in the current month , while the inverse relationship was observed at lag of one month ., Increasing population density , as anticipated , increases dengue incidence ., To investigate how SSTA influences climatic anomalies in Dhaka , standardized anomalies of temperature , relative humidity ( S1 Fig ) , and rainfall were calculated and plotted with SSTA over the study period ( Fig 3 ) ., Simple linear regression of temperature , relative humidity ( S1 Fig ) , and rainfall anomaly on SSTA at lag of zero and one month revealed a weak negative correlation between rainfall and SSTA ( Fig 4 ) even though the relationship is not temporally consistent ( Fig 3 ) presumably due to a non-linear relationship between them ., Model diagnostics were performed as follows ., Firstly , a model was run without the interaction terms and compared with model ( 1 ) ., The likelihood ratio test confirmed that the addition of interaction terms resulted in a significantly improved fit compared to the model without interactions ( p<0 . 000 ) ., The Pearson dispersion statistic ( 0 . 98 ) also provided evidence for the goodness-of-fit of the model ( 1 ) ., Secondly , residual analyses were performed to ensure that the model provided an adequate fit to the data ., Serial autocorrelation of the residuals was checked by examining a time plot and a partial autocorrelation plot of the residuals ( S2 and S3 Figs ) ., In addition , observed vs fitted plot of dengue cases was examined ( S4 Fig ) ., It is well established that temperature influences vector and virus biology and therefore dengue transmission ., Monthly changes in average temperature have been reported to be positively associated with dengue transmission in Puerto Rico 30 ., In addition to average temperature , temperature fluctuations also have an impact ., Large fluctuation around warmer temperature reduces transmission whereas around cooler temperature this speeds up the process and vice versa 7 , 8 , 10 ., However , studies of climate and dengue usually ignore diurnal temperature variation ., We found that dengue incidence in Dhaka was significantly influenced by mean temperature , DTR , and their synergistic effect , after adjusting for rainfall , anomalies in sea surface temperature , population density , autoregression and the long term temporal trend in dengue incidence ., Although mean temperature and DTR were positively associated with dengue incidence , the opposing direction of their interaction term suggested a negative synergy between these two variables ., This indicates that although increased mean temperature and reduced DTR or reduced mean temperature and increased DTR increase dengue incidence one month later , an increase or decrease in both lessen dengue incidence ., This is consistent with studies showing a positive association between DTR and dengue at low temperatures and a negative association at high temperatures 7 , 8 , 10 ., Use of mean temperature alone in predicting dengue outbreaks will therefore fail to capture the full complexity of the relationship between temperature and dengue transmission ., We demonstrated that increased incidence of dengue in Dhaka was associated with an increase in rainfall in the previous two months ., However , an earlier study in Dhaka identified a significant positive association only at lag of two months 22 ., The effect of rainfall on Ae ., aegypti breeding is lessened by the species’ egg laying in artificial containers filled with water by humans ., But Ae ., albopictus has also been found in Dhaka 31 ., Its dependence on rain-fed outdoor artificial containers as larval habitats might explain the positive association between rainfall and dengue incidence ., Such a relationship has also been reported in other countries 13 , 32 ., In Puerto Rico , rainfall has been proposed to have caused increases in dengue incidence by increasing Ae ., aegypti density , egg laying in water storage containers and discarded tires 33 ., In Thailand , monthly dengue incidence and epidemics of dengue have been associated with ENSO , which is believed to cause changes in temperature and relative humidity 34 ., At time lags of one to 11 months , both epidemics and monthly cases are correlated with El Niño , which is associated with higher temperature and in some places with lower relative humidity 34 ., A multivariate ENSO index , lagged at one to six months , alone explains a maximum 22% of the variations in monthly dengue cases 34 ., An increase in the number of dengue cases following El Niño was also observed in Mexico , French Guiana , Indonesia , Colombia , and Surinam 13 , 18 ., The role of ENSO in the inter-annual variability of monsoon rainfall in Bangladesh has been examined demonstrating that El Niño is generally associated with lower rainfall , whereas La Niña and sometimes moderate El Niño generate higher rainfall 35 ., However , the relationship is not consistent over time and ENSO is only partially responsible for the rainfall anomalies in Bangladesh ., Our study found a negative effect of SSTA on current dengue incidence together with a positive effect at a lag of one month ., Possible explanations for the negative association with current SSTA could be that the dry weather resulting from a strong El Niño or the heavy rainfall associated with a moderate El Niño both reduce adult mosquito survival 11 , 12 and thereby reduce DENV transmission ., Heavy rainfall , on the other hand , could increase transmission because people do not cover themselves in the post-rainfall humid weather resulting in increased human-mosquito contact ., The positive effect of SSTA on dengue incidence at a lag of one month is biologically plausible because moderate rainfall is needed for mosquito development , and is also consistent with our findings of a positive influence of rainfall on dengue transmission at a lag of one month ., However , heavy rainfall washes away mosquito larvae reducing vector numbers thereby transmission in the following month ., Consideration of the non-linear influence of ENSO on rainfall may provide a richer insight into the relationship between dengue and SSTA ., Socio-demographic and economic factors , as well as climate , powerfully influence dengue incidence ., A study projects the population at risk of dengue in 2050 under global climate change considering gross domestic product per capita ( GDPpc ) as an indicator of socio-economic development 19 ., The study reports 5 . 0% and 4 . 0% increases in the population at risk of dengue in 2050 compared to the baseline risk population in 2000 considering only the projected increase in population and the projected changes in both climate and GDPpc , respectively ., Positive but non-significant effects of population growth on dengue cases have also been reported in Mexico 13 ., In our study in Dhaka population density was used as a proxy for socio-demographic factors and was found to be positively associated with dengue incidence ., The strength of the present study is that we considered both small and large-scale climatic influences on dengue incidence along with the interaction between mean temperature and DTR and included population density in the model as a proxy for socio-demographic changes over time ., However , while we demonstrated significant associations between temperature and rainfall with dengue transmission we did not model non-linear relationships , and we excluded relative humidity from our model due to its strong correlation with mean temperature and DTR ., We used months as our temporal unit of study because daily data on dengue incidence were not available ., As a consequence , short-scale influences of climatic parameters on dengue incidence may not be fully captured by our model , and lag effects cannot be determined at a fine time-scale ., Another limitation of the model used here is that it did not allow for under-reporting from passive surveillance data or possible changes in the rate of under-reporting ., However , inclusion of a temporal trend variable in the model may indirectly capture variation in the rate of under-reporting ., In conclusion , our findings indicate that the association between weather and dengue transmission is complex , which is further confounded by socio-demographic factors like population density ., Models designed for forecasting should account for this complexity in order to minimize the risk of overestimation in relation to increasing mean temperature , thereby optimizing resource allocation in tropical overpopulated countries with limited resources . | Introduction, Materials and Methods, Results, Discussion | Local weather influences the transmission of the dengue virus ., Most studies analyzing the relationship between dengue and climate are based on relatively coarse aggregate measures such as mean temperature ., Here , we include both mean temperature and daily fluctuations in temperature in modelling dengue transmission in Dhaka , the capital of Bangladesh ., We used a negative binomial generalized linear model , adjusted for rainfall , anomalies in sea surface temperature ( an index for El Niño-Southern Oscillation ) , population density , the number of dengue cases in the previous month , and the long term temporal trend in dengue incidence ., In addition to the significant associations of mean temperature and temperature fluctuation with dengue incidence , we found interaction of mean and temperature fluctuation significantly influences disease transmission at a lag of one month ., High mean temperature with low fluctuation increases dengue incidence one month later ., Besides temperature , dengue incidence was also influenced by sea surface temperature anomalies in the current and previous month , presumably as a consequence of concomitant anomalies in the annual rainfall cycle ., Population density exerted a significant positive influence on dengue incidence indicating increasing risk of dengue in over-populated Dhaka ., Understanding these complex relationships between climate , population , and dengue incidence will help inform outbreak prediction and control . | The sensitivity of mosquito vector and dengue virus biology to diurnal temperature variability has been established , but this study is the first analyzing these relations with dengue occurrence ., We show that Dhaka’s tropical hot monsoon climate and small variation in daily temperature enhance dengue transmission one month later ., Large-scale climatic events like El Niño-Southern Oscillation and increasing population density of Dhaka also increase incidence ., Our results therefore enable us to accurately estimate dengue transmission dynamics in densely populated areas that are also vulnerable to global warming by considering diurnal variability ., Our approach reduces the chance of overestimating the effect of increasing temperature on dengue transmission intensity with the ultimate goal of outbreak prediction and control . | null | null |
journal.pntd.0007053 | 2,018 | A longitudinal systems immunologic investigation of acute Zika virus infection in an individual infected while traveling to Caracas, Venezuela | Zika virus ( ZIKV ) is an emerging arthropod-borne flavivirus ., It is primarily transmitted by Aedes sp ., mosquitos but can also be transmitted person to person vertically from mother to child , sexually and in blood during transfusions 1 ., Clinical manifestations occur in approximately 20% of infections and can include an acute onset low grade fever , pruritic erythematous macular papular rash , arthralgias and conjunctivitis 2 ., Clinically these symptoms can be confused with dengue virus ( DENV ) or chikungunya virus ( CHIKV ) infections that are transmitted by the same mosquito vectors and can co-circulate with ZIKV 3 ., During pregnancy , ZIKV can cause congenital Zika syndrome and other severe birth defects in fetuses 2 ., In adults , ZIKV is associated with life-threatening Guillain-Barré Syndrome ( GBS ) 4 , 5 ., The details of how ZIKV bypasses immune restriction to cause disease are still under investigation ., The relationship between flaviviruses and the immune system is complex 6 ., On one hand , the immune system can exacerbate viral pathogenesis ., For example , ZIKV , like DENV and West Nile virus ( WNV ) , infect innate immune white blood cells early in infection 7–11 ., Studies in ZIKV infected children identified monocytes , in particular CD14+CD16+ intermediate monocytes , and myeloid dendritic cells as the main targets of ZIKV infection in peripheral blood mononuclear cells ( PBMCs ) 9 ., These infected cells may act like a “Trojan horse” to increase spread of the virus to different tissue compartments ., Antibody ( Ab ) responses to flaviviruses are often cross-reactive and have the potential to mediate antibody-dependent enhancement ( ADE ) ., While there is no evidence that ADE alters ZIKV pathogenesis in humans , in a mouse model of ZIKV infection , administration of DENV or WNV convalescent plasma increased ZIKV morbidity and mortality through ADE 12 ., On the other hand , the development of protective adaptive immune responses is thought to be critical to clear ZIKV infection 6 ., Therefore , increasing our understanding of human immune responses to ZIKV infection can lead to better understanding of ZIKV clinical manifestations and pathogenesis and inform the development of vaccines ., Only a small number of studies have examined human responses to ZIKV infection in vivo ., Analysis of serum inflammatory markers during acute ZIKV infection identified some potential biomarkers associated with neurologic complications 13 and viremia plus moderate symptoms 14 ., Monoclonal Abs isolated from four donors infected with ZIKV demonstrated that neutralizing Abs primarily recognized the envelope protein domain III of ZIKV and that Abs recognizing different ZIKV epitopes could alternatively protect against ZIKV challenge or enhance subsequent DENV infection in mice 15 ., Another study tracking the development of Ab responses to ZIKV in three DENV-experienced and one DENV-naïve individual found that acute-phase Abs developing during ZIKV infection in DENV-experienced individuals were highly cross-reactive but poorly neutralizing 16 ., In a single flavivirus naïve individual , anti-ZIKV B-cell plasma neutralization activity and T-cell responses peaked later between day 15 and day 21 17 ., A large study examining T cell responses to ZIKV in DENV-naïve and DENV-immune patients revealed that DENV exposure prior to ZIKV infection influences the timing , magnitude , and quality of the T cell response 18 ., In another study that examined both innate and adaptive immune responses in 5 individuals infected with ZIKV , Lai et . al . observed that flavivirus-experienced individuals developed rapid cross-reactive antibody responses against both DENV and ZIKV as well as activated CD8+ T cell responses , albeit few ZIKV-specific CD8+ T cells were identified 19 ., These studies provide insight into human ZIKV infection , but our understanding remains limited due to the small number of reported cases ., Additionally , published reports have utilized conventional approaches to study the in vivo immune responses to ZIKV ., Combining these approaches with genome-wide next-generation sequencing ( NGS ) analyses could bring new insight into human ZIKV responses and inform direction and design of future studies of immune responses during infection in larger cohorts ., As a step towards improving our understanding of human immune responses to acute ZIKV infection through new approaches , we present a detailed immunologic characterization of the innate and adaptive temporal and cell type-specific responses to an acute ZIKV infection in a DENV-experienced patient ., This research study was approved by the UCSD IRB with Human Research Protections Program # 161060 ., Written informed consent was obtained from the adult human subject described in this report ., After obtaining written informed consent , blood was collected on five occasions d3 , d6 , d17 , d48 , and d240 post-onset of symptoms ( POS ) ., Urine was collected on d3 and d6 only ., Serum was isolated by collecting blood into a plain tube containing no anticoagulant , allowed to clot at room temperature for 20 minutes followed by centrifugation at 1500xg for 10 minutes in a refrigerated centrifuge ., Serum was frozen in single use aliquots at -80°C ., Peripheral blood mononuclear cells ( PBMCs ) were isolated from heparinized blood using Histopaque-1077 per manufacturers instructions and subjected to flow activated cell sorting ( FACS ) or cryopreserved in 5 million cell aliquots in 90% FBS + 10% DMSO ( Hybri-max Sigma ) using a Nalgene Mr . Frosty at -80°C for 24 hours before transfer to liquid nitrogen ., Cryopreserved cells were thawed rapidly to 37°C and slowly diluted with pre-warmed growth media , followed by gentle pelleting and resuspension in cold FACS staining buffer ., Five microliters of d3 POS serum or blood was inoculated into a T25 flask of C6/36 mosquito ( Aedes albopictus ) cells ., Supernatants ( 5 mL ) were harvested seven days after culture and titrated via BHK-21 cell-based focus forming assay ( FFA ) and anti-Flavivirus envelope ( E ) protein antibody clone 4G2 ., The urine culture supernatant had a titer of 2 . 0 x 104 focus forming units ( FFU ) /mL ., Infectious virus in the serum culture supernatant was undetectable ., Viral RNA from 0 . 2ml of C6/36 supernatant that was inoculated with d3 POS urine was extracted using the Roche High Pure Viral RNA Kit ( Roche ) and reversed transcribed using a primer specific method for ZikaBr ( Forward primer AGTGGAGACGATTGYTGTNGT , Reverse primer AACATGTCTTCTGTGGTCATCCA ) ( SuperScript III First-Strand Synthesis System for RT-PCR , Invitrogen ) ., cDNA was amplified using Taq polymerase ( Roche ) , cleaned using QIAquick PCR Purification Kit ( Qiagen ) and sequenced using BDT v3 . 1 on the ABI 3130xl Genetic Analyzer ., Forward and reverse sequences were used to make a contig and manually edited using Bioedit ref http://www . mbio . ncsu . edu/BioEdit/bioedit . html ., The Basic Alignment Search Tool ( BLAST ) ref: was then used with the resultant sequence ref: https://blast . ncbi . nlm . nih . gov/Blast . cgi ? PROGRAM=blastn&PAGE_TYPE=BlastSearch&LINK_LOC=blasthome which most closely aligned with other ZIKV NS5 sequences ., For phylogenetic analyses , RNA from ZIKV SD001 infected primary human macrophages were aligned to the human hg19 genome using STAR PMID: 23104886 ., Any unmapped reads were used as input for strand-specific de novo transcriptome assembly with Trinity PMID: 21572440 ., The longest assembled transcripts were approximately 9 kb , and corresponded to near full-length viral genomes ., The resulting alignment from ZIKV SD001 and 435 publicly available ZIKV sequences from NCBI viral genomes resource 20 were used to perform an approximate maximum likelihood phylogenetic tree with PhyML 21 ., The tree was rooted with ZIKV ( GenBank accession number KY241712 ) isolated in Asia ., For innate immune cell sorting ten million PBMCs were stained with antibodies against CD3 PE-Cy7 , CD19 PE-Cy7 CD20 PE-Cy7 , HLADR BV421 , CD11c AF700 , CD123 PE , CD14 AF488 , CD16 APC , CD56 APC-Cy7 , and Zombie Aqua Fixable viability dye and separated as shown ., For T cell sorting , five million cryopreserved PBMCs were stained with CD16 BV510 , CD56 BV510 , CD4 APC-eFluor780 , CD3 AF700 , CD8 BV785 , CD45RA BV570 , CCR7 PE-Cy7 , CXCR5 BV421 , CXCR3 BV605 , TCR V_24-J_18 BV711 , CD226 BB515 , CCR6 PerCP-Cy5 . 5 , CCR4 PE , CD25 PE-Dazzle 594 , and CD127 AF647 and sorted into CD3+ T cell CD4+ and CD8+ populations ., T cells were further analyzed for effector or memory phenotypes , CD4 T helper ( Th ) subsets based on the expression of chemokine receptors ( Th1: CCR6-CCR4-CXCR3+; Th2: CCR6-CCR4+CXCR3-; Th1/17: CCR6+CCR4-CXCR3+; and Th17: CCR6+CCR4+CXCR3- ) as well as the cytotoxicity marker CD226 ., Stained PBMCs were sorted in the La Jolla Institute ( LJI ) Flow Cytometry Core Facility on a FACSAria Fusion sorter ., Sequencing libraries were prepared using a low input RNA-seq prepared according to the Smart-seq2 method 22 with some modifications ., 5000–15 , 000 PBMCs ( pre-sort ) or FACS isolated cell populations were lysed in TRIzol and RNA extracted using Direct-zol RNA Microprep ( Zymo ) with on-column DNAseI treatment ., 10 μL purified RNA was mixed with 5 . 5 μL of SMARTScribe 5X First-Strand Buffer ( Clontech ) , 1 μL polyT-RT primer ( 2 . 5 μM , 5’-AAGCAGTGGTATCAACGCAGAGTAC ( T30 ) VN , 0 . 5 μL SUPERase-IN ( Ambion ) , 4 μL dNTP mix ( 10 mM , Invitrogen ) , 0 . 5 μL DTT ( 20 mM , Clontech ) and 2 μL Betaine solution ( 5 M , Sigma ) , incubated 50°C 3 min . 3 . 9 μL of first strand mix , containing 0 . 2 μL 1% Tween-20 , 0 . 32 μL MgCl2 ( 500 mM ) , 0 . 88 μL Betaine solution ( 5 M , Sigma ) , 0 . 5 μL ( 5 M , Sigma ) SUPERase-IN ( Ambion ) and 2 μL SMARTScribe Reverse Transcriptase ( 100 U/μL Clontech ) was added and incubated one cycle 25°C 3 min . , 42°C 60 min . 1 . 62 μL template switch ( TS ) reaction mix containing 0 . 8 μL biotin-TS oligo ( 10 μM , Biotin-5’-AAGCAGTGGTATCAACGCAGAGTACATrGrG+G-3’ ) , 0 . 5 μL SMARTScribe Reverse Transcriptase ( 100 U/μL Clontech ) and 0 . 32 μL SMARTScribe 5X First-Strand Buffer ( Clontech ) was added , then incubated at 50°C 2 min . , 42°C 80 min . , 70°C 10 min . 14 . 8 μL second strand synthesis , pre-amplification mix containing 1 μL pre-amp oligo ( 10 μM , 5’AAGCAGTGGTATCAACGCAGAGT-3’ ) , 8 . 8 μL KAPA HiFi Fidelity Buffer ( 5X , KAPA Biosystems ) , 3 . 5 μL dNTP mix ( 10 mM , Invitrogen ) and 1 . 5 μL KAPA HiFi HotStart DNA Polymerase ( 1U/μL , KAPA Biosystems ) , was added , then amplified by PCR: 95°C 3 min . , 5 cycles 98°C 20 sec , 67°C 15 sec and 72°C 6 min , final extension 72°C 5 min ., The synthesized dsDNA was purified using Sera-Mag Speedbeads ( Thermo Fisher Scientific ) with final 8 . 4% PEG8000 , 1 . 1M NaCl , then eluted with 13 μL UltraPure water ( Invitrogen ) ., The product was quantified by Qubit dsDNA High Sensitivity Assay Kit ( Invitrogen ) and libraries were prepared using the Nextera DNA Sample Preparation kit ( Illumina ) ., Tagmentation mix containing 11 μL 2X Tagment DNA Buffer and 1 μL Tagment DNA Enzyme was added to 10 μL purified DNA , then incubated at 55°C 15 min . 6 μL Nextera Resuspension Buffer ( Illumina ) was added and incubated at room temperature for 5 min ., Tagmented DNA was purified using Sera-Mag Speedbeads ( Thermo Fisher Scientific ) with final 7 . 8% PEG8000 , 0 . 98M NaCl , then eluted with 25 µL UltraPure water ( Invitrogen ) ., Final enrichment amplification was performed with Nextera primers , adding 1 μL Index 1 primers ( 100 μM , N7xx ) , 1 μL Index 2 primers ( 100 μM , N5xx ) and 27 μL NEBNext High-Fidelity 2X PCR Master Mix ( New England BioLabs ) , then amplified by PCR: 72°C 5 min . , 98°C 30 sec . , 6–12 cycles 98°C 10 seconds , 63°C 30 sec . , and 72°C 1 min ., Libraries were size selected , quantified using the Qubit dsDNA HS Assay Kit ( Thermo Fisher Scientific ) , pooled and sequenced on a Hi-Seq 2000 sequencer using single-end 50bp reads at a depth of 25 to 30 million single end reads per sample ., 50 , 000 FACS isolated classical monocytes or NK cells were lysed in 50 μl lysis buffer ( 10 mM Tris-HCl ph 7 . 5 , 10 mM NaCl , 3 mM MgCl2 , 0 . 1% IGEPAL , CA-630 , in water ) on ice and nuclei were pelleted by centrifugation at 500 RCF for 10 min ., Nuclei were then resuspended in 50 μl transposase reaction mix ( 1x Tagment DNA buffer ( Illumina 15027866 ) , 2 . 5 μl Tagment DNA enzyme I ( Illumina 15027865 ) , in water ) and incubated at 37°C for 30 min on a PCR cycler ., DNA was then purified with Zymo ChIP DNA concentrator columns ( Zymo Research D5205 ) and eluted with 10 μl of elution buffer ., DNA was then amplified with PCR mix ( 1 . 25 μM Nextera primer 1 , 1 . 25 μM Nextera index primer 2-bar code , 0 . 6x SYBR Green I ( Life Technologies , S7563 ) , 1x NEBNext High-Fidelity 2x PCR MasterMix , ( NEBM0541 ) ) for 8–12 cycles , size selected for fragments ( 160–290 bp ) by gel extraction ( 10% TBE gels , Life Technologies EC62752BOX ) and single-end sequenced for 51 cycles on a HiSeq 4000 or NextSeq 500 ., RNA-seq reads were aligned to the GRCh38/hg38 assembly of the human genome using STAR ( version 2 . 5 . 2a ) using default parameters 23 ., Gene expression values were calculated as fragments per kilobase per million mapped reads ( FPKM ) across GENCODE transcript exons ( release 24 ) 24 using HOMER 25 ., To remove possible contamination from genomic DNA in the RNA-seq samples , FPKM measurements were calculated for long introns ( >10 kb ) and the median intron FPKM per experiment was subtracted from each exon FPKM values to remove background signal ., Gene expression FPKM values across all samples set to a minimum of zero and then quantile normalized ., Only GENCODE transcripts with length greater than 300 bp were considered ., Log2 fold change ratios were calculated using a pseudo count by adding a FPKM of 4 to both numerator ( i . e . day 3 , 6 , 17 ) and denominator ( i . e . day 48/convalescent ) to reduce the impact of low expression noise and contamination on the lists of regulated genes ., Functional enrichment was performed using HOMER using pathway definitions from Gene Ontology and HALLMARK pathways from MSigDB 26 ., Promoter known motif enrichment was calculated using HOMER using sequence from -300 bp to +50 bp relative to annotated transcription start sites ., Hierarchical clustering of correlated gene expression profiles , motif enrichment , and GO/pathway function enrichment values were performed using Cluster 3 . 0 27 and visualized using Java TreeView 28 ., For ATAC-seq , fastq files were trimmed and aligned to hg38 using bowtie2 ., Reads mapping to Mitochondrial DNA were removed and PCR duplicates were removed ., Peaks were called using a standardized peak size using HOMER ( 300 bp ) ., To compare classical monocytes and NK cells the appropriate peak files were merged and differential peaks identified using getDifferentialPeaks command ( HOMER ) with threshold of fold change >3 and P-value < 0 . 001 ., Motif analysis was performed on differential peak files using findMotifsGenome . pl ( HOMER ) ., All human RNA-seq and ATAC-seq data described in this manuscript are available at the National Center for Biotechnology Information ( NCBI ) Gene Expression Omnibus ( GEO ) accession number GSE123541 ., Affymetrix gene expression microarray CEL files were downloaded from NCBI GEO for longitudinal DENV infection in humans ( GSE43777 ) and ZIKV infection in Rhesus macaques ( GSE93861 ) and processed into gene expression values using R/Bioconductor using GCRMA with default options ., For the human DENV infection data , only samples performed on whole genome HG-U133plus2 microarrays were used for the comparison ., Samples for the human DENV study were identified based on their annotated number of days since initial fever ( G1 , G2 , etc . ) and averaged to generate per day expression values ., Rhesus macaque ZIKV infection gene expression values were averaged based on the day post infection , and human orthologues were assigned using one-to-one orthologues defined by ENSEMBL BIOMART ( https://www . ensembl . org/biomart ) ., For each study , log2 activation ratios were calculated using the average expression for each day compared to the average of the convalescent samples ( human ) or pre-infection samples ( Rhesus ) ., Microarray and RNA-seq activation ratios were compared by linking the datasets using gene symbols , using data from the highest expressed isoform in the cases where multiple isoforms exist per gene ., Flow cytometry-based neutralization assay was used to evaluate SD001 serum neutralization of ZIKV ( strains FSS13025 and SD001 29 ) and DENV ( DENV1 strain West pacific 74 and DENV4 strain TVP-360 ) in vitro ., 2×104 FFU DENV or ZIKV were incubated with or without serial 3-fold dilutions ( starting at 1:10 ) of heat-inactivated SD001 serum in 96-well round bottom plates for 1-hour at 37°C ., U937 cells stably expressing DC-SIGN ( 1x105 ) were seeded in each well and incubated for 2 h at 37°C with occasional rocking ., After incubation , the plates were centrifuged for 5 minutes at 1500 rpm , supernatants aspirated and fresh medium added followed by incubation for 16 h at 37°C ., U937 cells were then fixed , permeabilized , stained with anti-CD209 PE and 4G2 FITC ( to detect ZIKV ) or 2H2 FITC ( to detect DENV ) and analyzed using an LSRII ., Percent inhibition was calculated by determining the relative infection in virus incubated with serial diluted patient serum ( tests ) versus no serum ( control ) ., Best fit curves and neutralizing titer 50 ( NT50 ) were determined using Prism 7 . 0 ( GraphPad ) ., Serum from 8 months prior to infection ( pre-infection ) as well as d3 , d6 , d17 , and d48 POS were prepared in duplicate using the Bio-Plex Pro Human Cytokine 27-plex Assay ( Bio-rad #M500KCAF0Y ) per manufacturers protocol and read using a Luminex machine ., Cytokine concentrations were calculated from standard curves generated using references included in the kit ., The following cytokines were measured FGF basic , Eotaxin , G-CSF , GM-CSF , IFN-γ , IL-1β , IL-1ra , IL-2 , IL-4 , IL-5 , IL-6 , IL-7 , IL-8 , IL-9 , IL-10 , IL-12 ( p70 ) , IL-13 , IL-15 , IL-17 , IP-10 , MCP-1 ( MCAF ) , MIP-1α , MIP-1β , PDGF-BB , RANTES , TNF-α , and VEGF ., A middle-aged , previously healthy , dengue virus ( DENV ) -experienced woman developed fatigue , an erythematous pruritic macular rash , and arthralgias six days after traveling to Caracas , Venezuela in March 2016 ( Fig 1A and 1B ) ., She presented on day 3 ( d3 ) post-onset of symptoms ( POS ) ., A comprehensive metabolic panel and complete blood count were within normal limits except for slight elevations in ALT ( 50 U/L , normal range 0–41 U/L ) and AST ( 44 U/L , normal range 0–40 U/L ) ., Serologic testing was consistent with acute flaviviral infection but did not differentiate between DENV and ZIKV infection ( Table 1 ) 30 ., A research-use nucleic acid amplification test ( NAAT ) ( Hologic ) was positive for ZIKV infection in d3 POS blood and urine samples ( Table 1 ) ., Blood and urine on d3 and d6 POS were negative for DENV , as determined via qRT-PCR ., Urine , from d3 and d6 POS , inoculated onto C6/36 cells produced infectious virus as measured by focus forming assay ( FFA ) ., Sequence analysis of C6/36 amplified virus was confirmed to be ZIKV using a validated population-based sequencing protocol for ZikaBr targeting ZIKV NS5 ., Phylogenetic analysis of the near complete viral genome ( >9kb ) showed the ZIKV San Diego isolate ( ZIKV SD001 29 ) was most closely related to other Latin American ZIKV isolates downloaded from Genbank ( Fig 1C ) 31 ., To characterize the systemic immune response to ZIKV infection , we first measured circulating serum cytokine levels ., Serum was collected on d3 , d6 , d17 and d48 POS ., These samples were compared to baseline pre-infection serum collected from this individual 8 months prior to infection ., We found that only a small number of cytokines , including IP-10 , MCP-1 and IL-1RA , showed dramatic increases during early infection ( Fig 2A ) ., Each of these cytokines peaked on d3 POS before returning toward baseline ., The levels of many inflammatory cytokines , including IFNγ and TNFα , did not change or minimally changed throughout infection ( S1 Fig ) ., To evaluate the cellular response to infection , we first performed RNA sequencing ( RNA-seq ) on PBMCs ., To identify induced and repressed genes during infection we compared transcriptomes at d3 , d6 and d17 POS with d48 ( convalescent ) ( Fig 2B ) ., Hierarchical clustering of normalized PBMC transcriptional profiles showed dynamic induction patterns with strong d3 up-regulation of many interferon-stimulated genes ( ISGs ) , which steadily declined at d6 and d17 ., Published human PBMC studies during acute DENV infections demonstrated sequential waves of gene expression with early induction of ISGs and inflammatory chemokines followed by a switch to induction of genes involved in cell proliferation 34 ., During ZIKV infection in this individual , there was similar strong induction of type I ISGs , exemplified by MX1 , OAS3 , RSAD2 , and IFI27 genes ( Cluster 2 ) , but minimal coincident induction of chemokines involved in leukocyte chemotaxis ( Cluster 1 ) ( Fig 2B ) ., This includes CXCL10 and CCL2 , that encode the chemokines IP-10 and MCP-1 , that were elevated at the protein level d3 POS ., Genes associated with cell differentiation and proliferation , such as BUB1 , DLGAP5 , PBK and CEP55 ( Cluster 3 ) were upregulated during DENV infection but not ZIKV , while EGR1 , HBEGF and MAFB ( Cluster 4 ) , were up-regulated at d17 POS during ZIKV infection ( Fig 2B ) ., To better understand if low-level cytokine gene induction in PBMCs was characteristic of ZIKV infection , we analyzed a published study where temporal gene expression profiles were measured in rhesus macaques following ZIKV infection 33 ., PBMCs from our patient and rhesus macaques showed similar early transcriptional upregulation of ISGs but minimal chemokine gene induction with the possible exception of CXCL10 in monkeys ( Fig 2C ) ., Analyzing PBMC transcription and serum cytokines provides important information about global immune responses but lacks cell population-level resolution ., To better understand how individual cell populations responds to ZIKV infection , we isolated three monocyte subsets; classical , intermediate , and non-classical; natural killer ( NK ) cells; two dendritic cell ( DC ) subsets; myeloid DCs ( mDCs ) and plasmacytoid DCs ( pDCs ) ; as well as CD4+ and CD8+ T cells at d3 , d6 , d17 and d48 POS using Fluorescence-activated cell sorting ( FACS ) ( S2 Fig ) ., RNA-seq transcriptional analysis of individual cell types and PBMCs together identified 1 , 147 genes induced at least 2-fold at d3 , d6 , or d17 when compared to d48 ( Fig 3A ) ., A similar analysis of PBMCs alone identified only 452 induced genes ( Fig 3A ) ., Innate immune cells ( monocytes and DCs ) induced the highest number of genes on d3 POS ( Fig 3B ) ., Genes up-regulated in innate immune subsets were most enriched for functional annotations associated with interferon ( IFN ) and immune responses at d3 and d6 POS ( Fig 3C ) ., Additionally , the promoters of genes induced at d3 and d6 POS in innate immune cells were most significantly enriched for ISRE , IRF-composite and STAT1 motifs ( Fig 3D ) ., Together , this data is consistent with early activation of type I IFN responses in innate immune populations through activation of interferon regulatory factors ( IRFs ) and interferon-stimulated gene factor 3 ( ISGF3 ) transcription factors 35 , 36 ., In contrast to innate immune cells , the peak of T cell gene up-regulation was delayed ( Fig 3B ) ., Genes induced in CD8+ T cells were functionally enriched for terms associated with cell cycle progression such as E2F and MYC targets and G2M checkpoint ( Fig 3C ) ., Additionally , the promoters of these induced genes were enriched for E2F , NFY and POU binding motifs where transcription factors involved in controlling cell cycle and cell differentiation can bind ( Fig 3D ) ., Like innate immune populations , NK cells responded rapidly to infection by inducing IFN pathways ( Fig 3C ) ., However , NK cells also activated cell cycle progression pathways like CD8+ T cells but did so earlier , d3 compared to d6 POS , during infection ( Fig 3C and 3D ) ., Although PBMC analysis identified fewer induced genes , PBMC functional and promoter motif enrichment analyses captured many core components observed in both individual innate immune and T cell analyses ( Fig 3C and 3D ) ., An unbiased analysis of gene expression profiles using hierarchical clustering of the top induced genes in all individual cell types and PBMCs revealed both temporal and cell type specific patterns of gene expression ( Fig 3E ) ., Gene expression at early time points , d3 and d6 POS , generally cluster together and apart from d17 responses ( Fig 3E ) ., The exception is T cells , where only d6 POS gene expression cluster in the early group ., The genes driving this difference are largely induced in a time dependent manner , with anti-viral genes ( Cluster B ) being up-regulated early and other immune pathway genes ( Cluster D ) later in infection ( Fig 3E ) ., Additionally , at d3 and d6 POS , transcriptional responses cluster by cell type suggesting early transcriptional responses are in part cell type specific ( Fig 3E ) ., Population-specific induction of genes is evident from genes that are up-regulated exclusively in pDCs ( Cluster A ) or NK and T cells ( Cluster C , Fig 3E ) ., In contrast to all other cell types , classical and intermediate monocytes cluster together based on day POS suggesting that gene induction in these two cell types is more dependent on time POS than cell type ., Many genes , including AIM2 , an ISG involved in inflammasome activation in macrophages , is induced in both time and cell-type specific manners ( Fig 3F and 3G ) ., Additionally , CXCL10 , CCL2 , and IL1RN that encode the cytokines , IP-10 , MCP-1 and IL1RA , were upregulated at least 2-fold in certain monocyte populations at d3 POS even though they were not significantly induced in PBMCs as a whole ( S3 Fig ) ., Chromatin accessibility is a major component of genome regulation ., Open regions of chromatin are putatively associated with genomic regulatory regions , including both promoters and enhancers ., The Assay for Transposase Accessible Chromatin using sequencing ( ATAC-seq ) can be used to identify transcription factors ( TFs ) involved in regulating important functions , such as differentiation and gene regulation through the analysis of open chromatin ., We did not obtain high quality d3 ATAC-seq data ., However , high quality ATAC-seq data were produced using samples from d6 and d17 POS ., Comparing ATAC-seq peaks in classical monocytes with NK cells on d6 POS we identified 13 , 792 and 13 , 200 peaks unique to classical monocytes and NK cells respectively ., De novo motif analysis of these peaks identified PU . 1 , CEBP and AP-1 in monocytes and ETS1 , RUNX and T-box in NK cells as the most enriched TF binding motifs ( Fig 4A ) ., Each of these TFs have been identified as important lineage-determining transcription factors ( LDTFs ) in monocytes and NK cells , respectively ., To investigate TFs that may be important during the cellular response to infection we examined dynamic changes in chromatin accessibility over time ., We identified 1 , 493 and 1 , 261 ATAC-seq peaks that were significantly upregulated at d6 POS compared to d17 POS in classical monocytes or NK cells respectively ., In addition to cell type specific LDTFs in monocytes and NK cells , motif analysis of upregulated ATAC-seq peaks at d6 POS demonstrated increased enrichment of PU . 1:IRF8 and bZIP TF binding motifs in classical monocytes ( Fig 4B ) ., In contrast , ISRE/IRF motifs were equally represented in regulated ATAC-seq peaks in classical monocytes and NK cells ( Fig 4B ) ., To help illustrate how these data characterize individual gene loci , we considered the open chromatin landscape at genes with both common and cell-type specific patterns of regulation ., Both classical monocytes and NK cells upregulated the ISGs IFIT2 and IFIT3 early in infection and ATAC-seq peaks were identified at sites containing ISRE motifs ( Fig 4C ) ., Although the ISRE associated peaks were common at these loci , the other ATAC-seq peaks were monocyte or NK specific and were associated with TF binding motifs enriched in the corresponding cell-type ., This suggests that although both cell types induce IFIT2 and IFIT3 they may utilize cell type specific TFs to help regulate gene expression ., Another ISG , APOBEC3A , was induced in monocytes but not in NK cells ( Fig 4D ) ., At this gene locus , the ATAC-seq peaks were all monocyte specific and were associated with monocyte-enriched TF binding motifs ( Fig 4D ) ., The gene MKI67 encodes the protein Ki-67 and is a marker of proliferation ., This gene was induced in NK cells early in infection but was never induced in classical monocytes ( Fig 4E ) ., The ATAC-seq peaks associated with this gene are NK-specific and associated with NK enriched TF binding motifs except for one common peak associated with an E2F motif ( Fig 4E ) ., These examples help illustrate how open chromatin patterns associated with cell-type specific transcription factors may play a role in defining common and cell-type specific patterns of gene expression ( Fig 4D and 4E ) ., We next evaluated the temporal development of adaptive immune responses ., Prior to the acute ZIKV infection , this individual had low but detectable neutralizing Abs to both DENV and ZIKV strains ( Fig 5A ) ., Neutralizing Ab titers to ZIKV and DENV rapidly increased after infection , peaking on d6 POS ( Fig 5A–5D ) ., The highest neutralizing titer 50 ( NT50 ) developed against the patient’s own virus followed by the related ZIKV FSS13025 ( Cambodia , 2010 ) ( Fig 5A and 5B ) 37 ., The NT50 also increased against both DENV1 and DENV4 but to a lesser degree than either ZIKV strain ., These results are consistent with the idea that ZIKV infection can induce cross-reactive neutralizing Ab responses to DENV especially in individuals with prior flavivirus experience with faster kinetics relative to naïve people 17 , 19 ., Lastly , we assessed the T cell response by flow cytometry ., Published studies have shown that the majority of DENV-specific and ZIKV-specific T cells display an effector or memory phenotype based on expression of CD45RA and CCR7 18 , 38 , 39 ., Moreover , in secondary DENV infections , the T cell response is associated with an expansion of T effector memory RA ( TEMRA ) and T effector memory ( TEM ) cells that can be more vigorous than in primary DENV infection 39 ., Accordingly , our data on bulk populations of unstimulated T cells showed higher proportions of CD4+ TEMRA cells and lower proportion of naïve CD8+ T cells ( TN ) at d6 POS as compared to 3 healthy DENV-naïve and 2 DENV-immune control individuals ( Fig 6A and 6B ) ., We also examined CD4 T helper ( Th ) subsets based on the expression of chemokine receptors ( Th1: CCR6-CCR4-CXCR3+; Th2: CCR6-CCR4+CXCR3-; Th1/17: CCR6+CCR4-CXCR3+; and Th17: CCR6+CCR4+CXCR3- ) ., No specific Th profile was observed in this individual ( S2 Table ) , consistent with the published observation that the majority of DENV-specific CD4+ T cells are not associated with common Th subsets 39 ., Studies of DENV-infected individuals have suggested that expanded CD4+ TEMRA cells can exhibit a virus-specific cytotoxic phenotype that has been associated with protection against severe DENV disease 39 , 40 ., Cytotoxic CD4 T cells are CD45RA+CCR7- ( TEMRA ) with increased expression of CD8α , cytotoxic effector molecules such as granzyme B and perforin , and CD226 , a co-stimulatory molecule that enhances CD8 effector and cytotoxic functions ., A CD4+ T cell population with low level CD8 expression ( CD4+CD8dim ) was detected in our individ | Introduction, Methods, Results, Discussion | Zika virus ( ZIKV ) is an emerging mosquito-borne flavivirus linked to devastating neurologic diseases ., Immune responses to flaviviruses may be pathogenic or protective ., Our understanding of human immune responses to ZIKV in vivo remains limited ., Therefore , we performed a longitudinal molecular and phenotypic characterization of innate and adaptive immune responses during an acute ZIKV infection ., We found that innate immune transcriptional and genomic responses were both cell type- and time-dependent ., While interferon stimulated gene induction was common to all innate immune cells , the upregulation of important inflammatory cytokine genes was primarily limited to monocyte subsets ., Additionally , genomic analysis revealed substantial chromatin remodeling at sites containing cell-type specific transcription factor binding motifs that may explain the observed changes in gene expression ., In this dengue virus-experienced individual , adaptive immune responses were rapidly mobilized with T cell transcriptional activity and ZIKV neutralizing antibody responses peaking 6 days after the onset of symptoms ., Collectively this study characterizes the development and resolution of an in vivo human immune response to acute ZIKV infection in an individual with pre-existing flavivirus immunity . | Zika virus ( ZIKV ) is an emerging flaviviral infection that causes significant clinical disease ., It is estimated that approximately one half of the world’s population is at risk for ZIKV infection ., There are only a limited number of studies describing the human immune response to ZIKV infection ., Carlin et al . combined conventional and genomic approaches to longitudinally analyze the innate and adaptive immune responses to acute ZIKV infection and its resolution in a person who was infected while traveling in Venezuela during the 2016 ZIKV epidemic year ., Genome-wide sequencing in individual cell types revealed that although many populations respond to interferon stimulation , only specific cell populations within peripheral blood mononuclear cells upregulate important inflammatory cytokine gene expression ., Additionally , analysis of open chromatin using ATAC-seq suggests that chromatin remodeling at sites containing cell-type specific transcription factor binding motifs may help us understand changes in gene expression ., Consistent with previous reports , this individual with prior exposure to dengue virus ( DENV ) , rapidly developed neutralizing anti-ZIKV responses that were cross-reactive with multiple DENV serotypes ., Collectively this study combines traditional and genomic approaches to characterize the cell-type specific development of an in vivo human immune response to acute ZIKV infection . | blood cells, dengue virus, innate immune system, medicine and health sciences, immune cells, immune physiology, pathology and laboratory medicine, cytokines, pathogens, immunology, microbiology, viruses, developmental biology, rna viruses, molecular development, white blood cells, animal cells, medical microbiology, gene expression, microbial pathogens, t cells, immune response, immune system, cell biology, flaviviruses, monocytes, nk cells, viral pathogens, physiology, genetics, biology and life sciences, cellular types, organisms, zika virus | null |
journal.pcbi.1003776 | 2,014 | The Fitness Landscape of HIV-1 Gag: Advanced Modeling Approaches and Validation of Model Predictions by In Vitro Testing | The ideal way to combat the spread of HIV-1 is with an effective prophylactic or therapeutic vaccine 1 , 2 ., One of the greatest challenges hindering the achievement of this goal is the incredible sequence diversity and mutability of HIV-1 3 , which can limit the effectiveness of the immune response 2 , 4 ., CD8+ T cells are instrumental in reducing viral load in HIV-1 acute infection 5 and in maintaining the viral set point during chronic HIV/SIV infection 6 , 7 ., However , HIV-1 is able to escape the CD8+ T cell response through mutations in or adjacent to HIV-1 epitopes that are presented by HLA class I molecules on the surface of the infected cells 7 ., One proposed strategy for realizing a potent prophylactic or therapeutic vaccine is to target CD8+ T cell responses to conserved regions of HIV-1 , aiming to reduce incidences of immune escape or , if escape occurs , to reduce viral fitness and lower the viral set point , thereby slowing disease course and reducing transmission at the population level 8 , 9 ., While escape mutations at highly conserved sites often damage the viability of virus 10 , this approach is confounded by the development of compensatory mutations which restore or partially restore viral fitness 9 ., Thus , to maximize the effectiveness of a vaccine-induced immune response one must look beyond conservation of single residues to identify regions where mutations are not only highly deleterious , but where further mutations elsewhere in the proteome are unlikely to restore lost fitness , but rather , lead to additional fitness costs due to deleterious synergistic effects ., Our group has developed computational models to identify such vulnerable regions of the HIV-1 proteome and to predict the fitness landscape of HIV-1 proteins , providing tools for designing vaccine immunogens that may limit both HIV-1 evasion of CD8+ T cell responses and the development of compensatory mutations 11 , 12 ., In an early qualitative study we identified groups of amino acids in HIV-1 Gag coupled by structural and functional constraints that cause these residues to co-evolve with each other , but evolve nearly independently of the other residues in the protein 12 ., In analogy with past studies on the economic markets and enzymes 13–15 , we termed these groups of residues “sectors” ., This analysis and human clinical data revealed one sector in Gag , which we termed sector 3 , where multiple mutations were more likely to be deleterious ., This group of residues is naturally targeted more by elite controllers 12 ., It is expected to be particularly vulnerable to CD8+ T cell responses that target multiple residues in it since multiple mutations within this sector are likely to significantly diminish viral fitness , thereby restricting available escape and compensatory paths 12 ., This approach , however , does not allow us to determine precisely which residues should be targeted , as it does not quantify the relative replicative viability of viral strains bearing specific mutations ., Nor does it identify viable escape routes that remain upon targeting residues in the vulnerable regions , or inform how best to block them ., To begin to address these issues , we developed a computational model , rooted in statistical physics , which aims to predict the viral fitness landscape ( viral fitness as a function of amino acid sequence ) from sequence data alone and applied it to HIV-1 Gag 11 ., Similar methods have previously been employed to study other complex biological systems , from describing the activity patterns of neuronal networks 16–19 to the prediction of contact residues in protein families 17 , 20 , 21 ., The idea underlying our approach is to first characterize the distribution of sequences in the population , which we expect to be correlated with fitness ( see below ) ., Due to the small number of available sequences compared to the size of the sequence space , direct estimation of the probability distribution characterizing the available sequences is precluded ., Thus , we instead aim to infer the least biased probability distribution of sequences that fits the observed frequency of mutations at each site , and all correlations between pairs of mutations ( the one- and two-point mutational probabilities ) ., Mathematically , “least biased” implies the distribution that has maximum entropy in the information-theoretic sense 22 ., The maximum entropy distribution that fits the one- and two-point mutational probabilities has a form reminiscent of that describing equilibrium configurations of an Ising model in statistical mechanics ., We generated such models using multiple sequence alignments ( MSA ) for the four subunit proteins of Gag in HIV-1 clade B 11 ( described in Supporting Information Text S1 , Section 1 ) ., This model assigns to each viral strain an “energy” ( E ) , which is inversely related to the probability of observing this sequence ., We expect more prevalent sequences to be more fit , consistent with expectations from simple models of evolution 23 though the precise correspondence between fitness and prevalence may have a more complicated dependence on factors such as the shape of the fitness landscape , as predicted by quasispecies theory 24 ., Furthermore , this expectation could be confounded by immune responses in the patients from whom the virus samples were collected , and phylogeny ., Recent analyses suggest ( described more fully in the discussion ) that in spite of these effects , at least for Gag proteins , the rank order of prevalence and in vitro replicative fitness should be similar 25 ., Strains with high E values are predicted to be less fit than strains with low E values ., Predictions of the model seemed to be in good agreement with experimental data on in vitro replicative fitness , as well as clinical observations on the frequency and impact of viral escape mutations 11 ., Our aim in the current work is twofold ., First , we present new advances in the inference and modeling of viral fitness landscapes that address previous theoretical and computational limitations ., Second , we describe new in vitro fitness measurements for viruses containing multiple Gag mutations , performed to further test fitness predictions using the improved computational methods ., To give a broad test of the predictive power of the fitness models , we have performed comparisons for HIV-1 strains containing multiple mutations predicted to harm HIV-1 viability as well as combinations predicted to be relatively fitness neutral ., We find that fitness measurements of these mutant strains are in good agreement with model predictions ., Our key hypothesis in formulating models of HIV fitness is that the prevalence of viruses with a given sequence , that is , how often the sequence is observed , is related to its fitness ., Simply , fitter viruses should be more frequent in the population than those that are unfit ., This hypothesis can be proven for some idealized evolutionary models 23 , but cannot be made exact for the complicated nonequilibrium host-pathogen riposte between humans and HIV ., However , our theoretical work , backed by extensive computational studies , suggests that the rank order of fitness and prevalence of strains should be strongly monotonically correlated , provided we compare sequences that are phylogenetically close 25 ., Thus , if we construct a model to predict the likelihood of observing different viral strains with given sequences , it can predict the relative fitness of the strains ., We achieved this goal by constructing a maximum entropy model for the probability of observing sequences in the MSA 26 ., The simplest model in this class is an Ising model , a simple model of interacting binary variables from statistical physics which has been widely applied to study collective behavior in complex systems ., The parameters of this Ising model are obtained by imposing the constraint that it reproduce the pattern of correlated mutations ( relative to the consensus sequence ) observed in a multiple sequence alignment ( MSA ) of HIV-1 amino acid sequences extracted from infected hosts ., Specifically , the parameters were chosen such that the frequency of mutations at each single residue and the frequency of simultaneous mutations at each pair of residues were the same in both the Ising model and the MSA ., Importantly , the model also reproduced higher order mutational correlations accurately , even though these mutational frequencies were not directly fitted 11 ., As described in our previous publication 11 , in the Ising model amino acid sequences in the MSA are compressed into binary strings by assigning a 0 to each position where the amino acid matches the consensus sequence ( “wild-type” ) , and a 1 to each position with a mismatch ( “mutant” ) ., While this binary approximation greatly simplified our modeling approach , the reduction in complexity has several drawbacks ., Firstly , there is a loss of residue-specific resolution ., The fitness predictions of our model are insensitive to the precise identity of mutant amino acids , and thus the model cannot resolve fitness differences between proteins containing different mutant amino acid residues in a particular position ., Secondly , for relatively conserved proteins such as HIV-1 Gag , where the number of viable amino acids at each position is rather limited , this binary simplification represents a reasonable approximation ., However , it is less justified for highly mutable proteins where the wild-type residue in each position is not the overwhelmingly most probable amino acid , as is the case for the HIV-1 Env protein ., In our original approach , we fit the Ising model parameters to precisely reproduce the observed one and two-residue mutational correlations within the MSA ., However , simultaneous mutations at certain pairs of residues were never observed ., This led to another deficiency in our original modeling approach in that pairs of mutations not observed in the MSA were predicted to be completely unviable ( E\u200a=\u200a∞ ) ., While it is possible that such mutant viral strains have exactly zero replicative fitness , it is more likely that they are highly unfit strains ( possessing non-zero replicative fitness ) that simply arise too seldom to be observed within our finite-sized MSA ., In this work , we present three significant advances of our original model to predict viral fitness , which also the aforementioned limitations ., First , we incorporate Bayesian regularization into our fitting procedure to eliminate the prediction of zero replicative fitnesses for mutations not present within our MSA ., Second , we implement a new algorithm for inferring an Ising model from sequence data , which dramatically accelerates the computation of model parameters ., Third , we relax the binary approximation to infer viral fitness landscapes that explicitly retain the amino acid identities at each position ., We achieve this by describing the viral fitness landscape using a multistate generalization of the Ising model known as the Potts model , another established and well-studied model in statistical physics 27 ., We also implement Bayesian regularization into the fitting of the Potts model parameters ., Inference of the parameters of the Ising models , commonly referred to as the inverse Ising problem , is a canonical inverse problem lacking an analytical solution that may be tackled in many ways 16 , 17 , 19–21 , 28 , 29 ., We improve upon our previous techniques described in 11 by incorporating regularization and implementing new inference algorithms , which greatly decrease the computational burden and accelerate model fitting ., To control the effects of undersampling and to improve the predictive power of the inferred fitness models , we incorporate Bayesian regularization into our inference algorithm 18 , 19 , 30 , 31 in the form of a Gaussian prior distribution for the model parameters describing pairwise couplings between residues ( see Text S1 , Sections 1 . 3 and 2 . 5 ) ., Regularization of this form is also known as Tikhonov regularization or ridge regression 32 ., With this addition , the probability of observing any sequence , including those containing pairs of mutations not observed in the MSA , is nonzero ., We have also computed a correction to the energy of each sequence to account for the possible bias that strains near fitness peaks are more likely to be observed than would be expected from their intrinsic fitness when sampled from a finite distribution ( see Text S1 , Section 3 . 2 ) ., In an algorithmic advance over our previous fitting procedure , we fit the parameters of our regularized Ising model using the selective cluster expansion algorithm of Cocco and Monasson 18 , 19 which identifies clusters of strongly interacting sites and iteratively builds a solution for the whole system by solving the inverse Ising problem for each cluster ., With this approach , we cut the CPU time necessary to infer the parameters of the Ising model from roughly 12 years 11 to 5 hours for p24 , an improvement by four orders of magnitude ., Roughly , we expect algorithm run-time to scale as O ( Nn exp ( n ) ) , where N is the system size ( number of amino acids ) and n is the size of a typical “neighborhood” of strongly interacting sites ., For a review and applications of this method see 18 , 30 ., Complete details of our modeling approach and numerical fitting procedures are provided in the Text S1 , Section, 1 . An ideal model of viral fitness would be able to capture the full ( unknown ) distribution of correlated mutations throughout the sequence , and thus reproduce the prevalence of every viral strain ., Sequences in the MSA represent a sample of the possible strains of the virus , providing information about the distribution of point mutations , pairs of simultaneous mutations , triplets of simultaneous mutations , and all higher orders ., However , since the number of available sequences in the MSA is very small compared to the size of the accessible sequence space , and because mutations at most sites are rare , higher order mutations will be severely undersampled ., Thus , following our previous approach we appeal to the maximum entropy principle to seek the simplest possible model capable of reproducing the single site and pair amino acid frequencies 11 , 22 , for which the problem of undersampling is less severe ., From this analysis , the Potts model is the least structured model capable of reproducing the one and two-position frequencies of amino acids observed within the MSA 21 ., To introduce the Potts model , we represent the sequence of a particular m-residue protein as a vector , , where the elements Ak can take on the q\u200a=\u200a21 integer values 1 , 2 , … , 21 denoting an arbitrary encoding of the 20 natural amino acids , plus a gap 21 ., In the Potts model the probability of observing a particular sequence is given by ( 1 ) In analogy with the statistical physics literature , we refer to E as a dimensionless “energy , ” the function as the Hamiltonian , and the normalizing factor Z as the partition function 27 ., The model is parameterized by a set of m q-dimensional vectors , , and a set of m ( m−1 ) /2 q-by-q matrices , ., The hi vectors give the contribution of the identity of each amino acid in each position to the overall sequence energy , and the Jij matrices give the contribution to the energy of pairwise interactions between amino acids in different positions ., To fit the Potts model , we implemented a generalization of the semi-analytical extension of the iterative gradient descent implemented by Mora and Bialek 11 , 17 ., This approach implements a multi-dimensional Newton search to iteratively adjust the model parameters until the predictions of the model for the one and two-position frequencies of amino acids reproduce those observed within the MSA ., In an advance over the original incarnation of this algorithm , we have derived closed form expressions for the gradients required by the Newton search , thereby obviating the need for their numerical estimation by finite differences ( which would result in a more computationally expensive and less numerically stable secant search procedure ) ., Our approach is semi-analytical in the sense that while we have analytical expressions for the Newton search gradients , we use a Monte Carlo procedure to numerically estimate the one and two-position amino acid frequencies predicted by the model at each stage of parameter refinement ., We are currently developing a Potts generalization of the cluster expansion algorithm 19 to accelerate fitting ., We incorporate Bayesian regularization into our fitting procedure in a precisely analogous manner to that described above for the regularized Ising model by introducing a Gaussian prior distribution over the Jij parameters ., Inference of the Potts model parameters for p24 required approximately 1 . 4 years of CPU time using a generalization of the gradient descent approach described in Ref ., 11 ., Fitting the model parameters by gradient descent is expected to scale as O ( ( Np ) 2 ) , where N is the number of amino acids in the protein , and p is the characteristic number of mutant residues observed at each position ., Full details of the fitting and regularization procedures are provided in Text S1 , Section, 2 . The code implementing the inverse Potts inference algorithm is also provided in Supporting Information Code S1 ., To test the accuracy of these models in predicting the fitness landscape of HIV-1 Gag , we performed in vitro experiments to measure the fitness of various Gag mutants ., Previously we had measured the in vitro replication capacities of 19 Gag p24 mutants , 16 of which contained single mutations in Gag p24 , and compared these with fitness predictions of our original Ising model 11 ., Here , we extend this work to measure the replication capacities of HIV-1 strains containing various combinations of mutations , predicted to be either harmful to HIV-1 viability or fitness-neutral , in Gag p24 and p17 and we compare measurements not only to the original Ising model described in Ref ., 11 , but also to regularized versions of Ising and Potts models that we have developed here ., Specifically we considered 17 mutations pairs , one triple , and 25 single mutations within these combinations , as listed in Table, 1 . These mutations were introduced into the widely used laboratory-adapted HIV-1 clade B reference strain NL4-3 ., The tested mutants can be divided into 4 categories , viz ., ( i ) Gag p24 pairs with high E values located within a group of co-evolving amino acids termed sector 3 ( cf . Ref 12 ) ,, ( ii ) HLA-associated Gag p24 pairs with high E values ,, ( iii ) Gag p24 pairs/triple with low E values , and, ( iv ) Gag p17 pairs ( Table 1 ) ., These mutation combinations were chosen according to E values predicted by the published Ising model 11 , where E>90 or E\u200a=\u200a∞ were considered high E values and E<15 were considered low E values ., Note that , due to the couplings between mutations at different sites , parameterized by the Jij in equation 1 , the E values depend not only on the specific mutations introduced but also on the sequence background ., The E values for mutations reported here are computed with the HIV-1 NL4-3 sequence background , which differs from the p17 and p24 MSA consensus sequences by 8 mutations ( R15K , K28Q , R30K , K76R , V82I , T84V , E93D , S125N ) and 2 mutations ( N252H , A340G ) , respectively ., The p24 region of Gag was focused on since this is the most conserved region of the protein ., First , we selected six mutation pairs , predicted to be unfavorable in combination , in sector 3 of Gag p24 since we previously found this to be an immunologically vulnerable group of co-evolving residues in which multiple mutations are not well-tolerated 12 ., Since it is desirable to identify low fitness/non-viable combinations of escape mutations for vaccine immunogen design aimed at reducing viral fitness or blocking viable escape pathways , we aimed to identify pairs of likely escape mutations with high E values ., Virus mutations that are statistically associated with the expression of specific host HLA class I alleles , which also restrict the same epitopes in which the mutations are found , are likely to be CD8+ T cell-driven escape mutations 33 ., We therefore tested five high E pairs of mutations located at HLA-associated Gag p24 codons ( HLA-associated variants defined in 34 , 35 ) in or next to optimal CD8+ T cell epitopes ( A-list epitopes from the Los Alamos HIV sequence database 36 ) that were restricted by the same HLA ., For comparison with high E mutation pairs , mutation combinations with low predicted E values were included in testing , comprising known favorable compensatory pairs in Gag p24 where 219Q compensates for the 242N escape mutant 37 and 147L compensates for the 146P escape mutant 10 , as well as one pair in sector 3 of Gag p24 and a Gag p24 triple mutant ., Additionally , for broader testing , two mutation pairs in Gag p17 were selected ., We note that the most commonly observed mutant amino acid at each codon was tested ., We introduced these mutation combinations into the HIV-1 NL4-3 plasmid by site-directed mutagenesis and their presence was confirmed by sequencing , as described previously 38 ., Generation of mutant viruses from mutated plasmids and the measurement of their replication capacities were performed as previously 11 , 38 ., Briefly , mutated plasmids were electroporated into an HIV-1-inducible green fluorescent protein reporter T cell line , harvested at ≈30% infection of cells , and the replication capacities of the resulting mutant viruses were assayed by flow cytometry using the same cell line ., Replication capacities were calculated as the exponential slope of increase in percentage infected cells from days 3–6 following infection at a MOI of 0 . 003 , normalized to the growth of wild-type NL4-3 ( RC\u200a=\u200a1 ) ., Three independent measurements were taken and averaged ., Mutant viruses were re-sequenced to confirm the presence of introduced mutations ., The values of E predicted by our original and new modeling approaches for the 43 HIV-1 NL4-3 Gag mutants tested here are shown in Table 2 ., Absolute comparison of the E values between the models are not meaningful , but the relative E values of mutants are generally in excellent concordance between models ( Pearsons correlation , r≥0 . 85 and p≤5 . 3×10−11 , two-tailed test ) ., The in vitro fitness measurements for all mutants , grouped according to categories , are shown in Figure 1 ., We initially compared our model predictions and fitness measurements for each category of mutant pairs to evaluate whether mutant combinations with high and low predicted E values corresponded to substantial fitness cost or little/no fitness cost , respectively ., Briefly , all Gag p24 sector 3 mutation pairs with high E values were not viable in our assay system , and were assigned a replication capacity of zero ( Figure 1A ) ., Similarly , with the exception of 315G331R , the five high E HLA-associated mutation pairs showed substantial reduction in replication capacity , to between 0–56% of wild-type levels ( Figure 1B ) ., Non-viable mutants ( RC\u200a=\u200a0 ) were those for which the generation of virus stocks from plasmids encoding these mutation pairs failed , or , in two instances – mutants 186I295E and 186I331R – were not viable unless further mutations developed , confirming unfavorability of the mutation combination ., Briefly , concentrated virus stocks for mutants 186I295E and 186I331R were harvested at >22 days post-electroporation compared with the median harvesting time of 6 days post-electroporation for all mutants ( at which time the 186I295E and 186I331R mutants had infected ≈1% cells ) ., Sequencing of these viruses revealed the presence of additional mutations and/or reversion of introduced mutations ., For mutant 186I295E , amino acid mixtures were detected at codons 63 ( Q/R ) , 177 ( D/E ) and 186 ( I/V ) , and for mutant 186I331R , mixtures were detected at codons 168 ( I/V ) and 331 ( K/R ) , as well as reversion of 186I to 186T ., On repeating virus generation for these mutants , additional mutations similarly developed – mixtures were observed at codons 214 ( R/K ) and 271 ( N/S ) for mutant 186I295E , and 232 ( R/M ) and 260 ( D/E ) for mutant 186I331R ., With the exception of 186I295E and 186I331R , sequencing confirmed that all mutant viruses had only the specific mutations introduced ., The spontaneous mutations 186V , 271S and 232M were not observed in the MSA and the new mutation combinations did not have lowered E values in any of the models , with the exception of the incomplete 186I331R260D combination ( complete observed combination 186I , 331R , 232R/M , 260D/E ) which displayed a slightly lower energy than 186I331R in the regularized Ising model only ( 11 . 5 vs . 13 . 7 ) ( data not shown ) ., Nevertheless , these observations confirm that 186I295E and 186I331R are unfit mutation combinations requiring compensatory paths to restore viability ., Taken together , the data on high E p24 mutants confirm mutation combinations predicted to be unfit , and also identify combinations of HLA-associated mutations in/next to optimal CD8+ T cell epitopes ( mutations likely to result in CD8+ T cell escape 33 ) that carry substantial fitness costs ., Those p24 mutation combinations , including known compensatory pairs , that were predicted to have low E values displayed replication capacities similar to that of wild-type NL4-3 , indicating that these combinations had little or no cost to HIV-1 replication capacity in accordance with predictions ( Figure 1C ) ., Similarly , all p17 mutants tested had replication capacities close to that of the wild-type NL4-3 virus , consistent with the predicted E values of all mutants except 86F92M ( Figure 1D ) ., Overall , for only two ( 86F92M and 315G331R ) of the 17 mutant pairs the fitness measurement did not correspond to the E value prediction of high or low fitness cost ., It should however be noted that the disparity between E values and measured replication capacities for these mutant pairs is somewhat mitigated in the regularized models ., The E values for the regularized Ising model for these mutants ( which were assigned an E value of infinity by the original Ising model ) are lower than those of other mutants previously assigned infinite energies , and the same is true for mutant 86F92M in the regularized Potts model ., Next , we assessed the relationship between fitness measurements and E values predicted by our original Ising , regularized Ising and regularized Potts models using Pearsons correlation tests ., There is a strong correlation between the metric of fitness ( values of E , Table 1 ) predicted by the original unregularized Ising model and our experimental measurements ( Pearsons correlation , r\u200a=\u200a−0 . 74 and p\u200a=\u200a3 . 6×10−6 , two-tailed ) ( Figure 2A ) , however this correlation out of necessity excludes mutants with E values equal to infinity ( n\u200a=\u200a13 ) ., The regularized Ising model allows for inclusion of these data points resulting in a stronger correlation between predictions and fitness measurements ( Pearsons correlation , r\u200a=\u200a−0 . 83 and p\u200a=\u200a3 . 7×10−12 , two-tailed ) ( Figure 2B ) , which is slightly improved by focusing on Gag p24 mutants only ( Pearsons correlation , r\u200a=\u200a−0 . 85 and p\u200a=\u200a1 . 4×10−11 , two-tailed ) ., There is also a strong agreement between the residue-specific Potts model energies and replication capacity ( Pearsons correlation , r\u200a=\u200a−0 . 73 and p\u200a=\u200a9 . 7×10−9 , two-tailed ) ( Figure 2C ) ., In practice , one may be concerned with a more coarse-grained measure of viral fitness: will a virus with a given sequence be able to replicate with similar efficiency to the wild-type , or will it be significantly impaired ?, To explore this point , we grouped the experimentally tested mutants into two categories , “fit” ( RC≥0 . 5 ) and unfit ( RC<0 . 5 ) , and tested the ability of the fitness landscape models to predict which class each sequence would belong to based on their E values ., This was accomplished by fitting a linear classifier to the data using logistic regression ( Text S1 , Section 3 . 1 ) ., The regularized Ising model E classifier is highly accurate ( 91% accuracy at optimal threshold , AUROC\u200a=\u200a0 . 93 ) – we observed a strong , significant difference in replication capacities between the mutants classified as unfit and those classified as fit ( Mann-Whitney U\u200a=\u200a32 , ) ( Figure 3A ) ., Specifically , four mutants ( 86F92M , 190I , 190I302R and 243P ) were not classified correctly ., However , 190I302R , which was classified as unfit ( E\u200a=\u200a8 . 6 ) , exhibited a fitness close to that of the 0 . 5 cutoff ( RC\u200a=\u200a0 . 56 ) and 243P , which displayed low fitness ( RC\u200a=\u200a0 . 36 ) , had a predicted E value ( E\u200a=\u200a7 . 4 ) bordering on the classifier E value ., The Potts model classifier also performs well ( 81% accuracy at optimal threshold , AUROC\u200a=\u200a0 . 80 ) , but provides a slightly weaker difference between the fit and unfit classes ( Mann-Whitney U\u200a=\u200a70 , ) ( Figure 3B ) ., Here , seven mutants were not classified correctly , including the same four not classified correctly by the regularized Ising model as well as mutants 174G , 181R , 269E and 315G331R ., Similar to mutant 243P , mutants 174G , 181R and 269E were unfit ( RC\u200a=\u200a0 ) but had a predicted E values ranging from 7 . 3 to 7 . 7 , fairly close to that of the classifier E value ., In this study , we have substantially advanced our modeling approaches and tested the predictive power of these models by in vitro fitness measurements of HIV encoding various mutation combinations in the Gag protein ., The in vitro functional data are overall in strong agreement with the viral fitness landscape models and support the capacity of these models to robustly predict both continuous and “coarse-grained” measures of HIV-1 in vitro replicative fitness ., Performance of the regularized Potts and regularized Ising models here is similar , which is not unexpected as Gag in general is not highly mutable and the mutants tested here were the most common ones , making the binary approximation a fairly good assumption ., Indeed , in instances where the binary approximation is valid , we might encounter poorer performance from the Potts model relative to the Ising due to a diminished ratio of samples ( i . e . , sequences in the MSA ) to parameters ( i . e . , h and J values ) making robust numerical fitting of the former more challenging than the latter ., It is nevertheless encouraging that we are capable of fitting a significantly more complicated Potts model that retains residue-specific resolution without compromising the fidelity of our predictions ., Improved inverse Potts inference methods which better meet these numerical challenges may also improve performance of the Potts model with respect to the Ising model results ., Simple theoretical analysis suggests that models which differentiate between different mutant amino acids at the same site , like the Potts model employed here , will be necessary to make fitness predictions for highly mutable proteins such as Env and Nef , or to predict the fitness of sequences containing sites with mutations to less frequently observed amino acids ., Using a simple toy model , we show in Text S1 , Section 3 . 4 that the binary approximation ( Ising model ) has several potential deficiencies compared to a Potts model ., In particular , the Ising model generically overestimates the fitness of mutant sequences , particularly for sequences containing uncommon mutations ., Also , in the Ising case the inferred interaction between mutations at different sites is dominated by the interaction between the m | Introduction, Methods, Results, Discussion | Viral immune evasion by sequence variation is a major hindrance to HIV-1 vaccine design ., To address this challenge , our group has developed a computational model , rooted in physics , that aims to predict the fitness landscape of HIV-1 proteins in order to design vaccine immunogens that lead to impaired viral fitness , thus blocking viable escape routes ., Here , we advance the computational models to address previous limitations , and directly test model predictions against in vitro fitness measurements of HIV-1 strains containing multiple Gag mutations ., We incorporated regularization into the model fitting procedure to address finite sampling ., Further , we developed a model that accounts for the specific identity of mutant amino acids ( Potts model ) , generalizing our previous approach ( Ising model ) that is unable to distinguish between different mutant amino acids ., Gag mutation combinations ( 17 pairs , 1 triple and 25 single mutations within these ) predicted to be either harmful to HIV-1 viability or fitness-neutral were introduced into HIV-1 NL4-3 by site-directed mutagenesis and replication capacities of these mutants were assayed in vitro ., The predicted and measured fitness of the corresponding mutants for the original Ising model ( r\u200a=\u200a−0 . 74 , p\u200a=\u200a3 . 6×10−6 ) are strongly correlated , and this was further strengthened in the regularized Ising model ( r\u200a=\u200a−0 . 83 , p\u200a=\u200a3 . 7×10−12 ) ., Performance of the Potts model ( r\u200a=\u200a−0 . 73 , p\u200a=\u200a9 . 7×10−9 ) was similar to that of the Ising model , indicating that the binary approximation is sufficient for capturing fitness effects of common mutants at sites of low amino acid diversity ., However , we show that the Potts model is expected to improve predictive power for more variable proteins ., Overall , our results support the ability of the computational models to robustly predict the relative fitness of mutant viral strains , and indicate the potential value of this approach for understanding viral immune evasion , and harnessing this knowledge for immunogen design . | At least 70 million people have been infected with HIV since the beginning of the epidemic and an effective vaccine remains elusive ., The high mutation rate and diversity of HIV strains enables the virus to effectively evade host immune responses , presenting a significant challenge for HIV vaccine design ., We have developed an approach to translate clinical databases of HIV sequences into mathematical models quantifying the capacity of the virus to replicate as a function of mutations within its genome ., We have previously shown how such “fitness landscapes” can be used to guide the design of vaccines to attack vulnerable regions from which it is difficult for the virus to escape by mutation ., Here , using new modeling approaches , we have improved on our previous models of HIV fitness landscape by accounting for undersampling of HIV sequences and the specific identity of mutant amino acids ., We experimentally tested the accuracy of the improved models to predict the fitness of HIV with multiple mutations in the Gag protein ., The experimental data are in strong agreement with model predictions , supporting the value of these models as a novel approach for determining mutational vulnerabilities of HIV-1 , which , in turn , can inform vaccine design . | immunodeficiency viruses, infectious diseases, medicine and health sciences, medical microbiology, hiv, viral pathogens, genetics, microbial pathogens, biology and life sciences, immunology, microbiology, computational biology, viral diseases | null |
journal.pbio.1000432 | 2,010 | Multiple Molecular Mechanisms Cause Reproductive Isolation between Three Yeast Species | Reproductive isolation preventing gene flow between diverging populations is crucial for the process of speciation 1 ., One of the general reproductive isolation mechanisms that lead to hybrid inviability or sterility is genetic incompatibility ( Dobzhansky-Muller incompatibility ) , which is caused by improper interactions between genetic loci that have functionally diverged in two different species 2 , 3 ., Since genetic incompatibility probably plays an important role at the incipient stage of speciation , identifying the incompatible loci and determining the selection forces underlying their functional divergence are vital for our understanding of how speciation occurs ., In the past two decades , scientists have discovered several genetic loci causing hybrid sterility or inviability 4 , 5 ., Genes involved in genetic incompatibility have been cloned from a variety of organisms , including flies , platyfishes , mice , and Arabidopsis 6 , 7 , 8 , 9 , 10 , 11 ., Nonetheless , most of the genes identified were from Drosophila , and in most cases , only one component of the incompatible genetic loci was cloned ., Systematic studies that involve more than two species in other organisms are still rare ., Bakers yeast , Saccharomyces cerevisiae , and its close relatives , the Saccharomyces sensu stricto yeasts , represent an interesting system for studying genetic incompatibility ., These yeasts can mate with each other freely under laboratory conditions ., Diploid hybrids collected from the wild or generated in the laboratory can reproduce asexually without showing any obvious defect ., However , the viability of hybrid gametes ( spores ) is very low ( about 0 . 5%–1% ) , suggesting that there is strong postzygotic reproductive isolation between these yeast species 12 ., Because yeast differs from flies in cellular complexity , life style , and population structure , studies in yeast will greatly expand our knowledge of speciation processes 13 ., Using chromosome replacement lines of hybrids of two yeast species , a previous study identified a strong incompatibility between a S . bayanus nuclear gene , AEP2 , and S . cerevisiae mitochondria that leads to interspecific F2 hybrid sterility 14 ., It was found that the 5′-UTR regions of a mitochondrion gene , OLI1 , have diverged dramatically between these two species ., Since interactions between the Aep2 protein and the 5′-UTR region of the OLI1 mRNA are essential for OLI1 translation , the incompatibility is probably caused by the failure of Sb-Aep2 to recognize the divergent 5′-UTR region of Sc-OLI1 ., The finding raises a few interesting questions: Does the cytonuclear incompatibility play a general role in yeast reproductive isolation , does the AEP2-OLI1 type of interaction ( activation of mRNA translation ) represent a common mode of cytonuclear incompatibility , and is there a specific selective force driving this type of cytonuclear evolution ?, The mitochondrion is a critical component of cellular energy production and several metabolic pathways ., In many organisms including yeast , proper mitochondrial functions are required for gamete development 15 , 16 , 17 ., The mitochondrion contains its own genome , though one in an advanced state of degeneration ., Most genes essential for mitochondrial functions have been transferred from the proto-mitochondrion genome to that of their host 18 , 19 ., As a consequence of these events , gene products from both mitochondrial and nuclear genomes are required for proper mitochondrial operations 20 , 21 ., In yeast , for example , the mitochondrial genome encodes only eight proteins , but it is estimated that ∼1 , 000 proteins function in mitochondria 22 ., Although these two genomes are under different mutation and selection pressures , they are constrained to evolve coordinately to maintain optimal functions 23; any change in mitochondria ( adaptive or drifted ) may require one or more consecutive changes in the nucleus 23 , 24 ., This type of interaction provides an ideal background for the evolution of Dobzhansky-Muller incompatibilities; when two populations containing well-adapted cytonuclear mutations mix their genomes , unmatched mitochondria and nuclei cause reduced hybrid fitness 25 ., Cytonuclear incompatibility has been observed in a wide range of organisms , including primates , amphibians , flies , wasps , a marine copepod , and a variety of plants 26 , 27 , 28 , 29 , 30 , 31 ., In yeast , cytonuclear incompatibility has been tested directly by transferring mitochondria from one species or strain into cells of another , which clearly showed a species barrier between mitochondrial and nuclear genomes 32 , 33 ., From these studies , we know that the deleterious effects of cytonuclear incompatibility can lead to reduced fitness , hybrid sterility , or inviability ., Nonetheless , molecular descriptions of such an intracellular conflict are rare , and its generality as an engine of speciation remains an open question ., Here , we present results of a systematic study aimed at understanding the role of cytonuclear incompatibility in postzygotic reproductive isolation ., We screened three sensu stricto yeasts , S . cerevisiae , S . paradoxus , and S . bayanus for incompatible genes causing F2 hybrid sterility and used the information about how these genes diverged in function to reconstitute the evolutionary history of the species ., We found that only a few strongly incompatible gene pairs have evolved between these species ., Two of them , MRS1 and AIM22 , were identified and the molecular mechanisms of incompatibility were characterized ., By analyzing the mutations of the MRS1 gene leading to its functional divergence between two species at the nucleotide level , we show that only three mutations make major contributions ., Finally , we show that the functional divergence of these incompatible genes is correlated with the phylogeny , suggesting that cytonuclear incompatibility not only represents a general mechanism of reproductive isolation but has also occurred repeatedly during yeast evolution ., A previous study has shown that incompatibility between a S . bayanus nuclear gene , AEP2 , and S . cerevisiae mitochondria causes interspecific F2 hybrid sterility 14 ., To examine whether nuclear-mitochondrial incompatibility represents a general mechanism of reproductive isolation in yeast , a systematic screen for such genes was conducted in three closely related yeast species: S . cerevisiae ( Sc ) , S . paradoxus ( Sp ) , and S . bayanus ( Sb ) ., Hybrid diploid strains between S . cerevisiae and S . paradoxus , or between S . cerevisiae and S . bayanus were induced to generate haploid spores containing different combinations of chromosomes from their parental species ., These spores were then assayed for cytonuclear incompatibility ., In hybrid diploids between species A and B , incompatibility could occur in two directions , between A-nucleus and B-mitochondria or between B-nucleus and A-mitochondria ., When generating the yeast hybrids , we deliberately removed one parental type of mitochondria ( by using ρ0 mutants , which lack mitochondrial DNA ) so that we could unambiguously assign the direction of incompatibility ., After sporulation of hybrids , viable spores were measured for their ability to grow on glycerol , a non-fermentable carbon source ( Figure 1 ) ., The results showed strong cytonuclear incompatibility in most of the interspecific hybrids; we observed 66%±5% , 78%±7% , and 32%±16% of respiration-proficient spores in the interspecific crosses between Sc-ρ0 and Sp , Sb-ρ0 and Sc , and Sc-ρ0 and Sb , respectively , while the intraspecific crosses generated almost 100% of respiration-proficient spores ( Figure 2 ) ., The percentage of viable spores that could grow on glycerol plates was used to estimate the number of strongly incompatible loci ( see Materials and Methods ) ., The results of this growth assay suggest that there are only one or few nuclear genetic loci strongly incompatible with mitochondria in the Sc-nucleus and Sp-mitochondria , the Sb-nucleus and Sc-mitochondria , and the Sc-nucleus and Sb-mitochondria pairs ., Although no strong incompatibility was detected for the Sp-nucleus and Sc-mitochondria pair , slow spore growth ( on glycerol plates ) was commonly observed , suggesting that some incompatibility may exist in this pair as well ., Because the incompatibility between the Sb-nucleus and Sc-mitochondria has been described earlier 14 , we focus here on the remaining two pairs ., In our experimental design , we constructed the hybrid diploids using spo11Δ mutants to prevent meiotic homologous recombination ., We expected that hybrid F1 haploids unable to respire ( see Figure 1 ) should carry a specific set of chromosomes containing the incompatible genes ., To identify the genes responsible for the observed cytonuclear incompatibility , the respiration-deficient clones were first examined for their chromosome contents using species-specific PCR ( see Materials and Methods ) ., In the cross between Sc-ρ0 and Sb , the haploids carrying Sb-mitochondria but unable to respire lacked Sb-Chromosome 6 , 9 , or both ( Table S1A ) ., When these hybrid clones were transformed with genomic DNA libraries to screen for those genes capable of rescuing the respiratory defect , Sb-MRS1 encoded on Sb-Chromosome 9 and Sb-AIM22 encoded on Sb-Chromosome 6 were isolated ., The results from the cross between Sc-ρ0 and Sp are more complex , as all the respiration-deficient clones did not contain the two Sp-Chromosomes 4 and 9 ( Table S1B ) , but were fully rescued by Sp-MRS1 alone , which is encoded on Sp-Chromosome 9 ., The reason why all the hybrid clones are missing Sp-Chromosome 4 is unclear ., One possibility is that Sp-Chromosome 4 is also incompatible with Sc-Chromosome 9 , an issue requiring further investigation ., To rule out the possibility that the respiration defects caused by Sc-MRS1 and Sc-AIM22 were simply due to nuclear-nuclear incompatibility in the hybrid haploid clones , we crossed Sp-mrs1Δ , Sb-mrs1Δ , and Sb-aim22Δ with Sc-ρ0 and examined the respiratory ability of the diploid cells ., All the diploid cells were still respiration-deficient , even though they contained a complete set of the S . cerevisiae genome ( Figure 3 ) ., This result demonstrated that Sc-MRS1 is incompatible with Sp-mitochondria and that both Sc-MRS1 and Sc-AIM22 are incompatible with Sb-mitochondria ., Yeast cells utilize non-fermentable carbon sources to induce meiosis ., Previous studies have shown that respiration-deficient cells were unable to sporulate 34 ., To confirm that indeed the cytonuclear incompatibility observed in our experiments contributes to reproductive isolation , the aforementioned diploid cells ( Sp-mrs1Δ×Sc-ρ0 , Sb-mrs1Δ×Sc-ρ0 , and Sb-aim22Δ×Sc-ρ0 ) were grown on sporulation medium and examined for their sporulation efficiency ., No ascus was observed in these cultures , while the control cultures sporulated efficiently ., Thus , cytonuclear incompatibility caused by Sc-MRS1 or Sc-AIM22 results in reproductive isolation between these yeast species ., Mrs1 is a mitochondrial protein required for excision of the aI5β intron in COX1 and the bI3 intron in COB 35 , 36 ., A previous study has shown that S . douglasii Mrs1 ( S . douglasii is a synonym of S . paradoxus ) is required to splice a S . douglasii-specific COX1 intron not existing in the S . cerevisiae COX1 37 ., We observed a similar result in S . bayanus ., Sc-MRS1 could not complement the respiratory defect of the Sp-mrs1Δ or Sb-mrs1Δ mutants ( Figure S1 ) ., When Sb-Mrs1 or Sp-Mrs1 were replaced by Sc-Mrs1 , the level of mature COX1 mRNA was drastically reduced but the COB mRNA was not affected ( Figure 4A ) ., We further analyzed the translation products from purified mitochondria and confirmed that only the Cox1 protein was missing ( Figure 4B ) ., By contrast , Sc-MRS1 transcription and protein transport into mitochondria appear to be normal in both S . paradoxus and S . bayanus ( Figure 5 ) ., Thus the incompatibility between the mitochondrial and nuclear genomes is most likely due to a change in the splicing specificity of Mrs1 that occurred after S . cerevisiae diverged from the common ancestor of S . cerevisiae and S . paradoxus ., In order to understand how the COX1 introns evolved in yeast , we compared COX1 intron patterns between S . cerevisiae , S . paradoxus , S . bayanus , S . servazzii , Candida glabrata , and Kluyveromyces thermotolerans ., S . servazzii and C . glabrata are species outside the sensu stricto complex and K . thermotolerans is a pre-WGD ( whole-genome duplication ) species ., The comparison indicates that the intron in the Sp- or Sb-COX1 gene incompatible with Sc-Mrs1 is an ancient intron ( Figure 4C ) ., Since this intron was eliminated only in the S . cerevisiae lineage , it is likely that the Sc-MRS1 gene product lost the ability to splice this intron after the intron loss event ( by adaptation or drift ) ., The fact that S . cerevisiae and S . paradoxus share a high degree of nucleotide sequence identity allowed us to determine the key mutations underlying the functional change of the MRS1 gene and to reconstruct the process of Mrs1 evolution ., To this end , chimeric proteins with regions from Sc-Mrs1 and Sp-Mrs1 were constructed and assayed for their ability to complement the respiratory defect of the Sp-mrs1Δ mutants ., We found that the functional difference between Sc-Mrs1 and Sp-Mrs1 is mainly determined by a region comprising 63 amino acids ( a . a . sites 179–241; Figure 6A ) ., Nine nonsynonymous changes have accumulated in this region since Sc-MRS1 and Sp-MRS1 diverged ( Figure S2A ) ., To determine which of these mutations led to the altered activity of Mrs1 , we introduced the S . paradoxus version of each of these sites into Sc-MRS1 and assayed their ability to rescue the Sp-mrs1Δ respiratory defect ., Analogous experiments were performed in the reverse direction by introducing Sc-specific amino acids into Sp-MRS1 ., Only mutations in three amino acids ( Sc to Sp: T201A , V211A , and M227I ) had obvious contributions ( Figure 6B and Figure S2 ) ., When all three mutations were combined together in a single mutant clone , it explained most of the effect ., The growth rate of cells carrying the mutant plasmid ( Sc-MRS1-n123 ) in a glycerol-containing medium is about 75% of that of wild type cells ., Interestingly , these three amino acids are all conserved in S . paradoxus , S . kudriavzevii , and S . bayanus but are changed in S . cerevisiae ., Among the other six nonsynonymous changes in this region , only two of them ( Sc to Sp: K186E and R223Q ) share the same pattern ., This observation is consistent with our hypothesis that the functional change of MRS1 occurred only after an ancestral COX1 intron was lost in S . cerevisiae ., Our results clearly demonstrate that the observed nuclear-mitochondrial incompatibility results from cumulative effects of multiple mutations ., On the other hand , they also suggest that only a small fraction of the nonsynonymous changes between species contributes to the incompatibility ., The Mrs1 protein does not contain any specific functional domain ., However , a recent study has used computer modeling to predict the Mrs1 protein structure 38 ., We examined the relative positions of these residues using the predicted structure ., All three residues ( a . a . 201 , 211 , and 227 ) were found to localize on the RNA-binding surface ( Figure S3 ) ., It is possible that these amino acid changes have altered the substrate specificity of Mrs1 that leads to the incompatibility ., S . cerevisiae nuclei can only support S . bayanus mitochondria if the cells contain a S . bayanus AIM22 gene ., AIM22 encodes a lipoate-protein ligase homologous to the bacterial lplA protein 39 , 40 ., In eukaryotic cells , lipoic acid has been shown to be an essential cofactor to a variety of mitochondrial proteins and lipoate-protein ligase ( together with other enzymes ) is required to lipoylate these mitochondrial targets 39 , 41 ., We have not investigated which mitochondrial protein or enzyme in Sb-mitochondria is incompatible with Sc-Aim22 ., However , our data indicate that the incompatibility is not caused by misregulation of the Sc-AIM22 transcription or failure to transport the Sc-Aim22 to Sb-mitochondria ( Figure 5 ) ., To investigate whether the AIM22 gene in the S . cerevisiae-S ., paradoxus branch has been under positive selection during evolution , we measured the ratio of nonsynonymous ( Ka ) to synonymous ( Ks ) nucleotide substitution rates between these species ., The Ka/Ks values of AIM22 in the S . cerevisiae–S ., paradoxus , S . cerevisiae–S ., bayanus , and S . paradoxus–S ., bayanus pairs are 0 . 13 , 0 . 12 , and 0 . 14 , showing no sign of positive selection ( Ka/Ks>1 ) ., We also ran a PAMLs branch-model analysis on the AIM22 gene 42 , 43 but could not detect any signature of significant positive selection ( Figure S4 ) ., Previous studies of AEP2 have shown that the nuclear-mitochondrial incompatibility is asymmetrical ., While Sb-Aep2 is completely incompatible with Sc-mitochondria , Sc-Aep2 retains partial compatibility with Sb-mitochondria 14 ., We also found that incompatibility caused by AIM22 and MRS1 only occurred in one direction ., Although Sc-AIM22 and Sc-MRS1 are not compatible with Sb-mitochondria , Sb-AIM22 could complement the Sc-aim22Δ mutant and both Sb- and Sp-MRS1 could rescue the Sc-mrs1Δ mutant ., Nuclear-mitochondrial incompatibility has been shown to occur commonly between different yeast species 32 , 33 ., However , it is unclear whether nuclear-mitochondrial incompatibility between different pairs of species has evolved at different periods of time in different species lineages ., To address this issue , we tested the compatibility between different orthologues of these incompatible genes and mitochondria from each species ., Different orthologous alleles of MRS1 , AIM22 , or AEP2 were transformed into the S . cerevisiae , S . paradoxus , or S . bayanus mutants in which the wild-type copy had been deleted ., The transformants were then tested for their ability to grow on glycerol plates ( Figure 7A and Figure S1 ) ., Information from these assays was used to deduce the time of occurrence of the functional change leading to incompatibility ., A clear correlation between the emergence of cytonuclear incompatibility and the phylogeny is observed ( Figure 7B ) ., Sc-MRS1 is incompatible with Sp-mitochondria or Sb-mitochondria , indicating that the functional change of Sc-MRS1 occurred only in the S . cerevisiae lineage ., On the other hand , the functional change of AIM22 represents a more ancient event in the common ancestor of S . cerevisiae and S . paradoxus , because Sc-AIM22 and Sp-AIM22 are exchangeable but neither of them is compatible with Sb-mitochondria ., Finally , the data suggest that Sb-AEP2 diverged in function only in the S . bayanus lineage since Sb-AEP2 is incompatible with either Sc-mitochondria or Sp-mitochondria ., These results provide evidence that nuclear-mitochondrial incompatibility has repeatedly arisen during the history of yeast evolution and probably represents an important reproductive isolation mechanism in yeast species ., Previous studies in Saccharomyces yeasts have suggested that the deleterious effect of DNA sequence divergence on meiotic recombination probably contributes in part to reproductive isolation during yeast evolution 12 , 44 , 45 ., Our results , on the other hand , suggest that nuclear-mitochondrial incompatibility is also a promising candidate for causing intrinsic hybrid dysfunction ., In fact , these two mechanisms are not mutually exclusive ., Genomes from different populations will accumulate enough DNA sequence divergence only after extended periods of allopatric isolation , but the effect of sequence divergence can be applied directly in the diploid F1 hybrid cells ., Cytonuclear incompatibility can be achieved by only a few mutations , and its deleterious effect can be carried on to the F1 gamete or F2 progeny ., In theory , cytonuclear incompatibility has a stronger impact on blocking gene flow between populations in the early stages of speciation , and reproductive isolation is reinforced later on when populations have accumulated enough DNA sequence divergence ., These two mechanisms can be complementary to each other in terms of their effects and evolutionary trajectories ., It will be interesting to investigate whether cytonuclear incompatibility exists between different populations of the same species ., Reproductive isolation resulting from genetic incompatibility has been discovered in a variety of organisms 7 ., Most of the examples characterized so far are caused by interactions between nuclear genes ., In yeast , this type of incompatibility has been investigated in a few studies , yet no strongly nuclear-nuclear incompatible genes were identified 14 , 46 ., On the other hand , cytonuclear incompatibilities were observed in hybrids between different yeast species or populations 47 ., Cytonuclear incompatibility probably represents a more general mechanism of reproductive isolation in yeast ., By analyzing the functions and the interacting components of the identified incompatible genes , we discovered that cytonuclear incompatibility could be achieved by multiple molecular mechanisms: intron splicing , protein lipoylation , and activation of mRNA translation ., This suggests that cytonuclear incompatibility in yeast can occur in various pathways by diverse molecular mechanisms ., Scientists usually compare orthologous sequences from different species and use the detected molecular signatures to infer the evolutionary process of a gene ., Since our data suggest that Mrs1 changed function only after S . cerevisiae diverged from S . paradoxus , it is reasonable to assume that the altered function originated from amino acid changes that occurred specifically in the S . cerevisiae lineage ., By comparing the coding regions of the MRS1 orthologues from S . cerevisiae , S . paradoxus , S . kudriavzevii , and S . bayanus , we observed 22 amino acids that are different in S . cerevisiae but are conserved in the other three species ., Interestingly , our functional assays showed that only three of these amino acid changes contribute significantly to the functional differences of Mrs1 ., The other mutations may have very minor effects unable to be detected by our functional assays or have been fixed simply by genetic drift ., In a previous study , Rawson and Burton have also observed that three amino acid changes in a nucleus-encoded cytochrome c ( CYC ) are responsible for cytonuclear incompatibility between different populations of a marine copepod , Tigriopus californicus 48 ., These results illustrate the importance of mapping the critical amino acid changes in order to understand how a gene evolved ., What is the major driving force underlying the evolution of mitochondria ?, It is known that the mitochondrial genome suffers a higher mutation load because it is constantly facing higher levels of oxidative reagents and its DNA protection system is more primitive as compared to the nuclear genome ., With a much smaller copy number of mitochondrial DNA in yeast ( 30–80 molecules of mitochondrial genomes in yeast cells compared to 1 , 000–5 , 000 molecules in animal cells ) , mutations may be fixed frequently in the mitochondrial genome by genetic drift ., Since wild yeast often propagate clonally in natural environments 13 , a founder cell with mild deleterious mutations in its mitochondrial genome may have a chance to accumulate suppressors in the mitochondrial or nuclear genomes to rescue the fitness before its progeny are outcompeted by cells from another population ., Alternatively , mitochondrial evolution may be driven by an “arms race” process between selfish mitochondrial DNA and the “wild-type” mitochondrial or nuclear genomes ., It is commonly observed in yeast that by manipulating the host replication or segregation systems , some mitochondrial genomes allow themselves to be inherited more efficiently , even though they may be carrying compromised respiratory functions 49 ., In a sexual population , the other mitochondrial or host nuclear genomes will be selected to counteract this selfish behavior or the deleterious effects carried by the selfish DNA ., Such an “arms race” may allow mitochondrial genomes to evolve faster than by genetic drift ( in a fashion similar to positive selection ) ., Incompatibilities driven by arms races between different genetic components have been suggested in a few recent studies 7 ., Among these genetic conflicts , most of them are caused by selfish elements manipulating segregation distortion 11 , 50 , 51 , 52 , 53 ., The “hybrid necrosis” phenotype observed in Arabidopsis probably results from recurrent conflicts between the host defense system and pathogens 10 ., The genetic conflict caused by selfish mitochondrial genomes may represent another type of arms race ., It will be interesting to investigate whether the arms race model can explain the cytonuclear incompatibilities observed in other organisms ., Finally , ecological adaptation may also contribute to mitochondrial evolution ., Evidence suggests that adaptive mutations have occurred in mitochondria in response to different environmental stresses that interfere with cellular energy demands 48 , 54 , 55 ., In yeast , it has been shown that S . bayanus grows much better than S . cerevisiae on media containing only non-fermentable carbon sources , with the opposite observed in fermentable media ., It has been speculated that the changes in Sb-AEP2 and Sb-OLI1 are a part of such ecological adaptation 14 ., Reciprocal crosses between species often generate asymmetrical hybrid viability or sterility , a general feature of intrinsic postzygotic isolation called Darwins corollary 56 , 57 , 58 , 59 , 60 ., The Dobzhansky-Muller model suggests that alleles causing reproductive isolation act asymmetrically 61 ., However , asymmetries in allele action do not necessarily lead to asymmetries in reproductive isolation 62 ., Incompatibility between autosomal loci affects both reciprocal crosses identically ., Asymmetric reproductive isolation is usually caused by incompatibility between autosomal loci and uniparentally inherited materials , such as cytoplasmic elements and sex chromosomes ., It has been suggested that cytonuclear incompatibility caused by different trajectories of mitochondrial evolution in different species may contribute to this phenomenon 59 , 63 ., Our results provide the molecular basis to support this hypothesis ., Since cytonuclear incompatibility can be achieved by multiple molecular mechanisms and evolve at different rates in different lineages , it can serve as a general mechanism of reproductive isolation and also create asymmetrical reproductive isolation between species ., Yeast strain genotypes are listed in Table S2 ., The parental S . cerevisiae strains ( JYL1127 and JYL1128 ) are isogenic with W303 ( MATa ura3-1 his3-11 , 15 leu2-3 , 112 trp1-1 ade2-1 can1-100 ) ., The parental S . paradoxus strains ( JYL1137 and JYL1138 ) are derived from YDG 197 and are a gift from Dr . Duncan Greig ( University College London , UK ) ., The parental S . bayanus strains ( JYL1030 and JYL1031 ) were derived from a strain ( S . bayanus #180 ) collected by Dr . Duccio Cavalieri ( University of Florence , Italy ) ., The strains JYL1157 , 917 , and 1256 were used for measuring hybrid fertility ., Substitutive and integrative transformations were carried out by the lithium acetate procedure 64 ., Media , microbial , and genetic techniques were as described 65 ., a and α cells of one species ( S . cerevisiae , S . paradoxus , or S . bayanus ) were crossed to α and a ρ0 cells of another species to generate F1 hybrid cells ., In both species , the SPO11 genes were deleted to prevent meiotic recombination between homologous chromosomes ., Strains from the first species also had a URA3 marker inserted near the centromere of a chromosome so that haploid spores could be efficiently selected on 5-FOA plates ., After the hybrid diploids were induced to sporulate , viable spores were tested for their respiratory ability ., Respiration-deficient clones were further crossed with ρ0 mutants of the parental species to confirm that the defect was caused by cytonuclear genomic incompatibility ( Figure 1 ) ., A genetic analysis was used to estimate the number of the nuclear genes that are incompatible with mitochondrial DNA from another species: in a cross between two spo11Δ mutants , meiotic recombination does not occur and homologous chromosomes segregate randomly 66 , 67 ., For a specific chromosome A , 25% of the spores will carry two A chromosomes ( of both parental types ) , 50% of them will carry one A chromosome , and 25% of them will have no A chromosome ( these cells will not survive so they will not be counted in our later analysis ) ., If only one gene on chromosome A is recessively incompatible with mitochondrial DNA ( in our case , all spores from a single cross carry the same parental type of mitochondrial DNA ) , we expect to see that 66% of the viable spores are Gly+ ( 1/3+2/3×1/2\u200a=\u200a66% ) ., The spores carrying two A chromosomes should be Gly+ because the incompatibility is recessive ., Half of the spores that carry only one A chromosome should be Gly+ if their A chromosome is from the same parent of the mitochondrial DNA ., If two genes on different chromosomes are involved , we expect to see that 43% of the viable spores are Gly+ ( 66%×66%\u200a=\u200a43% ) ., Epistatic effects are not taken into account in this analysis because considering such effects would make it impossible to estimate the involved locus number ., We examined the chromosome composition of respiration-deficient hybrid lines by PCR using species-specific primers for all 16 chromosomes ., Yeast genomic DNA libraries constructed from S . bayanus or S . paradoxus genomes were transformed into the Gly− clones to screen for incompatible genes ., Yeast strains were grown in 3 ml YPD liquid cultures at 30°C to stationary phase and total RNA was isolated using Qiagen RNeasy Midi Kits ( Qiagen , Valencia , CA ) ., Ten µg of total RNA was separated on a 1 . 3% agarose-formaldehyde gel and then transferred to a nylon membrane ( Millipore , Billerica , MA ) ., Northern blotting was performed as described 68 ., Because both COX1 and COB signals were difficult to be completely washed out and the background was high after the second hybridization , we used the same RNA samples to load three repeats on the same gel , cut the membrane after transferring , and hybridized each repeat with one specific probe ., The gene-specific primers for the DIG-labeled probes were described in Rodeheffer et al . 69 except for the probe of COX2 ., The primers for COX2 were 5′-TTAATGATAGTGGTGAAACTGTTG-3′ and 5′-CCAAAGAATCAAAATAAATGCTCG-3′ ., The probe generation and hybridization were as described in the Genius System Users Guide ( Roche , Indianapolis , IN ) ., Total RNA was isolated using Qiagen RNeasy Midi Kit ( Qiagen ) ., First-strand cDNA was synthesized using High Capacity cDNA Reverse Transcriptase Kit ( Applied Biosystems , Foster City , CA ) at 37°C for 2 h ., A 25-fold dilution of the reaction products was then subjected to real-time quantitative PCR analysis using gene-specific primers , the SYBR Green PCR master mix , and an ABI-7000 sequence detection system ( Applied Biosystems ) ., Data were analyzed using the built-in analysis program ., Yeast cells were cultured in selection medium at 30°C to the mid-exponential phase ., Mitochondria fractions were prepared as described 70 ., The post-mitochondrial supernatant ( PMS ) fractions were collected immediately following centrifugation of mitochondria ., The PMS was precipitated with ice-cold 10% | Introduction, Results, Discussion, Materials and Methods | Nuclear-mitochondrial conflict ( cytonuclear incompatibility ) is a specific form of Dobzhansky-Muller incompatibility previously shown to cause reproductive isolation in two yeast species ., Here , we identified two new incompatible genes , MRS1 and AIM22 , through a systematic study of F2 hybrid sterility caused by cytonuclear incompatibility in three closely related Saccharomyces species ( S . cerevisiae , S . paradoxus , and S . bayanus ) ., Mrs1 is a nuclear gene product required for splicing specific introns in the mitochondrial COX1 , and Aim22 is a ligase encoded in the nucleus that is required for mitochondrial protein lipoylation ., By comparing different species , our result suggests that the functional changes in MRS1 are a result of coevolution with changes in the COX1 introns ., Further molecular analyses demonstrate that three nonsynonymous mutations are responsible for the functional differences of Mrs1 between these species ., Functional complementation assays to determine when these incompatible genes altered their functions show a strong correlation between the sequence-based phylogeny and the evolution of cytonuclear incompatibility ., Our results suggest that nuclear-mitochondrial incompatibility may represent a general mechanism of reproductive isolation during yeast evolution . | Hybrids between species are usually inviable or sterile , possibly due to functional incompatibility between genes from the different species ., Incompatible genes are hypothesized to encode interacting components that cannot function properly when paired with alleles from another species ., To understand how incompatible gene pairs result in hybrid sterility or inviability , it is important to identify these genes and reconstruct their evolutionary history ., A previous study has shown that incompatibility between nuclear and mitochondrial genomes ( cytonuclear incompatibility ) causes hybrid sterility between two yeast species ., To expand on these findings , we screened three yeast species for genes involved in cytonuclear incompatibility , discovering two nuclear genes , MRS1 and AIM22 , which encode proteins that are unable to support full mitochondrial function in the hybrids ., Of these two genes , Mrs1 is required for removing a specific intron in the mitochondrial COX1 gene ., By comparing different yeast species , we find a clear coevolutionary relationship between Mrs1 function and the COX1 intron pattern ., We also show that changes in three amino acids in the Mrs1 RNA-binding domain are sufficient to make Mrs1 incompatible in hybrids ., Our results suggest that cytonuclear incompatibility may represent a general mechanism of reproductive isolation during yeast evolution . | evolutionary biology, genetics and genomics/microbial evolution and genomics, evolutionary biology/microbial evolution and genomics, evolutionary biology/evolutionary and comparative genetics | Incompatibility between nuclear and mitochondrial genomes in yeast species may represent a general mechanism of reproductive isolation during yeast evolution. |
journal.pbio.1000343 | 2,010 | Binding Site Turnover Produces Pervasive Quantitative Changes in Transcription Factor Binding between Closely Related Drosophila Species | Despite four decades of interest in the evolution of transcriptional regulation , we still have a poor understanding of the molecular bases for regulatory divergence and the constraints under which cis-regulatory sequences evolve ., Most regulatory sequences appear to be under strong selection to maintain their transcriptional output , and as a result , binding sites for the sequence-specific transcription factors that regulate mRNA synthesis are preferentially conserved 1 , 2 ., However , even in regulatory sequences with highly conserved function , transcription factor binding sites can be gained and lost over time at a high rate , leading to considerable differences in the composition and arrangement of binding sites between even closely related species 2–10 ., Whether and how this binding site turnover affects transcription factor binding , and what the consequences of changes in binding on transcription might be , remains unknown ., After years in which the study of regulatory evolution was primarily a computational exercise , a series of recent studies have compared genome-wide in vivo binding of transcription factors in the same conditions or tissues of related species 11–14 ., Among yeasts of the genus Saccharomyces 11 , 12 and between human and mouse 13 , 14 , a substantial fraction of experimentally observed interactions between transcription factors and DNA are species-specific ., While these differences could , in principle , be due to divergence of transcription factors and other trans-acting factors , binding differences appear to be driven primarily in cis 13 , suggesting that differences in the sequences , and not the factors binding to them , drive the divergence in binding ., Species-specific binding is generally associated with the gain/loss of sequence motifs recognized by the relevant factor 11 , 14 , although the correlations are weak ., Here we examine how the binding of a group of six factors that direct temporal and spatial patterns of gene expression along the anterior-posterior ( A-P ) axis during early development differs between Drosophila melanogaster and its sister species D . yakuba ., These two species , whose genomes have been fully sequenced 15 , 16 , diverged only five million years ago 17 ., They are separated by a molecular distance less than half that between mouse and human 18 , and D . yakuba orthologs of virtually all D . melanogaster genomic regions can be readily identified and aligned ., Though there are some subtle changes in the levels of expression of key regulators between these species ( our unpublished data ) , there is little difference in either their spatial expression patterns or those of their targets , a product at least in part of strong selection to maintain them 10 ., In our earlier work on the binding of these factors in D . melanogaster , we showed that they bind to an overlapping set of thousands of genomic regions in vivo 19 , 20 , as has subsequently been observed for many other animal transcription factors 21 ., A wealth of evidence suggests that , at least in D . melanogaster , and probably generally , only the several hundred most highly bound regions are directly involved in transcriptional regulation , with the remainder having a different , or more likely no , function 19 , 20 ., Thus these two fly species provide an ideal opportunity to study the effects of modest sequence divergence on transcription factor binding , its origins in changes in genomic sequence , and its functional consequences ., We expected binding differences between D . melanogaster and D . yakuba to be more modest than those observed between mouse and human , or between Saccharomyces species ., However , we hoped that the more modest differences in their genomes would improve our ability to associate sequence and binding divergence , and that our earlier work establishing the relationship for these factors between binding levels and regulatory function would provide an invaluable context for analyzing the functional consequences of the binding differences we observe ., Unlike in the yeast and mammalian studies described above , the gain or loss of bound regions between D . melanogaster and D . yakuba was rare , with fewer than 1% to 5% of peaks ( depending on the factor ) found in one species clearly absent or displaced in the other ( Table 2 ) ., The rate of gain/loss near known targets of the A-P factors was similar to the genome-wide rate ( Table 2 ) ., The measured binding at orthologous regions bound in both species varied considerably ( Figures 2 , S5 , and S6 ) both in the highly bound regions that our previous studies suggested are functional targets of these factors 19 , 20 and in the poorly bound regions that likely are not ., The more highly bound regions showed a greater total variation in binding ( Figure S7 ) , with the normalized divergence ( difference in binding over average binding level ) roughly constant across binding levels ( Figures 3 and S8 ) and relative to annotations ( Figure S9 ) ., The divergence was marginally lower within the 44 characterized D . melanogaster cis-regulatory modules ( CRMs ) known to be targeted by one or more of these factors ( correlation rA-P from 0 . 62 to 0 . 91 compared to 0 . 57 to 0 . 75 ) 27 and in peaks near genes ( within 10 Kb of the 5′ end ) known to be regulated by these A-P factors ( correlation rA-P from 0 . 59 to 0 . 92 , depending on the factor ) ., We sought to determine the extent to which sequence changes in the bound regions drove quantitative differences in binding ., We first examined overall measures of sequence divergence ., Levels of single-nucleotide divergence ( sequence identity ) and frequency of insertions and deletions in the 100 base pairs centered on the inferred peak of binding exhibited only low to moderate correlations with binding divergence ( 0 . 07 to 0 . 24; Figures S10 and S11 ) , consistent with our expectation that changes to specific short sequences , rather than entire regions , would have a disproportionate effect on binding ., We next sought to identify short sequences ( e . g . , transcription factor binding sites ) whose gain or loss was associated with changes in binding levels ., We devised an unbiased statistical approach that assessed the impact on binding of changes to a short sequence ( word ) by comparing the distribution of binding intensities in all bound regions where the word was conserved to the distribution in all bound regions where the word was present in one species but not the other ( defining bound regions as the 100 bp centered on peaks of maximal binding intensity ) ., If alterations to a word affect binding , then these distributions should be different ., We identified such words ( which we call divergence-driving words , or DDWs ) by comparing the conserved and non-conserved distributions for all 16 , 384 words of length 7 bp and picking those that showed a statistically significant difference ., We found DDWs for four of the six factors , and in each case , virtually all of these DDWs matched the known sequence specificities of the corresponding factor ( Figure 4 ) ., To quantify the fraction of binding divergence that is explained by the DDWs , we developed a method that used the gain and loss of DDWs to predict binding divergence between the species ., For each factor for which we had identified DDWs , we built a simple linear model relating the divergence of DDWs in a bound region to interspecies difference in binding at that bound region ., In the model , each divergent DDW in a bound region contributed a fixed amount to the predicted binding difference , with the effect of multiple divergence DDWs adding independently ., The contribution of each DDW was determined by a regression using the least angle regression method 28 with extensive cross-validation ( see Methods ) ., The correlations between predicted and observed divergence in binding of single factors across all peaks with at least one DDW in the two genomes ranged from 0 . 3 for HB to 0 . 41 for BCD ( Figures S12–S27 ) ., While far from perfect , these correlations demonstrate that changes in a highly restricted collection of sequences ( for example , BCD has only a single 7 bp DDW ) drive an appreciable fraction of binding divergence between species ., We additionally performed the same predictions using words derived from the in vitro factor binding specificities described by 29 ., The correlations between predictions and observations ranged from 0 . 18 for HB to 0 . 39 in BCD , similar to or lower than the correlations resulting from our DDWs ( unpublished data ) ., We investigated whether the lack of a strong relationship between probable enhancer function and quantitative conservation of binding was associated with similar trends at the sequence level ., For each factor for which we identified DDWs , we quantified motif enrichment and conservation as a function of the level of transcription factor occupancy in D . melanogaster ., Motif enrichment and conservation were elevated within bound regions above background levels across the genome ( Figure 5 ) ., The fraction of peaks with motifs showed a weak dependence on binding levels , with the most strongly bound regions exhibiting the greatest density of motifs ., The level of conservation of these motifs was weakly correlated with overall binding levels , consistent with our observation that quantitative divergence in binding strength decreased slightly near genes regulated by these factors ., In our initial comparison of binding between species , we noticed that increases in binding of a single factor were often correlated with increases in binding of many other factors ( Figures S28–S33 ) ., For example , changes in the binding of KR correlated with changes in the binding of other factors with r\u200a=\u200a0 . 36 ( KNI ) to 0 . 62 ( CAD ) , and such coordinated changes are recapitulated for all pairs of factors ., This widespread correlated change suggests a factor-independent mode of binding divergence ., To obtain an unbiased assessment of the extent of these correlated changes in binding , we quantified binding divergence for all six factors in all regions significantly bound by any factor and performed principal component analysis ( PCA ) , a method for analyzing variation between many factors simultaneously rather than only pairs of factors , on these data ( Figure 6A ) ., The first principal component , which represents the most significant axis of variation in the dataset , has the same direction and similar magnitude for all six factors , demonstrating that a pan-factor coordinated binding shift is the dominant driver of A-P factor binding divergence ( this principle component explains 38% of the overall variation in binding between the species ) ., A similar effect was observed when we performed PCA on the binding levels in each species independently ( Figure 6B and 6C ) , suggesting that a common effect is responsible for much of the variation in binding both between species and within a single genome ., The single-genome PCA revealed several interesting factor-specific correlations: increases in binding of the repressor GT are associated with decreases in binding of the activator HB ( PC2 in Figure 6B ) , increases in HB are associated with decreases in BCD ( PC3 in Figure 6B ) , etc ., As expected , given the overall similarity of binding between the species , the single-genome PCA analyses of D . melanogaster and D . yakuba yielded essentially identical results ., To investigate whether the features captured by these different principal components are related to specific sequences , we applied the same motif discovery method described above to projections of the binding data along each of the principal components shown in Figure 6A ., We discovered substantially more motifs in this analysis ( Figure 7 ) than in the single-factor analyses , likely because of the increased statistical power derived from considering all regions bound by any , as opposed to a single , factor ., Interestingly , one of the words whose divergence is associated with the first principal component is the “TAGteam” motif , CAGGTAG 30 , the binding site for Zelda , an activator of the early zygotic genome 31 ., Zeldas mechanism of action is unknown , but the strong correlation between gain and loss of its binding site with variation in changes in binding of all factors supports a direct or indirect role for Zelda in nucleosome positioning and chromatin remodeling ., Although D . melanogaster and D . yakuba are closely related , we were not always able to accurately identify orthologous sequences , largely due to ambiguities in the draft D . yakuba assembly ., Even where the orthology of regions was unambiguous , and despite this close evolutionary distance , base-level alignments were frequently uncertain ., Our analysis of sequence-specific effects required a precise alignment , and inevitable alignment errors will make nucleotide-level analysis of regulatory changes challenging for more distantly related species ( although the alignment accuracy estimates produced by FSA may help to identify reliably aligned loci ) ., Several aspects of this experiment should help direct future efforts to use comparative ChIP-Seq to study the relationship between sequence and binding divergence ., The widespread quantitative binding divergence between D . melanogaster and D . yakuba demonstrates that even relatively similar species can be used to study binding changes ., Indeed , given the magnitude of the binding divergence that we observe , we expect there to be quantitative differences between D . melanogaster and more closely related species , such as D . simulans , as well as among D . melanogaster individuals ., While comparisons with more distantly related species will likely reveal greater binding divergence , and will help explain how such divergence affects expression and phenotype , the difficulties with aligning genomes at this distance , and comparing embryonic stages , may render sequence-based analyses less powerful ., Even though we were working with very similar organisms , with similar timing and structure of embryonic development , there were undoubtedly subtle differences in our sampling of developmental stages in the two species ., Because transcription factor binding is dynamic , such sampling differences have the potential to manifest themselves as apparent interspecies differences in binding ., We do not believe this effect was significant in our data , however , as it is unlikely that this type of false-positive binding divergence would be associated with the specific sequence changes that we repeatedly observed ., Nonetheless , this will be a major difficulty in future studies , especially when developmentally and morphologically different organisms are compared , as precisely those changes that make such comparisons interesting also make them far more difficult ., Both D . melanogaster and D . yakuba embryos were collected from population cages for 1 h , and then allowed to develop to late stage 4 and early stage 5 before being harvested and fixed with formaldehyde ., The embryos from the two species developed very similarly , and the aging times to reach the desired age were 2 h for D . melanogaster embryos and 1 h and 45 min for D . yakuba embryos ., The staged embryos were harvested and cross-linked with formaldehyde , and the chromatin was isolated through CsCl gradient ultracentrifugation essentially as previously described 19 ., The chromatin used for immunoprecipitation was fragmented through sonication using a Branson Sonifier 450 to an average fragment size of 225 to 250 bp , which is shorter than the average size of chromatin used in our previous ChIP-chip experiments 19 ., ChIP was carried out using affinity purified rabbit polyclonal antibodies , and for two of the factors , HB and KR , two affinity purified antibodies that recognize non-overlapping parts of each factor were used ., These antibodies and the ChIP procedure were identical to those described in 19 ., The DNA libraries for sequencing were prepared from the ChIP reaction and from Input DNA following the Illumina protocol for preparing samples for ChIP sequencing of DNA using the reagents provided in the genomic-DNA or ChIP-DNA sample preparation kits , with some modifications ., Briefly , the DNA fragments were converted to phosphorylated blunt ends using T4 DNA polymerase , Klenow DNA polymerase , and T4 polymerase kinase , a 3′ A base overhang was added using Klenow DNA polymerase exo- ( 3′ to 5′ exo minus ) , and Illumina adapters were ligated to the fragments ., We carried out the PCR step for enrichment of adapter-modified DNA prior to the library size selection , and limited the amplification to 10–13 cycles to minimize the potential bias associated with PCR amplification ., After the amplification step , we size-selected DNA fragments of 150–500 bp ( including the adapter sequence ) for BCD , HB , GT , and KNI samples , and 200–500 bp for KR and CAD ., The DNA library was quantified by QPCR using ABI Power SYBR green PCR master mix and pair primers that match the adapter sequences ., We used a Solexa DNA library , which we generated with known concentration as a standard ., Due to the extreme sensitivity , the DNA used in the reactions ranged from 0 . 0001–0 . 01 ng ., The sequencing of the library DNA was performed on the Solexa/Illumina platform according to the manufacturers instruction ., Each library was analyzed in two lanes on the flow cell ., We used the Apr . 2006 assembly ( dm3 , BDGP Release 5 ) of the D . melanogaster genome , downloaded from http://hgdownload . cse . ucsc . edu/goldenPath/dm3/bigZips/chromFa . tar . gz , and the Nov . 2005 assembly ( droYak2 ) of the D . yakuba genome , downloaded from http://hgdownload . cse . ucsc . edu/goldenPath/droYak2/bigZips/chromFa . tar . gz ., We trimmed all sequenced tags to 20 bp and mapped the tags to the genomes using Bowtie v0 . 9 . 9 . 1 22 with command-line options ‘-v 1 -m 1’ , thereby keeping only tags that mapped uniquely to the genome with at most one mismatch ., Table 1 gives statistics on the total numbers of sequenced and mapped tags for all experiments ., Note that while we mapped tags to the entire genomes , we did not use the heterochromatic chromosomes or unassembled sequence for any analyses ., We used annotations from FlyBase r5 . 15 34 for analyses using genes in D . melanogaster ., We called peaks for each experiment using MACS v1 . 3 . 5 25 with the option ‘--pvalue 0 . 00001’ . We used total chromatin as background controls , and set the ‘--mfold’ option to the maximum value for which MACS could find a sufficient number of paired peaks ., In order to only consider peaks for which we could reliably assign orthology and to control for potential assembly errors in the draft D . yakuba genome , we used exonerate 35 to search for peaks whose associated sequence was duplicated in either genome ., For each peak , we ( 1 ) searched for duplicated sequence in the genome where the peak was called and ( 2 ) used the whole-genome alignment to pull out the orthologous sequence in the other genome and searched for duplicates of that sequence in the other genome , which frequently indicated a potential assembly error due to the unfinished nature of the D . yakuba assembly ., We discarded any peaks whose associated sequence was duplicated in either genome ., We used a large-scale orthology mapping created by Mercator 23 to identify syntenic regions of the genomes , which were each aligned with FSA v1 . 11 . 0 with the options ‘--exonerate --softmasked --refinement -1 --mercator cons seqs . fasta’ ., The resulting whole-genome alignment can be downloaded here: http://www . biostat . wisc . edu/~cdewey/data/fsa_mercator_alignments/drosophila_melanogaster-5 . 0-drosophila_yakuba-2 . 0-1 . 0 . tar . gz ., We first normalized the total number of sequenced tags to a fixed number for each experiment , the standard method of controlling for the variable success of amplification and sequencing ., This normalization , however , is insufficient for our purposes , since it does not take into account differences in genome size and background between the species ., We therefore performed an additional comparative normalization step ., Assuming that the total amount of binding near known regulatory targets of the six factors studied here ( A-P and D-V genes , as identified in 19 and listed below ) is constant , we scaled the total number of sequenced tags in D . yakuba for each factor such that the total difference in inferred binding strength across the 50 most highly bound peaks in each genome ( for a total of 100 ) within 10 kb of A-P targets was minimized ( using a least-squares linear regression ) ., This comparative normalization procedure assumes there are no differences in the total number of molecules bound to A-P targets in the two genomes ., Although this may not always be the case , we do not expect to see such global differences between such closely related species ., It is also possible that by using the 50 most highly bound peaks near known A-P target genes for normalization we would underestimate variation in these genes ., However , the effect of any single peak on the normalization was minimal , and the inferred divergence for any of these peaks did not change significantly when they were not included in the normalization ( unpublished data ) ., We assessed binding strength by estimating a fragment density by extending each sequenced tag to the average fragment length based on the selected size distribution ., We modified the SynPlot program 36 to display quantitative data along an alignment in order to create the plot in Figure 1 ., We compared binding between the two genomes as follows: Given a peak called in one genome , we used the whole-genome alignment to project the 100 bp containing the peak onto the other genome and computed the maximum binding strength within that homologous sequence in the other genome ., Note that therefore our maximum spatial resolution when assessing binding divergence is 50 bp , implying that if , for example , a binding site is present in D . melanogaster , and lost in D . yakuba but replaced by another site 30 bp away , then we will not detect any binding divergence if the two sites are bound at similar levels ., We labeled peaks that were within 10 Kb of a gene in D . melanogaster known to be regulated by A-P factors as A-P target loci ., We used the following list of genes: Brk , D , Doc1 , Doc2 , E ( spl ) , Kr , Phm , SoxN , Vnd , bowl , btd , cad , croc , dpp , ems , eve , fkh , ftz , gt , h , hb , hkb , ind , kni , knil , noc , nub , oc , odd , opa , os , pdm2 , pnr , prd , pxb , rho , run , salm , shn , sim , slp1 , slp2 , sna , sob , sog , ths , tld , tll , tsh , tup , twi , vn , wntD , zen ., We identified DDWs for each factor as follows ., For each word of a fixed length k , we identified all ( non-softmasked ) instances of the word ( on both strands ) within a 100 bp window centered on the empirical maximum of peaks called in D . melanogaster for that factor ., We then accumulated two distributions of binding strength divergence ( D . melanogaster − D . yakuba ) for the word , pcons and pdiv , with pcons consisting of instances where the word was exactly conserved in D . yakuba and pdiv consisting of instances where the word was diverged in D . yakuba ., We used a non-parametric statistical test , Kolmogorov-Smirnov test , to test for equality of distribution pcons ∼ pdiv ., If equality of distribution could be rejected with p<0 . 01 , then we called the word a candidate DDW ., We then performed the identical procedure in the opposite direction , wherein we examined peaks called in D . yakuba and assessed the conservation of words in D . melanogaster , and identified a second set of candidate DDWs ., We took the intersection of these two sets to obtain final lists of DDWs ., We performed this procedure to identify words of length k\u200a=\u200a6 and 7 ., We assessed whether sequence motifs matched the known DNA-binding specificities of A-P factors with position weight matrices ( PWM ) from 29 ., When creating Figures 4 and 7 , we said that a word matched the specificity for a factor if it matched a subsequence of the corresponding PWM with ln ( p value ) <−4 as reported by Patser 37 ., We used the Least Angle Regression ( LARS ) algorithm 28 , implemented in the package lars for R 38 , to learn a linear model of binding divergence using DDWs of length k\u200a=\u200a6 ., We performed 5-fold cross-validation to estimate the mean-squared prediction error ( MSE ) associated with each value of the lasso regularization parameter β and then chose the model given by the β that yielded the lowest MSE ., This cross-validation procedure helps to prevent the over-fitting characteristic of standard least-squares linear regression , making the correlations that we estimated robust to generalization error ., In order to ensure that ( 1 ) the DDWs that we identified truly have predictive value and ( 2 ) that the correlations reported are not due solely to base-composition effects , we randomly shuffled the nucleotides of each DDW to create a set of shuffled words with unchanged base composition , and then built a predictive model using these shuffled words ., Models constructed using these shuffled words had no predictive value , indicating that the correlations that we report for our DDWs are not statistical artifacts ., Figures S12–S27 show lasso variable selection curves and cross-validation curves for all values of β , as well as scatterplots of predicted and observed binding divergences , for predictive models constructed using our DDWs as well as their shuffled counterparts ., The cross-validation curves make clear that while the DDWs are correlated with binding strength , the shuffled words are not: MSE decreases as more DDWs are included into the model , indicating the gain and loss of these words correlates with changes in observed binding strength , whereas MSE increases as more shuffled words are included into the model , indicating that these words are uncorrelated with binding ., This provides clear evidence that our cross-validation procedure correctly chooses the model with the minimum generalization error , for example , that the models are not over-fit to the data ., We performed an identical analysis using words derived from the in vitro binding specificity data described in 29 ., We enumerated all k-mers that matched a subsequence of the corresponding PWM with ln ( p value ) <−8 as reported by Patser 37 , identifying four 6-mers for BCD , HB , and GT and sixteen 6-mers for KR , and then used the learning procedure described above to learn models of binding divergence using these words ., We calculated binding strengths of the six factors across all called peaks , subtracted the empirical means for each factor , and scaled the data for each factor such that it had unit variance ., We used the singular value decomposition routine in IT++ , a linear algebra library for C++ , to perform PCA , and created heatmaps of the PCA results using a modified version of the aspectHeatmap function in the ClassDiscovery package ., In order to confirm that the putative chromatin signal represented by the first principal component did reflect coherent increases and decreases in binding of all six factors in our data , we randomly interchanged the measured binding strengths for a single factor across called peaks while holding all others unchanged ( Figure S34 , panels A–F ) and similarly randomly interchanged the binding strengths of all factors ( Figure S34 , panel G ) , thereby removing spatial correlations between the binding of single factors and the other five ( Figure S34 , panels A–F ) and removing spatial correlations between the binding of any factors ( Figure S34 , panel G ) ., As expected , the chromatin signal disappeared after performing any of these transformations on the data ., We identified sequence motifs associated with interspecies divergence of each principal component using the same procedure described above , but with the data projected along the principal component of interest ., For each principal component , we accumulated the distributions pcons and pdiv across all peaks called for any of the six factors ., All sequence reads from the experiments described are available from the NCBIs GEO database with accession number GSE20369 ., Processed datasets , including mapped reads , called regions and peaks , D . melanogaster − D . yakuba alignments , and all software described here , are available at http://rana . lbl . gov/data/melyak . | Introduction, Results, Discussion, Methods | Changes in gene expression play an important role in evolution , yet the molecular mechanisms underlying regulatory evolution are poorly understood ., Here we compare genome-wide binding of the six transcription factors that initiate segmentation along the anterior-posterior axis in embryos of two closely related species: Drosophila melanogaster and Drosophila yakuba ., Where we observe binding by a factor in one species , we almost always observe binding by that factor to the orthologous sequence in the other species ., Levels of binding , however , vary considerably ., The magnitude and direction of the interspecies differences in binding levels of all six factors are strongly correlated , suggesting a role for chromatin or other factor-independent forces in mediating the divergence of transcription factor binding ., Nonetheless , factor-specific quantitative variation in binding is common , and we show that it is driven to a large extent by the gain and loss of cognate recognition sequences for the given factor ., We find only a weak correlation between binding variation and regulatory function ., These data provide the first genome-wide picture of how modest levels of sequence divergence between highly morphologically similar species affect a system of coordinately acting transcription factors during animal development , and highlight the dominant role of quantitative variation in transcription factor binding over short evolutionary distances . | The differentiation of cells , tissues , and organs during animal development is established by a process in which genes that control cell identity and behavior are turned on and off at specific times and places ., This process is choreographed , to a large extent , by a collection of proteins known as transcription factors that bind to specific sequences in DNA and thereby modulate the expression of neighboring genes ., Because of the central role that transcription factors play in shaping organismal form and function , they have long been suggested to be major players in phenotypic evolution ., However , we have a poor understanding of how changes to DNA affect transcription factor binding in living systems ., Here , we use a combination of biochemical and genomic techniques to compare , between two closely related species of fruit flies in the genus Drosophila , the binding of six transcription factors that help establish the characteristic segments that form along the anterior-posterior ( head to tail ) axis in developing flies ., We show that the patterns of transcription factor binding between these closely related species are broadly conserved , consistent with the nearly identical development and appearance of these species ., However , we also show that , whereas the DNA changes that have accumulated between these species in the five million years since their divergence—roughly one difference per 10 basepairs—have not altered the locations where these factors bind , they have had a considerable effect on the amount of factor bound at each site across a population of embryos ., We can trace these quantitative differences in binding to the gain and loss of the short sequences known to be preferentially recognized by these factors , giving us key insights into the effect that sequence changes have on the biochemical events that underlie animal development . | genetics and genomics/comparative genomics, genetics and genomics/functional genomics, computational biology/comparative sequence analysis, developmental biology/developmental evolution, computational biology/evolutionary modeling, computational biology/genomics | Genome-wide comparison of transcription factor binding between related Drosophila species highlights how sequence changes affect the biochemical events that underlie animal development. |
journal.pgen.1002791 | 2,012 | GWAS Identifies Novel Susceptibility Loci on 6p21.32 and 21q21.3 for Hepatocellular Carcinoma in Chronic Hepatitis B Virus Carriers | Hepatocellular carcinoma ( HCC ) is the sixth common cancer and the third common cause of cancer mortality worldwide 1 ., The incidence rate of HCC varies considerably in the world , with the highest in East , Southeast Asia and Sub-Saharan Africa , and China alone accounts for approximately half of HCC malignancies 1 , 2 ., Major risk factors for HCC are chronic infections with the hepatitis B or C viruses , and exposure to dietary aflatoxin B1 ., Hepatitis B virus ( HBV ) infection is particular important , because of its coherent distribution with the HCC prevalence 1 , 2 ., However , it is known that only a minority of chronic carriers of HBV develop HCC 3 , and the chronic HBV carriers with a family history of HCC have a two-fold risk for HCC than those without the family history 4 , strongly suggesting the importance of genetic susceptibility for HBV-related HCC ., A number of candidate genes were investigated by genetic association studies to evaluate their roles in the susceptibility to HCC 5 ., However , the findings from these studies are inconclusive due to moderate evidence and lack of independent validation ., Recently , a genome-wide association study ( GWAS ) of HBV-related HCC was performed 6 , in which 355 HBV–positive HCC patients and 360 chronic HBV carriers were used for the genome-wide discovery analysis , and the top 45 SNPs from the discovery analysis were further evaluated in additional 1 , 962 HBV–positive HCC patients and 1 , 430 controls ( both chronic HBV carriers and population controls ) as well as 159 trios ., The study identified KIF1B as a novel susceptibility locus ( top SNP rs17401966 ) on 1p36 . 22 ., Further study with better design and bigger sample size was recommended for identifying additional susceptibility loci for HCC 7 , 8 ., These motivate us to carry out a GWAS with a large sample size in Chinese population to discover novel susceptibility loci for HCC ., We performed a genome-wide discovery analysis by analyzing 523 , 663 common autosomal SNPs in two independent cohorts of the Han Chinese: 480 cases and 484 controls from central China and 1058 cases and 981 controls from southern China ( Table S1 and Figure S1 ) ., The principal component analysis ( PCA ) confirmed all the samples to be Chinese , but indicated moderate genetic mismatch between the cases and controls in the cohort of southern China ( Figure S2 ) ., To minimize the effect of population stratification , we performed the genome-wide association analysis using PCA-based correction for population stratification ., After the adjustment by the first principal component , the λgc of the genome-wide association results is 1 . 013 for the cohort of central China , 1 . 003 for the cohort of southern China and 1 . 012 for the combined samples ., Furthermore , for all the three genome-wide analyses of central , southern and combined samples , the quantile-quantile ( QQ ) plot of the observed P values revealed a good overall fit with the null distribution ( Figure S3 ) ., Taken together , these results clearly indicate that the final association results from our genome-wide discovery analysis are free of inflation effect due to population stratification ., The genome-wide discovery analysis revealed multiple suggestive associations ( P<10−5 ) on 2q22 . 1 , 6p21 . 32 , 11p15 . 1 and 20q12 ( Figure S4 and Table S2 ) ., To validate these findings , 39 SNPs were selected according to their overall association evidence in three GWAS analyses as well as their consistencies of association between the two independent GWAS samples ( Central and Southern China ) ( see the Methods for the selection criteria ) ., The 39 SNPs were genotyped in additional 2 , 112 HBV–positive HCC cases and 2 , 208 HBV carriers ( Phase I validation ) ( Table S1 ) ., Of the 39 SNPs , only 3 ( rs9272105 on 6p21 . 32 , rs11148740 on 13q21 . 32 and rs455804 on 21q21 . 3 ) were validated , showing consistent association between the GWAS discovery and Phase I validation samples ( Table S3 ) ., These 3 SNPs were then genotyped in additional 1 , 021 HBV–positive HCC cases and 1 , 491 HBV carriers ( Phase II validation ) ., The Phase II validation analysis ( Table 1 ) confirmed the associations at rs9272105 on 6p21 . 32 ( OR\u200a=\u200a1 . 41 , P\u200a=\u200a7 . 63×10−9 ) and rs455804 on 21q21 . 3 ( OR\u200a=\u200a0 . 83 , P\u200a=\u200a3 . 63×10−3 ) , but not the association at rs11148740 on 13q21 . 32 ( Table S3 ) ., For both rs9272105 and rs455804 , no heterogeneity of associations were observed among the GWAS and validation samples ( P>0 . 05 ) , and the associations in the combined GWAS and validation samples achieved genome-wide significance ( P<5 . 0×10−8 ) ( rs9272105: OR\u200a=\u200a1 . 30 , P\u200a=\u200a1 . 13×10−19 and rs455804: OR\u200a=\u200a0 . 84 , P\u200a=\u200a1 . 86×10−8 ) ( Table 1 ) ., As a replication , these two SNPs were genotyped in the fourth independent samples of 1 , 298 cases and 1 , 026 controls from central China , which further confirmed the associations at rs9272105 ( OR\u200a=\u200a1 . 25 , P\u200a=\u200a1 . 71×10−4 ) and rs11148740 ( OR\u200a=\u200a0 . 84 , P\u200a=\u200a6 . 92×10−3 ) ( Table 1 ) ., When combining all the five groups of samples , the two SNPs resulted in a 28% increased , and a 16% decreased risk for HCC development ( rs9272105: OR\u200a=\u200a1 . 28 , P\u200a=\u200a5 . 24×10−22 and rs455804: OR\u200a=\u200a0 . 84 , P\u200a=\u200a5 . 24×10−10 ) ( Table 1 ) , respectively ., The associations at the two SNPs remained genome-wide significant after adjusting for age , gender , smoking and drinking ( Table S4A ) ., Furthermore , stratification analysis by age , gender , smoking and drinking status revealed similar ORs for rs9272105 and rs455804 among subgroups , except that the association at rs9272105 showed a stronger effect in the non-smoking group than the smoking one ( OR\u200a=\u200a1 . 38 vs . 1 . 19 , P for heterogeneity\u200a=\u200a0 . 004 ) ( Table S4B ) ., Pair-wise interaction analysis among these two SNPs , smoking and drinking status did not reveal any significant interaction ( data not shown ) ., The samples used in the GWAS , validation and replication analyses are summarized in Table S1 , and the multi-stage design of the whole study is shown in Figure S5 ., We further investigated the association of HLA alleles in our GWAS samples through imputation ., After QC filtering ( see the Methods ) , 37 HLA alleles were successfully imputed , and 5 alleles showed nominal association ( P<0 . 05 ) ( Table S5 and Table 2 ) ., Further stepwise conditional analysis revealed that only two DRB1 alleles showed independent associations ( DRB1*0405: OR\u200a=\u200a0 . 69 , P\u200a=\u200a6 . 18×10−4; DRB1*0901: OR\u200a=\u200a0 . 82 , P\u200a=\u200a3 . 62×10−3 ) ( Table 2 ) ., Conditioning on rs9272105 could abolish the associations of the DRB1 alleles , and conditioning on the DRB1 alleles could weaken , but not eliminate , the association at rs9272105 ( Table 2 ) ., The haplotype analysis of rs9272105 and the two DRB1 alleles revealed consistent result , showing that both the DRB1 alleles sit on the haplotypes carrying the protective G allele of rs9272105 ( Table S6 ) ., Taking together , there seems to be additional risk effect beyond the ones carried by the DRB1 alleles ., We further explored whether the SNPs rs9272105 and rs455804 play any role in HBV infection ., First , we compared the frequencies of these 2 SNPs between 408 non-symptomatic HBV carriers and 521 symptomatic chronic HBV patients from southern China ( GWA scanned ) ., The analysis revealed a protective effect at rs9272105 ( OR\u200a=\u200a0 . 80 , P\u200a=\u200a1 . 67×10−2 ) on the development of symptomatic chronic hepatitis B , but no association at rs455804 ( Table S7A ) ., Furthermore , we genotyped these 2 SNPs in 1 , 344 individuals with HBV nature clearance and compared their frequencies with those in 4 , 183 asymptomatic HBV carriers ( all from the Central China ) ., The analysis also revealed a protective association at rs9272105 for HBV chronic infection ( OR\u200a=\u200a0 . 88 , P\u200a=\u200a3 . 78×10−3 ) ( Table S7B ) ., SNP rs9272105 is located between HLA-DQA1 and HLA-DRB1 on 6p21 . 32 ( Figure 1A ) ., SNP imputation in the GWAS discovery samples revealed additional SNPs showing association , but rs9272105 remained to be the top SNP within the region ( Figure 1A ) ., The residual association at rs9272105 after conditioning the association effects of the HLA alleles DRB1*0405 and *0901 suggests that there may be additional risk effect beyond the DRB1 alleles in Chinese population ., The associations of the DRB1 alleles revealed by this study are consistent with the previous reports that HLA-DQ/DR alleles associated with HCC risk 9 , 10 ., In addition , we investigated the previously reported HBV infection-associated SNPs rs3077 , rs9277535 , rs7453920 , and rs2856718 within the HLA DP/DQ region 11 , 12 with HCC development in our GWAS samples ., By imputation , we found the evidence of the association at rs9277535 with HCC ( rs9277535: OR\u200a=\u200a0 . 85 , P\u200a=\u200a7 . 9×10−3 ) ., However , there is no linkage disequilibrium ( LD ) between rs9277535 and our SNP rs9272105 ( r2\u200a=\u200a0 . 016 according the HapMap CHB+JPT samples ) , suggesting that the associations at rs9277535 and rs9272105 may be independent ., The HLA-DQ locus has also been shown to be associated with HCV-related HCC in a Japanese GWAS ( rs9275572 , OR\u200a=\u200a1 . 30 , P\u200a=\u200a9 . 38×10−9 ) 13 ., SNPs rs9275572 and rs9272105 are 79 kb away from each other and in weak LD ( D′\u200a=\u200a0 . 43 , r2\u200a=\u200a0 . 08 in the HapMap CHB samples ) ., The SNP rs9275572 did not show any association with HBV-related HCC in our GWAS discovery samples ( OR\u200a=\u200a0 . 93 , P\u200a=\u200a0 . 24 ) ( Table S8 and Figure S6B ) ., In addition to HLA-DQ , MICA ( rs2596542 ) on 6p21 . 33 and DEPDC5 ( rs1012068 ) on 22q12 . 3 were also identified as independent susceptibility loci for HCV-related HCC in Japanese population 13 , 14 ., But , our GWAS discovery analysis did not reveal any supportive evidence for these two loci ( rs2596542: OR\u200a=\u200a1 . 06 , P\u200a=\u200a0 . 36; and rs1012068: OR\u200a=\u200a1 . 06 , P\u200a=\u200a0 . 37 ) ( Table S8 and Figure S6C and S6D ) ., We also evaluated the power of our GWAS discovery samples and found that our samples should have sufficient power for detecting the previously reported associations at rs9275572 ( power\u200a=\u200a94% ) , rs2596542 ( power\u200a=\u200a92% ) and rs1012068 ( power\u200a=\u200a94% ) ., Taken together , the disparity of associations may suggest the different genetic background of the susceptibilities for HCV- and HBV-related HCC ., Further studies will be required to confirm the genetic heterogeneity of HCV- and HBV-related HCC ., The association of rs9272105 ( HLA-DQA1/DRB1 ) with HBV infection is consistent with the extensive reports on the association of HLA-DRB1 with HBV infection where both protective and risk DRB1 alleles for HBV infection and outcome were identified 11 , 12 , 15–19 ., Intriguingly , our study has revealed that the variant allele of rs9272105 showed a protective effect for HBV infection ( OR\u200a=\u200a0 . 88 ) and the progression to chronic symptomatic hepatitis B , but a risk effect for the development of HCC ( OR\u200a=\u200a1 . 30 ) ., Further studies will be needed to demonstrate whether the opposite associations of HBV infection and HBV-related HCC progression at rs9272105 are due to different causal variants within the HLA class II region ., SNP rs455804 is located within the first intron of GRIK1 that is the only gene within the LD region of the association ( Figure 1B ) , strongly implicating GRIK1 as a novel susceptibility gene for HBV-related HCC ., SNP imputation of the region did not reveal any SNPs that showed stronger association than rs455804 ., GRIK1 encodes CLUR5 , which is involved in the glutamate signaling , as one of the ionotropic glutamate receptor , kainite 1 protein ( GLUR5 ) , a subunit of ligand-activated channels and involved in glutamate signaling ., Our discovery of the association of GRIK1 with HCC has enhanced the emerging evidences for the important role of glutamate signaling pathway in cancer development ., Glutamate has been shown to play a central role in the malignant phenotype of gliomas through multiple molecular mechanisms 20 ., Inhibition of glutamate release and/or glutamate receptor activity can inhibit the proliferation and/or invasion of tumor cells in breast cancer 21 , laryngeal cancer 22 , and pancreatic cancer 23 , and ionotrpic glutamate receptor ( GLUR6 ) was also suggested to play a tumor-suppressor role in gastric cancer 24 ., Recently , the exome sequencing analysis revealed that GRIN2A ( encoding the ionotrpic glutamate receptor ( N-methyl D-aspartate ) subunit 2A ) was mutated in 33% of melanoma tumors , clearly indicating the involvement of glutamate signaling in melanoma development ., Finally , SNPs within GRIK1 have also been found significantly associated with paclitaxel response in NCI60 cancer cell lines , and may play a role in the cellular response to paclitaxel treatment in cancer 25 ., Consistent with the previous observations , our discovery of GRIK1 as a HBV-related HCC susceptibility gene has suggested the importance of glutamate signaling in HBV-related HCC development , and , although still speculative , has highlighted the glutamate signaling pathway as a potentially novel target for the treatment of HCC ., We also assessed the previously reported susceptibility locus KIF1B on 1p36 . 22 ( rs17401966 ) for HBV-related HCC 6 ., Our GWAS discovery analysis did reveal the consistent result for the association at rs17401966 , but the strength of association in our GWAS discovery sample ( OR\u200a=\u200a0 . 90 ) is much weaker than the previously reported one ( OR\u200a=\u200a0 . 61 ) ( Table S8 ) ., SNP imputation in our GWAS discovery samples did not reveal any stronger association than the association at rs17401966 within the LD region surrounding the 1p36 . 22 locus ( Figure S6A ) ., Previous studies have clearly shown the existence of subpopulation structure of Chinese Han population along the north-south axis , and further demonstrated that geographic matching can be used as a good surrogate for genetic matching , and PCA-based correction is very effective in controlling the inflation effect of population stratification 26 ., In the current study , all the cases and controls were matched by their geographic origin of residence ., Moreover , the GWAS discovery samples were from central and southern China , while all the validation and replication samples were from central China ., Our PCA analysis indicates that while there was mild population stratification in the sample of southern China , the cases and controls from central China were well matched without any indication of population stratification ., In our study , the PCA-based correction was used in the GWAS analysis , and all the validation and replication analyses were from central China ., Therefore , our findings should be free of adverse effect of population stratification in Chinese population ., In conclusion , the current GWAS identified two biologically plausible , novel loci on 6p21 . 32 and 21q21 for HBV-related HCC ., These findings highlight the importance of HLA-DQ/DR molecules and glutamate signaling in the development of HBV-related HCC ., The genome-wide discovery analysis was performed by genotyping 731 , 442 SNPs in 1 , 575 HBV positive HCC patients and 1 , 490 HBV positive controls derived from two independent case-control cohorts of 500 cases and 500 controls from Central China ( Shanghai ) and 1 , 075 cases and 990 controls from Southern China ( Guangdong ) ., The first stage validation samples included 2 , 112 HBV–positive cases and 2 , 208 HBV–positive controls recruited from Jiangsu ., The second stage validation samples consisted of 1 , 021 HBV–positive cases and 1 , 491HBV carriers recruited from Shanghai ., The replication samples of 1 , 298 HBV–positive cases and 1 , 026 HBV carriers were recruited from Central China ( Shanghai and Jiangsu ) ., ( Table S1 and Figure 1 ), All the samples are Han Chinese and partially participated in the previously published studies 27 , 28 ., The diagnosis of HCC was confirmed by a pathological examination and/or α-fetoprotein elevation ( >400 ng/ml ) combined with imaging examination ( Magnetic resonance imaging , MRI and/or computerized tomography , CT ) ., Because HCV infection is rare in Chinese , we excluded HCC with HCV infection ., Cancer-free HBV+ control subjects from central China were recruited from those receiving routine physical examinations in local hospitals or those participating in the community-based screening for the HBV/HCV markers and frequency-matched for age , gender , and geographic regions to each set of the HCC patients ., Almost all these community-based controls are asymptomatic HBV carriers ., Similarly , cancer-free control subjects from southern China are all HBV+ , and 408 of them were asymptomatic HBV carriers and 521 were symptomatic chronic hepatitis B patients ., All the HBV+ controls were positive for both HBsAg and antibody to hepatitis B core antigen ( anti-HBc ) , and negative for anti-HCV ., We also recruited a HBV natural clearance cohort form Jiangsu Province ( Zhangjiagang and Changzhou cities ) through a population based screening for the HBV/HCV markers in 2004 and 2009 , respectively ( 58 , 142 persons ) ., Subjects with HBV natural clearance were negative for HBsAg and anti-HCV , positive for both antibody to hepatitis B surface antigen ( anti-HBs ) and anti-HBc ., About 9 , 610 subjects with HBV natural clearance were identified ., No history of hepatitis B vaccination was reported for these people ., Then , we randomly selected 1 , 344 HBV natural clearance people without self-reported history of cancer in the current study ., The age for the 1 , 344 people were 52 . 6±10 . 2 years , and 217 ( 16 . 2% ) were females ., We collected smoking and drinking information through interviews ., Those who had smoked an average of less than 1 cigarette per day and less than 1 year in their lifetime were defined as nonsmokers; otherwise , they were considered as smokers ., Individuals were classified as alcohol drinkers if they drank at least twice a week and continuously for one year during their lifetime; otherwise , they were defined as nondrinkers ., At recruitment , the informed consent was obtained from each subject , and this study was approved by the Institutional Review Boards of each participating institution ., We performed standard quality control on the raw genotyping data to filter both unqualified samples and SNPs ., The samples with overall genotype completion rates <95% were excluded from further analysis ( 26 subjects ) ., Eight subjects were excluded as they showed discrepancy between the recorded and genetically inferred genders ., An additional 21 duplicates or probable familial relatives were excluded based on the IBD analysis implemented in PLINK ( all PI_HAT>0 . 25 ) ., SNPs were excluded when they fit the following criteria:, ( i ) not mapped on autosomal chromosomes;, ( ii ) had a call rate <95% in all GWA samples or in either of Central cohort study or Southern study samples;, ( iii ) had minor allele frequency ( MAF ) <0 . 05 in either of Central cohort study or Southern study samples; and, ( iv ) genotype distributions deviated from those expected by Hardy-Weinberg equilibrium ( P<1×10−5 in either of Central cohort study or Southern study samples ) ., We detected population outliers and stratification using a principal component analysis ( PCA ) based method ., After removing MHC SNPs on chromosome 6 from 25–37 Mb , PCA was performed by using common autosomal SNPs with low LD ( r2<0 . 2 ) in the reference samples of the HapMap project ( YRI ( n\u200a=\u200a90 ) , CEU ( n\u200a=\u200a90 ) , CHB ( n\u200a=\u200a45 ) and JPT ( n\u200a=\u200a44 ) ) as the internal controls and our 3 , 010 participants of the GWAS discovery samples ( after removal of samples with low call rates , ambiguous gender , and familial relationships ) ., Projection onto the two multidimensional scaling axes is shown in Figure S2A ., 7 outliers ( more than 6 standard deviations ) were identified and excluded ., Finally , 523 , 663 autosomal SNPs in 1 , 538 cases and 1 , 465 controls , consisting of 480 cases and 484 controls from Central China and 1 , 058 cases and 981 controls from Southern China , were retained for association testing ( Table S1 ) ., SNPs for the first stage validation were selected based on the following criteria:, ( i ) SNP had P joint≤1 . 0×10−4 in the analysis of the combined GWA samples or either the Central China sample or the Southern China sample , and had a consistent association in the two participant studies , meaning that the ORs from the two samples are both either above or below 1;, ( ii ) only SNP with the lowest P value was selected when multiple SNPs showed a strong LD ( r2≥0 . 8 ) ., As a result , a total of 39 SNPs were included in the first stage validation ., 3 SNPs that were significantly associated with HCC risk in the first validation stage were further genotyped in the second stage validation samples ., Genotyping in the two validation samples were done by using the iPLEX platform ( Sequenom ) or the TaqMan assays ( Applied Biosystems ) ., The primers and probes were available upon request ( Table S9 ) ., Laboratory technicians who performed genotyping experiments were blinded to case/control status ., For TaqMan assay , ten percent of random samples were repeated , and the reproducibility was 100% ., The 2 validated SNPs were genotyped in another independent replication using the same method ., Population structure was evaluated by the PCA in the software package EIGENSTRAT 3 . 0 26 ., PCA revealed one significant ( P<0 . 05 ) eigenvector which was included in the logistic regression with other covariates of age , gender , smoking and drinking status for both the genome-wide discovery analysis and the joint analysis of the combined discovery and replication samples ., Ancestral origin checking by PCA confirmed all the samples to be Han Chinese and further demonstrated moderate genetic stratification between the cases and the controls of the Southern cohort ( Figure S2 ) ., The genome-wide association analysis was therefore performed in logistic regression using PCA-based correction for population stratification and by treating the samples of two cohorts as independent studies ., The genomic-control inflation factor ( λgc ) after adjustment by the first PC was calculated for the Central cohort samples ( λgc\u200a=\u200a1 . 013 ) , the Southern cohort samples ( λgc\u200a=\u200a1 . 003 ) and the combined GWAS discovery samples ( λgc\u200a=\u200a1 . 012 ) ., Consistently , the QQ plot of the observed P values also showed a minimal inflation of genome-wide association results due to population stratification ( Figure S3 ) ., Statistical analyses were performed by using PLINK 1 . 07 29 and R 2 . 11 . 1 ., The Manhattan plot of −log10P was generated using Haploview ( v4 . 1 ) 30 ., Untyped genotypes were imputed in the GWAS discovery samples by using IMPUTE2 31 and the haplotype information from the 1000 Genomes Project ( ASN samples as the reference set ) and HapMap3 ( CHB and JPT samples as the reference samples ) ., The regional plot of association was created by using an online tool , LocusZoom 1 . 1 ., P value was two-sided , and OR presented in the manuscript was estimated by using additive model and logistic regression analyses if not specified ., To impute classical HLA alleles , we used 180 phased haplotypes from the HapMap CHB and JPT samples as our reference panel ., This panel comprised dense SNP data and HLA allele types at 4-digit resolution for the HLA class I ( HLA-A , B , C ) and II ( DQA1 , DQB1 and DRB1 ) genes as previously described 32 ., Genotypes , probability and allelic dosages were then imputed separately in the two discovery samples of Central and Southern Chinese using the BEAGLE program ., Association testing was performed by using a logistic regression model on the best-guessed genotypes and allelic dosages ., The results were checked for consistency between the two methods , and the results from best-guessed genotypes were presented . | Introduction, Results, Discussion, Methods | Genome-wide association studies ( GWAS ) have recently identified KIF1B as susceptibility locus for hepatitis B virus ( HBV ) –related hepatocellular carcinoma ( HCC ) ., To further identify novel susceptibility loci associated with HBV–related HCC and replicate the previously reported association , we performed a large three-stage GWAS in the Han Chinese population ., 523 , 663 autosomal SNPs in 1 , 538 HBV–positive HCC patients and 1 , 465 chronic HBV carriers were genotyped for the discovery stage ., Top candidate SNPs were genotyped in the initial validation samples of 2 , 112 HBV–positive HCC cases and 2 , 208 HBV carriers and then in the second validation samples of 1 , 021 cases and 1 , 491 HBV carriers ., We discovered two novel associations at rs9272105 ( HLA-DQA1/DRB1 ) on 6p21 . 32 ( OR\u200a=\u200a1 . 30 , P\u200a=\u200a1 . 13×10−19 ) and rs455804 ( GRIK1 ) on 21q21 . 3 ( OR\u200a=\u200a0 . 84 , P\u200a=\u200a1 . 86×10−8 ) , which were further replicated in the fourth independent sample of 1 , 298 cases and 1 , 026 controls ( rs9272105: OR\u200a=\u200a1 . 25 , P\u200a=\u200a1 . 71×10−4; rs455804: OR\u200a=\u200a0 . 84 , P\u200a=\u200a6 . 92×10−3 ) ., We also revealed the associations of HLA-DRB1*0405 and 0901*0602 , which could partially account for the association at rs9272105 ., The association at rs455804 implicates GRIK1 as a novel susceptibility gene for HBV–related HCC , suggesting the involvement of glutamate signaling in the development of HBV–related HCC . | Previous studies strongly suggest the importance of genetic susceptibility for hepatocellular carcinoma ( HCC ) ., However , the studies about genetic etiology on HBV–related HCC were limited ., Our genome-wide association study included 523 , 663 autosomal SNPs in 1 , 538 HBV–positive HCC patients and 1 , 465 chronic HBV carriers for the discovery analysis ., 2 , 112 HBV–positive HCC cases and 2 , 208 HBV carriers ( the initial validation ) , and 1 , 021 cases and 1 , 491 HBV carriers ( the second validation ) , were then analyzed for validation ., The fourth independent samples of 1 , 298 cases and 1 , 026 controls were analyzed as replication ., We discovered two novel associations at rs9272105 ( HLA-DQA1/DRB1 ) on 6p21 . 32 and rs455804 ( GRIK1 ) on 21q21 . 3 ., HLA-DRB1 molecules play an important role in chronic HBV infection and progression to HCC ., The association at rs455804 implicates GRIK1 as a novel susceptibility gene for HBV–related HCC , suggesting the involvement of glutamate signaling in the development of HBV–related HCC . | genome-wide association studies, cancer genetics, genetics, biology, genetics and genomics | null |
journal.pgen.1006648 | 2,017 | Adaptation of A-to-I RNA editing in Drosophila | Genomic mutations are the major sources for phenotypic changes and adaptation 1–4 ., In diploid multicellular organisms , a nonsynonymous DNA mutation ( a mutation that alters the amino acid sequence of a protein ) will permanently affect the protein products in all the cells ( soma or germline ) that express the mutant allele ., The “all-or-none” property of DNA mutations might incur pleiotropic effects that are antagonistic among cell types , tissues , developmental stages , sexes , or other aspects of life history 5–7 , which would constrain the available genetic diversity for a species ., Given the prevalence of pleiotropic effects in the genomes 8–10 , the sequence space might be inaccessible to many mutations , which potentially slows down the rate of phenotypic evolution and adaptation 11 ., However , the transcriptomic or proteomic diversity limited by mutation sequence space could be expanded by the alteration of RNA sequences in an epigenetic approach , such as RNA editing , which was hypothesized to facilitate adaptation 12–14 ., In addition , RNA editing has the advantage to quickly respond to environmental stress and adjust the activity of final protein products accordingly 15 , 16 ., RNA editing is an evolutionarily conserved mechanism that alters RNA sequences at the co-transcriptional or post-transcriptional level 13 , 17–23 ., Among various RNA editing systems in animals , the base substitution from adenosine ( A ) to inosine ( I ) , termed A-to-I editing , is the most common form 13 , 20 ., Due to the high level of structural similarity between inosine ( I ) and guanosine ( G ) , the cellular machineries , such as ribosomes , spliceosomes or the microRNA ribonucleoprotein complex ( miRNP ) , would recognize I as G during translation 13 , 20 , 24 , 25 , splicing 26–29 , microRNA target recognition 30–32 , or other RNA biological processes 14 ., Therefore , A-to-I RNA editing usually produces a change similar to an A-to-G DNA change in particular tissues or developmental stages , which potentially increases phenotypic plasticity without the alteration of genomic sequences 13 , 20 , 24 , 25 ., The adenosine deaminase acting on RNA ( ADAR ) family are the enzymes that convert adenosine ( A ) to inosine ( I ) in pre-mRNAs 33–36 ., Although multiple Adar genes are encoded in the genomes of mammals and worms , there is only one Adar locus in Drosophila 37 , 38 , which is predominately expressed in the nervous system 39 ., The substrates of ADAR are usually double-stranded RNAs 34 , 36 , 40–42 ., A-to-I editing plays essential roles in many biological processes 18 , 19 , 43–45 , and the abolition of Adar in D . melanogaster severely affects its viability and behavior 33 , 34 , 46 ., Previous studies have identified thousands of A-to-I editing sites in different developmental stages , adult heads or whole animals of Drosophila 47–52 ., In addition , A-to-I editing has been extensively characterized in other organisms , such as humans 53–58 , macaques 59 , 60 , mice 61 , worms 62 , and squids 63 ., Despite these intriguing advances , only a few examples of the advantageous effects conferred by RNA editing have been demonstrated 13 , 14 , 20 , 63 ., For example , the A-to-I editing events in Kv1 mRNA provide numerous adaptive amino acid changes that allow the octopus to adapt to extremely cold temperatures 64 ., The functional consequences of the majority of A-to-I editing events , however , remain to be explored ., In fact , comparative genomics has demonstrated that only a small fraction of the human A-to-I editing events were evolutionarily conserved 65–68 ., Furthermore , it was nicely demonstrated that the editing events in primate coding regions were generally non-adaptive 60 , 67 , 68 ., Nevertheless , the targets of RNA editing might have evolved rapidly across species because A-to-I editing in mammals predominantly occurs in repetitive sequences 53–55 , 61 , 65 , 66 , while the editing events in Drosophila are mainly located in coding regions of genes encoding neurotransmitters or ion channels 47–50 , 69 , 70 ., Therefore , the evolutionary forces acting on A-to-I RNA editing might be different between Drosophila 50 and primates 60 , 67 , 68 ., If A-to-I editing indeed facilitates adaptation by expanding proteomic diversity , we expect to observe predominant signals of adaptation in the editing sites ., A recent study 50 reported that although signals of positive selection could be found in genes of the nervous system , the A-to-I RNA editing events were overall subject to purifying selection in Drosophila ., Additionally , the overall effect of natural selection on the editome is different across Drosophila developmental stages 50 ., Despite these intriguing discoveries , it still remains a mystery whether or not we can find evidence to show that the whole editome is overall adaptive ., Specifically , we are interested in testing whether the nonsynonymous A-to-I editing events in Drosophila brains , the core component of the nervous system , are predominantly adaptive ., Furthermore , several other fundamental questions on editing deserve to be further investigated:, 1 ) Do editing sites preferentially increase sequence space of evolutionarily conserved genes ?, 2 ) Why does the global editome of different tissues or developmental stages show differential selective patterns ?, 3 ) How does temperature shape the global editomes ?, Answers to these questions will help understand the evolutionary principles and functional consequences of RNA editing ., In this study we addressed these questions by systematically sequencing the transcriptomes and deciphering A-to-I editing in the female and male brains of three Drosophila species at different temperatures ., With evolutionary analysis from different perspectives , we provided lines of evidence to demonstrate that the nonsynonymous editing events in coding regions are generally adaptive in brains of Drosophila ., Then we identified a set of gene candidates that had nonsynonymous editing events favored by natural selection ., Overall our results demonstrated that abundant nonsynonymous editing events in Drosophila brains were adaptive and maintained by natural selection during evolution ., To comprehensively characterize the A-to-I editing landscapes in brains of Drosophila , we set out to dissect the brains of 1- to 5-day-old or 1- to 14-day-old female and male adults of the inbred ISO-1 strain of Drosophila melanogaster that were constantly raised at 25°C , or raised at 25°C and treated at 30°C for 14 hours or 48 hours ( Table 1 ) ., Next we selected the poly ( A ) -tailed mRNAs , fragmented them , ligated the mRNA fragments with adaptors , and deep sequenced the transcriptome of each brain sample ( Materials and Methods ) ., We obtained 13 . 9–21 . 6M reads mapped on the reference genome ( see Table 1 and S1 Table for detailed statistics ) , and the median coverage on an exonic site in a library ranges from 5 to 9 reads ( S1 Fig ) ., As justified previously 71 , the mRNA fragmentation library preparation procedure we employ minimizes the bias of non-uniform sequencing read coverage along mRNAs , which would reduce the bias in detecting editing in 3 ends of mRNAs ., It is a challenging task to reliably distinguish the A-to-I editing events from SNPs 72–76 , therefore , the ISO-1 strain used in this study , which was inbred and sequenced to assemble the reference genome of D . melanogaster 77 , enables us to detect DNA-RNA differences with high accuracy and sensitivity ., We employed a two-step strategy to identify editing sites in the brains of D . melanogaster ( Fig 1A ) ., First , we used the GATK RNA-Seq variant calling pipeline 78 to identify the candidate A-to-I editing sites in each brain library ( i . e . , the A-to-G differences in the final sequencing results ) ., We identified 1 , 531 unique sites with A-to-G DNA-RNA differences in these brain libraries , and such differences accounted for 81 . 5% ( with a standard error of 0 . 97% ) of all the differences detected by the GATK pipeline in each library ( S2A Fig ) ., In contrast , the proportion of A-to-G DNA differences ( reference vs . alternative allele ) was only 9 . 9% out of all the mutations ( S2B Fig ) in the global populations of D . melanogaster 79 ., This comparison justified the reliability and accuracy of our procedures in defining the candidate A-to-I editing sites ., Second , we retrieved a total of 5 , 389 editing sites characterized in D . melanogaster in previous studies ( 972 in Graveley et al . 47 , 1 , 350 in Rodriguez et al . 49 , 3 , 581 in St . Laurant et al . 48 , and 1 , 298 in Yu et al . 50 ) ., Altogether , we obtained 5 , 925 unique candidate sites ( 986 sites overlapped between GATK and the four previous studies , S2 Table ) ., For each candidate site in a brain library k , we calculate the probability that the A-to-G difference ( if detected ) is caused by editing with Pk ( E1 ) = 1 − Pk ( E0 ) , where Pk ( E0 ) is the probability that the difference is solely caused by sequencing error ( ε ) , by incorporating the sequencing coverage ( Ck ) and the number of G allele ( Lk ) at that site ., Next , we utilize the multiple-sample information and calculate the joint probability that this site is edited in at least one library , P ( E1 ) = 1 − P ( E0 ) , where P ( E0 ) is the probability that the A-to-G differences observed in that site across all the applicable libraries are entirely caused by sequencing errors ( Materials and Methods ) ., Among the 5 , 925 candidate sites , we did not detect expression of 664 sites in any brain library ., For the remaining 5 , 261 expressed sites , we divided them into five exclusive classes with decreasing confidence based on P ( E1 ) , sequencing coverage , and the number of libraries in which the editing events were detected ., Briefly , Class I ( 1 , 702 sites ) were defined with the following criteria:, 1 ) at FDR of 0 . 001 ,, 2 ) the maximum sequencing coverage across all the libraries ( Cmax ) ≥ 10 ,, 3 ) the total coverage across all the libraries ( Ctotal ) ≥ 40 , and, 4 ) editing was detected in at least two libraries ., Among the remaining sites that had editing events detected in at least two libraries , we defined Class II ( 447 editing sites ) with these criteria:, 1 ) at FDR of 0 . 01 ,, 2 ) Cmax ≥ 5 and, 3 ) Ctotal ≥ 16 ( we also employed other cutoffs to define editing sites in Class I and II , and obtained results not very different from the results reported here; see S3 Table for details ) ., 824 sites do not meet the aforementioned two criteria but have P ( E1 ) > 0 . 99 , which suggests they might also be edited , although with lower confidence in brains of D . melanogaster ( Class III ) ., Moreover , 131 sites have editing detected in at least one library but have P ( E1 ) ≤ 0 . 99 ( Class IV ) ., Notably , we detected mRNA-Seq reads covering the remaining 2 , 157 candidate sites but none of them has editing events detected ( Class V ) ., The detailed information about these sites is presented in S2 Table ., It is not surprising that the sequencing coverage decreases in the order of Class I , II , and III in each library the median coverage in each library is 17 . 4±1 . 4 ( mean ± s . e . throughout this study ) , 3 . 87±0 . 35 and 1±0 raw reads , respectively; S3A Fig ., Interestingly , although the sites in Class II have significantly lower coverage compared to sites in Class I ( P < 0 . 05 in each library , Kolmogorov-Smirnov tests ) , the editing levels are even significantly higher in Class II than in Class I ( the median editing level in each library is 0 . 24±0 . 02 vs . 0 . 16±0 . 01 in Class II vs . I , S3B Fig ) ., From another perspective , 45 . 5% of the Class I sites were edited in all the eight brain libraries , meanwhile , only 8 . 5% of the Class II sites were edited in all the eight brain libraries ( P < 0 . 01 , Fisher’s exact test; S3C Fig ) ., Compared to Class I and II , sites in Class III have both lower coverage and editing levels ( S3A and S3B Fig ) ., Sites in Class IV are extremely lowly edited and sites in Class V do not have any editing event detected in our samples; however , these two classes do not have the lowest sequencing coverage compared to the other three classes ( the median coverage in a library is 18 . 6±1 . 1 and 6 . 1±0 . 35 for Class IV and V respectively , S3A Fig ) , suggesting they might have negligible editing in brains of D . melanogaster ., To estimate the false positive rates of the editing sites in each class , we analyzed the RNA-Seq datasets from paired samples of wild-type strain w1118 vs . Adar5G1 mutant of D . melanogaster as conducted previously 50 , 51 ., We found 1 , 145 , 161 , and 103 editing sites in Class I , II , and III respectively that have editing events detected in w1118 heads , and correspondingly , 33 , 2 , and 7 of these sites were detected in the heads of Adar5G1 mutant , yielding a false positive rate of 2 . 88% , 1 . 24% and 6 . 80% for Class I , II , and III , respectively ., Therefore , the sites in Class I and II captured the editing events in brains of D . melanogaster with high accuracy , and Class III sites were not considered in the down-stream analysis due to the high positive rate ., For the sites in Class I and II , we identified 1 , 630 ( 1 , 243 in Class I and 387 in Class II ) sites overlapped with previous studies 47–50 , and 519 sites ( 459 in Class I and 60 in Class II ) are novel in this study ( S2 Table ) ., It is not uncommon that many editing sites are not overlapping between studies in Drosophila 47–50: on average 30 . 4±3 . 6% of the editing sites are shared in pairwise comparisons ( ranging from 12 . 8% to 54 . 7% , S4 Table ) ; and we observed comparable proportions of shared sites between our study and the previous ones: 21 . 8% , 36 . 7% , 61 . 1% and 28 . 2% of the Class I+II sites in our study are overlapping with Graveley et al . 47 , Rodriguez et al . 49 , St . Laurant et al . 48 , and Yu et al . 50 , respectively ( S4 Table ) ., Importantly , when we pooled Class I and II together , we found the novel sites have comparable false positive rates as the common ones in the w1118 vs . Adar5G1 mutant analysis ( 8/242 = 3 . 31% vs . 27/1064 = 2 . 54% for the novel vs . common sites ) ., Furthermore , 111 of the novel sites are annotated in Ramaswami et al . 52 , which is independent from this study ., Taken together , we identified 2 , 114 “high-confidence” editing sites after combining sites in Class I and II ( 35 sites that have A-to-G difference in Adar5G1 mutants were removed ) , including 1 , 603 ( 75 . 8% ) sites overlapped with sites identified by previous studies 47–50 and 511 ( 24 . 2% ) novel sites ( Fig 1B ) ., The novel sites have slightly higher sequencing coverage than the common sites in the brain libraries ( the median coverage is 25 . 1±1 . 5 and 20 . 5±1 . 1 for the former and latter , respectively , P < 0 . 05 in each library , KS tests; S3D Fig ) , but generally lower editing levels ( the median is 0 . 18±0 . 006 vs . 0 . 36±0 . 006 , P < 10−16 in each library , KS tests; Fig 1C ) ., Moreover , compared to the common editing sites , the novel sites are generally edited in fewer brain samples: 42 . 9% of the common ones were detected in all the eight brain libraries , while only 20 . 0% of the novel sites were detected in all the libraries ( P < 10−16 , Fisher’s exact test; S3E and S3F Fig ) ., Altogether these results suggest that these novel editing sites are genuine but lowly edited in the brains , and were probably diluted in the samples of previous studies that were carried out in heads or whole flies 47–50 ., Among the 2 , 114 high-confidence sites ( Fig 1D ) , 235 ( 11 . 1% ) are in intergenic regions , 42 ( 2 . 0% ) are in ncRNAs , and 1 , 837 ( 86 . 9% ) are in 517 protein-coding genes , including 20 ( 0 . 95% ) in 5 UTRs , 550 ( 26 . 0% ) in introns , 414 ( 19 . 6% ) in 3 UTRs , 678 ( 32 . 1% ) nonsynonymous ( in CDS regions and causes amino acid changes when edited , abbreviated as N throughout this study ) , and 144 ( 6 . 8% ) synonymous ( in CDS regions but do not cause amino acid changes when edited , abbreviated as S ) , and one editing site ( chr3R:18806029 ) that putatively disrupts the stop codon of CG18208 ( UAG>UGG ) ., The detailed annotation in each library was presented in S5 Table ., The gene ontology analysis revealed that the high-confidence exonic editing sites were significantly enriched in genes that encode transporters , synaptic vesicles or neurotransmitters ( S6 Table and S7 Table for male and female brains , respectively; and the top 50 genes that had the largest number of editing sites were presented in S8 Table ) ., For the exonic editing sites , the editing levels ( averaged across libraries ) decrease in the order of N ( 0 . 319±0 . 010 ) , S ( 0 . 214±0 . 017 ) , 3 UTRs ( 0 . 168±0 . 008 ) , and 5 UTRs ( 0 . 133±0 . 020 ) , with levels in N sites significantly higher than those in the other three categories in the brains of D . melanogaster ( P < 0 . 001 , KS test; Fig 1D ) , suggesting that high levels of nonsynonymous editing events are overall favorable ., Among the 550 intronic editing sites , 167 of them might also be exonic due to alternative splicing ( we only used annotations of the canonical transcript and some intronic sites in the canonical transcripts might be coding in the non-canonical transcripts ) , and the coverage and editing levels ( 0 . 336±0 . 013 ) are comparable to the N sites ( Fig 1D ) ., Interestingly , the remaining 383 authentic intronic sites generally have significantly lower coverage than the coding regions ( Fig 1D ) , however , high editing levels in these sites ( 0 . 418±0 . 006 ) were observed , supporting previous results that editing is exerted co-transcriptionally 49 ., We uncovered a tendency that A-to-I editing events were more readily detected in the genes with higher expression levels ( or adenosine sites with higher mRNA-Seq coverage ) ., In each brain sample , when we grouped the expressed genes into 20 bins with increasing expression levels ( only genes with RPKM ≥ 1 were considered ) , we found a significant positive correlation between the editing density ( hereafter defined as the number of edited out of the total adenosine sites ) and the gene expression level ( P < 0 . 001 in each library; S4A Fig ) ., Similar patterns were observed if we grouped all the adenosine sites with increasing mRNA-Seq coverage in each sample ( only sites ≥ 5X coverage were considered; S4B Fig ) ., Analogous results were obtained if we weighted each editing site with its editing level ( “level-weighted density of editing sites” , see S4A and S4B Fig ) ., Consistent with previous observations 47–49 , we found the editing density was significantly increased from the 5 to 3 of pre-mRNAs ., After dividing the adenosine sites ( ≥ 5X coverage ) into 20 equal bins along their positions in pre-mRNAs , our meta-gene analysis indicated that the editing density in each bin was significantly positively correlated with the relative distance of that bin to the transcriptional start sites ( P < 0 . 005 in each library; S4C Fig ) ., Despite our experimental optimization , the poly ( A ) selection procedure still caused slightly increased coverage bias towards 3 ends of mRNAs ( S4D Fig ) ., However , we found the coverage difference between 5 and 3 of mRNAs was not the main cause of elevated editing density in the 3 ends of mRNAs with two analyses ., First , in each library , we split each gene ( RPKM ≥ 1 ) into two equal parts , calculated the RPKM values for each half-gene separately , ranked all the half-genes with increasing RPKM values , and grouped them into 20 bins ., Next , in each bin , we combined the 5 and 3 half-genes independently and calculated the editing density in the 5 and 3 half ., We found within each bin the editing density in the 3 half-genes are significantly higher than the 5 half genes ( P < 0 . 05 in each library; paired t tests , S4E Fig ) ., To further reduce the coverage variation within the half-genes , we ranked all the adenosine sites ( ≥ 5X ) with increasing coverage and binned them into 20 groups , and in each group we calculated the editing density in the 5 ( front ) half and 3 ( rear ) half of pre-mRNAs independently ., We also found the editing density were significantly higher for sites in the rear half compared to sites in the front half of pre-mRNAs ( P < 0 . 001 in each library; paired t tests , S4F Fig ) ., Taken together , the increasing editing density along mRNAs is not likely caused by detection bias , but more likely shaped by the recruitment of ADAR to the transcription elongation complex , as previous functional studies demonstrated 49 , 80 ., We predicted that 591 ( 50 . 1% ) of the 1 , 179 exonic editing sites were located in stable local hairpin structures of mRNAs ( Materials and Methods ) , such as Adar ( S5A Fig ) , rtp , DIP1 , rdgA , CG43897 , and CG42540 ( editing events in these genes were verified with Sanger sequencing; S5B and S5C Fig ) ., In contrast , we obtained only 363 exonic sites ( 332–393 sites within 95% CI ) located in stable hairpin structures after comprehensively folding all the transcripts expressed in brains and randomly sampling the equal amounts of editing sites ( Materials and Methods; Fig 1E ) ., Similar results were obtained when we focused on the N or S editing sites individually ( P < 0 . 002 in simulations for both cases; Fig 1E ) ., In addition , we found 181 intronic editing sites located in stable hairpin structures when we folded the pre-mRNA sequences ., Long-range pseudoknots are another class of RNA substrates recognized by ADAR 81 ., By extensively folding the flanking sequences of the editing sites ( see Methods for details ) , we inferred 260 ( 22 . 1% ) exonic editing sites that were located outside stable hairpin structures but were located in stems of long-range pseudoknots in pre-mRNAs of genes , such as the 3 UTR of Adar ( S5D Fig ) , nrm , B52 , nAchRbeta1 , CG8034 and roX1 ( S6 Fig; the editing events in nrm were verified by Sanger sequencing of the cDNA and genomic DNA , S6B and S6C Fig ) ., Taken together , our results systematically demonstrated that at least 874 ( 74 . 1% ) of the exonic A-to-I editing sites in the brains of D . melanogaster were located in pre-mRNA regions that formed stable secondary structures ., These results also well explain why the A-to-I editing sites are located in clusters , as commonly observed in previous studies 41 , 47–49 , 62 ., By clustering the editing sites with distances smaller than 100 nucleotides , we identified a total of 1 , 320 editing sites that form 413 clusters in brains of D . melanogaster ( S7 Fig ) , and unusually large editing clusters were frequently observed , such as in NaCP60E and CaMKII ( the Sanger verification was presented in S8 Fig ) ., To characterize the A-to-I editing events that were evolutionarily conserved ( i . e . , commonly observed ) across species , we deep sequenced the poly ( A ) -tailed transcriptomes of female and male brains of 1- to 5-day-old D . simulans and D . pseudoobscura that were accommodated at the same temperature conditions as D . melanogaster ( six libraries for each species ) ., The mapped reads range from 8 . 7–16 . 4M in each library of D . simulans , and 10 . 9–17 . 8M in D . pseudoobscura ( Table 2 , Table 3 and S1 Table for detailed information ) , and the median sequencing coverage on an exonic site in a library ranges from 5 to 9 reads in D . simulans , and ranges from 4 to 5 in D . pseudoobscura ( S1 Fig ) ., D . simulans diverged from D . melanogaster ~5 . 4 million years ago ( Fig, 2 ) while D . pseudoobscura diverged from D . melanogaster approximately 55 million years ago 82 ., Comparing A-to-I editing across these three species will help us understand the role of natural selection in shaping the brain editomes during evolution ., To exclude SNPs in the RNA editing characterization , we also deep sequenced the genomic DNA of the same strain of D . simulans ( the median coverage per site is 46 , totally 313 , 133 SNPs , S9A Fig ) and D . pseudoobscura ( the median coverage per site is 47 , totally 489 , 828 SNPs , S9B Fig ) and masked all the SNPs ( Materials and Methods ) ., For each high-confidence editing site in brains of D . melanogaster , we employed two complementary approaches to search for the orthologous sites in D . simulans and D . pseudoobscura ., First , we used liftOver 83 to convert the genomic coordinates of the orthologous sites between D . melanogaster and D . simulans , or between D . melanogaster and D . pseudoobscura , based on the pairwise genome alignments as previously conducted 51 ( termed “g_align” approach , Materials and Methods ) ., Second , we parsed out the genomic coordinates with the pairwise CDS alignments that were made based on the protein alignments between D . melanogaster and the other species ( “c_align” approach ) ., We pooled orthologous sites by the two approaches together ., For each site in each species , we calculated the joint probability that this site is edited in at least one library P ( E1 ) ., At FDR of 0 . 05 , we identified 996 sites edited in D . simulans ( S9 Table ) , and 451 sites edited in D . pseudoobscura ( S10 Table ) , and 367 sites edited in both D . simulans and D . pseudoobscura ( Fig 2 ) ., We present the editing sites evolutionarily conserved in the same gender under the same temperature conditions in D . simulans ( Table 2 and S11 Table ) and D . pseudoobscura ( Table 3 and S12 Table ) ., For the editing sites we characterized in the brains of D . melanogaster , 34 . 3–44 . 0% of them have editing events detected in the matched samples of D . simulans ( Table 2 ) , and 22 . 3–24 . 1% of them have editing in the matched samples of D . pseudoobscura ( Table 3 ) ., Note the proportion of editing sites in D . melanogaster that have editing events detected in brains of other species varies across libraries since we required the sites are edited in both paired samples ., In general , with divergence increases , the level of conserved editing sites decreased , suggesting the editing events are evolutionary dynamic ., Notably , for the 996 editing sites with conserved events in both D . melanogaster and D . simulans , and the 451 editing sites with conserved events in both D . melanogaster and D . pseudoobscura , we found 416 ( 41 . 8% ) and 78 ( 17 . 3% ) of them are located outside the coding regions , respectively ( Tables 2 and 3 ) , which is consistent with a recent study 84 and suggests a possible functional role for these sites , such as influencing alternative splicing 26–29 , microRNA targeting 30–32 , or other cellular processes related to RNAs 14 ., Comparing the editing sites with conserved and non-conserved events revealed two interesting features ., First , in each brain library , the editing levels are significantly higher in the sites with evolutionarily conserved editing events than in the remaining sites ( the mean level in a library is 0 . 340±0 . 008 vs . 0 . 252 ± 0 . 009 in the D . melanogaster/D . simulans comparison , and 0 . 323±0 . 012 vs . 0 . 187±0 . 009 in the D . melanogaster/D . pseudoobscura comparison; P < 0 . 01 in each comparison , KS tests , S10 Fig ) ., Second , the N sites are significantly enriched in the editing sites that are evolutionarily conserved: 72 . 9% ( 494 out of 678 ) N sites compared to 35 . 0% ( 502 out of 1436 ) of the remaining sites that have evolutionarily conserved events between D . melanogaster and D . simulans ( P < 10−10 , Fisher’s exact test ) , and 47 . 9% ( 325 out of 678 ) N sites compared to 8 . 77% ( 126 out of 1436 ) of the remaining sites that have evolutionarily conserved events between D . melanogaster and D . pseudoobscura ( P < 10−10 , Fisher’s exact test ) , suggesting the nonsynonymous editing events are generally maintained and regulated by different evolutionary forces compared to the other sites ., There are 84 editing sites that have editing events detected in both D . melanogaster and D . pseudoobscura but without editing events confidently identified in D . simulans ., Nevertheless , this does not necessarily mean these sites are not edited in D . simulans ( for 60 of these sites we did not find the orthologous sites in D . simulans , and for the 24 remaining sites , 10 of them have low level of editing but are undistinguishable from sequencing errors; S13 Table ) ., Sampling bias frequently causes the sites with low expression or low editing levels to yield no editing signals in the sequencing libraries ., Therefore , next we only focused on the sites with high sequencing coverage to explore the possible gain and loss patterns of editing events ., We obtained 87 editing sites that have minimal editing level of 0 . 05 in D . melanogaster and have at least 200 raw reads ( across all the libraries ) in both D . simulans and D . pseudoobscura ., We found 52 sites with editing events reliably detected in all the three species ., For each of the remaining 35 sites , in case no editing event was discovered at a site in a sample m in D . simulans ( or D . pseudoobscura ) , we calculate Pm ( D0 ) , the probability that this observation happens by sampling bias or because the editing signal was abolished by sequencing error ( ε ) , given a depth of Cm and an assumed editing level lm at that site ( Materials and Methods ) ., We assumed the orthologous sites in the other species have the same editing level as in D . melanogaster and calculated the joint probability P ( D0 ) that a site was edited despite zero edited allele was detected in all the libraries ., After correcting for multiple testing , at FDR of 0 . 05 , we found 20 sites with editing present in both D . melanogaster and D . simulans but absent in D . pseudoobscura , and 3 sites with editing specifically present in D . melanogaster ., The most parsimonious interpretation is that the brain editome in Drosophila is expanding during evolution ( Fig 2 ) ., We did not find any convincing case that editing was detected in both D . melanogaster and D . pseudoobscura but was absent in D . simulans , suggesting that the established editing events , at least for the set we examined here , are well maintained by natural selection during evolution ., In contrast to previous observations that nonsynonymous editing events were generally non-adaptive in mammals 67 , 68 and in Drosophila 50 , our analysis revealed the nonsynonymous editing events in Drosophila brains were predominantly adaptive ., The ratio of nonsynonymous ( N ) to synonymous ( S ) editing sites ( N/S ) in different brain samples of D . melanogaster ranges from 4 . 79 with 95% CI ( 3 . 85 , 6 . 20 ) to 6 . 25 ( 4 . 70 , 8 . 67 ) , all of which is significantly higher than the ratio expected under neutrality ( 3 . 80 ) that was calculated similarly as previously described 67 ( Materials and Methods; P < 0 . 03 in each library , Fisher’s exact tests; Table 1 ) ., In other words , in the brains of D . melanogaster , the rate of nonsynonymous A-to-I editing is significantly higher than the rate of synonymous editing ., Given the observed and expected N/S ratios under neutrality ( randomness ) , a conservative estimation is that 20 . 7% with 95% CI ( 1 . 3% , 38 . 7% ) of the N sites in the brains of D . melanogaster might be adaptive ( Table 1 ) ., Moreover , we obtained higher N/S ratios in each brain library when we increased the cutoff of editing level ( S11 Fig ) ., Our analysis is essentially the same as the classical dN/dS ( the number of nonsynonymous changes per nonsynonymous site over the number of synonymous changes per synonymous site ) test of DNA sequences in molecular evolution 85 , and provides compelling evidence that the nonsynonymous editing events in Drosophila brains are overall beneficial and favored by natural selection ., We observed significantly negative correlations between the sequencing coverage ( C ) and editing level ( l ) in each brain library or when we pooled the library together ( P < 10−10 in each case , S12 Fig ) ., These patte | Introduction, Results, Discussion, Materials and methods | Adenosine-to-inosine ( A-to-I ) editing is hypothesized to facilitate adaptive evolution by expanding proteomic diversity through an epigenetic approach ., However , it is challenging to provide evidences to support this hypothesis at the whole editome level ., In this study , we systematically characterized 2 , 114 A-to-I RNA editing sites in female and male brains of D . melanogaster , and nearly half of these sites had events evolutionarily conserved across Drosophila species ., We detected strong signatures of positive selection on the nonsynonymous editing sites in Drosophila brains , and the beneficial editing sites were significantly enriched in genes related to chemical and electrical neurotransmission ., The signal of adaptation was even more pronounced for the editing sites located in X chromosome or for those commonly observed across Drosophila species ., We identified a set of gene candidates ( termed “PSEB” genes ) that had nonsynonymous editing events favored by natural selection ., We presented evidence that editing preferentially increased mutation sequence space of evolutionarily conserved genes , which supported the adaptive evolution hypothesis of editing ., We found prevalent nonsynonymous editing sites that were favored by natural selection in female and male adults from five strains of D . melanogaster ., We showed that temperature played a more important role than gender effect in shaping the editing levels , although the effect of temperature is relatively weaker compared to that of species effect ., We also explored the relevant factors that shape the selective patterns of the global editomes ., Altogether we demonstrated that abundant nonsynonymous editing sites in Drosophila brains were adaptive and maintained by natural selection during evolution ., Our results shed new light on the evolutionary principles and functional consequences of RNA editing . | Adenosine-to-inosine ( A-to-I ) RNA editing is an evolutionarily conserved mechanism that alters RNA sequences at the co-transcriptional or post-transcriptional level ., RNA editing is hypothesized to facilitate adaptation in that it expands the transcriptomic and proteomic diversity ., However , evidence for adaptation of RNA editing at the whole editome level is still lacking ., In this study we systematically identified A-to-I RNA editing sites in female and male brains of three Drosophila species at different temperatures ., With evolutionary analysis from different perspectives , we provide lines of evidence to demonstrate that the nonsynonymous editing sites in Drosophila brains are generally adaptive ., The signals of adaptation for the editing sites are significantly enriched in genes related to chemical and electrical neurotransmission ., We show that the RNA editing events might interplay with gene expression plasticity in temperature stress responses ., Furthermore , we demonstrated that the expression level of Adar , together with the expression profiles of a set of genes that have editing sites favored by natural selection , were important in shaping the overall selective patterns of the global editomes at different developmental stages ( or tissues ) of D . melanogaster ., Altogether our results support the hypothesis that A-to-I editing provides a driving force for adaptive evolution in Drosophila from different aspects . | invertebrates, glycosylamines, computational biology, animals, invertebrate genomics, animal models, drosophila melanogaster, model organisms, experimental organism systems, genome analysis, drosophila, adenosine, research and analysis methods, genomic libraries, gene expression, evolutionary genetics, insects, animal genomics, arthropoda, biochemistry, rna, rna editing, nucleic acids, nucleosides, genetics, biology and life sciences, genomics, evolutionary biology, glycobiology, organisms | null |
journal.pbio.2003067 | 2,018 | Spliced integrated retrotransposed element (SpIRE) formation in the human genome | Long interspersed element-1 ( L1 ) is a non-long terminal repeat ( non-LTR ) retrotransposon that comprises approximately 17% of human genomic DNA 1 ., Over 99 . 9% of human L1s cannot retrotranspose due to 5′ truncations , internal DNA rearrangements , or point mutations that inactivate the L1-encoded proteins 1–4 ., However , the average diploid genome harbors approximately 80–100 full-length retrotransposition-competent L1s ( RC-L1s ) 5 , including a small number of expressed 6–8 , highly active ( i . e . , “hot” ) L1s 5 , 9–11 that can retrotranspose efficiently in cultured cells or cancers ., RC-L1 retrotransposition affects both intra- and interindividual human genetic variation ( reviewed in 12 , 13 ) and , on occasion , can lead to disease-producing mutations 14–16 ., Human RC-L1s are approximately six kilobases ( kb ) in length 17 , 18 ., They contain a 5′ untranslated region ( UTR ) that harbors both sense 19 and antisense 20 RNA polymerase II promoters ( Fig 1A ) as well as an antisense open reading frame ( ORF0 ) 21 , which encodes a protein that may mildly enhance L1 retrotransposition efficiency ., The 5′UTR is followed by two open reading frames ( ORF1 and ORF2 ) that are separated by a 63–base pair ( bp ) inter-ORF spacer that contains two in-frame stop codons 18 , 22 ( Fig 1A ) ., L1s end with a 3′UTR , which contains a conserved polypurine motif , a “weak” RNA polymerase II polyadenylation signal , and a variable length polyadenosine ( poly ( A ) ) tract ( Fig 1A ) 17 , 23–25 ., ORF1 encodes an approximately 40-kilodalton ( kDa ) protein ( ORF1p ) 26 that contains an amino-terminal coiled-coil ( CC ) domain required for ORF1p trimerization 27–29 , a centrally located noncanonical RNA recognition motif ( RRM ) domain 29 , 30 , and a carboxyl-terminal domain ( CTD ) harboring conserved basic amino acid residues 29–32 ( Fig 1A ) ., The RRM and CTD are critical for ORF1p nucleic acid binding 32–36; the nucleic acid chaperone activity is postulated to play a role in L1 integration 30 , 36 , 37 ., ORF2 encodes an approximately 150-kDa protein ( ORF2p ) 38–40 that contains conserved apurinic/apyrimidinic-like endonuclease ( EN ) 41 , 42 and reverse transcriptase ( RT ) domains 31 , 43 , 44 as well as a conserved cysteine-rich ( C ) domain 31 , 45 ( Fig 1A ) ., Biochemical activities contained within both ORF1p and ORF2p are required for canonical EN-dependent L1 retrotransposition in cultured human cells 31 , 41 ., A round of human RC-L1 retrotransposition begins with the internal sense-strand promoter initiating transcription at or near the first nucleotide of the 5′UTR 13 , 19 , 46 , 47 ., The resultant bicistronic L1 mRNA is exported to the cytoplasm , where it undergoes translation 22 , 46 , 48 , 49 ., Following translation , ORF1p and ORF2p preferentially bind to their encoding mRNA in cis to form a ribonucleoprotein particle ( RNP ) 33 , 35 , 50–53 ., The 3′ poly ( A ) tail of L1 mRNA is a critical Cis-acting determinant for recruitment of nascent ORF2p to L1 RNA 51 , 54 ., Components of the L1 RNP gain access to the nucleus by a mechanism that does not require nuclear envelope breakdown 55 , 127 ., L1 integration likely occurs by target-site primed reverse transcription ( TPRT ) 41 , 53 , 56 , 57 ., During TPRT , the L1 EN makes a single-strand endonucleolytic nick at a thymidine-rich sequence ( e . g . , 5′-TTTT/A-3′ , 5′-TTTC/A-3′ , etc . ) present on the “bottom” strand of a target site in genomic DNA to liberate a 3′ hydroxyl ( 3′-OH ) group 41 , 57 , 58 ., Microhomology-based annealing between the L1 poly ( A ) tail and thymidine residues at the L1 EN cleavage site in genomic DNA enhances the ability of the L1 ORF2p RT to use the resultant 3′-OH group as a primer to initiate reverse transcription of L1 mRNA 53 , 59 ., How TPRT is completed requires elucidation ., However , as demonstrated for the related R2 non-LTR retrotransposon from Bombyx mori 60 , it is possible that enzymatic activities associated with L1 ORF2p participate in both second-strand ( i . e . , “top” strand ) genomic DNA cleavage and second-strand L1 cDNA synthesis ., Although retrotransposition assays and biochemical studies revealed the L1-encoded proteins preferentially retrotranspose their encoding mRNA in cis 50 , 53 , 61 , 62 , L1 ORF1p and/or ORF2p can act in trans ( Trans-complementation ) to retrotranspose RNAs encoded by nonautonomous short interspersed elements ( SINEs; e . g . , Alu RNA 63 , 64 and SINE-R/VNTR/Alu SVA RNA 65–67 ) ., Additionally , the L1-encoded protein ( s ) can act in trans to retrotranspose noncoding RNAs 12 , 68–73 and cellular mRNAs , with the latter process leading to the formation of processed pseudogenes 50 , 62 , 73–77 ., The evolutionary success of L1 requires the faithful retrotransposition of full-length L1 mRNAs ., Previous studies have revealed the presence of functional splice donor ( SD ) , splice acceptor ( SA ) , and premature polyadenylation signals in primary full-length RC-L1 transcripts 24 , 78–81 ., Paradoxically , the use of these sites during posttranscriptional RNA processing leads to the production of truncated and/or internally deleted L1 mRNAs 24 , 78–81 , which could adversely affect L1 retrotransposition ., Thus , it is somewhat surprising that Cis-acting sequences that could negatively affect L1 retrotransposition have not been removed by negative selection during L1 evolution ., Here , we address how the retrotransposition of spliced L1 mRNAs leads to the generation of spliced integrated retrotransposed elements ( SpIREs ) ., We describe two classes of SpIREs: those that splice within the 5′UTR ( intra-5′UTR SpIREs ) and those that splice from within the 5′UTR into the ORF1 sequence ( 5′UTR/ORF1 SpIREs ) ., Additionally , we suggest a mechanism for why some apparently deleterious Cis-acting splice sites within L1 mRNA are conserved throughout L1 evolution ., Finally , we provide experimental evidence revealing that L1 splicing dynamics are altered by structural changes within the 5′UTR that allow L1s to evade host repression and that retrotransposition of the resultant spliced variants can lead to the generation of new classes of SpIREs ., Thus , these data provide a snapshot of how an “arms race” between L1 and host repressive factors may affect the evolutionary trajectory of L1 5′UTRs ., In sum , we conclude that SpIREs are deficient for retrotransposition and likely represent evolutionary “dead-ends” in the L1 retrotransposition process ., Using fosmid-based discovery methods , we previously identified a polymorphic L1 ( fosmid accession #AC225317 ) in the human population that contains a 524-nucleotide deletion within its 5′UTR 10 ., Upon closer inspection , we determined that this deletion likely resulted from the retrotransposition of a spliced L1 RNA that used a previously identified SD ( G98U99 ) 78 and an unreported SA ( A620G621 ) within the L1 5′UTR ( numbering based on L1 . 3 , accession #L19088; 9 , 82 ) ( Fig 1A and 1B ) ., The structure of this element resembled previous L1s characterized by Belancio and colleagues , supporting the hypothesis that spliced L1 transcripts can complete retrotransposition in the human genome 78 , 79 ., We named these L1s SpIREs to distinguish them from bona fide full-length genomic L1s ., The three SpIREs investigated here all use the same SD ( G98U99 ) but use different SA sequences that reside within either the L1 5′UTR ( SA: A620G621 or SA: A788G789 ) or L1 ORF1 ( SA: A974G975 ) ( Fig 1B , 1C and 1D ) ., We used the BLAST-like alignment tool ( BLAT ) ( https://genome . ucsc . edu ) 83 ( in which transposable element—derived DNAs are not masked ) to search the human genome reference ( HGR , GRCh38/hg38 ) for SpIRE G98U99/A620G621 sequences ( referred to as SpIRE97/622 ) ., The HGR contains an annotated record of L1s that have accumulated over evolutionary time ( i . e . , millions of years ) ; thus , searching the genome should reveal how SpIREs contribute to the genomic L1 repertoire ., We used a 100-nucleotide in silico probe that spans the intra-5′UTR splice junction present in SpIRE97/622 ( nucleotides 47–97 and 622–672 of L1 . 3 ) to query the HGR ., We identified 116 SpIRE97/622 sequences , which span the youngest L1PA1 subfamily ( also known as L1Hs , members of which are currently amplifying in the human population ) through the older L1PA6 subfamily ( which amplified approximately 27 million years ago MYA ) , but none in older ( L1PA7–L1PA17 , L1PB , and L1MA ) L1 subfamilies ( S1A Fig ) 84 , 85 ., Thus , 116 out of 6 , 609 ( about 1 . 8% ) of previously annotated full-length L1s in the L1PA1–L1PA6 subfamilies are actually SpIRE97/622 sequences ( S1 Fig; S1 Data; S1 Table ) ., Almost half of the SpIRE97/622 sequences we identified belong to the L1PA3 subfamily ( 53 sequences , comprising about 3 . 4% of previously annotated full-length L1s in that subfamily ) ( S1A Fig; S1 Data; S1 Table ) ., The L1PA1 subfamily harbors six SpIRE97/622 ( comprising about 2 . 0% of previously annotated full-length L1s in that subfamily ) and the L1PA6 subfamily contains only one SpIRE97/622 ( comprising 0 . 1% of previously annotated full-length L1s in that subfamily ) ( S1 Fig; S1 Data; S1 Table ) ., Seven SpIRE97/622 sequences could not be unambiguously assigned to a specific L1 subfamily and are classified as either L1PA2–L1PA3 or L1PA4–L1PA6 sequences ( S1 Fig; S1 Data; S1 Table ) 86 ., Given the above data , we used BLAT to search the HGR for additional L1s containing G98U99/A788G789 and G98U99/A974G975 splicing events identified by Belancio and colleagues ( referred to as SpIRE97/790 and SpIRE97/976 , respectively ) 78 , 79 ., These searches confirmed the presence of four previously identified SpIRE97/790 sequences in the L1PA1–L1PA2 subfamilies ( S1A Fig; S1 Data; S1 Table ) 78 ., We also discovered an additional SpIRE97/976 sequence in addition to the ten previously identified SpIRE97/976 sequences ( S1A Fig; S1 Data; S1 Table ) 78 , 79 ., In total , these three classes of SpIREs comprise a small but notable ( 131/6 , 609 or about 2% ) percentage of previously annotated full-length L1s from the L1PA1–L1PA6 subfamilies ., The SpIRE97/622 sequences discovered here represent the majority ( 116/131 or about 89% ) of identified SpIREs ., We next characterized the 131 SpIRE97/622 , SpIRE97/790 , and SpIRE97/976 sequences ., We first examined the post-integration ( i . e . , filled ) site of each SpIRE in the HGR sequence ., We then used the genomic sequences flanking each SpIRE to reconstruct a putative pre-integration ( i . e . , empty ) site ., Many of the SpIRE sequences , especially those from older L1 subfamilies , have degenerate poly ( A ) tails at their 3′ ends , which , in some cases , made it difficult to reconstruct the putative pre-integration site to bp resolution ( S1 Data; S1 Table ) ., These analyses revealed that SpIREs generally are flanked by target site duplications that ranged in size from about 6–25 bp , end in a 3′ poly ( A ) tract , and integrated into an L1 EN consensus cleavage site ( 5′-TTTT/A-3′ and variants of that sequence ) ( S1 Data; S1 Table ) ., Consistent with previous studies , approximately 39% ( 51/131 ) of the SpIREs are present within the introns of RefSeq ( https://www . ncbi . nlm . nih . gov/refseq/ ) 87 annotated genes 69 , 88 , 89 , and the majority ( 32/51 or about 63% ) of these SpIREs are present in the opposite transcriptional orientation of the annotated gene ( S1 Table ) 90 , 91 ., Other structural features of the SpIREs are shown in S1 Data and S1 Table ., In sum , our analyses strongly suggest that SpIREs represent a subset of genomic L1 insertions and retrotranspose by the canonical process of L1 EN-dependent TPRT ., The formation of SpIRE97/622 results in the deletion of five known transcription factor binding sites within the L1 5′UTR 47 , 92–97 ( Fig 1A ) ; thus , we hypothesized the SpIRE97/622 5′UTR would have reduced promoter activity ., To test this hypothesis , we subcloned the wild-type ( WT ) L1 . 3 9 , 82 or SpIRE97/622 5′UTR sequences upstream of a promoter-less firefly ( Photinus pyralis ) luciferase gene ( vector pGL4 . 11 ) , creating pPLWTLUC and pPL97/622LUC , respectively ( Fig 2A ) ., We then characterized the promoter activity of these 5′UTRs using functional assays ., We first conducted northern blot analyses using polyadenylated mRNAs isolated from untransfected HeLa-JVM cells and HeLa-JVM cells transfected with the luciferase expression vectors ( Fig 2 ) ., An RNA probe complementary to ribonucleotides 7–99 of the L1 5′UTR ( Fig 2A; purple line ) detected a strong signal at the expected size of about 2 . 7 kb in mRNAs derived from HeLa-JVM cells transfected with pPLWTLUC , but not in mRNAs derived from HeLa-JVM cells transfected with pPL97/622LUC or pGL4 . 11 or from untransfected HeLa-JVM cells ( Fig 2B , first panel ) ., Similar results were obtained using riboprobes complementary to either ribonucleotides 103–336 of the L1 5′UTR ( Fig 2A , red line; Fig 2B , second panel ) or the 3′ end of luciferase ( Fig 2A , blue line; Fig 2B , third panel ) ., These data are consistent with previously published findings 47 , which demonstrated that L1 transcription begins at or near the first nucleotide of the L1 5′UTR ., Control experiments verified the integrity and quality of the mRNAs ( Fig 2B , actin probe ) ., We were able to detect a faint band representing the predicted approximately 2 . 2 kb mRNA from HeLa-JVM cells transfected with pPL97/622LUC upon the prolonged exposure of the northern blots using probes complementary to ribonucleotides 7–99 of the L1 5′UTR , but not using a probe complementary to ribonucleotides 103–336 of the L1 5′UTR ( S2A Fig; purple arrow ) ., The absence of the predicted approximately 2 . 2-kb band in HeLa-JVM cells transfected with pPL97/622LUC using a probe complementary to the 3′ end of the luciferase gene is likely due to the limits of detection in our assay ( S2A Fig ) ., The origin of the approximately 2-kb transcript remains unclear ( Fig 2B , S2A Fig , orange arrow ) ; however , it could be representative of transcript initiation downstream of the canonical transcriptional start site within the 5′UTR 47 , 98 ., These data suggest that the SpIRE97/622 5′UTR retains weak promoter activity ., Because the splicing events that gave rise to SpIRE97/790 and SpIRE97/976 led to larger deletions of the 5′UTR when compared to SpIRE97/622 , we reasoned that they would lead to a similar , if not a greater , reduction in transcriptional activity; thus , they were not tested in this assay ., To corroborate the northern blot analyses , we conducted dual luciferase expression assays on whole cell lysates ( WCLs ) derived from HeLa-JVM cells co-transfected with firefly luciferase-based vectors ( pPLWTLUC , pPL97/622LUC , or pGL4 . 11 ) and a constitutively expressed Renilla ( Renilla reniformis ) luciferase internal control plasmid ( pRL-TK; Methods ) ., Consistent with the northern blot data , HeLa-JVM cells transfected with pPLWTLUC exhibited an approximately 267-fold increase of normalized firefly luciferase activity when compared to cells transfected with the promoter-less pGL4 . 11 vector ( Fig 2C; S2 Table ) ., By comparison , HeLa-JVM cells transfected with pPL97/622LUC exhibited only about a 7-fold increase of normalized firefly luciferase activity when compared to cells transfected with the promoter-less pGL4 . 11 vector ( Fig 2C; S2 Table ) ., Together , the above data suggest that the splicing event leading to the generation of SpIRE97/622 severely compromises its promoter activity ., Given that splicing reduces L1 promoter activity , we examined why the G98U99 SD may be conserved in the L1 5′UTR ., Previous studies revealed that a RUNX3 binding site within the 5′UTR is important for maximal L1 promoter activity 96 ., Interestingly , the SD site used to generate the three classes of SpIREs reported here is contained within the core sequence of a RUNX3 binding site that is conserved from the L1PA1–L1PA10 subfamilies ( Fig 1A; SD: G98U99; S1B Fig ) 84 ., Thus , we hypothesized that this SD is retained to maintain an active RUNX3 transcription factor binding site ., To test this hypothesis , we mutated the SD sequence within the WT 5′UTR ( U99C , creating pPLSDmLUC ) 99 and tested if this mutation affects 5′UTR promoter activity ., Northern blot analyses using the previously described riboprobes detected a signal at about 2 . 7 kb in mRNAs derived from HeLa-JVM cells transfected with pPLSDmLUC ., However , there is markedly less of this mRNA when compared to cells transfected with pPLWTLUC ( Fig 2B; about 18% of pPLWTLUC ) ., In contrast , mutating the SA site within the WT 5′UTR ( A620C , creating pPLSAmLUC ) did not drastically affect L1 promoter activity ( Fig 2B ) ., Thus , our data are consistent with previous findings 96 and suggest that the retention of the complete RUNX3 site containing the G98U99 SD is critical for L1 promoter activity ., We next sought to identify spliced L1 mRNAs that might have given rise to SpIREs ., To this end , we conducted end-point reverse transcription PCR ( RT-PCR ) experiments using poly ( A ) mRNAs isolated from HeLa-JVM cells transfected with a series of L1/firefly luciferase expression vectors ( S2B Fig; Methods ) ., The REV-LUC oligonucleotide ( S2B Fig , purple line ) was used to initiate L1/firefly luciferase first-strand cDNA synthesis; the cDNA products then were PCR amplified using FWD-5′UTR ( S2B Fig , red line ) and REV-LUC ( S2B Fig , purple line ) oligonucleotide primers ., The resultant cDNAs were separated on an agarose gel , cloned , and characterized using Sanger DNA sequencing ., Control experiments conducted in the absence of RT revealed that the characterized PCR products were derived from the amplification of cDNAs ( S2C Fig ) ., We detected the predicted full-length L1/firefly luciferase cDNA products from HeLa-JVM cells transfected with pPLWTLUC , pPLSDmLUC , and pPLSAmLUC ( Fig 2D , yellow “*” in lanes 1 , 3 , and 4 ) as well as the shorter predicted L1/firefly luciferase cDNA product from HeLa-JVM cells transfected with pPL97/622LUC ( Fig 2D , yellow “#” in lane 2 ) ., In agreement with our northern blot experiments ( Fig 2B ) , we did not detect cDNAs consistent with SpIRE97/622 splicing in pPLWTLUC transfected HeLa-JVM cells ( Fig 2D ) ., However , we did detect an L1/firefly luciferase cDNA that corresponds to the SpIRE97/790 splicing event from cells transfected with pPLWTLUC and pPLSAmLUC ( Fig 2D , yellow “+” , lanes 1 and 4; Fig 1C ) 78 ., Importantly , this product was not detected in HeLa-JVM cells transfected with either pGL4 . 11 or pPLSDmLUC or untransfected HeLa-JVM cells ., We next tested whether intra-5′UTR splicing affects L1 mRNA translation ., L1 sequences were cloned into an episomal pCEP4 expression vector that contains a hygromycin B resistance gene and a cytomegalovirus ( CMV ) early promoter , which augments L1 expression ., HeLa-JVM cells were transfected with a WT L1 ( pJM101/L1 . 3 ) , an L1 that contains a 5′UTR deletion ( pJM102/L1 . 3 ) , or an L1 containing the SpIRE97/622 deletion ( pPL97/622/L1 . 3 ) ( Fig 3A ) 9 , 50 ., Western blot analyses were conducted using WCLs that were derived from hygromycin-resistant HeLa-JVM cells transfected with the above constructs 9 days post-transfection ., An ORF1p polyclonal antibody ( α-N-ORF1p; directed against amino acids +31 to +49 in L1 . 3 100 UniProtKB accession #Q9UN81 ) detected an approximately 40-kDa product in cells transfected with pJM101/L1 . 3 , pJM102/L1 . 3 , and pPL97/622/L1 . 3 but not in cells transfected with the pCEP/GFP control ( Fig 3B ) ., HeLa-JVM cells transfected with pPL97/622/L1 . 3 exhibited a slight reduction in the steady-state level of ORF1p when compared to HeLa-JVM cells transfected with pJM101/L1 . 3 or pJM102/L1 . 3 ( Fig 3B ) ., Because a CMV promoter augmented L1 transcription , it is unlikely that this reduction is due to reduced L1 expression ., It is possible that the slight reduction in ORF1p is due to an alteration of the L1 5′UTR RNA secondary structure and/or minor changes in the stability of pPL97/622/L1 . 3 mRNA when compared to pJM101/L1 . 3 and pJM102/L1 . 3 mRNAs ., The splicing event yielding SpIRE97/976 results in an amino-terminal ORF1 deletion of 66 nucleotides , including the canonical ORF1p methionine start codon ( Fig 3C , black AUG , 40 kDa ) ., We hypothesized that ORF1p synthesis might initiate from two methionine codons ( AUG ) that are located in weak Kozak consensus sequences either 102 or 270 ribonucleotides downstream from the canonical AUG start codon ( Fig 3C ) 101 ., If the downstream methionine codons are used for translation initiation , we expect to detect amino terminal truncated ORF1 proteins of about 33 kDa and 27 kDa , respectively ., Western blot analyses were conducted as above using WCLs derived from hygromycin-resistant HeLa-JVM cells transfected with pJM101/L1 . 3 , an L1 containing the SpIRE97/976 deletion ( pPL97/976/L1 . 3 ) , or pCEP/GFP control vectors ( Fig 3A ) 22 ., As predicted , the α-N-ORF1p and α-C-ORF1p antibodies detected an approximately 40-kDa protein in WCLs derived from HeLa-JVM cells transfected with pJM101/L1 . 3 but did not detect a protein in WCLs derived from HeLa-JVM cells transfected with the pCEP/GFP control ( Fig 3D , left and right panels ) ., The α-N-ORF1p antibody detected an approximately 33-kDa protein in WCLs derived from HeLa-JVM cells transfected with pPL97/976/L1 . 3 ( Fig 3D , left panel ) , whereas the α-C-ORF1p antibody detected approximately 33-kDa and approximately 27-kDa proteins in the same extracts and an unknown cross-reacting protein at about 25 kDa ( Fig 3D , right panel ) ., Similar results were obtained when RNP extracts were used in western blot experiments , although western blots performed with the α-C-ORF1p antibody did not detect the cross-reacting approximately 25-kDa protein ( S3A Fig ) ., To confirm that the approximately 33-kDa and 27-kDa products were ORF1p derived , we introduced a T7-gene10 epitope tag to the 3′ end of ORF1 , creating pPL97/976/L1 . 3-T7 ., Western blots using a α-T7 antibody recapitulated our previous results and , similar to RNP preparations , did not identify the cross-reacting approximately 25-kDa protein ( S3B Fig ) ., Thus , the 5′UTR/ORF1 splicing event leads to the generation of an mRNA that , if translated , results in the synthesis of amino-terminal truncated derivatives of ORF1p ., Our data indicate that SpIRE97/622 contains a defective promoter and , if transcribed , SpIRE97/622 mRNA is translated at slightly lower levels than WT L1 mRNA ., Thus , we hypothesized that an intra-5′UTR spliced L1 mRNA would be capable of undergoing an initial round of L1 retrotransposition ., However , the resultant full-length retrotransposition events would contain a defective promoter , which may compromise subsequent retrotransposition ., To test the above hypothesis , we examined whether RNAs derived from a cohort of L1 expression constructs could retrotranspose using a cultured cell retrotransposition assay 31 ., The 3′UTR of each construct contains a retrotransposition indicator cassette ( mneoI ) ., The mneoI cassette consists of an antisense copy of a neomycin phosphotransferase gene whose coding sequence is interrupted by an intron that resides in the same transcriptional orientation as the L1 31 , 102 ., This arrangement ensures that the expression of a functional neomycin phosphotransferase gene will only be activated upon L1 retrotransposition , thereby conferring cellular resistance to the drug G418 31 , 102 ., Retrotransposition efficiency then can be quantified by counting the resultant numbers of G418-resistant foci 31 , 61 ., Consistent with previous reports ( e . g . , 31 , 41 ) , mRNAs derived from RC-L1s that contain both CMV and 5′UTR ( Fig 4A , pJM101/L1 . 3 , black bar; S3 Table ) , CMV only ( Fig 4A , pJM102/L1 . 3 , black bar; S3 Table ) , or 5′UTR only ( Fig 4A , pJM101/L1 . 3ΔCMV , gray bar; S3 Table ) promoters could efficiently retrotranspose ., By comparison , the pPL97/622/L1 . 3 expression construct produced mRNAs that could undergo efficient retrotransposition when a CMV promoter augmented L1 expression ( Fig 4A , black bar , about 70% the activity of pJM101/L1 . 3; S3 Table ) , but not when L1 expression was driven from the 5′UTR harboring the intra-5′UTR splicing event ( Fig 4A , pPL97/622/L1 . 3ΔCMV gray bar , about 7% the activity of pJM101/L1 . 3; S3 Table ) ., Consistent with this observation , control experiments revealed that an L1 lacking promoter sequences ( Fig 4A , pJM102/L1 . 3ΔCMV; S3 Table ) 50 was unable to retrotranspose ., Additional controls demonstrated that an L1 containing a missense mutation ( pJM105/L1 . 3; D702A ) that disrupts ORF2p RT activity 50 severely reduced L1 retrotransposition efficiency ( Fig 4A; S3 Table ) ., Thus , the data suggest that the SpIRE97/622 intra-5′UTR splicing event severely compromises L1 5′UTR promoter activity as well as subsequent rounds of L1 retrotransposition ., The retrotransposition of an mRNA derived from a 5′UTR/ORF1 splicing event would generate a SpIRE ( e . g . , SpIRE97/976 ) that contains a defective promoter and , if transcribed and translated , would produce amino-terminal truncated versions of ORF1p ., If the truncated version ( s ) of ORF1p were nonfunctional , we reasoned that the 5′UTR/ORF1 splicing event would lead to an L1 mRNA that is compromised for an initial round of retrotransposition in cis ., Indeed , RNAs derived from pPL97/976/L1 . 3 could not retrotranspose despite expression being driven by CMV ( Fig 4B; S4 Table ) ., We next hypothesized that a source of WT ORF1p would be required to act in trans to promote the retrotransposition of L1 mRNAs containing a 5′UTR/ORF1 splicing event ., To test this hypothesis , we co-transfected pPL97/976/L1 . 3 ( whose expression is augmented by a CMV promoter ) with a series of “driver” L1 expression plasmids that lack the mneoI retrotransposition indicator cassette 22 , 50 ., The co-transfection of pPL97/976/L1 . 3 with “driver” plasmids that express WT ORF1p , pJBM561 ( a monocistronic ORF1p expression vector ) , pJM101/L1 . 3NN , or pJM105/L1 . 3NN , resulted in low levels of pPL97/976/L1 . 3 RNA retrotransposition in trans ( Fig 4C; columns 1 , 2 , and 3 , respectively; S5 Table ) ., By comparison , the co-transfection of pPL97/976/L1 . 3 with “driver” plasmids that do not express ORF1p ( pORF2/L1 . 3NN a monocistronic ORF2p expression vector or pCEP4 ) did not support retrotransposition in trans ( Fig 4C; columns 4 and 5 , respectively; S5 Table ) 22 ., Thus , the expression of ORF1p , but not ORF2p , can promote low levels of retrotransposition of mRNAs derived from pPL97/976/L1 . 3 in trans ., RT-PCR experiments using L1/firefly luciferase expression vectors uncovered evidence of SpIRE97/790 splicing events ( Fig 2D ) ., Intriguingly , SpIRE97/790 sequences are only present in the L1PA1 and L1PA2 subfamilies ( S1A Fig; S1 Data; S1 Table ) 84 ., Indeed , the analysis of 1 , 000 genomes data 103 revealed that the L1PA1 SpIRE97/790-3 sequence ( S1 Data; S1 Table ) is polymorphic with respect to presence in the human population ( about 41% homozygous “filled”; 35% heterozygous; 24% homozygous “empty” ) , whereas L1PA2 SpIRE97/790 sequences appear to be fixed with respect to presence in humans ., Additionally , we identified four non-reference L1PA1 SpIRE97/790 sequences in data from the 1000 Genomes Project ( S1 Data; S1 Table ) ., Thus , SpIRE97/790 sequences may represent an evolutionarily younger SpIRE subfamily than the SpIRE97/622 and SpIRE97/976 sequences , which are predominantly found in older L1 subfamilies ( S1 Data; S1 Table ) ., Recently , an elegant study from the Haussler laboratory demonstrated that the Krüppel-associated Box-containing Zinc-Finger Protein 93 ( ZNF93 ) could bind within L1PA3 and L1PA4 5′UTRs to repress their expression 104 ., Intriguingly , a 129-bp deletion that eliminates the ZNF93 binding site within the L1PA2 and L1PA1 5′UTRs allowed them to evade ZNF93-mediated repression 104 ., This 129-bp sequence resides between a putative branch site and the SA sequence used to generate the spliced L1 RNA that gave rise to SpIRE97/790 sequences ( Fig 5A ) ., Thus , we hypothesized this 129-bp deletion may have altered L1 5′UTR splicing dynamics by relocating the SpIRE97/790 SA ( A916G917 in L1PA3 ) to a favorable splicing context in L1PA2 and L1PA1 subfamily members ( Fig 5A ) ., To test the above hypothesis , we generated L1/firefly luciferase expression vectors containing the 5′UTR of a “hot” L1 ( L1RP accession #AF148856 ) 106 or a version of the L1RP 5′UTR that includes the 129-bp L1PA4 sequence containing the ZNF93 binding site 104 upstream of a promoter-less firefly luciferase gene ( pGL4 . 11 ) , creating pJBMWTLUC and pJBMWT129PA4LUC , respectively ( Fig 5B , top panel ) ., We also created a control vector that has a “scrambled” version of the 129-bp L1PA4 sequence ( pJBMWT129SCRLUC ) 104 ( Fig 5B , top panel ) ., Dual luciferase assays using WCLs derived from HeLa-JVM cells co-transfected with pJBMWTLUC , pJBMWT129PA4LUC , pJBMWT129SCRLUC , or pGL4 . 11 and a constitutively expressed Renilla luciferase internal control plasmid ( pRL-TK; Methods ) revealed that pJBMWTLUC and pJBMWT129PA4LUC exhibited an increase ( about 345- or about 320-fold , respectively ) of normalized firefly luciferase activity , when compared to the promoter-less pGL4 . 11 vector ( Fig 5B , bottom panel; S6 Table ) ., By comparison , pJBMWT129SCRLUC exhibited a significant , though less pronounced , increase ( about 88-fold ) of normalized firefly luciferase activity ( Fig 5B , bottom panel; S6 Table ) ., Thus , in general agreement with previous studies 104 , the 129-bp L1PA4 insert does not negatively affect L1RP5′UTR transcriptional activity ., As an additional control , we confirmed that the 129-bp L1PA4 sequence did not significantly affect L1 activity using an EGFP-based retrotransposition assay ( Fig 5C; S7 Table ) 104 ., To test whether the presence or absence of the 129-bp L1PA4 sequence affects intra-L1 5′UTR splicing , we used a slightly modified version of the end-point RT-PCR strategy depicted in Fig 2D ., In agreement with experiments performed with pPLWTLUC ( Fig 2D ) , we detected the predicted full-length L1RP 5′UTR cDNAs as well as SpIRE97/790 spliced cDNAs in cells transfected with pJBMWTLUC ( Fig 5D , yellow “*” and yellow “+ , ” respectively , lane 3 ) ., By comparison , HeLa-JVM cells transfected with pJBMWT129PA4LUC yielded the predicted full-length 5′UTR L1 cDNA ( Fig 5C , yellow “**” lane 4 ) , but did not yield cDNAs corresponding to the SpIRE97/790 splicing event ., Instead , we detected a new spliced cDNA that used the same G98U99 SD and a new SA that resides within the 129-bp L1PA4 sequence ( A851G852 ) , which is not present in the WT L1RP sequence ( Fig 5A and 5D , lane 4 , yellow “@” ) ., Finally , we detected the predicted full-length L1RP 5′UTR cDNAs from cells transfected with pJBMWT129SCRLUC , as well as a biologically irrelevant product that utilized the same G98U99 SD and an SA that resides within the 129-bp L1PA4 scrambled sequence ( Fig 5D , lane 5 , yellow “***” and yellow “$ , ” respectively ) ., Thus , our data demonstrate that the loss of the 129-bp sequence from L1PA3 resulted in a new splicing pattern that led to the emergence of SpIRE97/790 sequences ( Fig 5A and 5D ) ., Finally , we examined whether the new cDNA detected from cells transfected with pJBMWT129PA4LUC corresponds to a SpIRE ., Indeed , a BLAT search of the human genome using an in silico probe that spans the intra-5′UTR splice junction present in this putative SpIRE ( nucleotides 47–97 and 853–903 of pJBMWT129PA4LUC ) yielded nine additional SpIRE97/853 sequences ( S1 Data; S1 Table ) ., These additional SpIREs retain L1 structural hallmarks ( S1 Data; S1 Table ) , indicating that canonical EN-dependent TPRT led to their generation ., The evolutionary success of L1 requires the continued retrotransposit | Introduction, Results, Discussion, Methods | Human Long interspersed element-1 ( L1 ) retrotransposons contain an internal RNA polymerase II promoter within their 5′ untranslated region ( UTR ) and encode two proteins , ( ORF1p and ORF2p ) required for their mobilization ( i . e . , retrotransposition ) ., The evolutionary success of L1 relies on the continuous retrotransposition of full-length L1 mRNAs ., Previous studies identified functional splice donor ( SD ) , splice acceptor ( SA ) , and polyadenylation sequences in L1 mRNA and provided evidence that a small number of spliced L1 mRNAs retrotransposed in the human genome ., Here , we demonstrate that the retrotransposition of intra-5′UTR or 5′UTR/ORF1 spliced L1 mRNAs leads to the generation of spliced integrated retrotransposed elements ( SpIREs ) ., We identified a new intra-5′UTR SpIRE that is ten times more abundant than previously identified SpIREs ., Functional analyses demonstrated that both intra-5′UTR and 5′UTR/ORF1 SpIREs lack Cis-acting transcription factor binding sites and exhibit reduced promoter activity ., The 5′UTR/ORF1 SpIREs also produce nonfunctional ORF1p variants ., Finally , we demonstrate that sequence changes within the L1 5′UTR over evolutionary time , which permitted L1 to evade the repressive effects of a host protein , can lead to the generation of new L1 splicing events , which , upon retrotransposition , generates a new SpIRE subfamily ., We conclude that splicing inhibits L1 retrotransposition , SpIREs generally represent evolutionary “dead-ends” in the L1 retrotransposition process , mutations within the L1 5′UTR alter L1 splicing dynamics , and that retrotransposition of the resultant spliced transcripts can generate interindividual genomic variation . | Long interspersed element-1 ( L1 ) sequences comprise about 17% of the human genome reference sequence ., The average human genome contains about 100 active L1s that mobilize throughout the genome by a “copy and paste” process termed retrotransposition ., Active L1s encode two proteins ( ORF1p and ORF2p ) ., ORF1p and ORF2p preferentially bind to their encoding RNA , forming a ribonucleoprotein particle ( RNP ) ., During retrotransposition , the L1 RNP translocates to the nucleus , where the ORF2p endonuclease makes a single-strand nick in target site DNA that exposes a 3′ hydroxyl group in genomic DNA ., The 3′ hydroxyl group then is used as a primer by the ORF2p reverse transcriptase to copy the L1 RNA into cDNA , leading to the integration of an L1 copy at a new genomic location ., The evolutionary success of L1 requires the faithful retrotransposition of full-length L1 mRNAs; thus , it was surprising to find that a small number of L1 retrotransposition events are derived from spliced L1 mRNAs ., By using genetic , biochemical , and computational approaches , we demonstrate that spliced L1 mRNAs can undergo an initial round of retrotransposition , leading to the generation of spliced integrated retrotransposed elements ( SpIREs ) ., SpIREs represent about 2% of previously annotated full-length primate-specific L1s in the human genome reference sequence ., However , because splicing leads to intra-L1 deletions that remove critical sequences required for L1 expression , SpIREs generally cannot undergo subsequent rounds of retrotransposition and can be considered “dead on arrival” insertions ., Our data further highlight how genetic conflict between L1 and its host has influenced L1 expression , L1 retrotransposition , and L1 splicing dynamics over evolutionary time . | transfection, luciferase, enzymes, messenger rna, enzymology, plasmid construction, sequence motif analysis, molecular biology techniques, dna construction, research and analysis methods, sequence analysis, sequence alignment, bioinformatics, proteins, oxidoreductases, gene expression, rna splicing, molecular biology, biochemistry, rna, rna processing, nucleic acids, database and informatics methods, genetics, biology and life sciences | null |
journal.pgen.1005932 | 2,016 | Mutational History of a Human Cell Lineage from Somatic to Induced Pluripotent Stem Cells | From the moment of fertilisation , as each cell divides random mutations occur which are fixed and inherited by daughter cells ., Most of these variants have little , if any , physiological consequence but contribute to genetic diversity within tissues ., A small proportion will contribute to pathogenic processes such as cancer 1 ., Whole genome sequence analysis of cancer genomes has revealed their mutational landscape 1–4 ., Cancers are clonally heterogeneous , like the somatic tissues from which they originate , and arise through a series of clonal expansions over decades often acquiring aberrant DNA repair processes 3 , 5 , 6 ., Thus , the extent to which mutational signatures in human cancers reflect normal non-pathological mutational patterns that have arisen in their normal non-cancerous somatic ancestors is obscure ., The mutations that have arisen in somatic cells throughout development and tissue homeostasis are generally difficult to identify in tissue biopsies because these are composed of heterogeneous polyclonal populations of cells ., To describe the landscape of mutations in normal somatic tissues , we sought to resolve the underlying heterogeneity of somatic tissues by reprograming the constituent cells into induced pluripotent stem cells ( iPSCs ) 7 , a process of single cell cloning that facilitates subsequent expansion ., Each clonal iPSC line generated from a heterogeneous polyclonal pool will carry a constellation of mutations reflecting both somatic and culture-induced mutations ., Indeed previous work has suggested that a proportion of iPSC mutations originate from the founder somatic cell 8 , 9 ., However although genome sequence analysis of these clones will reveal their mutational burden , it is not possible to definitively resolve the mutations which arose in vivo from those which arose during in vitro culture and reprogramming ( Fig 1A ) ., To confidently classify the origin of the mutations , we derived iPSC lines using monoclonal derived endothelial progenitor cells ( EPCs ) 10 ., The iPSCs isolated from a monoclonal source would share the mutations of the founder cell ( in vivo acquired somatic mutations ) and in addition carry culture-induced mutations as unique private mutations ., Sequencing of these iPSCs would allow interrogation of the number and pattern of somatic mutations present in vivo ( Fig 1A ) ., Fibroblasts and/or monoclonal EPC lines were derived from three individuals: a 65-year old alpha-1 antitrypsin deficiency male ( patient AATD 12 ) , a 22-year old healthy male ( S2 13 ) and a 57-year old healthy male ( S7 13 ) , which were reprogrammed into iPSCs ., The iPSC lines were initially screened using array-based comparative genomic hybridization ( CGH ) to select lines with the smallest number of copy number aberrations ( S1 Table ) ., In addition none of the lines selected had large scale loss of heterozygosity ( LOH ) through error-prone break recombination ( S1 Fig 14 ) ., Next we sequenced the protein-coding exons of these iPSC lines to determine the number and genomic location of their somatic mutations ( Fig 1B–1E and S11–S14 Tables ) ., Fibroblast-derived iPSCs from both individuals carried similar numbers of coding mutations , ranging between 14 and 28 single nucleotide variants ( SNV ) per line ( Fig 1B and 1C ) ., Consistent with a polyclonal origin , these SNVs were unique to each line and no shared SNVs were identified between lines from the same individual ( Fig 1B and 1C ) ., In contrast , monoclonal EPC-derived iPSC lines ( iPSC-2 , 3 , 4 and 5 from AATD and iPSC-RE2 , RE9 , RE14 , RE17 and RE19 from S7 ) carried fewer mutations , of which a subset was shared between them as well as with EPCs from the same individual ., None of the shared SNVs were detected in the corresponding fibroblasts or whole blood , indicating that these SNVs were somatically acquired by the EPCs in vivo ( Fig 1D and 1E ) ., In addition , private SNVs were detected which were unique to each monoclonal-derived iPSC line and these were not found in EPCs or the individual’s reference genome ., Deep sequencing of the donor EPC genome revealed that some of the mutations detected in the iPSCs were in fact present in the EPCs but at very low frequencies ( Fig 1D and 1E , orange boxes; S7 and S8 Tables ) , suggesting that these mutations were acquired by the EPCs during the in vitro expansion process , prior to reprogramming ., Notably no known driver mutations ( using COSMIC database ) , which could confer a selective advantage , were identified in any of the iPSC lines ., These results demonstrate that iPSCs derived from monoclonal somatic cells can be used to identify in vivo acquired somatic mutations ., The mutational burden of iPSCs reflects mutations accumulated in vivo in the ancestral somatic cell lineages and mutations acquired during in vitro cell culture and subsequent reprogramming ., The iPSCs from heterogeneous somatic cells usually do not share any mutations but the exome sequencing data demonstrated that by using monoclonal cell sources it is possible to resolve mutations acquired in vivo from those arising during in vitro cell culture ., Furthermore , identifying shared mutations in somatic cell lineages could be used to construct a cellular phylogenetic tree ., We therefore performed whole genome sequencing on the S7-derived monoclonal EPCs , 3 iPSC lines ( RE2 , RE11 and RE14 ) and fibroblasts , which were used as the reference genome ( S9 Table ) ., The total number of mapped bases obtained per sample was 108 . 1–122 . 8Gb with 33 – 37X sequence coverage ., We identified 463 SNVs in the monoclonal EPCs and 933 , 1119 and 840 in the iPSCs , respectively ( Fig 2A ) ., A proportion of the putative SNVs were validated using PCR amplicon re-sequencing ., This analysis revealed that we were able to detect SNVs with mutant allele frequencies of less than 30% with high specificity ( S10 Table ) , which most likely represent mutations acquired during the first few divisions after founder cells started dividing ( Fig 2B ) ., Amongst the SNVs called , 391 mutations were shared by all the iPSC lines and the monoclonal EPCs at a mutant allele frequency of approximately 50% , which is consistent with clonal mutations ( heterozygous SNVs in diploid chromosomes ) ., Therefore these 391 SNVs reflect the in vivo genetic divergence of the single EPC from fertilisation through development and adulthood ., Some SNVs were shared between the EPCs and only a subset of the lines ( Fig 2A ) , revealing the emergence of genetic differences during in vitro EPC culture ., The remaining SNVs were unique to each iPSC line and not present in the EPCs at a detectable frequency ., These private mutations in RE2 ( 506 SNVs ) , RE17 ( 419 SNVs ) and RE14 ( 719 SNVs ) represent in vitro SNVs acquired in the EPC culture and/or during reprogramming ( S2–S6 Tables ) ., The SNVs detected in the EPCs and iPSCs are a historical record of the phylogenetic lineage of the cells ( Fig 2C ) ., For the individual S7 , in the 57 years from fertilization to the point of derivation of the single EPC , 391 mutations had accumulated in vivo ., The single EPC was then expanded in vitro prior to reprogramming ., Following the first cell division of the EPC , one daughter cell ( A ) acquired at least 29 mutations and the other daughter cell ( B ) at least 9 mutations ., After daughter cell A divides , two further branches appear resulting in at least 7 mutations in one granddaughter cell ( A-1 ) and at least 1 mutation in the other ( A-2 ) ., The progeny of daughter cells A-1 , A-2 and B were the eventual substrates for the derived iPSC lines S7-RE2 , S7-RE17 and S7-RE14 , respectively ., The detailed mutation analysis we performed enabled us to estimate the in vitro mutation rate of the EPCs ., Apart from the 391 in vivo mutations , the clonal SNVs detected in the iPSCs were acquired during the EPC expansion and reprogramming and thus should be present in parental EPCs ., We sought to detect these sub-clonal mutations that are present in EPCs by deep sequencing and calculate a mutation rate during in vitro EPC expansion using a statistical model ( See Materials and Methods ) ., First , in order to ensure accuracy especially at the lower bound of allele frequencies , we investigated sequencing error rates ., Eight genomic regions ( S15 Table and S2 Fig ) were PCR-amplified from the AATD iPSC-B cells and sequenced on a MiSeq instrument ., Median error rates were 0 . 042–0 . 144% and 0 . 053–0 . 320% for the first and second reads respectively when the first and second reads were analysed separately ., However , median error rates were substantially improved ( 0 . 016–0 . 025% ) when consensus sequences were first generated from the first and second reads and then bases were counted ( S2 Fig ) ., We used this approach to accurately identify low-frequency subclonal mutations ., We amplified approximately 40% of the in vitro SNVs from genomic DNA derived from the S7 EPCs and performed deep sequence analysis ., Of this subset , we detected 60 , 51 and 58 SNVs in S7-RE2 , S7-RE14 , and S7-RE17 respectively to be present in the EPCs at allele frequencies between 41% and 0 . 05% ( Table 1 ) ., The sub-clonal SNVs in the EPCs were then used to calculate the mutation rate during in vitro culture , resulting in an estimated mutation rate of 14 . 0 ± 2 . 0 SNVs per cell per generation or 2 . 1 x 10−9 per nucleotide per generation ( see Materials and Methods ) ., Clinical use of iPSCs requires not only generation but also maintenance of iPSCs in cell culture ., We therefore sought to measure the rate of single nucleotide mutagenesis in iPSCs ., In order to calculate this precisely , we sub-cloned iPSCs from individuals S7 and S4 ( a 61-year old healthy female ) as well as H9 human embryonic stem ( ES ) cells 15 and grew these continuously for 60 divisions ., At the end of the expansion period , we sampled the population from each cell line by sequencing single cell sub-clones that had been expanded to provide an adequate DNA sample for whole genome sequencing ., Comparison of the DNA sequence from these sub-clones to its immediate parental population identified in vitro mutations acquired during 60 divisions ., All three lines had a similarly low mutation rate of 0 . 8–1 . 7 SNVs per cell per generation or 1 . 8 x 10−10 per nucleotide per generation ( Fig 3A and 3B ) ., Intriguingly , although both EPCs and pluripotent stem cells have a similar cell cycle time , the mutation rate in pluripotent stem cells was approximately tenfold lower than that in EPCs during in vitro culture ., Next , we sought to understand whether the patterns of the mutations could inform us of the mutagenic processes involved both in vivo and during in vitro cell culture ., We separated the S7 mutations into three groups that represented the continuous cellular lineage for this 57-year old man , from fertilisation to isolation of the single EPC ( in vivo ) , expansion of the EPCs and reprogramming ( in vitro somatic cells ) and finally maintenance of the iPSCs ( in vitro iPSCs ) ( Fig 4A ) ., Using a Bayesian Dirichlet process 16 , 17 we were able to model clusters of clonal and subclonal ( generated after the 1st cell division; <30% MAF ) SNVs for each cell population ., We explored the types of base substitutions seen in these groups of mutations and found variation in the overall mutation spectra ( Fig 4B ) ., There is a preponderance of C:G>T:A transitions in vivo and early in the cellular lineage ., In contrast , in vitro and later in the cellular lineage , there is a preponderance of C:G>A:T transversions ., To explore mutational processes in more detail , we conducted Non Negative Matrix Factorization ( NNMF ) analysis 4 ., Firstly , we found that the clonal mutations in S7-EPCs , representing somatic substitutions acquired in vivo , are associated with a signature that has been attributed to deamination of methylated cytosines , a process thought to occur in all cells ., This signature is similar to the mutations observed in germ cells , another example of in vivo mutations in normal cells ( Fig 4C ) ., Secondly , the mutation signatures acquired by the EPC population in vitro ( clonal S7REs ) were composed of a combination of deamination and C>A transversions ., We speculate that this latterly acquired signature represents damage accrued during culture and may be due to oxidative DNA damage 19 ., Thirdly , we detected a sharp increase in the proportion of mutations associated with C>A transversions in sub-clonal mutations in the iPSCs ( subclonal S7REs ) ., These sub-clonal mutations detected in iPSCs arise in the first few cell cycles after a clonal cell line appears ., Cells during this period are thought to be undergoing reprogramming , suggesting that iPSC reprogramming may stimulate a mutational process associated with C>A transversions ., Finally , the in vitro mutations of iPSCs ( maintenance cell culture ) were associated with both deamination of methylated cytosines and the C>A transversions , reinforcing the suggestion that it is a putative imprint of culture-related/oxidative damage in vitro ., We have extensively analysed a series of normal single-cell derived clones by whole genome and exome sequencing ., We report for the first time the number and characteristics of the acquired mutations in a monoclonal cell isolated from a healthy individual and subsequently derived iPSCs ., From this data we are able to reconstruct the mutational history of a cell beginning from the fertilised egg through to adulthood , then to reprogramming and maintenance of iPSCs in long-term culture , demonstrating how mutagenic processes evolve through that cellular lineage ., During first in vivo then in vitro cell divisions , there is a change in the mutation signatures , suggesting a proportional reduction in the contribution of deamination of methylated cytosines and a proportional increase in oxidative stress and DNA damage ., Finally , consistent with the expectation that an organism should protect its stem cells , we observed a ten-fold reduction in mutation rate in iPSCs , which mirrored that in human ES cells , which have not been subjected to reprogramming ., We find that reprogramming is mutagenic at the nucleotide level and , similar to previous reports 20 , 21 , not at the chromosomal level ., The nucleotide-level mutations are associated with a sharp increase in the proportion of mutations associated with oxidative DNA damage ., However established iPSCs seem to be substantially protected from DNA damage by their pluripotent state ., The increased DNA replication fidelity of iPSCs and ES cells may be due to the activity of homologous recombination throughout the cell cycle in pluripotent cells , whereas in somatic cells it is restricted to the stages of the cell cycle in which there is presence of replicated chromatin 22 , 23 ., Although in vitro culture of iPSCs has a reassuringly low mutation rate , the culture systems used altered the mutational spectrum , which shifted from predominantly C>T transitions to C>A transversions ., Over the relatively few generations we studied , we could not find any evidence of a selection sweep within the culture ., Notably we did not find any driver mutations in our analyses ., Understanding how mutations accrue through iPSC reprogramming and during maintenance cell culture is paramount to developing safe clinical therapies ., Furthermore the mutational signatures underlying normal development and tissue homeostasis provide insights into the biological processes occurring in normal cells ., Primary tissue samples and blood were obtained from a patient with alpha-1 antitrypsin deficiency ( patient 2 ) under the ethics approval REC No . 08/H0311/201 or adult cadaveric organ transplant donors referred to the Eastern Organ Donation Services Team ( part of NHS Blood and Transplant ) ., Ethics approval for the latter was obtained from Cambridgeshire Research Ethics Committee 3 ( REC No . 09/H306/73 ) ., All laboratory procedures were performed according to Standard Operating Protocols and safety assessments ., For each subject included in this study , around 3cm of skin was excised from the midline surgical incision ., The fat and dermal layers of the skin sample were removed and the skin was cut into approximately 1mm3 pieces ., These were dispersed evenly on a 10cm plate ( maximum 20 pieces ) and incubated with fibroblast growth media ( Knockout DMEM with 20% FBS ) ., At 21 days the fibroblasts were harvested using trypsin ., For each derivation , 100mL of blood was taken from the patient into two 50mL Falcon tubes each containing 5mL of 10% sodium citrate ., The sample was mixed by inversion and transporting to the laboratory on ice ., The blood samples were diluted 1:1 with Ca2+ and Mg2+ free PBS and 20mL was layered gently onto 15mL of Ficoll Paque Plus ( GE Healthcare ) and centrifuged at 400g for 35min ., The buffy coat containing the mononuclear cells was transferred into a new Falcon tube , diluted 1:1 with PBS and the cells were pelleted by centrifugation at 300g for 20min ., Cell pellets were re-suspended in 15mL of EPC media: EGM-2MV supplemented with growth factors ( Lonza ) supplemented with 20% FCS ( HyClone ) , and plated onto collagen coated T-75ml flasks ( BD Biosciences ) 10 ., The media was changed every 2 days and colonies started appearing from Day 10 ., After 21 days the EPCs were passaged using trypsin and re-plated into a new T-75 flask ( without collagen ) ., The cells were expanded through sequential passages in 1:3 ratios ., H9 hESCs were obtained from WiCell Research Institute ., Human iPSCs and ES cells were maintained as described previously 11 , 15 ., Briefly , the cells were cultured on irradiated mouse embryonic fibroblast ( MEF ) feeder layers in iPSC medium ( termed KSR + FGF-2 ) : Advanced DMEM/F12 ( Invitrogen ) supplemented with 20% Knockout Serum Replacement ( Invitrogen ) , 2mM L-glutamine ( Invitrogen ) , 0 . 1mM β-mercaptoethanol ( Sigma-Aldrich ) and 4ng/mL of recombinant human basic Fibroblast Growth Factor-2 ( R&D systems ) ., Medium was changed daily and the cells were passaged every 5–10 days depending on the confluence of the plates ., To split iPSCs and ES cells , the plates were washed in PBS and 3mL of each of collagenase and dispase was added ( Collagenase IV 1mg/mL , Invitrogen; Dispase 1mg/mL , Invitrogen ) ., For retroviral reprogramming , four pseudo-typed Moloney murine leukaemia retroviruses containing the coding sequences of each of human POU5F1 , SOX2 , KLF4 and MYC were obtained from Vectalys ., For each iPSC derivation , 1 x 105 primary cells ( fibroblasts or EPCs ) were plated one day before transduction ., The 4 viruses were added at a multiplicity of infection of 10 along with 10 μg/mL of polybrene ( Millipore ) ., The following day residual virus was washed off with PBS and the cells were re-fed with the fresh medium ., On day 5 after infection , the cells were re-plated using trypsin onto a 10cm dish of fresh MEF feeders and 2 days later , the medium was changed from primary cell-specific media to the iPSC medium ( KSR + FGF-2 ) ., The medium was changed every 2 days until colonies emerged after which the medium was changed daily ., For Sendai virus-mediated reprogramming , four viruses containing the coding sequences of human POU5F1 , SOX2 , KLF4 and MYC were obtained from DNAVec ., The protocol for reprogramming was identical to that of retroviruses except that 5 x 105 fibroblasts were used at a multiplicity of infection of three and polybrene was omitted ., The iPSC colonies were identified by their morphology and picked once they had reached sufficient size , typically from day 25 following transduction ., Each colony was first detached from the surrounding feeders by scoring around the circumference ., The colony was then split into quarters or eighths and the segments gently lifted off the plate and transferred to one well of a 12 well plate of fresh MEF feeders containing iPSC media ( KSR + FGF2 ) supplemented with ROCK inhibitor ( Y-27632 , Sigma ) 24 ., The majority of the iPSCs used in this study have been previously characterised in other publications 12 , 13 ., This was performed as described previously 11 ., Genomic DNA was extracted from cell pellets using the DNeasy Blood and Tissue kit ( Qiagen ) ., Short-insert 500bp whole genome libraries were constructed , flowcells prepared and sequencing clusters generated according to the manufacturer’s protocols and sequenced using the Illumina HiSeq2000 platform ( 100bp paired-end ) ., Short-insert paired-end reads were aligned to the reference human genome ( GRCh37/hg19 ) using the Burrows-Wheeler Aligner ( BWA ) 25 , duplicates removed ., The average sequence coverage was 34-fold ., Somatic base substitution mutations were called using CaVEMan ( Cancer Variants Through Expectation Maximization: http://cancerit . github . io/CaVEMan/ ) which provides a probabilistic estimate of a variant being a somatic mutation ., Only variants with likelihoods of 95% and above were included ., Post-hoc filters ( previously trained on 21 WGS cancers 3 ) that sought to remove systematic sequencing artifacts as well as artifacts that arise from mapping errors , were applied to reduce the false positive rate ., SNVs , for which PCR primers could be designed , were all analyzed by amplicon re-sequencing ., PCR primers were designed using BatchPrimer3 to amplify regions spanning SNVs ., PCR was performed with 5ng of genomic DNA ( Fibroblasts , EPCs and iPSCs ) used as a template with Phusion Hot Start DNA Polymerase with GC buffer in the following conditions: 98°C for 1 min , 35 cycles of 98°C for 15 sec , 58°C for 15 sec and 72°C for 30 sec , followed by the final extension , 72°C for 5 min ., PCR products were first pooled by sample and then purified with QIAquick PCR Purification Kit ( Qiagen ) ., Purified PCR products from A1ATD patient B-derived EPCs were converted to a 454 library by emulsion-PCR and sequenced using the 454 Titanium platform according to the manufacturer’s instruction ., Purified PCR products from the other samples were converted to an Illumina library by adaptor ligation and sequenced on either the MiSeq ( 150bp , paired end ) or the HiSeq2000 ( 100bp , paired end ) platforms ., Reads from the 454 platform were aligned to a reference constructed from PCR-amplified regions ., Paired end reads from the MiSeq or HiSeq2000 were first used to generate consensus sequences between each pair and then these were aligned to a reference using BWA SW 25 ., The number of reads reporting each of the four bases was counted using Samtool ., PCR primers were designed in a way that each SNV was located in a region where both Illumina reads could reach ., PCR and Illumina sequencing were performed as described above ., Fastq files ( 1 . fq and 2 . fq ) were first merged to generate consensus sequence reads ., In this process , base calls were accepted only when a sum of Q scores from both reads was higher than 40 and both reads reported the same base ., Reads were discarded if an overlapping region exhibited more than 10% mismatches between the two reads ., Consensus reads were subsequently mapped onto the reference sequence using BWA SW and the number of reads reporting each of the four bases was counted using Samtool ., Two-way contingency Chi-square tests were performed between the reads reporting reference and mutant variants and between fibroblasts and EPCs ., Multiple test correction was performed using the Bonferroni correction ., SNVs whose mutant read was significantly higher in EPCs were counted as subclonal mutations ., Analyses on the subclonal SNVs with less than 0 . 1% were shown in S16 Table ., It is not possible to subclone and serially expand EPCs therefore a statistical model was used to estimate the SNV mutation rate in EPCs ., We obtained 13 . 5 x 106 cells at the end of S7-EPC expansion , which represents that a single EPC underwent approximately 24 cell divisions ., When 5ng ( approximately 750 cells or 1 , 500 molecules ) were used as a template for each PCR , assuming that the sampling of DNA molecule follows the Poisson distribution , probability of sampling k number of DNA molecules carrying each SNV introduced at generation n is therefore given by, Pn ( X=k ) =λnkexp ( −λn ) k !, ,, where λn ( = 1500/2n+1 ) represents the mean molecule number of each mutation introduced at generation n in the 5ng DNA ., The total number of mutations that can be detected with amplicon re-sequencing is, ∑n=024Pn ( X>0 ) Mave=9 . 88Mave ,, where Mave is the average mutation rate , assuming that the mutation rate is similar throughout EPC culture ., Taking into account the numbers of sub-clonal EPC mutations detected ( SNVs detected in EPCs by deep sequencing; Table 1 ) and the 40% sampling for deep sequence analysis , we estimated mutation rate of 14 . 0 ± 2 . 0 SNVs per cell per generation or 2 . 1 x 10−9 per nucleotide per generation ., All work performed as part of this project was approved by an ethics committee under the REC Nos . 09/H306/73 and 08/H0311/201 ., The aCGH data has been deposited with the ArrayExpress under the accession number , E-MTAB-1319 ., Whole genome sequence data have been deposited with the European Genome-phenome Archive under the accession number EGAS00001000231 and exome data under the accession number EGAS00001000492 . | Introduction, Results and Discussion, Materials and Methods | The accuracy of replicating the genetic code is fundamental ., DNA repair mechanisms protect the fidelity of the genome ensuring a low error rate between generations ., This sustains the similarity of individuals whilst providing a repertoire of variants for evolution ., The mutation rate in the human genome has recently been measured to be 50–70 de novo single nucleotide variants ( SNVs ) between generations ., During development mutations accumulate in somatic cells so that an organism is a mosaic ., However , variation within a tissue and between tissues has not been analysed ., By reprogramming somatic cells into induced pluripotent stem cells ( iPSCs ) , their genomes and the associated mutational history are captured ., By sequencing the genomes of polyclonal and monoclonal somatic cells and derived iPSCs we have determined the mutation rates and show how the patterns change from a somatic lineage in vivo through to iPSCs ., Somatic cells have a mutation rate of 14 SNVs per cell per generation while iPSCs exhibited a ten-fold lower rate ., Analyses of mutational signatures suggested that deamination of methylated cytosine may be the major mutagenic source in vivo , whilst oxidative DNA damage becomes dominant in vitro ., Our results provide insights for better understanding of mutational processes and lineage relationships between human somatic cells ., Furthermore it provides a foundation for interpretation of elevated mutation rates and patterns in cancer . | The mutation load of human tissues is unknown and represents the genetic divergence from the fertilised egg ., Reprogramming of somatic cells generates induced pluripotent stem cells ( iPSCs ) , a cell type being considered for clinical applications ., We generated iPSCs from tissues of healthy individuals and used whole genome sequencing to identify in vivo mutations accrued in a somatic cell during the lifetime of the individual ., Next we identified in vitro mutations introduced during reprogramming and cell culture ., Each has a unique mutation signature suggesting different mutagenic processes ., Our study demonstrates the use of reprogramming as a tool to elucidate mutational processes within normal cells and highlights the importance of genetic characterisation of iPSCs prior to clinical translation . | sequencing techniques, medicine and health sciences, cell cycle and cell division, cell processes, fibroblasts, genome sequencing, mutation, stem cells, induced pluripotent stem cells, connective tissue cells, molecular biology techniques, research and analysis methods, artificial gene amplification and extension, animal cells, connective tissue, biological tissue, molecular biology, somatic mutation, cell biology, anatomy, gene identification and analysis, genetics, mutation detection, biology and life sciences, cellular types, polymerase chain reaction | null |
journal.pbio.2006738 | 2,018 | Evolutionary emergence of infectious diseases in heterogeneous host populations | Understanding the factors that govern the ability of pathogens to invade a new host population is of paramount importance to design better surveillance systems and control policies ., Mathematical epidemiology can provide key insights into these dynamics 1–4 ., For instance , simple deterministic models identified critical vaccination thresholds , above which pathogens are driven extinct , which informed policy guidelines for vaccination campaigns 1 ., However , chance events and rapid pathogen evolution can also play a critical role in determining the outcome of disease dynamics 2 , 4–6 ., For example , recent experimental studies indicated that the dramatic size of the 2013–2016 Ebola epidemic can at least be partially explained by the acquisition of genetic mutations that increased transmissibility to humans 7 , 8 ., Stochastic models of epidemiology can help to understand the emergence of evolving pathogen populations 5 , 9–13 ., These models , however , often make the unrealistic assumption that the pathogen is spreading in a well-mixed and homogeneous host population , in which all hosts are equally susceptible ., Although a handful of theoretical studies have shown that host heterogeneity could have an important impact on pathogen emergence , these models either relied on phenomenological or numerical approaches 12 , 13 , or assumed that the hosts only differ in their number of contacts but not their susceptibility to pathogens 11 ., Here , we extend this line of inquiry by, ( i ) building a mechanistic model of pathogen emergence in a diverse host population , in which only some hosts are resistant to the pathogen ,, ( ii ) deriving analytical expressions for the probability of evolutionary emergence of the pathogen , and, ( iii ) providing the first experimental test of theoretical predictions on pathogen evolutionary emergence using a bacteria–phage interaction ., We demonstrate that realistic increases in the diversity of host resistance alleles strongly reduce the probability of evolutionary emergence of novel pathogens , hence suggesting new strategies to manage the emergence of diseases ., Crucially , using bacteria with distinct Clustered Regularly Interspaced Short Palindromic Repeat ( CRISPR ) immunity and their lytic viruses ( bacteriophages ) 14–17 , we experimentally explore the effect of host population heterogeneity on the emergence and evolution of pathogens ., The experimental validation of our theoretical predictions with this microbial system confirms the ability of our mathematical model to capture the complexity of the interplay between the epidemiology and evolution of emerging pathogens in this model system ., In order to predict how the composition of host populations impacts the probability of pathogen emergence , we developed a branching process model 5 , 9–13 ., We aimed to capture host–pathogen interactions in which different groups of individuals within a host population each carry unique resistance alleles that recognize different pathogen epitopes , and pathogens can evade recognition by acquiring “escape” mutations in the corresponding epitopes ., In this model , we assume that the host population contains a fraction ( 1 − fR ) of individuals that are fully susceptible to the pathogen , while the remaining fraction fR of the population is resistant and composed of a mixture of n host types in equal frequencies , each of which has a different resistance allele ., The efficacy of resistance is assumed to be perfect ( we relax this assumption in section S1 . 2 of S1 Text ) ., Therefore , a pathogen with i escape mutations ( i between 0 and n ) can infect a fraction ( 1 − fR ) + fRi/n of the total host population ., We further assume that a host infected with a pathogen that does not carry escape mutations transmits at rate b and dies at rate d ., Host resistance prevents infection without affecting b or d ., Whereas escape mutations allow the pathogen to infect a larger fraction of the host population , they also carry a fitness cost , c which causes pathogens with i escape mutations to reproduce at rate bi = b ( 1 − c ) i ., The probability of acquiring an escape mutation is a function of n , the number of resistance alleles in the population , as well as i , the number of escape mutations already encoded by the pathogen ., The probability that a pathogen with i escape mutations will acquire an additional one equals ui , n = 1 − ( 1 − μ ) n−i , where μ is the pathogen mutation rate per target site ( a target site is a region of the pathogen genome where a point mutation or a deletion may allow escape from recognition by host immunity ) ., This simplifies to ui , n ≈ μ ( n − i ) when the pathogen mutation rate is assumed to be small ( note how the rate of escape mutations increases with ( n − i ) ) ., For the sake of simplicity , we assume that escape mutations cannot revert to the ancestral types ., These reversions are expected to have a negligible effect on the probability of evolutionary emergence when the target site mutation rate remains small 11 ., To account for the effect of spatial structure , we assume that when a pathogen is released from an infected host , it will land with probability ϕ on the same type of host ( i . e . , a host susceptible to this pathogen ) and with probability ( 1 − ϕ ) on a random host from the population , which may or not be of the same type ., The expected number of secondary infections caused by a pathogen with i escape mutations in an uninfected host population is given by its basic reproduction ratio:, Ri , n=bidFi , n, ( 1 ), where Fi , n = ( ϕ + ( 1 − ϕ ) ( fR i/n + ( 1 − fR ) ) ) is the effective fraction of hosts that can be infected by the focal pathogen ., A pathogen with n escape mutations has a basic reproduction ratio equal to R0 ( 1 − c ) n , where R0 = b/d refers to the basic reproduction ratio of the pathogen with 0 escape mutations in a fully susceptible host population ., Note , however , that a pathogen with 0 escape mutations introduced in a diverse host population has a basic reproduction ratio equal to R0 , n ≤ R0 ., The key question we wish to address with this model is how the composition and structure of the host population determines the ultimate fate of a pathogen ( i . e . , extinction versus emergence , see S1 and S2 Figs ) ., We detailed in the Materials and methods section the calculation of the probability of emergence , Pi , n , which is the probability that an inoculum of V0 pathogens with i escape mutations will not go extinct when introduced in a host population with n different resistance alleles ., To understand the role of pathogen evolution in this process , we also derive the probability of evolutionary emergence , which quantifies the importance of escape mutations to pathogen emergence ., Next , we wanted to experimentally explore the validity of the above predictions ., While this is challenging given the paucity of suitable empirical systems that are amenable to experimental manipulations in a timely fashion , we explored whether this could be achieved by studying the evolutionary emergence of “escape” phages against bacteria with a CRISPR–CRISPR-associated ( Cas ) system ., This immune defense provides full protection against a phage infection by adding phage-derived sequences ( known as “spacers” ) in a CRISPR locus carried by the bacterial host chromosome 14 ., This empirical system allowed us to overcome three important technical challenges ( see details of the experimental protocols in the Materials and methods section ) ., First , the stochastic nature of extinction requires a large number of replicate populations to measure a probability of emergence , which is possible using bacteria and phages in 96-well plates ., Second , by mixing bacteria with different and unique CRISPR resistance alleles , we could manipulate the fraction of resistant hosts and the diversity in resistance alleles without affecting other traits of the host 18 ., Third , unlike most other empirical systems , the mechanism of phage adaptation to CRISPR-based immunity is well known: lytic phages “escape” CRISPR resistance through mutation of their target sequence ( the “protospacer” ) 13 , 15 , 18 , 19 , 20 ., In order to validate the model using this empirical system , we used eight CRISPR-resistant clones ( also referred as bacteriophage-insensitive mutants BIMs ) of the gram-negative Pseudomonas aeruginosa strain UCBPP-PA14 , each of which carried a single and distinct spacer targeting the lytic phage DMS3vir ., Each of these spacers provides full resistance to infection ., For each of these eight CRISPR-resistant clones , the rate at which the phage acquires escape mutations was found to be approximately equal to 2 . 8*10−7 mutations/locus/replication , as determined using Luria-Delbrück experiments ( see section S2 . 1 . 6 in S1 Text and S9 Fig ) ., Using one of these BIMs , we first tested the theoretical prediction that the probability of emergence increases with the size of the virus inoculum ( V0 ) ., To this end , 96 replicate populations , each composed of an equal mix of sensitive bacteria and a CRISPR-resistant clone , were exposed with five different inoculum sizes of the phage ( corresponding to a mean V0 of approximately 0 . 3 , 3 , 30 , 300 , and 3 , 000 phages ) ., After 24 hours , we measured the fraction of phage-infected bacterial populations in which emergence had occurred ., Consistent with the model predictions , we observed that the larger the phage inoculum size , the higher the probability of pathogen emergence ( Fig 3 , dashed line ) ., In addition , we measured the fraction of viral populations in which the phages had evolved to escape CRISPR resistance ., Again , in accordance with the theory , we found that larger phage inocula were associated with an increased evolution of phage escape mutations ( Fig 3 , full line , Kendall , z = 3 . 416 , tau = 0 . 784 , p < 0 . 001 ) ., Furthermore , we obtained very similar results using a different empirical system consisting of the lytic phage 2972 and its gram-positive bacterial host Streptococcus thermophilus DGCC7710 ., In this experiment , 96 populations composed of sensitive bacteria and a CRISPR-resistant clone were infected with three different inoculum sizes of the phage ., As above , we found that a larger phage inoculum led to both a higher probability of emergence and a higher probability of evolutionary emergence ( S10 Fig ) ., Next , we tested the theoretical prediction that the probability of pathogen evolutionary emergence is highest in populations with an intermediate fraction of resistant hosts ( Fig 4 ) ., For each of the eight BIMs , we generated populations composed of sensitive bacteria and a variable proportion of CRISPR-resistant bacteria , ranging from 0% to 100% in 10% increments ., These populations were subsequently infected with V0 = 300 phages , and the fractions of emergence and evolutionary emergence were measured ., As expected , pathogen/phage emergence dropped when the proportion of host/bacteria resistance reached a threshold level ( S11 Fig ) ., Interestingly , examination of phage evolution among emerging phage populations also confirmed that the probability of observing escape mutations is maximized for intermediate proportions of host resistance ( Fig 4 ) ., Again , we obtained very consistent results with phage 2972 and S . thermophilus ( S10 Fig ) ., We noticed substantial variation among CRISPR-resistant hosts in the observed frequencies of escape phage evolution ( Fig 4 ) ., Variations in phage mutation rates are unlikely to explain this variability because , as pointed out above , we failed to detect significant variations in the rate of escape mutations to the different CRISPR-resistant hosts ( see S9 Fig ) ., Variations in the fitness cost associated with these mutations could , however , explain the observed variations in the final frequency of escape mutations ( see S5 Fig ) ., Finally , we experimentally explored the effect of resistance allele diversity on evolutionary emergence for a fixed proportion of host resistance ( fR = 0 . 5 ) ., To this end , we generated bacterial populations that were composed of sensitive bacteria and an equal mix of one , two , four , or eight CRISPR-resistant clones ., In this case , as expected , an inoculum size of 300 phages always led to pathogen emergence , but increasing host diversity had a strong negative effect on the ability of the phage to evolve to escape host resistance ( Fig 5 ) ., We also found higher probabilities of observing multiple escape mutations in the low diversity treatment ( Kendall , z = −4 . 8771 , Tau = −0 . 3259 , p = 1 . 07*10−6 ) , which further supports the prediction that host diversity hampers the evolution of the phage population ., The emergence and re-emergence of pathogens has far-reaching negative impacts on wildlife , agriculture , and public health ., Unfortunately , pathogen emergence events are notoriously difficult to predict and we need good biological models to experimentally explore the interplay between epidemiology and evolution taking place at the early stages of an epidemic ., Here , we used a combination of diverse theoretical and experimental analyses to examine how the composition of a host population impacts the probability of pathogen emergence and evolution ., Our theory is tailored to the biology of CRISPR–phage interactions , and subsequent validation using this experimental system demonstrates the predictive power of this theoretical framework ., However , we suggest that this framework may be suitable for predicting pathogen emergence whenever hosts recognize specific pathogen epitopes and resistance can be overcome by epitope mutations ., For instance , the specificity of the host–parasite interaction driven by CRISPR immunity ( S12 Fig ) is akin to the classical gene-for-gene system described in plant pathosystems 21 ., However , host immunity may not always be perfect , which will impact both the dynamics and the evolution of the pathogen population 22–24 ., To further generalize our findings , we derived the probability of pathogen emergence when immunity is imperfect ( see section S1 . 2 in S1 Text ) ., Note , however , that this should be considered separately from the more complex epidemiological dynamics that occur when the probability of a successful infection depends on the pathogen dose or when the pathogen causes immunosuppression , both of which can cause emergence to become dependent on the pathogen population density 25 , 26 ., Our framework provides several insights on emergence and re-emergence in both the presence and absence of pathogen evolution ., For instance , this model captures how the composition and diversity of the host population impacts the emergence of a nonevolving pathogen ., In this context , a larger proportion of resistant hosts decreases pathogen emergence , but this effect is weaker in spatially structured populations in which transmission is more likely to occur between the same host types , which allows for pathogen persistence in sensitive subpopulations ., This effect is akin to the effect of the spatial distribution of suitable habitats on extinction thresholds 27–30 and consistent with earlier work that shows that host composition and spatial structure impact the growth rate of bacteriophage ϕ6 31 ., In the context of an evolving pathogen , our theory helps to explain the general observation that evolutionary emergence and the spread of escape mutations is maximal for an intermediate proportion of resistant hosts in the population 32 ., Specifically , this is because increasing host resistance in the population has two opposite effects:, ( i ) the influx of new mutations decreases because the ancestral pathogen cannot replicate on resistant hosts , and, ( ii ) selection for escape mutations increases ., Second , our model predicts that diversity in host resistance alleles decreases the probability of evolutionary emergence ., Even though larger host diversity increases the number of adaptive mutations for the pathogen ( i . e . , a larger number of targets of selection ) , each mutation is associated with a smaller fitness advantage ( i . e . , a smaller increase in the fraction of the host population that can be infected ) ., The theory presented here therefore helps to explain previous empirical data on the impact of host CRISPR diversity on the evolution of escape phages 18 ., The link between host biodiversity and infectious diseases has attracted substantial attention recently 33–43 ., Several studies support the “dilution effect” hypothesis , which postulates that host diversity limits disease spread 39 , 40 , 43 ., For example , host diversity may limit the spread of a pathogen by increasing the fraction of bad-quality hosts in the population 43 ., Indeed , increasing the fraction of resistant hosts ( but not the diversity of resistance alleles ) decreases the basic reproduction ratio of the wild-type pathogen 44 , 45 ., In addition , host diversity per se may also limit disease spread , and several studies have shown the negative effect of host diversity on the deterministic growth rate of the pathogen under specific patterns of host–parasite specificity 35 , 46 , 47 ., Notwithstanding these important insights , what sets our theoretical model apart is its ability to understand the factors that impact the initial pathogen emergence , rather than the downstream spread of a pathogen once it has already emerged ., Studying this requires stochastic models , which are critical to model the probability of rare events , for example , pathogen spillover across species , including at the human-animal interface 48 , 49 , 4 , 50 , the emergence of drug resistance 51 , 52 , the evolution of vaccine resistance 53 , and the reversion of live vaccines 54–58 ., In all these public health issues , understanding pathogen emergence requires models accounting for the stochastic nature of epidemiological and evolutionary dynamics ., The present study focuses on the effect of the diversity of host resistance when each resistant host carries a single resistance allele ( i . e . , a single spacer in CRISPR ) ., Our joint theoretical and experimental approach could be readily extended to evaluate the impact of the accumulation of multiple resistance alleles in a single host genotype rather than mixing multiple genotypes with a single resistance allele in the host population ., The impact of such alternative strategies on the durability of resistance and on disease spread is particularly relevant in agriculture 59 , 60 ., Our work provides a theoretical framework to study these different issues , and our experimental model system can be used to evaluate the ability of different control strategies to limit pathogen adaptation and emergence ., We detail the derivation of the probability of pathogen emergence presented in the main text ( the main parameters of the model are listed in S1 Table ) ., We are interested in the ultimate fate ( extinction or not ) of a single pathogen with i escape mutations dropped into a very large host population with a proportion fR of resistant hosts ., This resistant population is composed of an equal frequency of n different resistance genotypes ., This free infectious particle first has to infect a host to avoid extinction , and the probability of ultimate extinction of this pathogen is, Qi , n= ( 1-fR ) qi , n+fR ( inqi , n+n-in ), ( 5 ), where qi , n is the probability of ultimate extinction of the pathogen when it is currently infecting a host ., Next , we focus on the probability qi , n ( t ) at time t that a pathogen with i mutations in an infected host will ultimately go extinct ., In a small interval of time , dt , four different events may take place ., First , the pathogen may transmit to a new host without additional escape mutations ., Second , after a mutation event , the pathogen may transmit a pathogen with i + 1 escape mutations to a new host ., Third , the infected host ( and the pathogen in the host ) may die ., Fourth , nothing may happen during this interval of time dt ., Collecting these different terms allows us to write down recursions for the probability qi , n ( t ) , at time t , as a function of the probability qi , n ( t + dt ) and qi+1 , n ( t + dt ) , at time t + dt:, qi , n ( t ) =Ai , ndtqi , n ( t+dt ) qi , n ( t+dt ) ︸reproductionwithoutmutation+Bi , ndtqi , n ( t+dt ) qi+1 , n ( t+dt ) ︸reproductionwithmutation+ddt︸death+qi , n ( t+dt ) ( 1−Ai , ndt−Bi , ndt−ddt ) ︸noevent, ( 6 ), with: Ai , n = bi ( 1 − ui , n ) Fi , n Bi , n = biui , nFi+1 , n Fi , n = ( ϕ + ( 1 − ϕ ) ( fR i/n + ( 1 − fR ) ) ) ., The above calculation is based on the assumption that the pathogen never reaches a high prevalence and that the composition of the host population remains constant ( i . e . , Fi , n is assumed to remain constant ) ., In other words , the probabilities qi , n ( t ) are assumed to be invariant with time ., We can thus set qi , n ( t ) = qi , n ( t + dt ) to obtain a recursion equation that allows us to derive qi , n from qi+1 , n ., The first term of this recursion gives the probability of extinction , qn , n that a pathogen with n escape mutations ( a pathogen fully adapted to the novel host population ) will go extinct ., The heterogeneity of the environment has no impact on a fully adapted pathogen , and its probability of extinction is simply the extinction probability of the birth–death process:, qn , n=1/ ( R0 ( 1-c ) n ), ( 7 ), Next , to derive qn−1 , n from qn , n , we need the recursion equation for qi , n ., However , we have to distinguish two different scenarios ., First , if Ai , n = 0 , for example , the case of a cell infected by a fully maladapted pathogen ( i . e . , i = 0 ) in a well-mixed population with no susceptible hosts ( i . e . , ϕ = 0 , fR = 1 ) , we find:, qi , n=dd+Bi , n ( 1-qi+1 , n ), ( 8 ), Second , in the more general scenario , in which Ai , n > 0 , we have, qi , n=Ci , n--4dAi , n+Ci , n22Ai , n, ( 9 ), with: Ci , n = Ai , n + Bi , n ( 1 − qi+1 , n ) + d ., Knowing qn , n and the above recursion equations , we can derive qn−1 , n and next qn−2 , n… until we get q0 , n ., We are particularly interested in q0 , n and Q0 , n because these quantities measure the probability of extinction of a pathogen with no escape mutations ( in an infected host or as an infectious particle , respectively ) ., Ultimately , we obtain the probability of emergence of an inoculum of V0 propagules of pathogen with no escape mutations ( when n = 1 , this yields Eq 2 in the main text ) :, P0 , n=1- ( Q0 , n ) V0, ( 10 ), We show in Fig 2 how the diversity of host resistance affects the probability of pathogen emergence through a reduction of evolutionary emergence ., In S4 Fig , we illustrate the interaction between host diversity and spatial structure in pathogen emergence ., We show that more spatial structure decreases the impact of host diversity on evolutionary emergence and increases the overall probability of pathogen emergence ., To study the impact of the host population composition on the probability of evolutionary emergence , we used two different microbial systems:, ( i ) the gram-negative P . aeruginosa and its lytic phage DMS3vir , and, ( ii ) the gram-positive S . thermophilus and its lytic phage 2972 ., All the resistant bacteria ( i . e . , BIMs ) derived from the phage-sensitive wild-type strains P . aeruginosa UCBPP PA14 and S . thermophilus DGCC7710 rely on CRISPR-Cas immunity for complete resistance against the corresponding phage 14 , 61 ., For all treatments , we performed 96 replicate infections of the corresponding host populations ., We manipulated the composition of the host populations by mixing overnight cultures of sensitive bacteria and BIMs in the proportions indicated in the text , figures , and figure legends ., Each replicate population was inoculated 1:100 into fresh growth media and infected with a quantity V0 of phages ( the inoculum size ) , as indicated in the text , figures , and figure legends ., After 23 hours , we monitored within each population, ( i ) the occurrence of phage epidemics ( i . e . , an emergence ) and, ( ii ) the presence of escape mutants ( i . e . , an evolutionary emergence ) ., A detailed description of these experiments is provided in section S2 of S1 Text . | Introduction, Results, Discussion, Materials and methods | The emergence and re-emergence of pathogens remains a major public health concern ., Unfortunately , when and where pathogens will ( re- ) emerge is notoriously difficult to predict , as the erratic nature of those events is reinforced by the stochastic nature of pathogen evolution during the early phase of an epidemic ., For instance , mutations allowing pathogens to escape host resistance may boost pathogen spread and promote emergence ., Yet , the ecological factors that govern such evolutionary emergence remain elusive because of the lack of ecological realism of current theoretical frameworks and the difficulty of experimentally testing their predictions ., Here , we develop a theoretical model to explore the effects of the heterogeneity of the host population on the probability of pathogen emergence , with or without pathogen evolution ., We show that evolutionary emergence and the spread of escape mutations in the pathogen population is more likely to occur when the host population contains an intermediate proportion of resistant hosts ., We also show that the probability of pathogen emergence rapidly declines with the diversity of resistance in the host population ., Experimental tests using lytic bacteriophages infecting their bacterial hosts containing Clustered Regularly Interspaced Short Palindromic Repeat and CRISPR-associated ( CRISPR-Cas ) immune defenses confirm these theoretical predictions ., These results suggest effective strategies for cross-species spillover and for the management of emerging infectious diseases . | The probability that an epidemic will break out is highly dependent on the ability of the pathogen to acquire new adaptive mutations and to induce evolutionary emergence ., Forecasting pathogen emergence thus requires a good understanding of the interplay between the epidemiology and evolution taking place at the onset of an outbreak ., Here , we provide a comprehensive theoretical framework to analyze the impact of host population heterogeneity on the probability of pathogen evolutionary emergence ., We use this model to predict the impact of the fraction of susceptible hosts , the inoculum size of the pathogen , and the diversity of host resistance on pathogen emergence ., Our experiments using lytic bacteriophages and CRISPR-resistant bacteria support our theoretical predictions and demonstrate that manipulating the diversity of resistance alleles in a host population may be an effective way to limit the emergence of new pathogens . | genome engineering, medicine and health sciences, ecology and environmental sciences, pathology and laboratory medicine, engineering and technology, bacteriophages, pathogens, synthetic biology, synthetic bioengineering, crispr, viruses, mutation, evolutionary emergence, synthetic genomics, bioengineering, synthetic genome editing, ecological metrics, pathogenesis, evolutionary immunology, species diversity, point mutation, ecology, host-pathogen interactions, genetics, biology and life sciences, evolutionary biology, evolutionary processes, organisms | null |
journal.pgen.1001244 | 2,010 | GC-Rich Sequence Elements Recruit PRC2 in Mammalian ES Cells | Polycomb proteins are epigenetic regulators required for proper gene expression patterning in metazoans ., The proteins reside in two main complexes , termed Polycomb repressive complex 1 and 2 ( PRC1 and PRC2 ) ., PRC2 catalyzes histone H3 lysine 27 tri-methylation ( K27me3 ) , while PRC1 catalyzes histone H2A ubiquitination and mediates chromatin compaction 1 , 2 ., PRC1 and PRC2 are initially recruited to target loci in the early embryo where they subsequently mediate lineage-specific gene repression ., In embryonic stem ( ES ) cells , the complexes localize to thousands of genomic sites , including many developmental loci 3–5 ., These target loci are not yet stably repressed , but instead maintain a “bivalent” chromatin state , with their chromatin enriched for the activating histone mark , H3 lysine 4 tri-methylation ( K4me3 ) , together with the repressive K27me3 6 , 7 ., In the absence of transcriptional induction , PRC1 and PRC2 remain at target loci and mediate repression through differentiation ., The mechanisms that underlie stable association of the complexes remain poorly understood , but likely involve interactions with the modified histones 8–12 ., Proper localization of PRC1 and PRC2 in the pluripotent genome is central to the complex developmental regulation orchestrated by these factors ., However , the sequence determinants that underlie this initial landscape remain obscure ., Polycomb recruitment is best understood in Drosophila , where sequence elements termed Polycomb response elements ( PREs ) are able to direct these repressors to exogenous locations 13 ., PREs contain clusters of motifs recognized by DNA binding proteins such as Pho , Zeste and GAGA , which in turn recruit PRC2 14–17 ., Despite extensive study , neither PRE sequence motifs nor binding profiles of PRC2-associated DNA binding proteins are sufficient to fully predict PRC2 localization in the Drosophila genome 1 , 16 , 18 , 19 ., While protein homologs of PRC1 and PRC2 are conserved in mammals , DNA sequence homologs of Drosophila PREs appear to be lacking in mammalian genomes 13 ., Moreover , it remains controversial whether the DNA binding proteins associated with PRC2 in Drosophila have functional homologs in mammals ., The most compelling candidate has been YY1 , a Pho homolog that rescues gene silencing when introduced into Pho-deficient Drosophila embryos 20 ., YY1 has been implicated in PRC2-dependent silencing of tumor suppressor genes in human cancer cells 21 ., However , this transcription factor has also been linked to numerous other functions , including imprinting , DNA methylation , B-cell development and ribosomal protein gene transcription 22–26 ., Recently , researchers identified two DNA sequence elements able to confer Polycomb repression in mammalian cells ., Sing and colleagues identified a murine PRE-like element that regulates the MafB gene during neural development 27 ., These investigators defined a critical 1 . 5 kb sequence element that is able to recruit PRC1 , but not PRC2 in a transgenic cell assay ., Woo and colleagues identified a 1 . 8 kb region of the human HoxD cluster that recruits both PRC1 and PRC2 and represses a reporter construct in mesenchymal tissues 28 ., Both groups note that their respective PRE regions contain YY1 motifs ., Mutation of the YY1 sites in the HoxD PRE resulted in loss of PRC1 binding and partial loss of repression , while comparatively , deletion of a separate highly conserved region from this element completely abrogated PRC1 and PRC2 binding as well as repression 28 ., In addition to these locus-specific investigations , genomic studies have sought to define PRC2 targets and determinants in a systematic fashion ., The Ezh2 and Suz12 subunits have been mapped in mouse and human ES cells by chromatin immunoprecipitation and microarrays ( ChIP-chip ) or high-throughput sequencing ( ChIP-Seq ) 3–5 , 29 ., Such studies have highlighted global correlations between PRC2 targets and CpG islands 5 , 30 as well as highly-conserved genomic loci 4 , 7 , 31 ., Recently , Jarid2 has been shown to associate with PRC2 and to be required for proper genome-wide localization of the complex 32–35 ., Intriguingly , Jarid2 contains an ARID and a Zinc-finger DNA-binding domain ., However , it is unclear how Jarid2 could account for PRC2 targeting given the lack of sequence specificity and the low affinity of its DNA binding domains 33 , 36 ., In summary , a variety of sequence elements including CpG islands , conserved elements and YY1 motifs have been implicated in Polycomb targeting in mammalian cells ., Causality has only been demonstrated in two specific instances and a unifying view of the determinants of Polycomb recruitment remains elusive ., Here we present the identification of multiple sequence elements capable of recruiting PRC2 in mammalian ES cells ., This was achieved through an experimental approach in which engineered bacterial artificial chromosomes ( BACs ) were stably integrated into the ES cell genome ., Evaluation of a series of modified BACs specifically identified a 1 . 7 kb DNA fragment that is both necessary and sufficient for PRC2 recruitment ., The fragment does not share sequence characteristics of Drosophila PREs and lacks YY1 binding sites , but rather corresponds to an annotated CpG island ., Based on this result and a genome-wide analysis of PRC2 target sequences we hypothesized that large GC-rich sequence elements lacking transcriptional activation signals represent general PRC2 recruitment elements ., We tested this model by assaying the following DNA sequences:, ( i ) a ‘housekeeping’ CpG island which was re-engineered by removal of a cluster of activating motifs; and, ( ii ) two large GC-rich intervals from the E . coli genome that satisfy the criteria of mammalian CpG islands ., We found that all three GC-rich elements robustly recruit PRC2 in ES cells ., We propose that a class of CpG islands distinguished by a lack of activating motifs play causal roles in the initial localization of PRC2 and the subsequent coordination of epigenetic controls during mammalian development ., To identify DNA sequences capable of recruiting Polycomb repressors in mammalian cells , we engineered human BACs that correspond to genomic regions bound by these proteins in human ES cells ., We initially targeted a region of the human Zfpm2 ( hZfpm2 ) locus , which encodes a developmental transcription factor involved in heart and gonad development 37 ., In ES cells , the endogenous locus recruits PRC1 and PRC2 , and is enriched for the bivalent histone modifications , K4me3 and K27me3 ( Figure 1A ) ., We used recombineering to engineer a 44 kb BAC containing this locus and a neomycin selection marker ., The modified BAC was electroporated into mouse ES cells , and individual transgenic ES cell colonies containing the full length BAC were expanded ( Figure S1 ) ., Fluorescent in situ hybridization ( FISH ) confirmed integration at a single genomic location ( Figure S2 ) ., We used ChIP and quantitative PCR ( ChIP-qPCR ) with human specific primers to examine the chromatin state of the newly incorporated hZfpm2 locus ., This analysis revealed strong enrichment for K27me3 and K4me3 ( Figure 1B ) ., In addition , we explicitly tested for direct binding of the Polycomb repressive complexes using antibody against the PRC1 subunit , Ring1B , or the PRC2 subunit , Ezh2 ., We detected robust enrichment for both complexes in the vicinity of the hZfpm2 gene promoter ( Figure 1B ) ., To confirm this result and eliminate the possibility of integration site effects , we tested two additional transgenic hZfpm2 ES cell clones with unique integration sites as well as a fourth transgenic ES cell line containing a distinct Polycomb target locus , Pax5 ., In each case , we observed a bivalent chromatin state analogous to the endogenous loci ( Figure S3 ) ., Similar to endogenous bivalent CpG islands , we found the Zfpm2 CpG island was DNA hypomethylated ( Figure S4 ) ., These results suggest that DNA sequence is sufficient to initiate de novo recruitment of Polycomb in ES cells ., A key function of Polycomb repressors is to maintain a repressive chromatin state through cellular differentiation ., To determine if the integrated BAC is capable of maintaining K27me3 , the hZfpm2 transgenic ES cells were differentiated to neural progenitor ( NP ) cells in vitro 38 ., ChIP-qPCR analysis revealed continued enrichment of K27me3 but loss of K4me3 ( Figure 1C ) , a pattern frequently observed at endogenous loci that are not activated during differentiation 39 ., This indicates that DNA sequence at the hZfpm2 locus is sufficient to initiate K27me3 chromatin modifications in ES cells , and maintain the repressive chromatin state through neural differentiation ., We next sought to define the sequences within the hZfpm2 BAC required for recruitment of Polycomb repressors ., First , we re-engineered the 44 kb hZfpm2 BAC to remove 20 kb of flanking sequences that contained distal non-coding conserved sequence elements ( Figure 1A ) ., When we integrated the resulting 22 kb construct into ES cells we found that it robustly enriches for PRC1 , PRC2 , K4me3 and K27me3 ( Figure 1B ) ., Hence , these particular distal elements do not appear to be required for the recruitment of the complexes ., Next , we considered the necessity of the CpG island which corresponds to the peak of Ezh2 enrichment in ChIP-Seq profiles ( Figure 1A ) ., We excised a 1 . 7 kb fragment containing the CpG island , and integrated the resulting BAC ( ΔCGI ) into ES cells ., The ΔCGI BAC failed to recruit PRC1 or PRC2 , and showed significantly reduced K27me3 levels relative to the other constructs ( Figure 1B ) ., This suggests that the CpG island is essential for recruitment of Polycomb proteins to the hZfpm2 locus ., We next asked whether the hZfpm2 CpG island is sufficient to recruit Polycomb repressors to an exogenous locus ., To test this , we selected an unremarkable gene desert region on human chromosome 1 that shows no enrichment for PRC1 , PRC2 or K27me3 in ES cells ( Figure 2A ) ., We also verified that the gene desert BAC alone does not show any enrichment for K27me3 or Ezh2 when integrated into ES cells ( Figure 2B ) ., Using recombineering , we inserted the 1 . 7 kb sequence that corresponds to the hZfpm2 CpG island into the gene desert BAC ., The resulting construct was integrated into mouse ES cells and three independent clones were evaluated ., ChIP-qPCR analysis revealed strong enrichment for K27me3 , K4me3 and PRC2 over the inserted CpG island ( Figure 2C , Figure S5 ) ., In contrast , we observed relatively little enrichment for the PRC1 subunit Ring1B ( Figure 2C ) ., We confirmed the specificity of these enrichments with primers that span the boundary between the insertion and adjacent gene desert sequence ., Notably , K27me3 enrichment was detected across the gene desert locus up to 2 . 5 kb from the inserted CpG island ( Figure 2C ) ., This indicates that the localized CpG island can initiate K27me3 that then spreads into adjacent sequence ., Lastly we found no YY1 enrichment across the CpG island by ChIP-qPCR ( Figure S5 ) ., Together , these data suggest that the hZfpm2 CpG island contains the necessary signals for PRC2 recruitment but is insufficient to confer robust PRC1 association ., The functionality of a CpG island in PRC2 recruitment is consistent with prior observations that a majority of PRC2 sites in ES cells correspond to CpG islands 4 , 5 and with the striking correlation between intensity of PRC2 binding and the GC-richness of the underlying sequence ( Figure 2D ) ., We therefore considered whether specific signals within the Zfpm2 CpG island might underlie its capacity to recruit PRC2 ., First , we searched for sequence motifs analogous to the PREs that recruit PRC2 in Drosophila ., We focused on motifs recognized by YY1 , the nearest mammalian homolog of the Drosophila recruitment proteins ., Notably , both of the recently described mammalian PREs contain YY1 motifs 27 , 28 ., The 44 kb hZfpm2 BAC contains 11 instances of the consensus YY1 motif ., However , none of these reside within the CpG island ( Figure S6 ) ( see Methods ) ., We also examined YY1 binding directly in ES cells and NS cells using ChIP-Seq ., Consistent with prior reports , YY1 binding is evident at the 5′ ends of many highly expressed genes , including those encoding ribosomal proteins , and is also seen at the imprinted Peg3 locus ( Figure 2E , Table S1 ) 26 ., However , no YY1 enrichment is evident at the Zfpm2 locus ., Moreover , at a global level , YY1 shows almost no overlap with PRC2 or PRC1 , but instead co-localizes with genomic sites marked exclusively by K4me3 ( Figure 2F , Figure S6 , and Table S1 ) ., Thus , although YY1 may contribute to Polycomb-mediated repression through distal interactions or in trans , it does not appear to be directly involved in PRC2 recruitment in ES cells ., We previously reported that CpG islands bound by PRC2 in ES cells could be predicted based on a relative absence of activating transcription factor motifs ( AMs ) in their DNA sequence 5 ., We reasoned that transcriptional inactivity afforded by this absence of AMs is a requisite for PRC2 association 40 , 41 ., This could explain why PRC2 is absent from a majority of CpG islands , many of which are found at highly active promoters ., Consistent with this model , when we examined a recently published RNA-Seq dataset for poly-adenylated transcripts in ES cells , we found that virtually all of the high-CpG promoters ( HCPs ) lacking Ezh2 are detectably transcribed ( Figure S7 ) ., The small proportion of HCPs that are neither Ezh2-bound nor transcribed may reflect false-negatives in the ChIP-Seq or RNA-Seq data ., Alternatively , these HCPs tend to correspond to CpG islands with relatively low GC-contents and lengths and may therefore have insufficient GC-richness to promote PRC2 binding ( Figure S7 ) ., Thus , correlative analyses implicate large GC-rich elements that lack transcriptional activation signals as general PRC2 recruitment elements in mammals ., To obtain direct experimental support for the general sufficiency of large GC-rich elements lacking AMs in PRC2 recruitment , we carried out the following experiments ., First , we tested whether a K4me3-only CpG island could be turned into a PRC2 recruitment element by removing activating motifs ., We targeted a 1 . 3 kb CpG island that overlaps the promoters of two ubiquitously expressed genes – Arl3 and Sfxn2 ., Neither gene carries K27me3 in ES cells , or in any other cell type tested ( Figure S8 , and data not shown ) ., This CpG island was selected as it has many conserved AMs clustered in one half of the island ( Figure 3A ) ., We hypothesized that the portion of the Arl3/Sfxn2 CpG island lacking AMs would , in isolation , lack active transcription and recruit PRC2 ., In contrast , we predicted that the half containing multiple AMs would lack Polycomb ., To test this , we generated two additional BAC constructs containing the respective portions of the Arl3/Sfxn2 CpG island positioned within the gene desert , and integrated these constructs into ES cells ( Figure 3A ) ., ChIP-qPCR shows that the portion of the CpG island lacking AMs is able to recruit PRC2 and becomes enriched for K27me3 ( Figure 3B ) ., In contrast , the AM-containing portion shows no enrichment for K27me3 or Ezh2 , but is instead marked exclusively by K4me3 , similar to the endogenous human locus ( Figure 3C , Figure S8 ) ., Thus , a GC-rich sequence element with no known requirement for Polycomb regulation can recruit PRC2 when isolated from activating sequence features ., Next , we tested whether even more generic GC-rich elements might also be capable of recruiting PRC2 in ES cells ., Here , we focused on sequences derived from the genome of E . coli , reasoning that there would be no selection for PRC2 recruiting elements in this prokaryote given the complete lack of chromatin regulators ., We arbitrarily selected three 1 kb segments of the E . coli genome ., Two with GC contents above the threshold for a mammalian CpG island but that each contained few AMs , and one AT rich segment as a control ( Table S3 ) ., We recombined each segment into the gene desert BAC and integrated the resulting constructs into ES cells ., ChIP-qPCR confirmed that both GC-rich E . Coli segments recruit Ezh2 and form a bivalent chromatin state ( Figure 4A , 4B , Figure S9 ) ., Notably , the GC-rich segment also enriches for Jarid2 , a PRC2 component with DNA binding activity ( Figure S10 ) ., In contrast , the AT-rich segment did not recruit Ezh2 or enrich for either K4me3 or K27me3 ( Figure 4C , Figure S9 ) ., Together , our findings suggest that GC-rich sequence elements that lack signals for transcriptional activation have an innate capacity to recruit PRC2 in mammalian ES cells ., Several lines of evidence suggest that the initial landscape of Polycomb complex binding is critical for proper patterning of gene expression in metazoan development 1 , 2 , 13 ., Failure of these factors to engage their target loci in embryogenesis has been linked to a loss of epigenetic repression at later stages ., Accordingly , the determinants that localize Polycomb complexes at the pluripotent stage are almost certainly essential to the global functions of these repressors through development ., We find that DNA sequence is sufficient for proper localization of Polycomb repressive complexes in ES cells , and specifically identify a CpG island within the Zfpm2 locus as being critical for recruitment ., We provide evidence that GC-rich elements lacking activating signals suffice in general to recruit PRC2 ., This includes demonstrations, ( i ) that a motif devoid segment of an active ‘housekeeping’ CpG island can recruit PRC2; and, ( ii ) that arbitrarily selected GC-rich elements from the E . coli genome can themselves mediate PRC2 recruitment when integrated into the ES cell genome ., Several possible mechanistic models could explain the causality of GC-rich DNA elements in PRC2 recruitment ( Figure 5 ) ., First , we note that CpG islands have been shown to destabilize nucleosomes in mammalian cells 42 ., At transcriptionally inactive loci , this property could increase their accessibility to PRC2-associated proteins with DNA affinity but low sequence specificity , such as Jarid2 or AEBP2 32–35 , 43 ( Figure S10 ) ., Although this association would be abrogated by transcriptional activity at most CpG islands , those lacking activation signals would remain permissive to PRC2 association ( Figure 5 ) ., In support of this model , PRC2 targets in ES cells are also enriched for H2A . Z and H3 . 3 , histone variants linked to nucleosome exchange dynamics 44 , 45 ., Alternatively or in addition , targeting could be supported by DNA binding proteins with affinity for low complexity GC-rich motifs or CpG dinucleotides , such as CXXC domain proteins 46 ., Localization may also be promoted or stabilized by long and short non-coding RNAs 47–50 as well as by the demonstrated affinity of PRC2 for its product , H3K27me3 11 , 12 ., Notably , PRC2 recruitment in ES cells appears distinct from that in Drosophila , as we do not find evidence for involvement of PRE-like sequence motifs or mammalian homologues such as YY1 ., It should be emphasized that PRC2 localization does not necessarily equate with epigenetic repression ., Indeed virtually all PRC2 bound sites in ES cells , and all CpG islands tested here , are also enriched for K4me3 , and presumably poised for activation upon differentiation ., Epigenetic repression during differentiation may require PRC1 and thus depend on additional binding determinants ., YY1 remains an intriguing candidate in this regard , given prior evidence for physical and genetic interactions with PRC1 51 , 52 ., YY1 consensus motifs are present in the Polycomb-dependent silencing elements recently identified in the MafB and HoxD loci ., Interestingly , the HoxD element combines a CpG island with a cluster of conserved YY1 motifs ., Mutation of the motifs abrogated PRC1 binding but left PRC2 binding intact ., Still , the fact that only a small fraction of documented PRC2 and PRC1 sites have YY1 motifs or binding suggests that this transcription factor may act indirectly and/or explain only a subset of cases ., Nonetheless , it is likely that a fully functional epigenetic silencer would require a combination of features , including a GC-rich PRC2 element as well as appropriate elements to recruit PRC1 ., Further study is needed to expand the rules for PRC2 binding to include a global definition of PRC1 determinants and ultimately , to understand how the initial landscape facilitates the maintenance of gene expression programs in the developing organism ., BAC constructs CTD331719L ( ‘Zfpm2 44’ ) , CTD-2535J16 ( ‘Pax5’ ) and CTD-3219L19 ( ‘Gene Desert’ ) were obtained from Open Biosystems ., Recombineering was done using the RedET system ( Open Biosystems ) in DH10B cells ., Homology arms 200–500 bp in length were PCR amplified and cloned into a PGK; Neomycin cassette ( Gene Bridges ) ., This cassette was used to recombineer all BACs to enable selection in mammalian cells ., The 22 kb hZfpm2 BAC was created by restricting the hZfpm2 BAC at two sites using ClaI , and re-ligating the BAC lacking the intervening sequence ., The CpG island was excised from the 22 kb hZfpm2 BAC by amplification of flanking homology arms , and cloned into a construct containing an adjacent ampicillin cassette ( Frt-amp-Frt; Gene Bridges ) ., After recombination , the ampicillin cassette was removed using Flp-recombinase and selection for clones that lost ampicillin resistance ( Flp-706; Gene Bridges ) ., PCR across the region confirmed excision of the CpG island ., For the Gene Desert BACs , the Zfpm2 , Arl3 , Sfxn2 and E . coli CpG islands were amplified with primers containing XhoI sites and cloned into the Frt-amp-Frt vector that contains homology arms from the Gene Desert region ., The final constructs were confirmed by sequencing across recombination junctions ., All primers used for CpG islands and recombineering homology arms are listed in Table S2 ., ES cells ( V6 . 5 ) were maintained in ES cell medium ( DMEM; Dulbeccos modified Eagles medium ) supplemented with 15% fetal calf serum ( Hyclone ) , 0 . 1 mM ß-mercaptoethanol ( Sigma ) , 2 mM Glutamax , 0 . 1 mM non-essential amino acid ( NEAA; Gibco ) and 1000U/ml recombinant leukemia inhibitory factor ( ESGRO; Chemicon ) ., Roughly 50 µg of linearized BAC was nucleofected using the mouse ES cell nucleofector kit ( Lonza ) into 106 mouse ES cells , and selected 7–10 days with 150 µg/ml Geneticin ( Invitrogen ) on Neomycin resistant MEFs ( Millipore ) ., Individual resistant colonies were picked , expanded and tested for integration of the full length BAC by PCR ., Differentiation of hZfpm2 ES cell clone 1 into a population of neural progenitor ( NP ) cells was done as previously described 53 ., FISH analysis was done as described previously 54 ., DNA methylation analysis was done as previously described 55 and primers used to amplify bisulfite treated DNA are listed in Table S2 ., For each construct , between one and three ES cell clones were expanded and subjected to ChIP using antibody against K4me3 ( Abcam ab8580 or Upstate/Millipore 07-473 ) , K27me3 ( Upstate/Millipore 07-449 ) , Ezh2 ( Active Motif 39103 or 39639 ) , or Ring1B ( MBL International d139-3 ) as described previously 5 , 7 , 39 ., ChIP DNA was quantified by Quant-iT Picogreen dsDNA Assay Kit ( Invitrogen ) ., ChIP enrichments were assessed by quantitative PCR analysis on an ABI 7500 with 0 . 25 ng ChIP DNA and an equal mass of un-enriched input DNA ., Enrichments were calculated from 2 or 3 biologically independent ChIP experiments ., For K27me3 , and Ezh2 enrichment , background was subtracted by normalizing over a negative genomic control ., Error bars represent standard error of the mean ( SEM ) ., We confirmed that the human specific primers do not non-specifically amplify mouse genomic DNA ., Primers used for qPCR are listed in Table S2 ., Genomewide maps of YY1 binding sites were determined by ChIP-Seq as described previously 39 ., Briefly , ChIP was carried out on 6×107 cells using antibody against YY1 ( Santa Cruz Biotechnology sc-1703 ) ., ChIP DNA was used to prepare libraries which were sequenced on the Illumina Genome Analyzer ., Density profiles were generated as described 39 ., Promoters ( RefSeq; http://genome . ucsc . edu ) were classified as positive for YY1 , H3K4me3 or H3K27me3 if the read density was significantly enriched ( p<10−3 ) over a background distribution based on randomized reads generated separately for each dataset to account for the varying degrees of sequencing depth ., ChIP-Seq data for YY1 are deposited to the NCBI GEO database under the following accession number GSE25197 ( http://www . ncbi . nlm . nih . gov/projects/geo/query/acc . cgi ? acc=GSE25197 ) ., Sites of Ezh2 enrichment ( p<10−3 ) were calculated genomewide using sliding 1 kb windows , and enriched windows within 1 kb were merged ., DNA methylation levels were calculated using previously published Reduced Representation Bisulphite Sequenced ( RRBS ) libraries 55 ., Composite plots represent the mean methylation level in sliding 200 bp windows in the the 10 kb surrounding the TSSs of the indicated gene sets ., YY1 motifs were identified using the MAST algorithm 56 where a match to the consensus motif was defined at significance level 5×10−5 ., Candidate CpG islands for TF motif analysis were identified by scanning annotated CpG islands ( http://genome . ucsc . edu ) for asymmetric clustering of motifs related to transcriptional activation in ES cells 5 ., Motifs shown in Figure 3A and Figure S6 are from UCSCs TFBS conserved track ., GC-rich elements from the E . coli K12 genome were selected by calculating %GC and CpG O/E in sliding 1 kb windows ., Sequences matching the criteria for mammalian CpG islands while simultaneously being depleted of motifs related to transcriptional activation 5 were chosen for insertion into mouse ES cells ., Transcriptionally inactive HCPs were selected based on a lack of transcript enrichment by both expression arrays 39 and RNA-Seq data 57 ., In the case of RNA-Seq , each gene was assigned the maximum read density within any 1 kb window of exonic sequence ., To ease analysis of promoter CpG island statistics , only HCPs containing a single CpG island were considered . | Introduction, Results, Discussion, Methods | Polycomb proteins are epigenetic regulators that localize to developmental loci in the early embryo where they mediate lineage-specific gene repression ., In Drosophila , these repressors are recruited to sequence elements by DNA binding proteins associated with Polycomb repressive complex 2 ( PRC2 ) ., However , the sequences that recruit PRC2 in mammalian cells have remained obscure ., To address this , we integrated a series of engineered bacterial artificial chromosomes into embryonic stem ( ES ) cells and examined their chromatin ., We found that a 44 kb region corresponding to the Zfpm2 locus initiates de novo recruitment of PRC2 ., We then pinpointed a CpG island within this locus as both necessary and sufficient for PRC2 recruitment ., Based on this causal demonstration and prior genomic analyses , we hypothesized that large GC-rich elements depleted of activating transcription factor motifs mediate PRC2 recruitment in mammals ., We validated this model in two ways ., First , we showed that a constitutively active CpG island is able to recruit PRC2 after excision of a cluster of activating motifs ., Second , we showed that two 1 kb sequence intervals from the Escherichia coli genome with GC-contents comparable to a mammalian CpG island are both capable of recruiting PRC2 when integrated into the ES cell genome ., Our findings demonstrate a causal role for GC-rich sequences in PRC2 recruitment and implicate a specific subset of CpG islands depleted of activating motifs as instrumental for the initial localization of this key regulator in mammalian genomes . | Key developmental genes are precisely turned on or off during development , thus creating a complex , multi-tissue embryo ., The mechanism that keeps genes off , or repressed , is crucial to proper development ., In embryonic stem cells , Polycomb repressive complex 2 ( PRC2 ) is recruited to the promoters of these developmental genes and helps to maintain repression in the appropriate tissues through development ., How PRC2 is initially recruited to these genes in the early embryo remains elusive ., Here we experimentally demonstrate that stretches of GC-rich DNA , termed CpG islands , can initiate recruitment of PRC2 in embryonic stem cells when they are transcriptionally-inactive ., Surprisingly , we find that GC-rich DNA from bacterial genomes can also initiate recruitment of PRC2 in embryonic stem cells ., This supports a model where inactive GC-rich DNA can itself suffice to recruit PRC2 even in the absence of more complex DNA sequence motifs . | molecular biology/histone modification, developmental biology/stem cells, genetics and genomics/gene expression, developmental biology/cell differentiation, genetics and genomics/epigenetics, molecular biology/chromatin structure | null |
journal.pcbi.1005713 | 2,017 | A conceptual and computational framework for modelling and understanding the non-equilibrium gene regulatory networks of mouse embryonic stem cells | Our principal aim is to capture and integrate the results of multiple high-throughput experiments using a logical and transparent computational framework ., This would allow us to model protein expression , particularly TF expression , across multiple layers of stem cell regulation ., However , this first requires a sound theoretical framework to understand and predict how regulatory layers self-organize and interact , both within individual cells and between multiple cell types within larger cell assemblies ., Here , we begin to address the problem by describing a novel computational concept derived from the hypothesis of exploratory behaviour described ( see Exploratory hypothesis of stem cell fate computation , above ) and from branching process theory ( see Branching Processes , above ) ., The utility of such a computational approach relies on the scale-invariant nature of the reproducing units ., By its very nature , critical dynamics describes multiple parallel processes that propagate in some way , but because these processes occur within a bounded system , not all can realize their full potential to propagate ., Unlike self-organized critical systems , which typically incorporate local stability thresholds , the propagation of a branching process in a critical-like system depends on other processes propagating at that time ( although of course there may also be local stability thresholds involved ) ., The key point is that the critical-like state is achieved or computed via direct interaction among branching processes ., In contrast , in self-organized critical systems all of the branching processes are temporally separated such that they cannot interact directly , only indirectly via local stability thresholds ., In our model , branching processes are used conceptually to define boundaries of information flow ., ChIP-Seq data are used to capture the entire genome of a pluripotent stem cell , where the expression of each pluripotency TF is defined in terms of a branching process that propagates through time , interacting and competing with others ., In this view , fluctuation in the expression of pluripotency genes when mouse ESC are withdrawn from self-renewal conditions ( 2i ) is not trivial: it is central to model dynamics ., They are expected to determine differentiation trajectories ., When we consider the expression of a single TF as a distinct branching process , the population of TF molecules can be thought of as the backbone of the branching process as it propagates through time ., This can be likened to propagation of a family name through the male offspring; female offspring traditionally fail to propagate the family surname , similar to bursts of TF expression that activate target genes that do not feed back into the transcriptional network ., In the context of TF expression , the branching process effectively defines how the population of TF molecules reproduces via the entire regulatory network ., In this sense , the trajectories of individual ESCs are intrinsically knowable and able to be calculated from patterns of competition and interference among cascades of gene expression bounded by the cistrome of each TF ., Genome location data , which describe interactions between TFs and other genes at genome-wide scales , can be used to simulate these branching processes and estimate patterns of interference that give rise to individual cell trajectories ., From the simulation of the TFBP , we can demonstrate the existence of mcrit values , values of the branching parameter below which the simulation rapidly dissipates and above which supercritical branching can take place; see Greaves et al 30 for details ., The model includes the following parameters and values: The determination of mcrit for Nanog , Sox2 and Oct4 is found in Greaves et al 30 ., ( The experiments were repeated with the new multi-cistrome code used in this paper , run with a single cistrome; the same results were found . ), The determination of mcrit for Nanog is illustrated in Fig 1 ., The value of mcrit for the Nanog cistrome ( 8 ) is found to be somewhat higher than that for the other core pluripotency cistromes of Oct4 and Sox2 ( 6 and 7 respectively ) ., The observed mcrit results are summarised in the final table below ., An infinite limit model discussed in 30 is used to calculate an estimate of mcrit; in this limit the system tips from fully dissipated to supercritical immediately ., Fig 1 is not such a step function ., The finite sized , noisy , system tips sharply from dissipated to ignited at the experimentally observed value mcrit , but then requires a somewhat higher value of m to become fully saturated ., The observed mcrit is slightly higher than the infinite limit value ( Table 1 ) , due to noise ., Importantly , even though the system is noisy , it still maintains supercriticality in the face of fluctuations in all runs with mcrit ≤ m ., We have also verified that that mcrit is not affected by whether the initial value s0 = r or s0 = 0 . 5r and that results are scalable i . e . we can alter the values of p and r in a cistrome , providing that the ratio p / r remains constant and obtain the same results 30 ., In addition to these previously existing findings , we have run similar simulations for the cMyc cistrome to determine its value of mcrit ( see Fig 2 and Table 1 ) ., This cistrome has a relatively low value of mcrit ( of 4 ) due to its relatively high number of active sites compared to the other cistromes investigated ., The multi-cistrome simulation shows that one active cistrome can “ignite” a fully dissipated cistrome , and drive it to criticality ., This is illustrated in two cases ., The critically branching Oct4 cistrome can drive fully dissipated Nanog cistrome to criticality ( Fig 3 ) ., The figure shows the first of 200 simulation runs carried out with these initial conditions ( the others runs are qualitatively similar ) ., The inset shows the first 25 timesteps ., Similarly , the critically branching Nanog cistrome can drive a fully dissipated Oct4 cistrome to criticality ( Fig 4 ) ., These show in each that , although the cistrome begins the simulation fully dissipated , it is swiftly ignited to sustainable branching by the critically branching cistrome to which it is coupled ., This is a non-trivial result because:, ( i ) the driving cistrome is working just at the critical value of m needed to sustain itself;, ( ii ) the driving cistrome incurs a “cost” to do so , because some of the TPs it produces are transferred to the other initially dissipated cistrome , rather than being used to maintain its own activity ., In addition to igniting a dissipated cistrome , an effect of coupling cistromes is the reduction of mcrit ., In 30 we use an infinite limit model to calculate an estimate of mcrit in a single cistrome ., We here use a similar approach to calculate an estimate in the reduction of mcrit in coupled cistromes ., This infinite limit case is essentially noiseless , with each TF site being ignited the average number of times ., Consider the case of two cistromes , X and Y . At time t let there be stX sites active in cistrome X . In the TFBP model , each of these active sites emits mX TFs , so a total of stX × mX TFs are emitted ., Let each of these TFs be absorbed by a separate site with uniform probability ., There are three cases ( three kinds of sites ) : Similar arguments , mutatis mutandis , hold for TFs emitted by cistrome Y . So at timestep t+1 , the number of active sites in cistrome X is those activated from cistrome Y plus those activated from cistrome X, st+1X=stXmX ( rX−σ ) pX+stYmYσpY, A similar equation holds for st+1Y ., At the critical tipping point , st+1 = st . We take these two equations , eliminate sX / sY , then solve for mX , to get, mX=pX ( pY−mYrY+mYσ ) ( pY−mYrY ) ( rX−σ ) +mYrXσ, This gives the infinite limit predicted value of mX in the case of a given mY ., If we substitute mY = pY/rY , the infinite limit single cistrome critical value for Y , we get mX = pX/rX ., That is , in the infinite limit , the critical values are unchanged ., Alternatively , if we substitute σ = 0 ( isolated cistromes ) , we also recover the original predicted value of mX ., However , if we substitute the observed critical value in the finite sized noisy case for mY , and the experimentally derived value of σ ( shown in Table 2 ) we get a different prediction for mX , as shown in Table, 3 . Table 3 shows the prediction that critical value for cistrome X should fall when coupled with cistrome Y run at its observed critical value ., Under the assumptions used to generate Table 3 , cistrome Y is being run at a higher rate than it needs in the infinite limit , and the excess TF production can be used to lower cistrome X’s required rate ., The question is: does this reduction carry over in the finite case , or does the presence of noise , requiring a higher than predicted rate to sustain , negate any such change ?, Our simulation results , for several of these cases , are presented below ., Figs 5–10 show the behaviours of various combinations of coupled cistromes ., Each plot shows the first 250 of 1000 timesteps performed , and shows the first run from a set of 200 runs performed; the others runs are similar ., In each case , the critically branching cistrome ( s ) can drive the branching process in the other cistrome , and the sub-critically branching cistrome dissipates the branching process in the critically branching cistrome ( s ) ., Coupling the Nanog cistrome to that of Oct4 ( with Oct4 having its m value fixed at its observed value mcrit = 6 ) reduces mcrit for the Nanog cistrome by one ( Fig 5 ) ., Similar results hold for coupling the Nanog cistrome to Sox2 ., Coupling the Oct4 cistrome to that of Nanog ( with Nanog having its m value fixed at its mcrit = 8 ) reduces mcrit for the Oct4 cistrome by one ( Fig 6 ) ., Coupling the Nanog cistrome to both the Oct4 and the Sox2 cistromes reduces mcrit for the Nanog cistrome by 2 ( Fig 7 ) ., The c-Myc cistrome is extensively overlapped with the core pluripotency cistromes ., Coupling the Nanog cistrome to the c-Myc cistrome alone reduces mcrit for the Nanog cistrome by 2 ( Fig 8 ) ., Hence the c-Myc cistrome has twice the effect on Nanog mcrit as either Oct4 or Sox2 alone ., Similar results hold for Oct4 ., Coupling the Oct4 cistrome to that of cMyc cistrome reduces the value of mcrit for the Oct4 cistrome from 6 to 4 ( Fig 9 ) ., Coupling the cMyc cistrome to that of Oct4 cistrome leaves the value of mcrit for the cMyc cistrome unchanged at 4 ( Fig 10 ) ., The various reductions in mcrit caused by coupling cistromes are summarised in Table, 4 . Similar to the single cistrome case , the observed values of mcrit are a little higher than the infinite case predictions ., In all cases but one the observed coupled value is nevertheless lower than the observed uncoupled value , demonstrating that the reduction can be maintained in the presence of noise ., In the case of Oct4 driving cMyc , there is no reduction in the observed value of mcrit , possibly because the value is so low in the first place , and so any reduction would be proportionally smaller ., We have taken a first step towards the creation of a multi-layered model of the stem cell regulatory network and in our opinion , these results are interesting and begin to tease out the utility of the Branching Process Theory in understanding Stem Cell fate computation ., We have demonstrated that a regulatory network may self-organise from one equilibrium state to another , through ignition of a coupled dissipated cistrome ., Understanding the complex dynamics of such self-organising changes of state possible with communicating cistromes may ultimately give insight into how a pluripotent state can tip to a differentiated state , and may help in understanding how to control and even reverse such tipping ., However , our model does not , as yet , capture all the biologically-relevant dynamics ., For example , in our current model , a TF can interact with genes in another cistrome only in a stimulatory manner ., We need to include the possibility of inhibitory TF binding of TFs to their binding sites on the various gene segments in the cistromes modelled ., We therefore need to develop a Domain Specific Language to allow a standardised way of describing the TF circuitry ( the pattern of excitatory and inhibitory relationships between individual TFs ) to be modelled in the simulation ., The current model is essentially ‘blind’ to the identity of any TFs produced via transcription of the gene segments in any given cistrome , and also ignores the potential requirement for multiple TF to bind to a gene segment in order to activate it ( or inhibit it ) ., There are also other considerations that we need to take on board in order to capture further biologically-relevant dynamics ., We also need to accommodate combinatorial binding of TFs to gene segment promoter sequences , as a gene may require binding of specific combinations of TFs to their binding sites in order to promote transcription of the gene segment ., Such combinatorial TF binding to enhancer sites can impart transcriptional synergy in a future multicellular model ., We need to include expression data to factor in the strength of activation: currently each gene segment is assumed to produce the same amount of TF each timestep ., We also need to include some concept of TF half-life into the model as currently all TFs are assumed to survive and remain bound to their binding sites for a single simulation time-step only ., As epigenetic histone modifications may help to shape the circuitry of self-organization , it would also be useful to be able to take into account epigenetic considerations and the effect of enhancer sites within the cistromes modelled ., Our model is currently aspatial , in that it lacks any detail of three-dimensional genomic architecture , which affects how TFs access their binding sites ., Inclusion of histone data to incorporate spatial effects is a future goal ., Additionally , we remain uncertain of how the degree of cistrome overlap ( number of shared segments ) determines the extent of the effect of cistrome coupling on reduction of mcrit ., More experimental data on this overlap , particularly in the case of multiple cistromes , is needed to investigate this ., The model presented represents a novel example of self-organization that may apply to other complex systems of interest from a theoretical point of view because it helps to demonstrate how distributed interactions among units result in higher order emergent behaviours ., Such complexity could provide dynamic templates of organization upon which natural selection builds additional elaborations 5 ., However , even this extremely pared down implementation of the TFBP demonstrates the ability of coupled TFBPs ( equivalent to transcription patterns ) to influence and modulate each other’s branching behaviour via constructive interference as suggested by the theoretical model proposed by Halley and Winkler 5 ., The code for the simulation , batch scripts for running the simulation on an SGE enabled compute cluster , Python scripts for generating real or synthetic cistromes , and example R scripts for processing simulation results into graphical form , are all available on GitHUB at: github . com/CellBranch/CellBranch, In Greaves et al . 30 we detail the development of a simulation of a single cistrome branching process using the iterative CoSMoS simulation design protocol 40 , taking the Transcription Factor Branching Process ( TFBP ) Model 2 , 3 , 5 as our initial domain model ., That simulation was written as object-oriented code in Java , using the MASON simulation development environment ., The reader is referred to Greaves et al . 30 for the relevant design details , and assumptions made about the domain ., The context of the simulation development remains the same as in our previous work , i . e . the investigation of the TFBP approach to modelling stem cell fate computation ., Fig 11 shows how the ChIP-Seq data is used to produce a model of a single cistrome ., Fig 12 shows how this single cistrome is used to underpin the TFBP model ., In Greaves et al . 30 we have reasoned that if the activities of single TFs can be adequately described as a critical-like Branching Processes , as suggested by our results in that publication , then their interplay should define a critical-like genome-wide interference pattern ., This pattern would then in some way , capture the nature of the entire pluripotency TF regulatory network 3 ., We now wish to gain a deeper understanding of the behaviour of constructive interference between interfering branching processes ., In particular , we aim to characterise TF branching processes and how they might propagate in the presence of cross-cistrome communication ., So we now discuss the extension of the earlier , simple model Greaves et al . 30 , to a model of two or more interacting , branching cistromes , to enhance the biological relevance of the simulation ., We also need to allow for segments that are shared by multiple cistromes and to specify TF sharing algorithms ., These refinements of the simulation require us to make further assumptions about system behaviour and to revise exiting ones ., Most obviously , this includes assumptions about how cistromes communicate throughout the simulation ., By communication , we mean the transfer of TF branching behaviour products from one cistrome to the appropriate binding sites on another ., We assume that the inclusion of constructive interference of TFBPs is a sensible increment in the model of the system , but acknowledge that it does not yet yield a simulation of full biological relevance ., In Greaves et al . 30 we use the language of the most abstract of our models of the system , our ‘sparking posts’ model , but here we use the language of the more biological abstraction , i . e . cistromes and segments and ‘transcription’ ., In our new multi-cistrome model , the TF binding sites in a cistrome can be occupied by TFs transcribed from genes in the same cistrome or those from another cistrome ., We have found it useful to implement the model so that we can distinguish the separate origins of incoming TFs–same cistrome or different cistrome , with those TF incoming from a different cistrome being bound to a binding site that we nicknamed the ‘spark bucket’ in our abstract sparking posts model , but which is merely the equivalent of a second TFBS for the appropriate incoming TF in our abstract biological model ., We start our augmentation of the simulation by examining the case of cooperation between two TFs at a genome-wide scale ., This is still far from a biologically realistic system as very many TFs will intercommunicate to regulate cellular systems i . e . multiple TFBPs will interfere constructively and destructively such that we would expect to ultimately be able to generate the interference patterns predicted to underpin cell circuitry self-organization ., Strictly speaking , in biology , any segment shared by two or more cistromes will have a TF binding site for each of the TFs produced by the cistromes which share the gene segment ( remembering that in our simplified model cistrome X branching gives rise to TF X at all times if a TF product binds to a binding site in cistrome X ) ., However , since the model allows a segment in cistrome X to produce TF products when we have either a TF derived from cistrome X or a TF derived from any other coupled cistrome , then it is not necessary for us to include the possibility of TF transfer from more than one cistrome to the segment in Cistrome X . Again , we believe that the distinction between this description and the one we actually employ will be subtle , but acknowledge that it should be fully tested before we progress the model further ., We also assume that TFs can remain bound to their binding sites for one simulation time-step only ., At each time-step a segment in a particular cistrome , let us call this cistrome X for sake of argument , can receive a TF molecule resulting from branching within Cistrome X . It can also receive a TF molecule generated by branching within another cistrome , say Y , which shares this particular segment with Cistrome X . We have further assumed that in any single time-step , only one TF from another cistrome can be transferred to a corresponding binding site in Cistrome X . Our implementation model is illustrated in Fig 13 , and is an implementation of this description of TF transfer between cistromes ., In our model , each gene segment has two TF binding sites: one for TF derived from the same cistrome and a secondary one for incoming TF from another cistrome ., If either of these two sites is occupied at the end of a simulation step , then that gene segment will be activated and a TF product will be released to propagate the appropriate TF branching process ., Fig 11 outlines four special cases of inter-cistrome communication in our model: Alternatively , we could have assumed that if a segment in Cistrome X has a TF molecule resulting from branching within Cistrome X bound to it , then no TFs resulting from branching in other cistromes can be accepted by the segment in Cistrome X . However , we have decided to reject this description of the system as in the biological system , communication between cistromes will occur via shared gene segments and these segments will therefore have binding sites for both the TFs concerned ., This model has limitations from the point of view of studying biologically relevant systems , because we have deferred the inclusion of destructive interference between branching processes to a later increment in order to permit the acquisition of a fuller understanding of this simplified representation of the system prior to adding another layer of complexity to the model ., We also , at this point , acknowledge that another equally valid assumption about inter-cistrome communication would be that a TF transfer cannot occur between cistromes if the destination cistrome already has any TF bound to a site in the segment under consideration ., i . e . in Fig 11C , Cistrome X would not be able to transfer its bound TF to Cistrome Y’s second TFBS ., We believe that this alternative model will not substantially alter the qualitative nature of the results obtained from the simulation and we mean to undertake this verification prior to any further expansion of our computational model ., c-Myc is a TF that is connected with approximately 30 , 000 target areas throughout the genome ., Thus , c-Myc represents an evolutionarily ancient undercurrent that underpins circuitry self-organization and cell behaviour in many different ways 41 , 42 ., c-Myc overlaps core pluripotent TF cistromes to roughly the same extent as they overlap each other ( refer to Tables 1 and 2 ) , but with mcrit ( 4 ) being half that for the Nanog cistrome ( 8 ) and lower than that for both Oct4 and Sox2 ( 6 and 7 respectively ) ., We tested the extended simulation by first using it to replicate the results obtained from the single cistrome simulation presented in Greaves et al . 30 ., We then ran simulations in which two or more cistromes were coupled under a variety of starting conditions , e . g . one or more saturated cistromes coupled to an initially dissipated cistrome ., Visualisation of results was carried out using R scripts to present plots of the proportion of ‘red’ segments activated at a given timestep or the proportion of ‘red’ segments activated at the end of the simulation ( t = 1 , 000 ) against the value of the branching parameter , m . | Introduction, Results, Discussion, Model | The capacity of pluripotent embryonic stem cells to differentiate into any cell type in the body makes them invaluable in the field of regenerative medicine ., However , because of the complexity of both the core pluripotency network and the process of cell fate computation it is not yet possible to control the fate of stem cells ., We present a theoretical model of stem cell fate computation that is based on Halley and Winkler’s Branching Process Theory ( BPT ) and on Greaves et al . ’s agent-based computer simulation derived from that theoretical model ., BPT abstracts the complex production and action of a Transcription Factor ( TF ) into a single critical branching process that may dissipate , maintain , or become supercritical ., Here we take the single TF model and extend it to multiple interacting TFs , and build an agent-based simulation of multiple TFs to investigate the dynamics of such coupled systems ., We have developed the simulation and the theoretical model together , in an iterative manner , with the aim of obtaining a deeper understanding of stem cell fate computation , in order to influence experimental efforts , which may in turn influence the outcome of cellular differentiation ., The model used is an example of self-organization and could be more widely applicable to the modelling of other complex systems ., The simulation based on this model , though currently limited in scope in terms of the biology it represents , supports the utility of the Halley and Winkler branching process model in describing the behaviour of stem cell gene regulatory networks ., Our simulation demonstrates three key features:, ( i ) the existence of a critical value of the branching process parameter , dependent on the details of the cistrome in question;, ( ii ) the ability of an active cistrome to “ignite” an otherwise fully dissipated cistrome , and drive it to criticality;, ( iii ) how coupling cistromes together can reduce their critical branching parameter values needed to drive them to criticality . | Pluripotent stem cells possess the capacity both to renew themselves indefinitely and to differentiate to any cell type in the body ., Thus the ability to direct stem cell differentiation would have immense potential in regenerative medicine ., There is a massive amount of biological data relevant to stem cells; here we exploit data relating to stem cell differentiation to help understand cell behaviour and complexity ., These cells contain a dynamic , non-equilibrium network of genes regulated in part by transcription factors expressed by the network itself ., Here we take an existing theoretical framework , Transcription Factor Branching Processes , which explains how these genetic networks can have critical behaviour , and can tip between low and full expression ., We use this theory as the basis for the design and implementation of a computational simulation platform , which we then use to run a variety of simulation experiments , to gain a better understanding how these various transcription factors can combine , interact , and influence each other ., The simulation parameters are derived from experimental data relating to the core factors in pluripotent stem cell differentiation ., The simulation results determine the critical values of branching process parameters , and how these are modulated by the various interacting transcription factors . | gene regulation, regulatory proteins, dna-binding proteins, cell differentiation, simulation and modeling, developmental biology, stem cells, transcription factors, research and analysis methods, cell potency, animal cells, genetic interference, proteins, gene expression, pluripotency, biochemistry, cell biology, gene regulatory networks, genetics, biology and life sciences, cellular types, computational biology | null |
journal.pcbi.1000352 | 2,009 | Statistical Methods for Detecting Differentially Abundant Features in Clinical Metagenomic Samples | The increasing availability of high-throughput , inexpensive sequencing technologies has led to the birth of a new scientific field , metagenomics , encompassing large-scale analyses of microbial communities ., Broad sequencing of bacterial populations allows us a first glimpse at the many microbes that cannot be analyzed through traditional means ( only ∼1% of all bacteria can be isolated and independently cultured with current methods 1 ) ., Studies of environmental samples initially focused on targeted sequencing of individual genes , in particular the 16S subunit of ribosomal RNA 2–5 , though more recent studies take advantage of high-throughput shotgun sequencing methods to assess not only the taxonomic composition , but also the functional capacity of a microbial community 6–8 ., Several software tools have been developed in recent years for comparing different environments on the basis of sequence data ., DOTUR 9 , Libshuff 10 , ∫-libshuff 11 , SONs 12 , MEGAN 13 , UniFrac 14 , and TreeClimber 15 all focus on different aspects of such an analysis ., DOTUR clusters sequences into operational taxonomic units ( OTUs ) and provides estimates of the diversity of a microbial population thereby providing a coarse measure for comparing different communities ., SONs extends DOTUR with a statistic for estimating the similarity between two environments , specifically , the fraction of OTUs shared between two communities ., Libshuff and ∫-libshuff provide a hypothesis test ( Cramer von Mises statistics ) for deciding whether two communities are different , and TreeClimber and UniFrac frame this question in a phylogenetic context ., Note that these methods aim to assess whether , rather than how two communities differ ., The latter question is particularly important as we begin to analyze the contribution of the microbiome to human health ., Metagenomic analysis in clinical trials will require information at individual taxonomic levels to guide future experiments and treatments ., For example , we would like to identify bacteria whose presence or absence contributes to human disease and develop antibiotic or probiotic treatments ., This question was first addressed by Rodriguez-Brito et al . 16 , who use bootstrapping to estimate the p-value associated with differences between the abundance of biological subsytems ., More recently , the software MEGAN of Huson et al . 13 provides a graphical interface that allows users to compare the taxonomic composition of different environments ., Note that MEGAN is the only one among the programs mentioned above that can be applied to data other than that obtained from 16S rRNA surveys ., These tools share one common limitation — they are all designed for comparing exactly two samples — therefore have limited applicability in a clinical setting where the goal is to compare two ( or more ) treatment populations each comprising multiple samples ., In this paper , we describe a rigorous statistical approach for detecting differentially abundant features ( taxa , pathways , subsystems , etc . ) between clinical metagenomic datasets ., This method is applicable to both high-throughput metagenomic data and to 16S rRNA surveys ., Our approach extends statistical methods originally developed for microarray analysis ., Specifically , we adapt these methods to discrete count data and correct for sparse counts ., Our research was motivated by the increasing focus of metagenomic projects on clinical applications ( e . g . Human Microbiome Project 17 ) ., Note that a similar problem has been addressed in the context of digital gene expression studies ( e . g . SAGE 18 ) ., Lu et al . 19 employ an overdispersed log-linear model and Robinson and Smyth 20 use a negative binomial distribution in the analysis of multiple SAGE libraries ., Both approaches can be applied to metagenomic datasets ., We compare our tool to these prior methodologies through comprehensive simulations , and demonstrate the performance of our approach by analyzing publicly available datasets , including 16S surveys of human gut microbiota and random sequencing-based functional surveys of infant and mature gut microbiomes and microbial and viral metagenomes ., The methods described in this paper have been implemented as a web server and are also available as free source-code ( in R ) from http://metastats . cbcb . umd . edu ., To account for different levels of sampling across multiple individuals , we convert the raw abundance measure to a fraction representing the relative contribution of each feature to each of the individuals ., This results in a normalized version of the matrix described above , where the cell in the ith row and the jth column ( which we shall denote fij ) is the proportion of taxon i observed in individual j ., We chose this simple normalization procedure because it provides a natural representation of the count data as a relative abundance measure , however other normalization approaches can be used to ensure observed counts are comparable across samples , and we are currently evaluating several such approaches ., For each feature i , we compare its abundance across the two treatment populations by computing a two-sample t statistic ., Specifically , we calculate the mean proportion , and variance of each treatment t from which nt subjects ( columns in the matrix ) were sampled:We then compute the two-sample t statistic:Features whose t statistics exceeds a specified threshold can be inferred to be differentially abundant across the two treatments ( two-sided t-test ) ., The threshold for the t statistic is chosen such as to minimize the number of false positives ( features incorrectly determined to be differentially abundant ) ., Specifically , we try to control the p-value—the likelihood of observing a given t statistic by chance ., Traditional analyses compute the p-value using the t distribution with an appropriate number of degrees of freedom ., However , an implicit assumption of this procedure is that the underlying distribution is normal ., We do not make this assumption , but rather estimate the null distribution of ti non-parametrically using a permutation method as described in Storey and Tibshirani 21 ., This procedure , also known as the nonparametric t-test has been shown to provide accurate estimates of significance when the underlying distributions are non-normal 22 , 23 ., Specifically , we randomly permute the treatment labels of the columns of the abundance matrix and recalculate the t statistics ., Note that the permutation maintains that there are n1 replicates for treatment 1 and n2 replicates for treatment 2 ., Repeating this procedure for B trials , we obtain B sets of t statistics: t10b , … , tM0b , b\u200a=\u200a1 , … , B , where M is the number of rows in the matrix ., For each row ( feature ) , the p-value associated with the observed t statistic is calculated as the fraction of permuted tests with a t statistic greater than or equal to the observed ti:This approach is inadequate for small sample sizes in which there are a limited number of possible permutations of all columns ., As a heuristic , if less than 8 subjects are used in either treatment , we pool all permuted t statistics together into one null distribution and estimate p-values as: Note that the choice of 8 for the cutoff is simply heuristic based on experiments during the implementation of our method ., Our approach is specifically targeted at datasets comprising multiple subjects — for small data-sets approaches such as that proposed by Rodriguez-Brito et . al . 16 might be more appropriate ., Unless explicitly stated , all experiments described below used 1000 permutations ., In general , the number of permutations should be chosen as a function of the significance threshold used in the experiment ., Specifically , a permutation test with B permutations can only estimate p-values as low as 1/B ( in our case 10−3 ) ., In datasets containing many features , larger numbers of permutations are necessary to account for multiple hypothesis testing issues ( further corrections for this case are discussed below ) ., Precision of the p-value calculations is obviously improved by increasing the number of permutations used to approximate the null distribution , at a cost , however , of increased computational time ., For certain distributions , small p-values can be efficiently estimated using a technique called importance sampling ., Specifically , the permutation test is targeted to the tail of the distribution being estimated , leading to a reduction in the number of permutations necessary of up to 95% 24 , 25 ., We intend to implement such an approach in future versions of our software ., For complex environments ( many features/taxa/subsystems ) , the direct application of the t statistic as described can lead to large numbers of false positives ., For example , choosing a p-value threshold of 0 . 05 would result in 50 false positives in a dataset comprising 1000 organisms ., An intuitive correction involves decreasing the p-value cutoff proportional to the number of tests performed ( a Bonferroni correction ) , thereby reducing the number of false positives ., This approach , however , can be too conservative when a large number of tests are performed 21 ., An alternative approach aims to control the false discovery rate ( FDR ) , which is defined as the proportion of false positives within the set of predictions 26 , in contrast to the false positive rate defined as the proportion of false positives within the entire set of tests ., In this context , the significance of a test is measured by a q-value , an individual measure of the FDR for each test ., We compute the q-values using the following algorithm , based on Storey and Tibshirani 21 ., This method assumes that the p-values of truly null tests are uniformly distributed , assumption that holds for the methods used in Metastats ., Given an ordered list of p-values , p ( 1 ) ≤p ( 2 ) ≤…≤p ( m ) , ( where m is the total number of features ) , and a range of values λ\u200a=\u200a0 , 0 . 01 , 0 . 02 , … , 0 . 90 , we computeNext , we fit with a cubic spline with 3 degrees of freedom , which we denote , and let ., Finally , we estimate the q-value corresponding to each ordered p-value ., First , ., Then for i\u200a=\u200am-1 , m-2 , … , 1 , Thus , the hypothesis test with p-value has a corresponding q-value of ., Note that this method yields conservative estimates of the true q-values ,, i . e ., ., Our software provides users with the option to use either p-value or q-value thresholds , irrespective of the complexity of the data ., For low frequency features , e . g . low abundance taxa , the nonparametric t–test described above is not accurate 27 ., We performed several simulations ( data not shown ) to determine the limitations of the nonparametric t-test for sparsely-sampled features ., Correspondingly , our software only applies the test if the total number of observations of a feature in either population is greater than the total number of subjects in the population ( i . e . the average across subjects of the number of observations for a given feature is greater than one ) ., We compare the differential abundance of sparsely-sampled ( rare ) features using Fishers exact test ., Fishers exact test models the sampling process according to a hypergeometric distribution ( sampling without replacement ) ., The frequencies of sparse features within the abundance matrix are pooled to create a 2×2 contingency table ( Figure 2 ) , which acts as input for a two-tailed test ., Using the notation from Figure 2 , the null hypergeometric probability of observing a 2×2 contingency table is: By calculating this probability for a given table , and all tables more extreme than that observed , one can calculate the exact probability of obtaining the original table by chance assuming that the null hypothesis ( i . e . no differential abundance ) is true 27 ., Note that an alternative approach to handling sparse features is proposed in microarray literature ., The Significance Analysis of Microarrays ( SAM ) method 28 addresses low levels of expression using a modified t statistic ., We chose to use Fishers exact test due to the discrete nature of our data , and because prior studies performed in the context of digital gene expression indicate Fishers test to be effective for detection of differential abundance 29 ., The input to our method , the Feature Abundance Matrix , can be easily constructed from both 16S rRNA and random shotgun data using available software packages ., Specifically for 16S taxonomic analysis , tools such as the RDP Bayesian classifier 30 and Greengenes SimRank 31 output easily-parseable information regarding the abundance of each taxonomic unit present in a sample ., As a complementary , unsupervised approach , 16S sequences can be clustered with DOTUR 9 into operational taxonomic units ( OTUs ) ., Abundance data can be easily extracted from the “* . list” file detailing which sequences are members of the same OTU ., Shotgun data can be functionally or taxonomically classified using MEGAN 13 , CARMA 32 , or MG-RAST 33 ., MEGAN and CARMA are both capable of outputting lists of sequences assigned to a taxonomy or functional group ., MG-RAST provides similar information for metabolic subsystems that can be downloaded as a tab-delimited file ., All data-types described above can be easily converted into a Feature Abundance Matrix suitable as input to our method ., In the future we also plan to provide converters for data generated by commonly-used analysis tools ., Human gut 16S rRNA sequences were prepared as described in Eckburg et al . and Ley et al . ( 2006 ) and are available in GenBank , accession numbers: DQ793220-DQ802819 , DQ803048 , DQ803139-DQ810181 , DQ823640-DQ825343 , AY974810-AY986384 ., In our experiments we assigned all 16S sequences to taxa using a naïve Bayesian classifier currently employed by the Ribosomal Database Project II ( RDP ) 30 ., COG profiles of 13 human gut microbiomes were obtained from the supplementary material of Kurokawa et al . 34 ., We acquired metabolic functional profiles of 85 metagenomes from the online supplementary materials of Dinsdale et al . ( 2008 ) ( http://www . theseed . org/DinsdaleSupplementalMaterial/ ) ., As outlined in the introduction , statistical packages developed for the analysis of SAGE data are also applicable to metagenomic datasets ., In order to validate our method , we first designed simulations and compared the results of Metastats to Students t-test ( with pooled variances ) and two methods used for SAGE data: a log-linear model ( Log-t ) by Lu et al . 19 , and a negative binomial ( NB ) model developed by Robinson and Smyth 20 ., We designed a metagenomic simulation study in which ten subjects are drawn from two groups - the sampling depth of each subject was determined by random sampling from a uniform distribution between 200 and 1000 ( these depths are reasonable for metagenomic studies ) ., Given a population mean proportion p and a dispersion value φ , we sample sequences from a beta-binomial distribution Β ( α , β ) , where α\u200a=\u200ap ( 1/φ−1 ) and β\u200a= ( 1−p ) ( 1/φ−1 ) ., Note that data from this sampling procedure fits the assumptions for Lu et al . as well as Robinson and Smyth and therefore we expect them to do well under these conditions ., Lu et al . designed a similar study for SAGE data , however , for each simulation , a fixed dispersion was used for both populations and the dispersion estimates were remarkably small ( φ\u200a=\u200a0 , 8e-06 , 2e-05 , 4 . 3e-05 ) ., Though these values may be reasonable for SAGE data , we found that they do not accurately model metagenomic data ., Figure 3 displays estimated dispersions within each population for all features of the metagenomic datasets examined below ., Dispersion estimates range from 1e-07 to 0 . 17 , and rarely do the two populations share a common dispersion ., Thus we designed our simulation so that φ is chosen for each population randomly from a uniform distribution between 1e-08 and 0 . 05 , allowing for potential significant differences between population distributions ., For each set of parameters , we simulated 1000 feature counts , 500 of which are generated under p1\u200a=\u200ap2 , the remainder are differentially abundant where a*p1\u200a=\u200ap2 , and compared the performance of each method using receiver-operating-characteristic ( ROC ) curves ., Figure 4 displays the ROC results for a range of values for p and a ., For each set of parameters , Metastats was run using 5000 permutations to compute p-values ., Metastats performs as well as other methods , and in some cases is preferable ., We also found that in most cases our method was more sensitive than the negative binomial model , which performed poorly for high abundance features ., Our next simulation sought to examine the accuracy of each method under extreme sparse sampling ., As shown in the datasets below , it is often the case that a feature may not have any observations in one population , and so it is essential to employ a statistical method that can address this frequent characteristic of metagenomic data ., Under the same assumptions as the simulation above , we tested a\u200a=\u200a0 and 0 . 01 , thereby significantly reducing observations of a feature in one of the populations ., The ROC curves presented in Figure 5 reveal that Metastats outperforms other statistical methods in the face of extreme sparseness ., Holding the false positive rate ( x-axis ) constant , Metastats shows increased sensitivity over all other methods ., The poor performance of Log-t is noteworthy given it is designed for SAGE data that is also potentially sparse ., Further investigation revealed that the Log-t method results in a highly inflated dispersion value if there are no observations in one population , thereby reducing the estimated significance of the test ., Finally , we selected a subset of the Dinsdale et al . 6 metagenomic subsystem data ( described below ) , and randomly assigned each subject to one of two populations ( 20 subjects per population ) ., All subjects were actually from the same population ( microbial metagenomes ) , thus the null hypothesis is true for each feature tested ( no feature is differentially abundant ) ., We ran each methodology on this data , recording computed p-values for each feature ., Repeating this procedure 200 times , we simulated tests of 5200 null features ., Table 1 displays the number of false positives incurred by each methodology given different p-value thresholds ., The results indicate that the negative binomial model results in an exceptionally high number of false positives relative to the other methodologies ., Students t-test and Metastats perform equally well in estimating the significance of these null features , while Log-t performs slightly better ., These studies show that Metastats consistently performs as well as all other applicable methodologies for deeply-sampled features , and outperforms these methodologies on sparse data ., Below we further evaluate the performance of Metastats on several real metagenomic datasets ., In a recent study , Ley et al . 35 identified gut microbes associated with obesity in humans and concluded that obesity has a microbial element , specifically that Firmicutes and Bacteroidetes are bacterial divisions differentially abundant between lean and obese humans ., Obese subjects had a significantly higher relative abundance of Firmicutes and a lower relative abundance of Bacteriodetes than the lean subjects ., Furthermore , obese subjects were placed on a calorie-restricted diet for one year , after which the subjects gut microbiota more closely resembled that of the lean individuals ., We obtained the 20 , 609 16S rRNA genes sequenced in Ley et al . and assigned them to taxa at different levels of resolution ( note that 2 , 261 of the 16S sequences came from a previous study 36 ) ., We initially sought to re-establish the primary result from this paper using our methodology ., Table 2 illustrates that our method agreed with the results of the original study: Firmicutes are significantly more abundant in obese subjects ( P\u200a=\u200a0 . 003 ) and Bacteroidetes are significantly more abundant in the lean population ( P<0 . 001 ) ., Furthermore , our method also detected Actinobacteria to be differentially abundant , a result not reported by the original study ., Approximately 5% of the sample was composed of Actinobacteria in obese subjects and was significantly less frequent in lean subjects ( P\u200a=\u200a0 . 004 ) ., Collinsella and Eggerthella were the most prevalent Actinobacterial genera observed , both of which were overabundant in obese subjects ., These organisms are known to ferment sugars into various fatty acids 37 , further strengthening a possible connection to obesity ., Note that the original study used Students t-test , leading to a p-value for the observed difference within Actinobacteria of 0 . 037 , 9 times larger than our calculation ., This highlights the sensitivity of our method and explains why this difference was not originally detected ., To explore whether we could refine the broad conclusions of the initial study , we re-analyzed the data at more detailed taxonomic levels ., We identified three classes of organisms that were differentially abundant: Clostridia ( P\u200a=\u200a0 . 005 ) , Bacteroidetes ( P<0 . 001 ) , and Actinobacteria ( P\u200a=\u200a0 . 003 ) ., These three were the dominant members of the corresponding phyla ( Firmicutes , Bacteroides , Actinobacteria , respectively ) and followed the same distribution as observed at a coarser level ., Metastats also detected nine differentially abundant genera accounting for more than 25% of the 16S sequences sampled in both populations ( P≤0 . 01 ) ., Syntrophococcus , Ruminococcus , and Collinsella were all enriched in obese subjects , while Bacteroides on average were eight times more abundant in lean subjects ., For taxa with several observations in each subject , we found good concordance between our results ( p-value estimates ) and those obtained with most of the other methods ( Table 2 ) ., Surprisingly , we found that the negative binomial model of Robinson and Smyth failed to detect several strongly differentially abundant features in these datasets ( e . g the hypothesis test for Firmicutes results in a p-value of 0 . 87 ) ., This may be due in part to difficulties in estimating the parameters of their model for our datasets and further strengthens the case for the design of methods specifically tuned to the characteristics of metagenomic data ., For cases where a particular taxon had no observations in one population ( e . g . Terasakiella ) , the methods proposed for SAGE data seem to perform poorly ., Targeted sequencing of the 16S rRNA can only provide an overview of the diversity within a microbial community but cannot provide any information about the functional roles of members of this community ., Random shotgun sequencing of environments can provide a glimpse at the functional complexity encoded in the genes of organisms within the environment ., One method for defining the functional capacity of an environment is to map shotgun sequences to homologous sequences with known function ., This strategy was used by Kurokawa et al . 34 to identify clusters of orthologous groups ( COGs ) in the gut microbiomes of 13 individuals , including four unweaned infants ., We examined the COGs determined by this study across all subjects and used Metastats to discover differentially abundant COGs between infants and mature ( >1 year old ) gut microbiomes ., This is the first direct comparison of these two populations as the original study only compared each population to a reference database to find enriched gene sets ., Due to the high number of features ( 3868 COGs ) tested for this dataset and the limited number of infant subjects available , our method used the pooling option to compute p-values ( we chose 100 permutations ) , and subsequently computed q-values for each feature ., Using a threshold of Q≤0 . 05 ( controlling the false discovery rate to 5% ) , we detected 192 COGs that were differentially abundant between these two populations ( see Table 3 for a lisitng of the most abundant COGs in both mature and infant microbiomes . Full results are presented as supplementary material in Table S1 ) ., The most abundant enriched COGs in mature subjects included signal transduction histidine kinase ( COG0642 ) , outer membrane receptor proteins , such as Fe transport ( COG1629 ) , and Beta-galactosidase/beta-glucuronidase ( COG3250 ) ., These COGs were also quite abundant in infants , but depleted relative to mature subjects ., Infants maintained enriched COGs related to sugar transport systems ( COG1129 ) and transcriptional regulation ( COG1475 ) ., This over-abundance of sugar transport functions was also found in the original study , strengthening the hypothesis that the unweaned infant gut microbiome is specifically designed for the digestion of simple sugars found in breast milk ., Similarly , the depletion of Fe transport proteins in infants may be associated with the low concentration of iron in breast milk relative to cows milk 38 ., Despite this low concentration , infant absorption of iron from breast milk is remarkably high , and becomes poorer when infants are weaned , indicating an alternative mechanism for uptake of this mineral ., The potential for a different mechanism is supported by the detection of a Ferredoxin-like protein ( COG2440 ) that was 11 times more abundant in infants than in mature subjects , while Ferredoxin ( COG1145 ) was significantly enriched in mature subjects ., A recent study by Dinsdale et al . profiled 87 different metagenomic shotgun samples ( ∼15 million sequences ) using the SEED platform ( http://www . theseed . org ) 6 to see if biogeochemical conditions correlate with metagenome characteristics ., We obtained functional profiles from 45 microbial and 40 viral metagenomes analyzed in this study ., Within the 26 subsystems ( abstract functional roles ) analyzed in the Dinsdale et al . study , we found 13 to be significantly different ( P≤0 . 05 ) between the microbial and viral samples ( Table 4 ) ., Subsystems for RNA and DNA metabolism were significantly more abundant in viral metagenomes , while nitrogen metabolism , membrane transport , and carbohydrates were all enriched in microbial communities ., The high levels of RNA and DNA metabolism in viral metagenomes illustrate their need for a self-sufficient source of nucleotides ., Though the differences described by the original study did not include estimates of significance , our results largely agreed with the authors qualitative conclusions ., However , due to the continuously updated annotations in the SEED database since the initial publication , we found several differences between our results and those originally reported ., In particular we found virulence subsystems to be less abundant overall than previously reported , and could not find any significant differences in their abundance between the microbial and viral metagenomes ., We have presented a statistical method for handling frequency data to detect differentially abundant features between two populations ., This method can be applied to the analysis of any count data generated through molecular methods , including random shotgun sequencing of environmental samples , targeted sequencing of specific genes in a metagenomic sample , digital gene expression surveys ( e . g . SAGE 29 ) , or even whole-genome shotgun data ( e . g . comparing the depth of sequencing coverage across assembled genes ) ., Comparisons on both simulated and real dataset indicate that the performance of our software is comparable to other statistical approaches when applied to well- sampled datasets , and outperforms these methods on sparse data ., Our method can also be generalized to experiments with more than two populations by substituting the t-test with a one-way ANOVA test ., Furthermore , if only a single sample from each treatment is available , a chi-squared test can be used instead of the t-test ., 27 ., In the coming years metagenomic studies will increasingly be applied in a clinical setting , requiring new algorithms and software tools to be developed that can exploit data from hundreds to thousands of patients ., The methods described above represent an initial step in this direction by providing a robust and rigorous statistical method for identifying organisms and other features whose differential abundance correlates with disease ., These methods , associated source code , and a web interface to our tools are freely available at http://metastats . cbcb . umd . edu . | Introduction, Materials and Methods, Results, Discussion | Numerous studies are currently underway to characterize the microbial communities inhabiting our world ., These studies aim to dramatically expand our understanding of the microbial biosphere and , more importantly , hope to reveal the secrets of the complex symbiotic relationship between us and our commensal bacterial microflora ., An important prerequisite for such discoveries are computational tools that are able to rapidly and accurately compare large datasets generated from complex bacterial communities to identify features that distinguish them ., We present a statistical method for comparing clinical metagenomic samples from two treatment populations on the basis of count data ( e . g . as obtained through sequencing ) to detect differentially abundant features ., Our method , Metastats , employs the false discovery rate to improve specificity in high-complexity environments , and separately handles sparsely-sampled features using Fishers exact test ., Under a variety of simulations , we show that Metastats performs well compared to previously used methods , and significantly outperforms other methods for features with sparse counts ., We demonstrate the utility of our method on several datasets including a 16S rRNA survey of obese and lean human gut microbiomes , COG functional profiles of infant and mature gut microbiomes , and bacterial and viral metabolic subsystem data inferred from random sequencing of 85 metagenomes ., The application of our method to the obesity dataset reveals differences between obese and lean subjects not reported in the original study ., For the COG and subsystem datasets , we provide the first statistically rigorous assessment of the differences between these populations ., The methods described in this paper are the first to address clinical metagenomic datasets comprising samples from multiple subjects ., Our methods are robust across datasets of varied complexity and sampling level ., While designed for metagenomic applications , our software can also be applied to digital gene expression studies ( e . g . SAGE ) ., A web server implementation of our methods and freely available source code can be found at http://metastats . cbcb . umd . edu/ . | The emerging field of metagenomics aims to understand the structure and function of microbial communities solely through DNA analysis ., Current metagenomics studies comparing communities resemble large-scale clinical trials with multiple subjects from two general populations ( e . g . sick and healthy ) ., To improve analyses of this type of experimental data , we developed a statistical methodology for detecting differentially abundant features between microbial communities , that is , features that are enriched or depleted in one population versus another ., We show our methods are applicable to various metagenomic data ranging from taxonomic information to functional annotations ., We also provide an assessment of taxonomic differences in gut microbiota between lean and obese humans , as well as differences between the functional capacities of mature and infant gut microbiomes , and those of microbial and viral metagenomes ., Our methods are the first to statistically address differential abundance in comparative metagenomics studies with multiple subjects , and we hope will give researchers a more complete picture of how exactly two environments differ . | computational biology/metagenomics | null |
journal.pgen.1000895 | 2,010 | Candidate Causal Regulatory Effects by Integration of Expression QTLs with Complex Trait Genetic Associations | The biological interpretation of genome-wide association study ( GWAS ) signals 1–5 is very challenging since most candidate loci fall either in gene deserts or in regions with many equally plausible causative genes ., Following the concurrent progress in understanding the genetic basis of regulatory variation 6–9 , differential gene expression has been proposed as a promising intermediate layer of information 10 to aid this interpretation 11 ., Most commonly , interrogating the GWAS SNPs themselves for significant associations with gene expression 12–13 has been employed to explain some of the GWAS results ., However , the ubiquity of regulatory variation throughout the human genome 6 , 14 makes coincidental overlaps of eQTLs and complex trait loci very likely ., This likelihood is a direct consequence of the correlation structure in the genome ( linkage disequilibrium - LD ) , which makes functionally unrelated variants statistically correlated ., As sample sizes increase , allowing the discovery of larger numbers of eQTLs of smaller effect size and as the expression experiments will be performed in a larger variety of tissues , we can envisage that almost every gene will have an associated eQTL under a certain condition ., Consequently , the probability that any of these will map to a genomic region where a GWAS SNP also resides is very high ., Therefore , it is important to emphasize that while it is very tempting to infer potential causal mechanisms based on such overlaps , this would be a naïve inference in the absence of additional supporting evidence for causality ., In the long run , this will not only be an issue for gene expression , but also for any other cellular phenotype ., Association studies for intermediate phenotypes with possible relevance to complex traits are underway and their results will overlap some of the GWAS signals ., The biological meaning of these overlaps will again need to be evaluated in the context of the genomes correlation structure ., It is not evident though how to model each genomic region with overlapping association signals in the absence of information about the history of the region ., Accounting for the historical parameters of a region under the coalescent , while desirable , is computationally and practically not feasible since the human population history is too complex to properly model and small deviations or slightly incorrect assumptions could create false signals or reduce power ., In order to distinguish such accidental colocalizations 15–16 from true sharing of causal variants , we propose here an empirical methodology instead ., This directly combines eQTL and GWAS data while accounting for the LD of the region harbouring the GWAS SNP ., We demonstrate the value of the approach by predicting the regulatory impact of several GWAS variants in cis and trans and we also show that the correlation strength ( r2 , D ) between the GWAS SNP and the eQTL is not a sufficient predictor of regulatory mediated disease effects ., To identify likely causal effects ( not variants since we do not have full sequencing data ) associated with complex traits and diseases we took advantage of published association data catalogued in the NHGRI 17 database and gene expression data generated in LCLs derived from HapMap 3 individuals ( see Methods ) ., In this study , we limited the expression analysis to the 109 CEU individuals , as they are the closest in ancestry to the majority of individuals in published GWAS studies ., We used the NHGRI database ( accessed 02 . 03 . 09 ) to extract 976 GWAS SNPs with minor allele frequency ( MAF ) >5% that were also genotyped in the HapMap 3 CEU , thus allowing to test the exact GWAS SNPs for associations with differential gene expression in LCLs ., In total we examined 17673 genes ., In order to discover eQTLs , we used Spearman Rank Correlation ( SRC ) ., This method 14 captures the vast majority of associations discovered with standard linear regression ( LR ) models , with the additional advantage that its not affected by outliers and hence has more power and allows direct comparison of nominal P-values ., We looked for both proximal ( cis ) and distal ( trans ) effects as follows: variants within 1Mb on either side of the transcription start site ( TSS ) of a gene are considered to be acting in cis , while those at least 5 Mb downstream or upstream of the TSS or on a different chromosome are considered to be acting in trans ., In order to assess the overall impact of the currently known GWAS SNPs on expression , we contrasted their cis and trans effects to those of a random set of SNPs , representing the null ., In a QQ plot ( Figure 1 ) , we compare the distributions of the best cis and trans association p-values per SNP for the 976 GWAS SNPs ( observed ) to 1000 sets of most significant p-values of 976 random SNPs each ( expected ) ., The 1000 random sets of 976 SNPs were sampled to have identical MAF distribution to the GWAS SNPs ., In cis , we observe a much stronger regulatory signal in the GWAS data compared to random ( Figure 1 ) ., The significant difference between the two becomes apparent above a −log10 ( P-value ) =\u200a4 ., In trans , we also detect a more significant regulatory signal for GWAS SNPs compared to random , however not as strong as in cis ., This is to be expected given that the much greater statistical space were exploring in trans limits the power to detect such effects ., Nevertheless , despite their confinement to one tissue type - LCLs , these comparisons support the overall explanatory potential of regulatory variation for the biological effects of GWAS variants ., As expected given the tissue nature , the phenotypes responsible for this enrichment are immunity related ( Figure 2 ) ., To identify the subset of causal effects from the regulatory enrichment observed , we focused only on the genomic regions harbouring either cis or trans eQTLs ., We split the genome into recombination hotspot intervals based on genome-wide estimates of hotspot coordinates from McVean et . al . 18 Limiting the search space for causal effects to these intervals is a reasonable conventional approach , as few or no recombination events are expected between the reported associated SNPs and the functional variants they are tagging ., Given the abundance of cis eQTLs in the genome , mere interval overlap in not sufficient to claim that a colocalized cis eQTL and a GWAS SNP are tagging the same functional variant ., However , if the GWAS SNP and the eQTL do tag the same causal SNP , we expect that removing the genetic effect of the GWAS SNP will have a marked consequence on the eQTL association ., Starting from this hypothesis , we developed an empirical method to uncover regulatory mediated associations with complex traits ., For all genes with a significant cis eQTL ( 0 . 05 permutation threshold as defined in Stranger et . al . 2007 , see Methods ) in a given interval , we create corrected phenotypes from the residuals of the standard LR of the GWAS SNP against normalized expression values of the gene for which we have an eQTL ., The residuals capture the remaining unexplained expression variance after the removal of the GWAS SNP effect ., We redo the SRC analysis with the pseudo phenotype and retain the adjusted association P-value ., Depending on the internal LD structure of the hotspot interval , the correlation between the GWAS SNP and the eQTL will vary , hence so will the P-values after and before correction ., One way to assess the relevance of the GWAS SNP to the eQTL is to compare its correction impact to that of all other SNPs in the interval ., For this purpose , we define a Regulatory Trait Concordance ( RTC ) Score for each gene-GWAS SNP combination as follows , taking into account the ranking of the correction with respect to all SNPs in the interval ( RankGWAS SNP ) and the total number of tested SNPs ( NSNPs ) ., The rank denotes the number of SNPs which when used to correct the expression data , have a higher impact on the eQTL ( smaller adjusted P-value ) than the GWAS SNP ( i . e . RankGWAS SNP\u200a=\u200a0 if the GWAS SNP is the same as the eQTL SNP , RankGWAS SNP\u200a=\u200a1 if of all the SNPs in the interval , the GWAS SNP has the largest impact on the eQTL ) ., Given this , the RTC Score will always be in the range ( 0 , 1 , with values close to 1 indicating that the GWAS effect is the same as the eQTL effect ., We investigated the properties and robustness of the RTC score under the null hypothesis ( H0: eQTL and GWAS are tagging two different causal SNPs ) and the alternative hypothesis ( H1: same causal SNP ) ., For this purpose , we have simulated causal SNPs ( cSNP ) , eQTLs and dSNPs ( see Methods ) varying the LD levels between them as well as the LD pattern of the hotspot interval where they reside ., We have then masked the cSNPs and calculated the RTC score under these different LD scenarios for both hypotheses ., The RTC score is uniformly distributed under the null , when the simulated causal eQTL SNP ( c-eQTL ) and the causal disease SNP ( c-dSNP ) are different ( Figure 3 , left panel ) ., Under the H1 on the other hand , the RTC score is right skewed , with a clear enrichment for values close to 1 recovering the single causal SNP effect ( Figure 3 , middle panel ) ., The simulations show that the complexity and variability of the LD structure in the genome impede the simple use of correlation metrics to infer shared causal effects ., The statistical correlation ( r2 ) between the eQTL and the dSNP is not on its own sufficient to predict whether they tag the same cSNP ( Figure 4 ) ., The RTC outperforms r2 as it is able to recover causal effects even for low correlated pairs ., The historical correlation metric between eQTLs and dSNPs ( D ) is also not fully predictive of high RTC scores ( Figure 5 ) ., We observe from the H0 simulation results that D is not correlated with RTC , meaning that when the eQTL and dSNP tag different functional variants , the RTC score is not high just because D is high ., In addition , while high RTC scoring cases cluster much tighter around high D values under the H1 compared to r2 previously , a high D is not sufficient to predict causal effects ., That is because it would be impossible to distinguish causal from coincidental effects given a perfect historical correlation scenario ., Finally , we investigated the effect of the overall LD pattern in a region of interest on the RTC ., For this purpose , we calculated the median r2 of each hotspot interval and checked its relationship to the RTC score under the null and alternative hypothesis ., It is expected that RTC will perform better in intervals with overall low LD , where the correlation between the eQTL and other non-disease SNPs will decay much faster , making the correction for the dSNP stand out ., However , we confirm that the LD of the region does not determine high scores by itself ., Intervals of low LD where different c-eQTLs and c-dSNPs reside have a uniform distribution of RTC scores ( Figure 6 , left panel ) ., As expected , we do observe from the H1 simulations that we have most power in intervals with low median r2 ( Figure 6 , right panel ) ., As a positive control , we tested the method first on intervals harbouring already identified regulatory associations ., We used published cis eQTLs ( 10−3 permutation threshold ) discovered in the same tissue as the HapMap CEU eQTLs ( LCLs ) but derived from an independent set of samples: 75 individuals of Western European origin from the GenCord project 19 ., In this experiment , we considered the GenCord eQTLs as the equivalent of GWAS SNPs and we limited our analysis to intervals with cis eQTLs in both datasets ., Furthermore , we conditioned the associated genes for the same interval to be identical in the two expression datasets , expecting thus a common functional variant ., As a result of this filtering , we tested SNPs in 157 hotspot intervals , associated with differential expression levels of 154 genes ., As expected from the H1 simulations , the RTC Score distribution after correcting for the GenCord eQTLs is right-skewed ( Figure 3 , right panel ) , suggesting that the scoring method is sensitive to associations tagging the same functional variant ., We detect 33 SNP-probe pairs with an RTC Score of 1 out of the total 185 tested pairs ., Given the marked difference in genotyping density between HapMap and GenCord ( ∼1 . 2 millionSNPs versus ∼400 , 000 SNPs respectively ) and our hypothesis that the 157 overlapping intervals share the same functional variant , we expect approximately 3 times more perfect scoring cases ( 99 pairs with RTC Score\u200a=\u200a1 ) than what we observe , had individuals from both datasets been equally densely genotyped ., We use the degree of sharing between the eQTLs in the two datasets to derive a reasonable , yet conservative threshold: currently , 105 SNP-probe pairs pass the 0 . 9 RTC threshold , making it thus a suitable stringent cut-off for calling significant discoveries ., We then applied the scoring method on the NHGRI GWAS SNPs ., The 976 common GWAS SNPs map to 784 hotspot intervals ., Of these , we focused the cis analysis on GWAS intervals ( N\u200a=\u200a130 ) where at least one significant cis eQTL at a 0 . 05 permutation P-value threshold also resides ( Dataset S1 ) ., For the trans analysis , we ordered all 784 GWAS intervals by their most significant trans eQTL and kept the topmost 50 intervals for further examination ( Dataset S2 ) ., Table 1 summarizes our most confident cis results ordered by RTC Score ., We detect SNP-gene combinations passing the 0 . 9 threshold for 28 intervals out of the 130 , twice as many than expected by chance ( 13 expected top 10% scoring intervals under the uniform distribution ) ., Our method confirms prior results in the literature suggestive of disease effects mediated through expression ( ORMDL3 for asthma risk 13 , C8orf13 locus for lupus risk 20 , SLC22A5 for Crohns disease 12 , 21 ) ., In addition , we detect several other yet unknown candidate genes for a variety of conditions ., An interesting example of a novel cis regulatory mediated effect is the one for Crohns disease with gene SLC38A3 , member 3 of the solute carrier family 38 ., Independent studies detected significant Crohns associations of two SNPs in the same hotspot interval on chromosome 3 ( rs3197999 12 , a non-synonymous SNP in gene MST1 and rs9858542 1 , 22 , a synonymous SNP in nearby gene BSN ) ., Suggestive literature evidence in addition to the disease associated non-synonymous SNP made MST1 the most attractive candidate gene out of the many present in that region 23 ., However , our data supports an additional regulatory component underlying the susceptibility locus ., For both GWAS SNPs , SLC38A3 is the highest scoring candidate in the region ( RTC Score: 0 . 92 ) ., Interestingly , this is functionally similar to another Crohns susceptibility gene SLC22A5 confirmed with our method ( RTC Score: 1 . 0 ) and also encoding a sodium dependent multi-pass membrane protein ( solute carrier family protein ) ., The observed direction of effect is the same for both genes ( eQTLs associate with low expression levels ) as in previous expression datasets 12 and suggests a possible involvement of this gene family in the disease ., This is in agreement with recent studies reporting that disease causative genes are functionally more closely related 24 ., The tissue under investigation is LCLs so we expect GWAS signals of immunity related traits ( comprising here autoimmune disorders and diseases of the immune system e . g . AIDS progression ) to more likely show an overlap with eQTLs ., In order to evaluate the relevance of our results , we analyzed the distributions of the best RTC Scores per GWAS SNP stratified by the immunity relatedness of the complex trait they associate with ( Figure 7 ) ., We observe a significant overrepresentation of high-scoring genes ( >\u200a=\u200a0 . 9 ) for immunity related traits compared to non-immunity related ones ( Fishers Exact Test , P-value\u200a=\u200a0 . 0125 ) 25 ., This suggests that the scoring scheme predicts regulatory effects of the relevant phenotypes ., In addition , we observed that for GWAS signals with RTC score >0 . 9 , only 10% of the nearest gene to the GWAS SNP was also the eQTL gene ., These however , correspond as expected to instances when the eQTL gene is also the nearest gene to the eQTL itself ., If that is not the case , the inference of relevance of a gene simply based on its proximity to the GWAS SNP is not informative ., Even if the causal SNP is not cis-regulatory , using gene expression to determine its downstream targets , coupled with information about the biological pathways these targets act in could help interpret the primary GWAS effect ., We investigated this hypothesis in the topmost 50 GWAS intervals ordered by their trans eQTL significance ., For each interval , we apply the RTC Scoring scheme on the subset of genes in the whole genome with a notable effect in trans ( SRC nominal P-value <10−5 ) ., These signals amount to a total of 552 genes ., We obtain SNP-gene combinations passing the 0 . 9 Score threshold for 24 of the 50 tested intervals ( corresponding to a total of 85 genes ) ., Six of these intervals contain GWAS SNPs associated with immunity related traits ( Table 2 ) ., While not statistically significant - unsurprisingly given that were only testing a small subset of the total GWAS intervals - these examples support the usefulness of the trans approach ., As hypothesized , for the same complex trait associated SNP we can discover several potential candidate genes in trans , throughout the genome ., Some of these are biologically plausible results and merit further investigation ., However , many trans candidates are hard to interpret at this stage given their incomplete annotation and further functional studies will need to be performed for validation ., The power to detect significant associations between genotyped SNP proxies and a phenotype depends on the correlation between those proxies and the functional variant 26 ., Just like for the simulated data , we tested whether the correlation between a GWAS SNP and its colocalizing eQTL is sufficient for predicting a shared causal effect ., For both the cis and the trans analysis , we observe that the r2 between the eQTL and the disease SNP is not a direct predictor of the RTC Score , and in several cases we predict that even pairs with low r2 are likely tagging the same functional effect ( Figure 8 , top panel ) ., The reason for this is that many of the high scoring pairs with poor statistical correlation ( low r2 ) are actually historically correlated ( D\u200a=\u200a1 ) ., Nevertheless , D is not very informative either ( Figure 8 , bottom panel ) , the main problem here being that in regions with generally high D among many SNPs , one cannot determine which of the pairs actually represents a common functional variant ., Another metric of potential predictive value is the fraction of eQTL variance explained by the dSNP ., Figure 9 indicates the relationship between the RTC score and the fraction of explained variance at the eQTL left unexplained after the dSNP correction ( ratio of linear regression adjusted R∧2 after and before correction ) ., As expected given the definition of the RTC , the highest density of good scoring results is registered for dSNPs that explain most of the eQTL variance ., However , RTC outperforms the variance metric , scoring high even when thats not the case and thus making the setting of a threshold on the explained variance not sufficiently informative either ., To aid the functional interpretation of complex trait association signals , we describe here an empirical methodology that directly integrates eQTL and GWAS data while correcting for the local correlation structure in the human genome ., As regulatory variants are pervasive throughout the genome , coincidental overlaps of eQTLs and GWAS SNPs are very likely ., Hence , current methods that limit themselves to asking whether disease intervals also harbour eQTLs are unreliable for distinguishing trait relevant regulatory effects from other eQTLs ., Our methodology addresses and helps resolve this issue ., This approach is not limited to gene expression , but could be generalized to any other phenotype ., As new methods are developed and larger cohorts become available , various intermediate cellular phenotypes are interrogated via association studies with the hope to find explanatory links between genotypic variation and complex trait predisposition ., However , the biological interpretation of these discoveries will also be hardened by the presence of tight LD ., It is therefore necessary to evaluate them in a conservative manner , correcting for the local correlation structure in each genomic interval with overlapping association signals ., In this paper , we discover causal regulatory effects and their affected candidate genes in cis and to some extent in trans by assessing the impact on the expression phenotype of the removal of the GWAS SNP effect ., We compute a score ( RTC ) for each individual genomic interval that assesses the likelihood that the eQTL and the GWAS SNP are tagging the same functional variant ., By ranking the effect of the removal of the GWAS SNP in comparison to the outcome for any other SNP in the region and by accounting for the number of SNPs tested , we produce a score comparable across intervals ., We evaluate the performance of the score in various simulated LD scenarios and we present its robustness by its expected uniform distribution when the eQTL and GWAS SNP are tagging different functional variants ., In comparison , we investigate how well do current SNP correlation metrics ( r2 , D ) perform on their own ., We show that the LD between the GWAS SNP and its colocalizing eQTL is not a good predictor of a shared functional effect ., This is very important especially since most of the current replication and follow-up studies only focus on variants highly correlated ( r2>0 . 8 ) with the initial discoveries ., It is important to stress at this point that neither the eQTL nor the GWAS SNP is likely the causal variant ., Therefore , what really matters is not the statistical correlation between two proxies but the correlation between each of the proxies and the causal variant , whose frequency is unknown ., In any case , no obvious combination of LD measures can substitute the RTC scoring scheme and we thus conclude that many interesting candidate genes would be missed if one were to rely solely on correlation-based approaches ., In this paper , we also explore the explanatory potential of regulatory variation given the currently published GWAS data ., We observe a significant overrepresentation of eQTLs among GWAS SNPs , especially affecting genes in cis ., Long-range trans effects are also present but less prevalent , possibly due to lower power to detect such associations ., As expected given the tissue the expression data was measured in ( LCLs ) , we observe a significant abundance of cis regulatory causal effects for immunity related traits ., Our result reinforces the necessity to expand the tissue diversity 27–28 of genome-wide expression studies in order to facilitate such discoveries for a wider range of human conditions ., By applying the RTC method on the NHGRI GWAS SNPs , we are able to confirm previously suspected regulatory mediated disease effects and discover novel candidate genes affected by GWAS SNPs ., We provide a list of follow-up candidate genes affected in cis and in addition , we show the utility of genome-wide expression data irrespective of the nature of the primary SNP effect by predicting clusters of genes affected in trans ., The individual examination of the candidates prioritized with our approach will undoubtedly assist the biological interpretation of the ever-increasing list of GWAS signals ., As associations with more intermediate cellular phenotypes will be reported , the integration of all these signals will be crucial for understanding the biology of complex traits ., RNA levels were measured in lymphoblastoid cell lines ( LCLs ) derived from the HapMap 3 individuals using a whole-genome expression array ( Illumina Sentrix WG-6 , Version 2 ) as previously described 14 ., Each sample had two technical replicates ., We analyzed here only expression data from the CEU , a HapMap 3 population of 109 unrelated individuals of Northern European ancestry ., The mapping of Illumina probes to unique Ensembl gene IDs resulted in 21 , 811 probes corresponding to 17 , 673 autosomal genes available for association analysis ., 1 , 186 , 075 SNPs ( MAF >5% ) genotyped in the same individuals were used in the eQTL analysis ., The log2 transformed raw intensity values were normalized as follows: quantile normalization of sample replicates ( two intensity values per Illumina probe ) followed by median normalization across all individuals ., All SNPs from the catalogue of genome-wide association studies maintained by the National Human Genome Research Institute ( NHGRI www . genome . gov/26525384 ) and published by 02 . 03 . 2009 were downloaded ., Of these , only the 976 unique common variants ( MAF >5% ) genotyped in the HapMap 3 CEU samples were kept for analysis ., Associations between SNP genotypes and normalized expression values were conducted using Spearman Rank Correlation ( SRC ) ., For the cis analysis , we considered only SNPs within a 1MB window from the TSS of genes , while in trans we test all SNPs further than 5MB away from the genes TSS and all SNP-gene pairs on different chromosomes ., We assess the statistical significance of the cis associations using permutations as previously described 7 , 14 ., We call a cis eQTL significant if the nominal association P-value is greater than the 0 . 01 tail of the minimal P-value distribution resulting from the SNPs associations with 10 , 000 permuted sets of expression values for each gene ., We mapped all common autosomal CEU HapMap 3 SNPs ( 1 , 186 , 075 SNPs ) to recombination hotspot intervals as defined by McVean et . al . 18 For the cis analysis we selected the 130 hotspot intervals where at least one significant cis eQTL and a GWAS SNP colocalize while for the trans , we analyzed a subset of 50 of the total 784 unique intervals ( where the 976 GWAS SNPs map to ) ., These are the topmost intervals ordered by their most significant trans eQTL ( nominal SRC P-value ) ., For both the cis and trans GWAS analysis , the best P-value associations per SNP were stored ., The set of the most significant P-values of the 976 GWAS SNPs was compared to 1000 sets of most significant P-values of 976 random SNPs ., The 1000 random sets of 976 SNPs each were conditioned to have the same MAF distribution as the 976 GWAS set ., The QQ plot showing the abundance of regulatory signal in GWAS data is the median QQ plot of 1000 ( GWAS , random SNPs ) comparisons ., It shows the distribution of the −log10 quantile values of the GWAS best associations ( observed ) versus the median of the corresponding 1000 −log10 quantile values from each of the 1000 random SNP sets ( expected ) ., In order to assess the significance of the observed versus expected median QQ plot , we superimpose the upper limit of the 95% confidence interval ., This is calculated from the sorted 0 . 95 quantiles of 10000 pairs of 976 random SNPs each ., We assess the likelihood of a shared functional effect between a GWAS SNP and an eQTL by quantifying the change in the statistical significance of the eQTL after correcting for the genetic effect of the GWAS SNP ., We redo the SRC association of the eQTL genotype with the residuals from the standard LR of the “corrected-for” SNP against normalized expression values ., We account for the LD structure in each hotspot interval separately by ranking ( RankGWAS SNP ) the impact on the eQTL ( quantified by the adjusted association P-value after correction ) of the GWAS SNP correction to that of correcting for all other SNPs in the same interval ., By taking into account the total number of SNPs in the interval ( NSNPs ) , we can compare this ranking across different genes and intervals ., For this purpose we define the regulatory trait concordance ( RTC ) Score ranked below ranging from 0 to 1 , with values closer to 1 indicating causal regulatory effects ., We investigate the properties of the RTC score with respect to different correlation metrics under the null hypothesis ( H0: eQTL and dSNP tag different functional variants ) and the alternative hypothesis ( H1: eQTL and dSNP tag the same functional variant ) ., We use the HapMap3 CEU cis eQTLs ( 315 genes at 10−3 permutation threshold ) to create a list of causal SNPs ( cSNP ) ., For the H0 , we call these cSNPs causal eQTL SNPs ( c-eQTL ) ., For each c-eQTL , we sample a different causal disease SNP ( c-dSNP ) from the same interval , with the requirement that its MAF comes from a distribution identical to that of the 976 NHGRI GWAS SNPs ., Subsequently , we sample up to five eQTL-dSNP pairs per interval where the eQTLs and dSNPs are the topmost correlated ( r2 ) SNPs with the c-eQTL and the c-dSNP respectively ., After sampling , we exclude cases where the eQTL and dSNP are identical , as these contradict the H0 . ., c-eQTL-c-dSNP-eQTL-dSNP quartets mapping to 287 unique hotspot intervals were sampled and tested under H0 ., Under the H1 , we sample up to five eQTL-dSNP pairs for each hotspot interval harbouring a cSNP as follows: the eQTLs are chosen as the top most significant SNPs per eQTL gene - excluding the cSNP; the dSNPs are randomly sampled from the same hotspot interval such that the r2 between each of them and the cSNP is in the range 0 . 5 , 0 . 9 ., At any stage of the 5-step iteration per cSNP , the dSNP must be different from the cSNP and the eQTLs sampled up to that point ., cSNP-eQTL-dSNP trios mapping to 290 unique hotspot intervals throughout the genome were sampled and tested under the H1 ., We use the LD values ( r2 ) of all pairwise SNP combinations per interval to calculate the median r2 , an estimate of the LD extent per region ., To perform a control experiment where the trait is gene expression , we used cis eQTLs ( 10−3 P-value permutation threshold ) detected in LCLs derived from 75 unrelated individuals of Western European origin from the GenCord project 19 ., Hotspot intervals ( N\u200a=\u200a157 ) where both a HapMap and a GenCord eQTL associating with the same Ensembl gene reside were analysed with the RTC Scoring scheme . | Introduction, Results, Discussion, Methods | The recent success of genome-wide association studies ( GWAS ) is now followed by the challenge to determine how the reported susceptibility variants mediate complex traits and diseases ., Expression quantitative trait loci ( eQTLs ) have been implicated in disease associations through overlaps between eQTLs and GWAS signals ., However , the abundance of eQTLs and the strong correlation structure ( LD ) in the genome make it likely that some of these overlaps are coincidental and not driven by the same functional variants ., In the present study , we propose an empirical methodology , which we call Regulatory Trait Concordance ( RTC ) that accounts for local LD structure and integrates eQTLs and GWAS results in order to reveal the subset of association signals that are due to cis eQTLs ., We simulate genomic regions of various LD patterns with both a single or two causal variants and show that our score outperforms SNP correlation metrics , be they statistical ( r2 ) or historical ( D ) ., Following the observation of a significant abundance of regulatory signals among currently published GWAS loci , we apply our method with the goal to prioritize relevant genes for each of the respective complex traits ., We detect several potential disease-causing regulatory effects , with a strong enrichment for immunity-related conditions , consistent with the nature of the cell line tested ( LCLs ) ., Furthermore , we present an extension of the method in trans , where interrogating the whole genome for downstream effects of the disease variant can be informative regarding its unknown primary biological effect ., We conclude that integrating cellular phenotype associations with organismal complex traits will facilitate the biological interpretation of the genetic effects on these traits . | Genome-wide association studies have led to the identification of susceptibility loci for a variety of human complex traits ., What is still largely missing , however , is the understanding of the biological context in which these candidate variants act and of how they determine each trait ., Given the localization of many GWAS loci outside coding regions and the important role of regulatory variation in shaping phenotypic variance , gene expression has been proposed as a plausible informative intermediate phenotype ., Here we show that for a subset of the currently published GWAS this is indeed the case , by observing a significant excess of regulatory variants among disease loci ., We propose an empirical methodology ( regulatory trait concordance—RTC ) able to integrate expression and disease data in order to detect causal regulatory effects ., We show that the RTC outperforms simple correlation metrics under various simulated linkage disequilibrium ( LD ) scenarios ., Our method is able to recover previously suspected causal regulatory effects from the literature and , as expected given the nature of the tested tissue , an overrepresentation of immunity-related candidates is observed ., As the number of available tissues will increase , this prioritization approach will become even more useful in understanding the implication of regulatory variants in disease etiology . | genetics and genomics/genetics of disease, genetics and genomics/complex traits, genetics and genomics/gene expression | null |
journal.pntd.0000861 | 2,010 | Trachoma Prevalence and Associated Risk Factors in The Gambia and Tanzania: Baseline Results of a Cluster Randomised Controlled Trial | Trachoma is caused by ocular infection with serovars A , B , Ba or C of the bacterium Chlamydia trachomatis ., It is the leading infectious cause of blindness worldwide 1 with an estimated 40 . 6 million people suffering from active trachoma ( trachomatous inflammation , follicular ( TF ) and/or intense ( TI ) ) and 8 . 2 million having trichiasis 2 ., As part of the “SAFE” ( Surgery , Antibiotics , Facial cleanliness , Environmental improvement ) trachoma control strategy , the World Health Organization ( WHO ) recommends mass antibiotic treatment annually for at least three years of all individuals in any district or community where the prevalence of TF in children aged 1–9 years is at least 10% ., After three or more years of A , F and E interventions , the prevalence is reassessed and a decision is made regarding the need to continue or cease treatment 3 ., Mass antibiotic treatment aims to clear infection from the community , most of which is found in children 4 ., Trachoma is endemic in both The Gambia and Tanzania , with estimated active trachoma prevalences in children aged 1–9 years of 10 . 4% and 27% , respectively 5 , 6 ., Accordingly , they have both recently qualified for a donation of the antibiotic azithromycin for mass treatment by Pfizer via the International Trachoma Initiative ., Given the different endemicities of these two countries , one in which trachoma is almost disappearing and one in which trachoma shows only modest signs of being reduced , the question of whether the same risk factors are predictive of trachoma is of interest ., In addition , since the presence of trachoma clinical signs is often poorly correlated with that of ocular C . trachomatis infection 5 , 7 , 8 , 9 , 10 , 11 , the risk factors for these markers of trachoma may also differ ., Studies have shown that although young age is a common risk factor for active trachoma , other risk factors may be setting-specific ., Furthermore , few studies have simultaneously reported risk factors for active trachoma and ocular C . trachomatis infection within the same setting 12 , 13 ., Information on risk factors can contribute to our understanding of trachoma transmission within the study area , and the targeting of trachoma control interventions can be aided through knowledge of risk factors ., We aimed to assess the prevalence of , and risk factors for , both active trachoma and ocular C . trachomatis infection pre-treatment in The Gambia and Tanzania , as part of the Partnership for the Rapid Elimination of Trachoma ( PRET ) cluster randomised controlled trial ., The aims of PRET are to test the impact on the prevalence of active trachoma and ocular C . trachomatis infection , as detected by Amplicor PCR , after three years in communities mesoendemic for trachoma ( between 20% and 50% TF ) or hypoendemic ( between 10% and 20% TF ) , when communities are randomised to different mass treatment population coverage levels and a different number of rounds of treatment , with a graduation rule if the prevalence of TF or detected ocular C . trachomatis infection falls below 5% ( Stare et al . submitted ) ., The data presented here are from the baseline surveys of PRET , where data on the prevalence of TF and evidence of ocular C . trachomatis infection were collected , and risk factors for these outcomes were obtained in a standardised fashion ., The research was done in accordance with the declaration of Helsinki ., Ethical approval was obtained from the London School of Hygiene & Tropical Medicine ( LSHTM ) , UK , Ethics Committee; The Gambia government/Medical Research Council ( MRC ) Joint Ethics Committee , The Gambia; the Johns Hopkins Institutional Review Board; and the Tanzanian National Institute for Medical Research ., Oral consent was obtained from the village leaders , and written ( thumbprint or signature ) consent from the childs guardian at the time of examination , which was signed by an independent witness ., In The Gambia , 48 census Enumeration Areas ( EAs ) , designed to have similar population sizes of between 600–800 people , were randomly selected from within 4 strata consisting of the following districts: Foni Bintang and Foni Kansala in Western Region , and Central Baddibu and Lower Baddibu in North Bank Region ( 12 EAs per district ) ( Figure 1 ) ., In Tanzania , 32 communities ( geographically distinct areas within a village with an average population of approximately 1500 people ) were selected in Kongwa district , Dodoma region ( Figure 2 ) ., Tanzanian communities were selected based on having an active trachoma prevalence above 20% in preliminary surveys and were therefore not randomly selected as they were in The Gambia ., A week-long workshop was conducted in February 2008 to standardise all fieldwork methods , including trachoma grading , photography , sample collection , form filling , facial cleanliness status grading , and data entry ., For trachoma grading , graders were standardised against a senior grader ( RB ) every day by examining participants in the field ., A kappa of >0 . 6 for TF grading was required between the senior grader and the graders in the final grading exam ., All other procedures had to be performed correctly five times in the field under observation by senior investigators before certification was given ., Fieldwork in The Gambia took place between 19th May 2008 and 29th July 2008 ., In Tanzania , data collection was between 15th May and 1st November 2008 ., Data were entered into a customised database ( MS Access v2007 ) developed at the Dana Center , Johns Hopkins University ., Key fields were double-entered by different entry clerks ., Reports of discrepant , missing or query entries were generated in the database and resolved by reference to the forms , or in some cases by return field visits ., Further queries of data inconsistencies were produced in statistical packages ( Stata v10 , STATA Corp . , College Station , TX , USA for the Gambian data; SAS v9 . 2 , SAS Institute Inc . , Cary , NC , USA for the Tanzanian data ) prior to analysis ., All queries were verified against the original paper forms ., The analyses presented here were conducted using Stata , v10 ., Baseline characteristics of household attributes and population size were summarised for both countries ., Evidence of variability between communities ( clusters ) and households was assessed using random effects logistic regression models assuming a 3-level hierarchy to the data structure ( community , household and individual ) in null regression models ., Univariate associations with TF and ocular C . trachomatis infection in children aged 0–5 years were tested using random effects logistic regression , accounting for between-cluster and between-household variation ( variance ) , comparing models with and without covariates using the likelihood ratio test ( LRT ) ., Multivariate model building for TF and C . trachomatis infection in both countries employed the same stepwise strategy; age and sex were considered a priori risk factors and included in all models ., Covariates associated with TF or evidence of C . trachomatis infection at the 10% significance level in univariate analyses were added in turn ( a forward stepwise approach ) and covariates retained in the model if the LRT p-value was ≤0 . 1 ., In The Gambia , the final multivariate model also adjusted for district to account for sampling stratified by district ., In The Gambia , 5033 children aged 0–5 years were examined ., In Tanzania , 3198 children were examined but ocular C . trachomatis data were missing from 76 of these children ., In The Gambia , there were 9 households with missing data for awareness of a village face-washing education programme ., In Tanzania , the number of missing values was 5 for household head education , 6 for time to water , 20 for latrine access , and 693 ( of which 677 were recorded as “unknown” ) for knowledge of a face-washing health education programme ., Community randomisation units were larger in Tanzania than in The Gambia , containing more , smaller households , as seen from the total population size and average household sizes ( Table 1 ) , although similar proportions of the total population were children aged under 10 years ., Household heads in The Gambia had less formal education than in Tanzania , whereas latrines and water were less easily accessible in Tanzania ., Around one third of households in both countries reported awareness of receiving community face-washing health education programmes ., The low prevalence of Amplicor positives in The Gambia provided little power for formal risk factor analyses ( Table 5 ) ., Chi-squared tests of association suggested that ocular discharge was a possible risk factor for an Amplicor positive result ( p\u200a=\u200a0 . 044 ) and that prevalence varied by district ( p<0 . 001 ) ., In Tanzania , Amplicor positivity was associated in univariate analyses with being aged ≥1 year , having ocular or nasal discharge , flies on the childs face , lack of household head education , and poor access to water or a latrine ( Table 5 ) ., In multivariate models , being Amplicor positive was only significantly related to being aged 2–5 years , having discharge , and a head of household educational level of less than 7 years , and possibly poor access to water ( Table 5 ) ., Other factors were not related to evidence of C . trachomatis infection ., In summary , this study showed that despite different prevalences of active trachoma and evidence of infection between the Tanzanian and Gambian study sites , the risk factors for TF were similar ., The risk factors for being Amplicor positive in Tanzania were similar to those for TF , whereas in The Gambia , only ocular discharge was associated with evidence of C . trachomatis DNA , suggesting that at this low endemicity , this may be the most important risk factor ., The lack of an association between being Amplicor positive and having TF in The Gambia highlights the poor correlation between the presence of trachoma clinical signs and evidence of C . trachomatis infection in this setting . | Introduction, Methods, Results, Discussion | Blinding trachoma , caused by ocular infection with Chlamydia trachomatis , is targeted for global elimination by 2020 ., Knowledge of risk factors can help target control interventions ., As part of a cluster randomised controlled trial , we assessed the baseline prevalence of , and risk factors for , active trachoma and ocular C . trachomatis infection in randomly selected children aged 0–5 years from 48 Gambian and 36 Tanzanian communities ., Both childrens eyes were examined according to the World Health Organization ( WHO ) simplified grading system , and an ocular swab was taken from each childs right eye and processed by Amplicor polymerase chain reaction to test for the presence of C . trachomatis DNA ., Prevalence of active trachoma was 6 . 7% ( 335/5033 ) in The Gambia and 32 . 3% ( 1008/3122 ) in Tanzania ., The countries corresponding Amplicor positive prevalences were 0 . 8% and 21 . 9% ., After adjustment , risk factors for follicular trachoma ( TF ) in both countries were ocular or nasal discharge , a low level of household head education , and being aged ≥1 year ., Additional risk factors in Tanzania were flies on the childs face , being Amplicor positive , and crowding ( the number of children per household ) ., The risk factors for being Amplicor positive in Tanzania were similar to those for TF , with the exclusion of flies and crowding ., In The Gambia , only ocular discharge was associated with being Amplicor positive ., These results indicate that although the prevalence of active trachoma and Amplicor positives were very different between the two countries , the risk factors for active trachoma were similar but those for being Amplicor positive were different ., The lack of an association between being Amplicor positive and TF in The Gambia highlights the poor correlation between the presence of trachoma clinical signs and evidence of C . trachomatis infection in this setting ., Only ocular discharge was associated with evidence of C . trachomatis DNA in The Gambia , suggesting that at this low endemicity , this may be the most important risk factor ., ClinicalTrials . gov NCT00792922 | Trachoma is caused by Chlamydia trachomatis and is the leading infectious cause of blindness ., The World Health Organizations ( WHO ) control strategy includes antibiotic treatment of all community members , facial cleanliness , and environmental improvements ., By determining how prevalent trachoma is , decisions can be made whether control activities need to be put in place ., Knowing what factors make people more at risk of having trachoma can help target trachoma control efforts to those most at risk ., We looked at the prevalence of active trachoma and C . trachomatis infection in the eyes of children aged 0–5 years in The Gambia and Tanzania ., We also measured risk factors associated with having active trachoma or infection ., The prevalence of both active trachoma and infection was lower in The Gambia ( 6 . 7% and 0 . 8% , respectively ) than in Tanzania ( 32 . 3% and 21 . 9% , respectively ) ., Risk factors for active trachoma were similar in the two countries ., For infection , the risk factors in Tanzania were similar to those for TF , whereas in The Gambia , only ocular discharge was associated with infection ., These results show that although the prevalence of active trachoma and infection is very different between the two countries , the risk factors for active trachoma are similar but those for infection are different . | evidence-based healthcare/clinical decision-making, ophthalmology/eye infections, infectious diseases/neglected tropical diseases, public health and epidemiology/epidemiology, public health and epidemiology/infectious diseases, infectious diseases/bacterial infections, infectious diseases/epidemiology and control of infectious diseases | null |
journal.pntd.0007067 | 2,019 | DNA plasmid coding for Phlebotomus sergenti salivary protein PsSP9, a member of the SP15 family of proteins, protects against Leishmania tropica | As a vector-borne parasitic and tropical disease , leishmaniasis has been reported in 98 countries ( in 4 continents ) with an annual incidence level of 0 . 9–1 . 6 million individuals , and 350 million individuals are at the risk of this infection 1 , 2 ., The Leishmania parasite , as the causative agent of this disease , is delivered to the host during blood-feeding by infected female sand flies , namely , the Phlebotomus species in the Old World ( Asia , Africa and Europe ) and the Lutzomyia species in the New World ( Central and South America ) ., Various clinical symptoms , in visceral , mucocutaneus and cutaneous forms have been reported for leishmaniasis ., The most commonly observed type of this disease is CL , which results in ulcers and permanent scars on the affected regions of the body and serious disabilities for the patient ., Clinical forms of leishmaniasis are geographically distributed depending on the availability of the pathogenic species of Leishmania and their vectors ., For example , CL has been mainly observed in Afghanistan , Brazil , Iran , Iraq and Syrian Arab Republic 1 ., The fact that infection with certain species , including L . major or exposure to live Leishmania ( leishmanization ) results in long-term protection , in addition to the high costs of current treatments , drug toxicity and parasite resistance emergence intensify the need for producing a vaccine for this disease 3–5 ., However , despite numerous investigations , no human vaccine for Leishmania is yet available 6 ., Studies on development of Leishmania vaccines have focused on Leishmania antigens ., In the past two decades , researchers started to examine sand fly salivary proteins 7 , which are transmitted along with the parasite during blood feeding , as an alternative or a component of a vaccine against leishmaniasis ., Different molecules are present in sand fly saliva which alter the hemostatic , inflammatory and immune response of the host , thereby facilitating parasite infection 8 ., Simultaneous inoculation of Leishmania parasite and sand fly saliva has been reported to increase the parasite burden and cause exacerbation of the generated lesions 7 , 9–12 ., However , immunization of animals with salivary components 13–15 or salivary gland homogenate ( SGH ) 12 , 16 , 17 from sand flies or exposing them to the bites of uninfected sand flies 18–25 resulted in protection against infection with Leishmania 26 ., A cell-mediated immune response ( CIR ) , as a form of a DTH response to sand fly salivary proteins , was shown to provide protection against Leishmania infection ., This anti-salivary immunity appears to be important , considering that the parasite is unavoidably naturally injected along with salivary proteins into the biting site ., Hence , similar to an adjuvant , a Th1 anti-saliva immunity can accelerate the induction of a protective Th1 immunity against Leishmania 27 ., In Iran , zoonotic CL is caused by L . major , which is usually transmitted by the Ph . papatasi vector ., On the other hand , anthroponotic CL is caused by L . tropica mainly transmitted by the Ph . sergenti sand fly 28 ., According to the literature , exposing mice to the bites of uninfected sand flies or immunizing them with Ph . papatasi SGH can protect them against L . major infection caused by needle inoculation 7 or the bites of infected sand flies 13 ., Immunization with PpSP15 , a salivary protein from Ph . papatasi , provides protection in mice against L . major infection 20 , 21 , 29 , 30 ., The immunity induced by Ph . sergenti salivary proteins and their effects on L . tropica infection have not been previously reported ., Therefore , in this work , the immunogenicity of Ph . sergenti salivary components were assessed and we tested whether immunizing BALB/c mice with Ph . sergenti SGH or DNA plasmids coding for Ph . sergenti salivary proteins can lead to an appropriate immunity to control L . tropica infection ., Apyrogenic deionized water ( Milli-Q System , Millipore , Molshem , France ) was used to prepare all needed solutions ., The Endo-Free Plasmid Mega kit , RNeasy Mini Kit , Quanti Nova SYBR Green Master Mix and Anti-His antibody were purchased from QIAGEN , Germany ., TRIzol Reagent and SuperScript III First-Strand Synthesis system were from Thermo Fisher Scientific ( Invitrogen Company , USA ) ., The materials needed for PCR reaction , agarose gel electrophoresis and enzymatic digestion were all provided by Roche Applied Sciences , Germany ., Diaminobenzidine ( DAB ) powder , Bovine Serum Albumin ( BSA ) and acrylamide were provided by Merck , Germany ., Horseradish peroxidase-conjugated goat anti-mouse IgG , Urea , Ponceau-S , Sodium dodecyl sulfate ( SDS ) , Tris-base , Tris-HCL , M199 medium , RPMI-1640 , DMEM , gentamicin , kanamycin , L-glutamine , hemin , HEPES , adenosine and Ficoll-400 were purchased from Sigma , Germany ., Fetal Calf Serum ( FCS ) and Schneider insect media were provided by Gibco ( Life Technologies , Germany ) ., Goat anti-mouse IgG1-HPR and IgG2-HPR were obtained from Southern Biotech , Canada ., Peroxidase Substrate System as an ELISA substrate was from KPL ( ABTS , USA ) ., Linear Polyethylenimine was purchased from Polyscience , Germany ., Protran Nitrocellulose Transfer Membranes was from Schleicher & Schuell BioScience , Germany ., GF-1 Tissue DNA Extraction Kit was purchased from Vivantis Technologies , Malaysia ., The Agilent RNA 6000 Nano reagent kit were purchased from Agilent Technologies , USA ., The Agilent 2100 Bioanalyzer instrument ( Agilent Technologies , USA ) , ELISA reader ( Tecan , USA ) , NanoDrop ( Nanodrop , ND-1000 , USA ) and Speed Vac ( Thermo Scientific , USA ) devices were also used in this investigation ., Ph . sergenti were kept in the insectary at the Laboratory of Malaria and Vector Research , National Institutes of Health , NIH ( Rockville , MD , USA ) ., The salivary glands were dissected from 5- to 7-day-old and non-blood fed female sand flies and then transferred to PBS for subsequent storage at −70 °C ., After disruption by ultra-sonication and centrifugation , the produced supernatant was collected and dried in a Speed Vac device and reconstituted before use ., Female BALB/c mice , ( 6–8 weeks old ) with weight range of 18–20 g were obtained from Pasteur Institute of Iran ., The animals under investigation were maintained , handled , anesthetized and euthanized under the approval of Institutional Animal Care and Research Advisory Committee of Pasteur Institute of Iran ( ethical code: IR . RII . REC . 1394 . 0201 . 6417 , dated 2015 ) ., All experiments were designed and carried out according to the Specific National Ethical Guidelines for Biochemical Research ( 2005 ) by the Research and Technology Deputy of Ministry of Health and Medicinal Education ( MOHM ) of Iran ., Mice were euthanized through cervical dislocation method ., Animals were anesthetized via intraperitoneal ( i . p . ) administration of Xylazine/Ketamine anesthetizing cocktail 31 in order to minimize suffering animals under investigation ., The mammalian codon optimized nucleotides encoding the N-terminus to the stop codon of the most abundant secreted proteins from Ph . sergenti salivary gland proteins were cloned in a modified mammalian expression plasmid ( VR1020-TOPO ) through T/A cloning strategy and topoisomerase technology and then transformed into the DH5α strain ., The VR1020-TOPO plasmid has features such as a CMV promoter , the signal-secretory peptide of tissue plasminogen activator ( TPA ) , replacing the sand fly specific-secretory signal peptide and a 6×His-tag , downstream from the target insert ., This modified vector enables efficient production of the secreted proteins in animal tissues and other mammalian-based expression systems ., The fourteen cloned transcripts used in this study are PsSP7 ( HM560864 , D7-related proteins ) , PsSP9 ( HM569364 , PpSP15-like protein ) , PsSP14 ( HM560870 , PpSP15-like protein ) , PsSP15 ( HM560868 , PpSP15-like protein ) , PsSP20 ( HM560866 , yellow-related proteins ) , PsSP26 ( HM569362 , yellow-related proteins ) , PsSP40 ( HM560860 , apyrase ) , PsSP41 ( HM560862 , apyrase ) , PsSP42 ( HM560861 , apyrase ) , PsSP44 ( HM569368 , PpSP32-like protein ) , PsSP52 ( HM537134 , antigen 5-related proteins ) , PsSP54 ( HM569365 , PpSP15-like protein ) , PsSP73 ( HM569367 , unknown ) , and PsSP98 ( HM569366 , unknown ) ., In order to purify all recombinant plasmids and the empty control plasmid ( VR1020 ) , an Endo-Free Plasmid Mega kit was used according to the manufacturer’s protocol ., In order to confirm the expression of the 14 plasmids harboring each Ph . sergenti salivary protein , COS-7 cells ( ATCC CRL-1651 ) were used as an expression host ., Briefly , COS-7 cells were cultured in six-well plates ( Greiner ) in complete RPMI medium supplemented with 10% FCS at 37 °C in the presence of 5% CO2 ., For cell transfection , we used PEI/DNA complexes , prepared through mixing Linear Polyethylenimine ( LINPEI , MW = 25 kDa , 10 μM ) and 5 μg of each recombinant plasmid or VR1020 ( as the control ) 32 ., Expression of Ph . sergenti salivary proteins in transfected COS-7 cells were confirmed by western blot analysis ., In brief , forty eight hours after transfection , the supernatant of transfected COS-7 cells were harvested , then mixed with SDS-PAGE sample buffer and boiled for 5 min , and run on a 12 . 5% or 15% SDS-PAGE gel ., Proteins were then transferred from gel to nitrocellulose membranes by electro-blotting ., Free binding sites on nitrocellulose membrane were blocked with blocking solution ( PBS with 0 . 05% Tween 20 and 2 . 5% BSA ) for 2 hours ., After three washes , the membrane was incubated with HRP-conjugated goat anti-mouse IgG ( 1:2000 ) for 2 hours at room temperature and visualized using the 3 , 30-diaminobenzidine substrate ( DAB ) ., In order to assess the immunogenic characteristics of DNA plasmids coding for Ph . sergenti salivary proteins , 6–8 weeks old BALB/c mice ( 6 mice in each experimental group ) were i . d . immunized three times at two-week intervals using 30-gauge needle in the right ear ., For this purpose , we used 10 μg of plasmid ( either the empty plasmid control or a recombinant plasmid encoding a Ph . sergenti salivary protein ) , an equivalent of 0 . 5 Ph . sergenti salivary gland pair or PBS; all in a total volume of ~10 μl with PBS ., ELISA microplates were coated with 100 μl of SGH diluted to two pairs of SGHs/ml in coating buffer ( Na2CO3 0 . 02 M , NaHCO3 0 . 45 M , pH 9 . 6 ) overnight at 4 °C ., Following rinsing with PBS-0 . 05% Tween , wells were blocked using 100 μl of 1% BSA in PBS for 2 h at 37 °C ., Wells were incubated for 3 h with sera from mice immunized with recombinant or control plasmids obtained two weeks after the last immunization and diluted ( 1:50 ) in PBS-0 . 05% Tween-1% BSA ., After one more washing step , wells were incubated with Horseradish peroxidase-conjugate goat anti-mouse IgG diluted ( 1:5000 ) in PBS-0 . 05% Tween-1% BSA for 2 h at 37 °C ., Following another washing step , plates were incubated with Peroxidase Substrate System ( KPL ) as the substrate for 30 min at 37 °C ., After stopping the reactions with 1% SDS , the absorbance was measured at 405 nm using an ELISA reader ., The cut off value was determined by measuring anti-saliva IgG of plasmid control mice group ( mean + 3 SD ) ., Two weeks after the last step of immunization , the animals were inoculated i . d . into the left ear dermis with Ph . sergenti SGH ( 0 . 5 salivary gland pair per mouse ) using a 30-gauge needle ., Forty eight hours later , we measured ear thickness ( DTH response ) using a digital caliper ( with a resolution of 0 . 01 mm ) ., For histopathological assessment at this time point , after fixing the dissected ears in 10% phosphate-buffered formalin , they were processed , embedded in paraffin , and then the 5-μm sections prepared by microtome were stained using hematoxylin and eosin ( H & E ) and analyzed using light microscopy ., For morphometric analyses , inflammatory cells were counted in three fields/section using a 400× magnification , covering a total area of 710 mm2 ., After screening the Th1 DTH response in BALB/c mice which was triggered by the saliva proteins of Ph . Sergenti , the animals ( 10 mice per group ) were immunized three times with two-week intervals using the selected plasmid encoding Ph . sergenti salivary protein , SGH , empty plasmid or PBS in the right ear using a 30-gauge needle ., Two weeks after the last immunization , animals were challenged i . d . into the left ear dermis with 107 metacyclic L . tropica parasites plus Ph . sergenti SGH ( 0 . 5 salivary gland pair ) using a 30-gauge needle in an almost 10 μl total volume ., The L . tropica parasite named as MOHM/IR/09/Khamesipour-Mashhad was isolated from patient in city of Mashhad , Iran , in 2009 ( provided as gift by Dr . Ali Khamesipour ) ., L . tropica promastigotes were cultured in M199 medium supplemented with 10% hi-FCS ., The Ficoll-400 step-gradient was used to isolate the L . tropica metacyclic promastigotes 33 ., To mimic the natural model of infection and examine whether immunity to the salivary proteins of the sand fly can protect the animals against CL , mice were infected by injection of parasites together with Ph . sergenti SGH into their ear dermis ., The ear thickness was monitored and measured weekly using a digital caliper ., To measure the disease burden ( area under the curves , AUC ) , the ear thickness of each individual immunized mouse was recorded once per week ., A disease course curve for each mouse in the experimental and control groups was separately obtained ., Prism ( GraphPad Software ) was used to calculate AUC ., Cytokine profiles were analyzed at two time points: once after Ph . sergenti SGH inoculation ( to screen the Th1 DTH response against each DNA plasmid ) , and once after infectious challenge with L . tropica plus Ph . sergenti SGH ., For this purpose , total RNA was extracted from the mouse dLN using TRIzol reagent and then RNeasy Mini Kit ., The quality of extracted RNA was confirmed using an Agilent RNA 6000 Nano reagent kit ., For first-strand cDNA synthesis , approximately 2 μg of RNA reverse-transcribed in a total volume of 20 μl using SuperScript III reverse transcriptase according to the manufacturer’s instructions ., For quantification of gene expression , the 1:10 diluted cDNA was subjected to the reaction containing 5 pmol of each forward and reverse primer and 12 . 5 μl Quanti Nova SYBR Green Master Mix in a 25 μl total volume ., Real-time PCR reactions were performed in duplicates on an Applied Biosystems 7500 instrument ., Thermal cycles with an initial incubation step at 95 °C for 5 min followed by 45 cycles at 95 °C for 10 s , at 60 °C for 15 s , and at 72 °C for 35 s ., The mRNA levels of each target gene were normalized to that of HPRT ., The results are shown in fold change compared to the PBS control ., Gene expression was analyzed based on the comparative method ., The cycle threshold ( Ct ) values for cytokines were normalized to the expression of HPRT based on the following formulation: ΔCt = Ct ( target gene ) −Ct ( HPRT gene ) ., We obtained the fold change using 2−ΔΔCt , in which ΔΔCt = ΔCt ( test ) −ΔCt ( control ) 34 ., We used the following primers for real-time PCR: HPRT ( Forward: 5′-GTCCCAGCGTCGTGATTAG-3′; Reverse: 5′-GAGCAAGTCTTTCAGTCCTGTC-3′ ) ; IFN-γ ( Forward: 5′-TCTGAGACAATGAACGCTACAC-3′; Reverse: 5′-CTTCCACATCTATGCCACTTGAG-3′ ) ; IL-5 ( Forward: 5′-TGACAAGCAATGAGACGATGAG-3′; Reverse: 5′-CTCCAATGCATAGCTGGTGA-3′ ) ., Quantification of the parasites in the infected ear of animals in different groups was performed using Real-time PCR at one and two months after challenge ., After euthanizing animals in each group ( 5 mice per group ) the genomic DNA was extracted from each infected ear using GF-1Tissue DNA Extraction Kit according to manufacturer’s instruction ., DNA concentration was measured by a NanoDrop device ., The following primer set was used to target a part of L . tropica kinetoplastid minicircle DNA: ( KDNA1F: ( 5-GGGTAGGGGCGTTCTGC-3 ) and KDNA1R ( 5-TACACCAACCCCCAGTTTGC-3 ) ) 35 , 36 ., The absolute copy number corresponding to the target sequence was measured on an Applied Biosystems 7500 real time PCR system ., Standard L . tropica genomic DNA was used in 10-fold dilution corresponding to 2×108 to 2×101 parasites for drawing the standard curve ., To quantify the parasites in tissues , 50 ng of DNA was applied to a reaction containing 5 pmol of each of the forward and reverse primers and 12 . 5 μl of Quanti Nova SYBR Green Master Mix in a 25 μl total volume ., The PCR program was as the following: at 95 °C for 5 min; 40 cycles at 95 °C for 10 s , at 60 °C for 15 s , and at 72 °C for 35 s ., All reactions were performed in duplicate ., We replicated each measurement twice and averaged the obtained results ., Shapiro-Wilk test was used to check the distribution of DTH response , Antibody response , IFN-γ , IL-5 and ratio of IFN-γ to IL-5 mRNA expression in dLN as well as parasite burden and disease burden ( area under the curves , AUC ) ., Due to the non-normality of all data , non-parametric van der Waerden’s normal score test and Dunn’s multiple comparison tests were used to compare the distribution of these variables between experimental and control group ., To assess the effect of variables on ear thickness , we used Linear Mixed Models for repeated measured data ., The significant interaction between time and group factors implied that the experimental groups were changing with time , with different manners ., A statistically significant interaction effect may indicate that the overall patterns of differences at the level of main effects are not likely to be consistent across all groups ., The p values below 0 . 05 were considered as significant ., Statistical analyses were carried out using Stata ( 14 . 0 ) ( StataCorp . 2015 . Stata Statistical Software: Release 14 . College Station , TX: StataCorp LP ) and R 3 . 4 . 3 ( R Core Team ( 2017 ) ., R: A language and environment for statistical computing ., R Foundation for Statistical Computing , Vienna , Austria ) using the Rfit , PMCMR , nparcomp and multcomp packages ., Before screening the 14 different plasmids coding for Ph . sergenti salivary proteins in animals , we tested the protein expressions of these DNA constructs in COS-7 cells ., The construct has a signal secretory peptide and we added to each transcript a histidine tag ., We observed expression of protein in COS-7 cells from all plasmids ( Fig 1 ) ., The 14 different DNA plasmids coding for Ph . sergenti salivary proteins were screened to select a plasmid that can induce a Th1 cellular immune response ., We also compared these plasmids to Ph . sergenti salivary gland homogenate ., The parameters for selection were, 1 ) The presence of a distinct DTH response associated with mononuclear cell infiltration in the ear and, 2 ) increased levels of IFN-γ and low levels of IL-5 expression in dLN in comparison with the empty plasmid control group ., Surprisingly , immunization with the whole Ph . sergenti salivary gland homogenate ( SGH ) did not induce a detectable DTH response ( Fig 2A ) ., In contrast , seven plasmids induced a positive DTH and showed greater ear thickness than the control plasmid that produced a median of 0 . 18 mm ( Fig 2A ) ., The median ranks of DTH responses were in the following descending order: PsSP40 ( 0 . 26 mm ) , resulting in the greatest measurable response in animal skin; PsSP52 ( 0 . 24 mm ) , PsSP44 ( 0 . 23 mm ) , PsSP9 ( 0 . 23 mm ) , PsSP26 ( 0 . 22 mm ) , PsSP41 ( 0 . 22 mm ) and PsSP42 ( 0 . 21 mm ) , as the smallest detectable levels ., Groups immunized with PsSP40 ( p <0 . 01 ) , PsSP52 ( p <0 . 01 ) , PsSP9 ( p <0 . 01 ) , PsSP44 ( p <0 . 01 ) , PsSP26 ( p <0 . 01 ) , PsSP41 ( p = 0 . 04 ) and PsSP42 ( p = 0 . 03 ) showed significantly higher DTH responses compared to the control plasmid ( Fig 2A and S1 Table ) ., Specific total IgG antibody responses against Ph . sergenti SGH in the sera of all groups was measured ., In comparison with the control groups ( PBS and VR1020 ) , the Ph . sergenti SGH-immunized mice group produced higher levels of anti-saliva-IgG than the cut off value ( median at OD405 nm = 0 . 40 , S1 Table and Fig 2B ) ., Moreover , the group immunized with PsSP26-encoding plasmid had the highest level of total IgG antibody production against Ph . sergenti SGH in comparison with the control plasmid group ( median at OD405 nm = 1 . 53 , p<0 . 01 ) ., In addition , the two other groups encoding the PsSP7 and PsSP44 proteins also produced higher levels of total IgG than the cut off value ( PsSP44 with median OD405 nm = 0 . 30 , p = 0 . 04 and PsSP7 with median OD405 nm = 0 . 45 , p = 0 . 07 ) as demonstrated in Fig 2B and S1 Table ., Based on histological analysis 48 h after Ph . sergenti SGH injection in the ear dermis of plasmid-immunized animals , PsSP9- and PsSP40-immunized groups were characterized by a robust mononuclear infiltration mainly containing macrophages , lymphocytes and a lower number of neutrophils compared to the PBS- and plasmid- control groups ( Fig 3 ) ., The number of inflammatory cells recruited in the Ph . sergenti SGH-immunized mice group was moderate and there were fewer cells than in the PsSP9- and PsSP40-immunized groups ( Fig 3 ) ., Furthermore , the number of inflammatory cells recruited in PsSP41- and PsSP52-immunized mice was similar to that of Ph . sergenti SGH-immunized mice ., No detectable increase in the number of inflammatory cells recruited in PsSP42-immunized mice was observed , compared to the PBS- and plasmid- control groups ( Fig 3 ) ., Based on the outcome of the DTH response , the antibody response ( as shown in Table, 1 ) and histological analysis against various salivary proteins , PsSP9 , PsSP40 , PsSP41 , PsSP42 , and PsSP52 plasmids were selected for further evaluation of the cellular immune response in BALB/c mice ., Forty-eight hours after Ph . sergenti SGH injection , mice immunized with the above-mentioned plasmids were euthanized and IFN-γ and IL-5 cytokine expression was evaluated in their dLN by Real-time PCR ( Fig 4 ) ., PsSP9 immunized mice was the only group that induced a statistically significant increase in IFN-γ mRNA expression , compared to the control plasmid group ( median of fold change = 22 . 30 , p = 0 . 04 , Fig 4A , S2 Table ) , and was the group exhibiting the lowest level of IL-5 ( Fig 4B , S2 Table ) , which translated to the highest ratio of IFN-γ to IL-5 expression ( median of fold change = 17 . 12 , p = 0 . 16 , Fig 4C , S2 Table ) ., In PsSP52- , PsSP41- and PsSP42-immunized mice , IL-5 expression was significantly higher compared to the control plasmid group ( p<0 . 01 in PsSP52 , p = 0 . 01 in PsSP41 and p = 0 . 03 in PsSP42 , Fig 4B , S2 Table ) ., Furthermore , neither the IFN-γ expression nor the ratio of IFN-γ/IL-5 expression in these groups was significantly different from the control plasmid group ( Fig 4A and 4C , S2 Table ) ., The data also revealed that there was no significant difference in terms of IL-5 and IFN-γ expression between the PsSP40-immunized mice and the control plasmid group ( Fig 4A and 4B , S2 Table ) ., The exact p values and median ( Q1-Q3 ) for the six tested samples are presented in S2 and S3 Tables ., Based on the findings that the PsSP9-immunized group induced a Th1 immune response , we examined whether immunization with DNA coding for PsSP9 salivary protein can lead to protection of animals against L . tropica infection ., BALB/c mice were immunized intradermally in their ears three times ( with two-week intervals ) with either Ph . sergenti SGH , plasmid encoding a PsSP9 salivary protein , PBS , or empty plasmid ., Two weeks after the last immunization , animals were challenged with metacyclic forms of L . tropica plus Ph . sergenti SGH ., PsSP9-immunized mice had significantly smaller nodules , indicated by ear thickness measurements , compared to the control plasmid group which showed a significantly greater ear thickness ( Fig 5A , S4 and S5 Tables ) ., The disease burden was determined based on the area under the curves ( AUC ) as shown in Fig 5B ., The data indicated that there was a significant reduction in the disease burden after PsSP9 immunization , in comparison with the plasmid control group ( p = 0 . 02 , Fig 5B , S6 Table ) ., One-month post-infection , the parasite load in the ear of PsSP9-immunized mice was significantly lower than the control plasmid group ( 7 . 2×105 parasites in the control plasmid group and 1 . 4×105 parasites in the PsSP9-immunized group , p = 0 . 02 , Fig 5C , S7 Table ) ., At two months post challenge , the parasite load in the ear in PsSP9-immunized mice remained lower than that in the plasmid control mice group , but the difference was not statistically significant ( 1 . 0×106 parasites in control plasmid group and 2 . 7×105 parasites in PsSP9-immunized group , p = 0 . 36 , Fig 5D ) ., The reduced parasite load correlated with the observed lower ear thicknesses in PsSP9-immunized mice ., SGH-immunized mice showed no significant differences in ear thickness or disease burden compared to the control groups ( Fig 5A and 5B , S4–S6 Tables ) ., In addition , at one and two month post-infection , there were no statistically significant differences in the ear parasite burden in SGH-immunized mice compared with the control groups ( Fig 5C and 5D , S7 Table ) ., The cytokine profile expression in the dLN was assessed at one and two months post-challenge ., At one-month post-challenge , no significant differences in IFN-γ and IL-5 expression levels or in the ratio of IFN-γ to IL-5 expression were observed in any of the groups ( Fig 6A–6C , S8 Table ) ., At two months post-infection , IFN-γ expression was higher in the dLN of PsSP9-immunized mice than in the control plasmid group , although the difference was not statistically significant ( median of fold change = 13 . 29 , p = 0 . 05 , Fig 6D , S9 Table ) ., At this time point , the expression of IL-5 in dLN of PsSP9-immunized mice was not different from that in the control plasmid group ( Fig 6E , S9 Table ) ., Importantly , a significant increase in the ratio of IFN-γ to IL-5 expression was observed in PsSP9-immunized mice compared with the control plasmid group ( median of fold change = 17 . 06 , p = 0 . 04 , Fig 6F , S9 Table ) ., Exact p values and median ( Q1 , Q3 ) for the 5 studied samples are presented in S4–S9 Tables ., It has been previously reported that bites of Ph . papatasi 13 and Ph . duboscqi 15 , 24 , as well as the injection of Ph . papatasi 7 , Lu ., longipalpis 17 , and Lu ., whitmani SGH 16 , protected rodents against leishmaniasis , manifested by reduced lesion size and decreased parasite burden ., Moreover , exposure to sand fly SGH has been reported to result in a DTH response which is associated with the enhanced production of IFN-γ and IL-12 13 , 37 or a higher IFN-γ to IL-4 ratio 7 ., The current study demonstrated that immunization with whole Ph . sergenti SGH did not confer protection against L . tropica infection ., Indeed , the ear thickness , disease burden and parasite load in the Ph . sergenti SGH-immunized mice were similar to those of the mice in the control groups ., The lack of protection in the Ph . sergenti SGH-immunized mice against infection correlates with a low ratio of IFN-γ to IL-5 production in the dLN during infection ., The current results were similar to the findings of Moura et al . , who reported that the immunization of BALB/c mice with Lu ., intermedia SGH resulted in a non-protective Th2 immunity 38 ., In contrast to whole SGH immunization , a single sand fly salivary protein was shown to influence the outcome of leishmaniasis ., Moura et al . observed that DNA immunization with Lu ., intermedia salivary protein Linb-11 resulted in an intense protective cell-mediated immunity against infection with L . braziliensis 23 in contrast to the whole Lu ., intermedia saliva that exacerbated L . braziliensis infection 38 ., The current results from our work also suggest that distinct Ph . sergenti salivary proteins induce different immune responses ., Immunization with 14 plasmids coding for Ph . sergenti secreted salivary proteins resulted in the identification of 9 salivary proteins which produced either a positive DTH response , an antibody response , or both responses in BALB/c mice ( Table 1 ) ., The triggering of a humoral immunity by such proteins is not necessary for anti-leishmaniasis protection ., As demonstrated by Gomes et al . , LJM19 was the only saliva protein that induced strong DTH in hamster without any detectable antibodies and DNA immunization with this plasmid protected the animal against VL 19 ., Salivary proteins that result in a Th1 type cellular immune response can be considered as suitable candidate for producing an anti-leishmaniasis vaccine ., In this study , the type of immune response generated by the protective plasmid PsSp9 was characterized by a DTH response with no detectable antibody response and a significantly high ratio of IFN-γ to IL-5 expression in the dLN compared to the control group ., The current study reveals that DNA immunization with the Ph . sergenti salivary protein PsSP7 can induce an antibody response , but no detectable DTH response ., Moreover , DNA immunization with PsSP41 , PsSP42 and PsSP52 produced a DTH response but induced high IL-5 expression in the dLN after Ph . sergenti SGH inoculation ., Therefore , DNA immunization with these proteins shifted the immune response to a Th2 type ., Hence , application of such salivary proteins of Ph . sergenti ( i . e . PsSP7 , PsSP26 , PsSP41 , PsSP42 , PsSP44 and PsSP52 ) may exacerbate L . tropica infection or cause no protective impact ., Interestingly , DNA immunization with Ph . sergenti salivary protein PsSP9 ( among the 14 tested Ph . sergenti salivary proteins ) produced a DTH with high mononuclear infiltration in the ear , a high ratio of IFN-γ to IL-5 expression in the dLN and no detectable antibodies 48 h after Ph . sergenti SGH inoculation ., Therefore , this protein directed the immunity toward a Th1 response and was chosen as a candidate for producing an experimental vaccine against L . tropica ., The mice immunized with PsSP9 displayed a smaller ear thickness , lower disease burden , and a lower parasite load in the ear in addition to a high IFN-γ to IL-5 expression ratio in the dLN compared with the empty control plasmid group two months after challenge ., The draining lymph node is a part of the immune system , where the induction of specific immunity against an antigen occurs and where effector cells migrate to the skin and contribute to protection 39 ., The protection observed in the PsSP9-immunized mice against CL can be explained by the anti-PsSP9 immunity at the parasite transmission site in the ear dermis , which may facilitate direct parasite killing by macrophage activation through IFN-γ ., In fact , IFN-γ acts to restrict Leishmania growth in macrophages of mice and humans as well as the progression of leishmaniasis 40 ., Vinhas et al . 41 demonstrated that the peripheral blood mononuclear cells ( PBMC ) of individuals who were subjected to uninfected Lu ., longipalpis sand fly bites , exhibited IFN-γ expression after SGH stimulation ., The production of IFN-γ was also associated with L . chagasi parasite killing in a macrophage-lymphocyte autologous culture 41 ., This implies that after PsSP9 DNA immunization , the effector cells ( IFN-γ+ ) might have migrated to the site of L . tropica and Ph . sergenti SGH inoculation , activating the macrophages and causing parasite killing which led to smaller lesions and reduced parasite load ., The presence of an immune response against PsSP9 at the site of the parasite plus SGH injection in the dermis of the ear may act as an adjuvant to accelerate the triggering of a proper host immune response against Leishmania ., Macrophage activation induces inflammatory responses and prevents the growth of Lei | Introduction, Materials and methods, Results, Discussion | The vector-borne disease leishmaniasis is transmitted to humans by infected female sand flies , which transmits Leishmania parasites together with saliva during blood feeding ., In Iran , cutaneous leishmaniasis ( CL ) is caused by Leishmania ( L . ) major and L . tropica , and their main vectors are Phlebotomus ( Ph . ) papatasi and Ph . sergenti , respectively ., Previous studies have demonstrated that mice immunized with the salivary gland homogenate ( SGH ) of Ph . papatasi or subjected to bites from uninfected sand flies are protected against L . major infection ., In this work we tested the immune response in BALB/c mice to 14 different plasmids coding for the most abundant salivary proteins of Ph . sergenti ., The plasmid coding for the salivary protein PsSP9 induced a DTH response in the presence of a significant increase of IFN-γ expression in draining lymph nodes ( dLN ) as compared to control plasmid and no detectable PsSP9 antibody response ., Animals immunized with whole Ph . sergenti SGH developed only a saliva-specific antibody response and no DTH response ., Mice immunized with whole Ph . sergenti saliva and challenged intradermally with L . tropica plus Ph . sergenti SGH in their ears , exhibited no protective effect ., In contrast , PsSP9-immunized mice showed protection against L . tropica infection resulting in a reduction in nodule size , disease burden and parasite burden compared to controls ., Two months post infection , protection was associated with a significant increase in the ratio of IFN-γ to IL-5 expression in the dLN compared to controls ., This study demonstrates that while immunity to the whole Ph . sergenti saliva does not induce a protective response against cutaneous leishmaniasis in BALB/c mice , PsSP9 , a member of the PpSP15 family of Ph . sergenti salivary proteins , provides protection against L . tropica infection ., These results suggest that this family of proteins in Ph . sergenti , Ph . duboscqi and Ph . papatasi may have similar immunogenic and protective properties against different Leishmania species ., Indeed , this anti-saliva immunity may act as an adjuvant to accelerate the cell-mediated immune response to co-administered Leishmania antigens , or even cause the activation of infected macrophages to remove parasites more efficiently ., These findings highlight the idea of applying arthropod saliva components in vaccination approaches for diseases caused by vector-borne pathogens . | Leishmaniasis is a vector-borne disease transmitted to humans by an infected sand fly bite , through which Leishmania parasites and saliva are co-delivered into the host skin ., Despite the numerous studies performed in this area , no vaccine is yet available to control this neglected disease in humans ., During the past two decades , saliva of sand flies has been tested for possible application as a vaccine against leishmaniasis ., Exposure to specific salivary proteins or sand fly bites can induce a protective cell-mediated immunity ., Immunization with Ph . papatasi saliva or recombinant PpSP15 has been previously reported to provide protection against L . major infection ., In this study , the efficiency of immunization with Ph . sergenti saliva or plasmid coding for Ph . sergenti salivary proteins in protecting the BALB/c mice against L . tropica was explored ., Here we show that although immunization with whole saliva induces a humoral response , this immune response is unable to protect mice against infection; in contrast , immunization with a plasmid coding for Ph . sergenti PsSP9 salivary protein induces a Th1 immune response characterized by a strong DTH response , no detectable antibody response , and a high expression ratio of IFN-γ to IL-5 in lymph nodes ., The Th1 induced immunity in PsSP9-immunized mice correlates with the protection observed against infection with L . tropica , which was associated with a significant reduction in ear thickness and decrease in parasite load in the ear of protected animals in comparison to control group ., Our data suggest that instead of using the whole sand fly salivary proteins , a more effective approach is the use of a single Ph . sergenti salivary protein . | medicine and health sciences, ears, immunology, sand flies, parasitic diseases, parasitic protozoans, otology, ear infections, protozoans, leishmania, insect vectors, antibody response, digestive system, infectious diseases, exocrine glands, head, otorhinolaryngology, disease vectors, immune response, eukaryota, anatomy, salivary glands, biology and life sciences, species interactions, organisms | null |
journal.pcbi.1004356 | 2,015 | How Co-translational Folding of Multi-domain Protein Is Affected by Elongation Schedule: Molecular Simulations | While in vitro folding dynamics of single-domain proteins has been relatively well understood by now1 , 2 , several additional factors make in vivo protein folding much more difficult to characterize ., About 70% of proteins have multiple domains and inter-domain interactions often cause many metastable intermediates and can hamper folding to the native states 3 , 4 ., Cellular environment is highly crowded by macromolecules , which affects folding kinetics and could cause aggregation 5–7 ., To circumvent some of these difficulties , several types of molecular chaperones facilitate folding 8 ., During protein synthesis in ribosome , nascent polypeptides start folding co-translationally 9 ., Co-translational folding ( CTF ) has been suggested for in vivo folding mechanism since 1960’s 10 and there is no room to doubt its relevance both in bacteria and in eukaryotic cells 11 ., Many elements in the CTF have been characterized 12 ., First of all , many proteins , once denatured in a test tube , do not refold with high probability , whereas they fold in the CTF condition ., Thus , as a rule of thumb , the CTF condition facilitates correct folding of many proteins 13 , 14 ., Ribosome is not just a machine for synthesis , but also helps folding of nascent chains at the exit tunnel and on the surface 15 ., The translation elongation is not at uniform rate , but there are some regions on mRNA where the elongation is markedly slowed down 16–18 ., This so-called elongation attenuation can be realized by a few mechanisms ., Most notably , for a given codon , the elongation rate is affected by its cognate tRNA binding kinetics , thus depending on the concentration of the cognate tRNA 19 ., The concentration of cognate tRNAs are highly correlated with the frequencies of codon usage for each of species ., There are some codons , of which the cognate tRNAs have markedly low concentration19 ., These rare codons sometime appeared in mRNA as a cluster , which often leads to translational attenuation ., On top , some portions of mRNA form partial secondary structures , which may slow down the elongation contributing to the elongation attenuation as well20 ., It was anticipated that the locations of the attenuation might have evolved to facilitate the CTF ., Some of them appear near domain boundaries of multi-domain proteins 21 ., By synonymous substitution of rare codons , one can speed up the translation elongation at a certain position without changing amino acid sequence , which led to reduce or impair functions and/or protease resistance for some proteins , such as an acetyl-transferase 22 and SufI 16 ., SufI in E . coli was recently used to test the role of translational attenuation in the CTF 16 ., SufI , an about 450-residue protein , is made of three domains; N- ( blue in Fig 1A ) , M- ( green ) , and C- domains ( red ) , in order ., Zhang et al . first identified three clusters of rare codons , two of which indeed exhibited elongation attenuation 16 ., Synonymous substitutions of some rare codons in these regions led to reduction or impair of protease resistance ., Separately , using a cell-free system , they also increased the concentrations of the corresponding tRNAs , which showed the similar results to the above synonymous substitution experiment ., It should also be noted that , they found no interactions of SufI with molecular chaperones ., Thus , these experiments provide us an unambiguous evidence of biological importance of the elongation attenuation for efficient folding in the CTF condition ., These experimental data can be complemented with theoretical and computational analysis to deepen our understanding on the CTF mechanisms ., Previously , lattice Monte Carlo simulations 23 , 24 and statistical theories 25 , 26 addressed physical aspects of CTF mechanisms ., Coarse-grained molecular dynamics ( CG MD ) was used to investigate interaction with ribosome in the CTF 15 , 27–29 ., These works helped understanding general and conceptual aspects of the CTF , but they were not specific enough to compare with experimental data of specific substrate proteins ., It is time to start computational study of CTF for a specific protein , of which clear experimental data are available ., This enables us to address structural aspects of CTF mechanisms , which is indeed the purpose of this work and we chose SufI for it ., Since the CTF becomes non-trivial primarily for relatively large and multi-domain proteins ( SufI has three domains and is about 450 residue long ( Fig 1A ) ) , all-atom MD simulations are not feasible for this problem at the moment ., By now , no all-atom MD simulation for folding to the native structure of multi-domain proteins was reported ., To overcome size and time scale limit in all-atom MD simulations , protein folding simulations have commonly performed by coarse-grained ( CG ) models that are based on the energy landscape theory 30 , 31 ., In particular , these simulations include medium-to-large proteins , such as multi-domain proteins32–34 ., Yet , to address mechanisms of the CTF and , in particular , an impact of elongation attenuation by CGMD simulations , technically , there are two major issues ., First , we need to realize misfolding as well as correct folding in a well-balanced manner ., Thus , the CG model needs to be calibrated so that an energy landscape is globally funneled in one hand and modestly rugged in the other hand ., There have been a considerable number of studies towards hybrid modeling of structure-based potentials for globally funneled landscape with sequence-dependent terms for modestly rugged surfaces 35 ., Yet , it should be noted that , currently , there is no established manner to balance the two aspects ., Thus , here we unavoidably take a heuristic and empirical approach ., Second , we need to design a scheme that mimics co-translational folding in silico ., Quantitative kinetic measurements and detailed mechanisms of translation attenuation are not available at the moment , which led us to take a rather simplistic modeling of CTF scheme ., Albeit these limitations , with the current CGMD , we can simulate complete folding and misfolding events of full-length SufI hundreds of times in scheme that mimics the CTF ., In this paper , we first describe computational modeling of CGMD for the CTF ., Then , we performed the CTF and , as a control , the refolding simulations of SufI , comparing these results ., Characteristics of misfolded structures are then analyzed ., Next , folding networks for these simulations clarify impacts of CTF and the elongation attenuation on folding reaction mechanisms ., Finally , the correlation between the degree of folding and the translation elongation time was investigated ., In the current CG modeling , each amino acid is represented by one bead located at the Cα position ., For folding simulations by CGMD , the so-called perfect-funnel model , or often called Go model , has been widely used giving many insightful lessons for folding dynamics 36–39 ., However , the perfect-funnel approximation may not be sufficient to study CTF dynamics where successful folding competes with misfolding , or non-native traps ., The latters are , by definition , not realized by the perfect funnel approximation ., To this end , here we developed a hybrid CG model where we added a generic hydrophobic ( HP ) interaction potential VHP to the Go model potential VGo; the latter is responsible for globally funnel-like shape of the landscape , while the former makes the landscape modestly rugged leading to many metastable non-native traps ., Concretely , the entire potential function of a protein is aVGo+bVHP ., The Go potential was parameterized based on the atomic interaction at the native structure , called the AICG model developed by Li et al 32 ., The HP interaction is a generic many-body potential that estimates how a hydrophobic residue is buried by other residues 40 ., The HP interactions were applied not only natively interacting pairs , but also any residues ., Detailed potential functions are described in Materials and Methods , Coarse-grained model ., As is well-known , proteins in vivo are gradually synthesized by ribosome from their N-termini and released from the ribosome exit tunnel , which we try to mimic in a simple manner ., In the protocol , amino acids are added one by one to the C-terminus of the nascent polypeptide chain with certain “translation” rates ( Fig 1B and 1C . See Materials and Methods , Coarse-grained model for details ) ., To investigate effects of elongation attenuation , we employed the following three translation rate schemes ( Fig 1D ) :, 1 ) The uniformly fast translation scheme ( a dashed line , designated as CTFfast ) ,, 2 ) the uniformly slow translation scheme ( a dotted line , CTFslow ) , and, 3 ) the non-uniform codon-based translation scheme ( a solid line in Fig 1D and 1C , CTFcodon ) that is dependent on the cognate tRNA concentration ., We note that , in our scheme , the in silico translation rate is not proportional to the translation rate predicted from cognate tRNA concentrations ., The translation attenuation was linked to a cluster of rare codons , which implies that the attenuation is a collective phenomenon and possesses distinct phases ., Thus , using a threshold of the predicted translation rate , we introduced a two-phase approximation where the in silico translation is either normal or slowed ., The slowed translation phase , of which translation speed is 100 times slower than the normal case , corresponds to the translation attenuation ., Since there is no quantitative kinetic measurement on the attenuation , this two-phase approximation and use of slowing factor 100 are rather simple , possibly over-simplified , schemes ., Yet , we consider it qualitatively captures some of the major features of the translation attenuation ., As far as the slowed phase is sufficiently slower that the normal phase , we expect qualitatively similar results ., The relation between the translational time scales and the inherent folding time scales is of crucial importance , which will be discussed at the end of the results ., Detailed in silico elongation scheme is described in Materials and Methods , Translational elongation scheme ., Additionally , Vtunnel was introduced to mimic the ribosome steric effect that is realized by a combination of a wall and a tunnel ( Fig 1E ) ., Note that we did not include any molecular representation of ribosome and thus the tunnel is merely to restrict the nascent chain in a confined geometry ., During elongation , a polypeptide chain is tethered to the base of the tunnel ., On average , about 28 residues resided in the tunnel ( S1 Fig ) ., After completing the elongation , the chain is released from the base ., We note that the exit tunnel was included to account for the gap between the residue at the catalytic center and the segment that can fold ., The codon-based translation rate is based on the codon ( sequence ) at the catalytic center ., In principle , some alpha helical structures can be formed in the exit tunnel depending of the sequence ( although , retrospectively , we did not find it ) ., For comparison , we also performed folding simulations of SufI in a refolding scheme , where a full-length polypeptide chain started folding from denatured conformations obtained by high temperature simulations ., No wall-and-tunnel potential Vtunnel was utilized in this scheme ., MD simulations were performed at 0 . 82TF* , where TF* is an upper limit of denaturation temperature in our CG model ., To determine the temperature , starting from the native state of SufI , we performed unfolding simulations for 1 x 108 time steps at many temperatures ., The lowest temperature at which we observed unfolding was defined as the upper limit of denaturation temperature TF* ., ( S2 Fig ) ., We note that , even with the CG modeling , accurately calculating the denaturation temperature is a formidable task for this size of proteins; using the standard replica-exchange method or multi-canonical ensemble method , we did not succeeded to obtain the reversible folding/unfolding trajectories ., First we compare a representative folding trajectory via the codon-based co-translational folding ( CTFcodon ) scheme with that via the refolding scheme ., Fig 2 illustrates folding time courses quantified as the so-called Q-score defined as the fraction of formed contacts that exist at the native structure , together with some representative snapshots ., In the refolding trajectory shown in Fig 2A , the protein first acquired one globular region , which roughly corresponds to the N-domain ., After a while , another globular region was formed , which contains , roughly , the C-domain and a half of M-domain ., They gradually coalesced and made a single globular structure , which was a deep misfolded trap; the protein stayed in this trap until the end of the simulation ., On the other hand , the CTFcodon trajectory in Fig 2B showed markedly different time course ., A cooperative folding of the N-domain at ~ 0 . 2 × 108 time step is followed by the folding of M-domain at ~ 1 . 2 × 108 time step ., Subsequently , at ~ 1 . 7 × 108 time step , the protein folded to near native structure in which the C-domain is partly misfolded ., Finally , at around 1 . 9 × 108 time step , it quickly transited into the native-like conformation ., More quantitatively , we repeated folding simulations of SufI 100 times both in the refolding and the CTFcodon schemes ., In each trajectory , we judged whether the protein is folded or not by a set of native-ness scores , Q-scores , at the final 100 structures of the simulation ( 0 ≤ Q ≤ 1 . Q = 1 at the native structure . We have both a generous and a stringent criteria for the judgment of folding . See Materials and Methods , Criteria for folding for more detail ) ., Using a stringent criterion of folding , of 100 trajectories we found 18 successful folding cases in the refolding scheme ( Table 1 ) ., Whereas , the CTFcodon resulted in 35 cases of correct folding ., To clarify the statistical significance of the difference , we computed the histograms of Q-scores of the final structures in each scheme ( S3 Fig ) ., The difference in Q-score probability distributions was tested by the Kolmogorov-Smirnov test , which gave p-value of 0 . 000174 ( Table 2 , See also Table 3 for pairwise Mann-Whitney U tests ) ., Thus , we conclude that the codon-based CTF simulation can fold SufI with significantly higher probability than the refolding can do ., We then investigate effects of translational attenuation regions in SufI sequence , that was studied in experiments 16 ., Experimentally , accelerating translation at certain slow translating regions , either by synonimous substitutions or by increasing concentrations of the rare tRNAs , inpaired SufI functions , most likely , due to misfolding ., To test this idea in simulations , we conducted folding simulations by the CTF scheme , in which the chain is elongated with a uniform and fast rate across the entire chain ( CTFfast ) ., In the same way as the CTFcodon case , we repeated the CTFfast simulations 100 times ., Using the same criteria for the judgment of folding , i . e . , Q-scores , we found only 20 cases of successful folding , which is much fewer than the CTFcodon scheme ., The statistical analysis of the distribution suggested that the difference is significant ( p = 0 . 00822 ) ., Actually , the result by the CTFfast scheme is statistically indistinguishable to that by the refolding scheme ( p = 0 . 556 ) ., This is consistent with the experiment of Zhang et al 16 ., Experimentally , lowering temperature could rescue the low-folding yield of the impaired folding scheme , which we now test in simulations ., For the purpose , we performed folding simulations of SufI by the CTF where the elongation is slow and is in a uniform rate entirely ( CTFslow ) ., Of 100 simulations , we found 25 successful foldings by the same criteria as above ., The statistics test resulted in no significance between th CTFslow and the CTFcodon schemes , while a subtle p value , p = 0 . 14 for the comparison between the slow and the fast CTF schemes ., To understand the CTF , comparison between the translation time scale and the folding time scale is of central importance ., To estimate relevant folding time scales , for individual domains , we performed kinetic folding simulations ., Time required to reach structures that have Q > 0 . 5 was computed for each of domains ( S4 Fig ) ., First , the M-domain is rather unstable and we could not observe successful folding of the standalone M-domain ., The time scale for rough folding of N-domain τN−fold was 1 . 4 × 107 time steps , which is longer than that of the C-domain , τC−fold = 3 . 6 x 106 time steps ., Interestingly , τN−fold is longer than the time to complete translation by the CTFfast scheme , τtranslation−fast ~ 4 . 4 × 106 , but is comparable to that by the CTFslow scheme , τtranslation−slow ~ 1 . 3×107 ., Importantly , when the time for completion of the translation of N-domain is comparable to or longer than the average folding time of N-domain , the success ratio of SufI is high ., To understand why the codon-based CTF can facilitate folding of SufI , we now look into misfolded structures ., For each of the four folding schemes , we analyzed probabilities of misfolding of individual domains at the ends of simulations ( Fig 3A . Statistical test given in Tables 4–9 ) ., Here , the misfolded state was judged by the Q-scores of individual domains ( To help understanding of typical Q-scores in SufI , we tabulated Q-scores of individual domains as well as those of interface for every snapshots in Fig 2 as S1 Table ) ., Clearly , misfolding in the N-domain and the M-domain occurred with the highest probability by the refolding scheme , which is followed by the CTFfast scheme ., The CTFcodon showed the smallest probabilities of misfolding for these domains ., Of the four schemes , the rank order in misfolding of N- and M-domains is well ( anti- ) correlated with the probability of successful folding of the full-length SufI ., ( Table, 1 ) In particular , probabilities of misfolding of the M-domain are markedly different between the refolding and the codon-based CTF ., We note that the M-domain is not very stable and cannot fold as an isolated domain ( S4 Fig ) ., Folding of M-domain is achieved by structural support of the N-domain ., In CTF schemes , when M-domain is synthesized and released from the exit tunnel , the N-domain has large chance to be folded ., The folding of the C-domain is not much different among the four schemes ., We now show some representative misfolded structures ( Fig 3B ) ., A conformation in Fig 3B, ( i ) taken from a refolding trajectory , is misfolded in the N-domain , while the M- and C-domains are well-folded ( The non-native contact map is given in S6 Fig, ( i ) ) ., In this structure , C-terminal end of the N-domain is unfolded and is flipped out to the left side in the figure ( See the block arrow ., Also , see S6 Fig, ( i ) for many non-native contacts in C-terminus of the N-domain ) ., With this flipped-out segment , three domains coalesced to form near-native domain-domain interfaces ., Once the interfaces are firmly formed , the protein is topologically trapped and an escape event from this trap was not realized ., Fig 3B, ( ii ) illustrates a case where C-terminus segment of the M-domain was entangled with the C-domain ( the block arrow ., See also the non-native contact map S6 Fig, ( ii ) ) ., Again , the domain-domain interfaces are near-native like , which makes an escape from this trap difficult ., The right cartoon of Fig 3B, ( iii ) shows the case that a N-terminal segment of the C-domain , 314–340 residues ( shown in red-and-gray striped pattern with block arrow ) goes through different paths from the native structure ( the left cartoon of Fig 3B, ( iii ) ) ., Next , we investigated the ensemble of folding pathways for the CTF and the refolding schemes ., To clarify folding pathways , we drew folding networks where nodes represent discretized conformational states and links represent transitions between the states41 , 42 ., Conformational states were discretized by the native-ness scores ( Q-scores ) and by the non-native contact scores ( N-scores ) ( See Materials and Methods , Discretization of states by Q-scores of parts for more details ) ., For each domain and each interface between domains , we defined Q-score and N-score ( we have six Q- and N-scores , in total ) ., As usual , the Q-score measures fraction of formed native contacts ., The N-score is defined as the number of non-native contacts normalized by its maximal number observed ., Each Q-score is categorized into 5 classes , while each N-score is divided into 3 classes ., Together , we have as many as 56 × 36 ~ 1 . 1 × 107 states ( nodes ) ., To simplify the network , we removed any loops that go from a node and return to the same node later ., All 100 trajectories were used to draw a network for each folding scheme ., We depict folding networks of SufI for four different folding schemes ( Figs 4 and S5 ) ., Comparing the folding networks of the refolding ( Fig 4A ) and the CTFcodon ( Fig 4B ) schemes , we found , first of all , that the network for the refolding has much more nodes ( 3284 nodes ) than the CTFcodon has ( 820 nodes ) ., By refolding , the protein exhibited much more divergent conformational states , many of which are characterized by low Q-scores and high N-scores ., Second , while the refolding scheme did not show any dominant pathways , the CTFcodon has a clear folding route from the top in the figure to the bottom ., Obviously , the CTF enforced SufI to fold vectorially from N-terminal , which provided constraints to the order of domain folding events ., In contrast , the refolding scheme made a protein fold freely from any segments resulting into diverse transitions ., The CTF restricts kinetics of proteins and reduces conformational ensemble being observed , and are consistent with earlier theoretical works23 ., The folding network for the CTFfast scheme ( S5A Fig ) apparently looks similar to that of the refolding ., The number of nodes found was 3108 , which is only slightly fewer than that of the refolding network , i . e . 3284 ., The transitions are diverse with no dominant pathway to the native state ., On the other hand , the slow CTF scheme showed the folding network ( S5B Fig ) rather similar to that in the CTFcodon scheme ., The number of nodes found in the slow CTF was 1096 , which is slightly larger than that found in the CTFcodon , i . e . 820 ., We see a single and nearly identical folding route in these two schemes ., It is interesting to ask to what extent the translation rate is designed ( optimized ) , via codon usage , to facilitate folding ., To this end , here we investigate the correlation , if any , between a putative translation rate and the degree of folding ., For the former , we simply use the translation rate , in arbitrary unit , predicted by an algorithm proposed in Spencer et al ( Fig 5B ) 43 ., This translation rate is encoded in the codon usage as well as tRNA concentrations and other factors , but not apparently dependent on the physical chemistry of folding ., For the degree of folding acquisition , we defined the progress of native-ness ΔQi in a nascent chain of the length i as, ΔQi=〈〈Q〉L=i−〈Q〉L=i−1〉100trajectories, where 〈Q〉L=i is the average Q-score when the nascent chain has the length i and 〈 〉100 trajectories means the average over 100 trajectories in the slow CTF scheme ., If ΔQi is high at i-th residue , a nascent chain gains Q-score without disturbance from more C-terminal region of the chain ., Note that if we used the codon-based CTF scheme in calculation of ΔQi , it would naturally correlate with the translation rate ., Importantly , however , we did not bias the CTF by the codon usage ., Instead , we used a uniform and slow CTF scheme ., Thus , ΔQi is not directly related to the difference in the translation rate , but is a purely physicochemical quantity determined by the amino acid sequence ., We note that ΔQi was smoothed by a window average of the 5-residue windows to reduce the noise ., The ΔQi profile shown in Fig 5A exhibits several peaks ., First , we focus on the peaks that correspond to folding of M-domain because it is the most difficult event ., We find a high ΔQi region around 280–310 , which well correlates with a translational attenuation region , 33-40kDa region ( 281–326 residues , grey shaded in Fig 5 ) ., Experimentally , synonymous substitutions of rare codons in this region reduced resistance to a protease 16 ., The other translational attenuation experimentally tested is 25-28kDa ( 214–240 residues ) , in which synonymous substitution of two leucine codons impaired the protease resistance of SufI ., In Fig 5A , we see peak in the ΔQi profile at ~245 ., More quantitatively , by using 200th-350th residues , we computed the correlation between the ΔQi profile and the translation rate profile ( Fig 5C ) finding the correlation coefficient 0 . 51 ., Thus , they are indeed , albeit modestly , correlated ., The highest peak of the ΔQi profile in Fig 5A is located at 166-th residue , which corresponds to the situation that the N-domain ( 1–143 residues ) is mostly released from the ribosome exit tunnel ., ( Remember that the average number of residues in the exit tunnel is 28 as in S1 Fig ) ., However , the translation profile in Fig 5B does not indicate any attenuation in this region ., It seems that misfolding in the N-domain is not very probable in any CTFs and thus translational attenuation at this point is not required for successful folding ., Comprehensively performing molecular simulations of co-translational folding ( CTF ) and refolding of SufI , we elucidated mechanisms of how translational attenuation can facilitate correct folding from structural perspectives ., First , coarse-grained simulations showed that the codon-based CTF , CTFcodon , exhibited higher probability of correct folding than the refolding did ., When the translational attenuation is removed , the CTFfast simulations resulted in the success rate similar to that by the refolding scheme ., When the elongation was uniformly slowed down , the CTFslow simulation gave essentially the same results as those of CTFcodon ., These are all consistent with recent experiments ., On top , the simulations provided much of structural and mechanistic insights ., Specifically for SufI , we found that the M-domain is least stable and can fold only when it is supported by the pre-folded N-domain ., Once a segment of the M-domain is entangled with either N- or C-domain , an escape from the trap was difficult ., Combining molecular simulations with biochemical experiments provided detailed mechanistic understanding of CTFs ., A recent theoretical study suggested that , under certain situations , fast translation can coordinate folding to the native structure 44 ., Apparently , this is not the case in our SufI simulations ., Whether slower or faster translation facilitates the correct folding depends on the folding kinetic network as was shown in 44 ., We need some more investigations for specific proteins , through which we know which scenarios are more common ., We note that the current CG modeling has some limitations ., One of the major limitations is on the time scales ., Using the CG modeling , one cannot easily estimate the absolute time scales of folding and translation ., Using a low viscosity in Langevin dynamics and structure-based potentials , we speeded up the folding kinetics some orders of magnitude ., Translation kinetic parameters in the normal and slowed phases are not accurately known ., This makes quantitative comparison difficult ., Another limitation is the balance between the structure-based potential and the sequence-dependent terms , which was determined empirically here ., Accurate modeling of these balances is highly desired in future work ., In this study , we studied folding of a three-domain protein SufI 45 ( Fig 1 , PDB code: 2UXT ) ., Starting with the PDB structure 2UXT , we removed the His-tag and modeled missing residues by MODELLER 46 , resulting in the 443-residue long protein model ., The model structured was refined by the energy minimization with AMBER 47 ., Using Pfam’s 48 , we defined three domains; N-terminal domain as 1–143 , the middle ( M- ) domain as 160–300 , and the C-terminal domain as region 314–443 ., Segments between two domains are termed linkers ., The linker between M- and C- domains are rather long and extended ., In the simulation , one residue is represented by one CG particle which locates at Cα position ., We used our in-house developing software CafeMol for all the simulations 49 ., The potential energy function consists of the native-based AICG2+ potential ( VGo ) and non-local many body hydrophobic interaction potential ( VHP ) ., The total energy Vtotal for the refolding simulation is given as, Vtotal=aVGo+bVHP, where a and b are coefficients to control the balance between two terms ., The potential for the CTF simulations is written as, Vtotal=aVGo+bVHP+Vtunnel, The native-based potential VGo is defined as 32:, VGo=∑kb ( bi−b0i ) 2+∑Vθi ( θi ) +∑Vφ , i ( φi ) +∑εloc , i , i+2exp− ( ri , i+2−r0 , i , i+2 ) 2/2wi , i+22+∑εloc , i , i+3exp− ( φi , i+3−φ0 , i , i+3 ) 2/2wi , i+32+∑i>j+3nativeεnon−loc , i , j5 ( r0 , ijrij ) 12−6 ( r0 , ijrij ) 10+∑i>j+3non−nativeεex ( drij ) 12, The first term keeps virtual bonds between consecutive amino acids , the second and the third terms represent statistical potential for virtual bond-angles and virtual dihedral-angles 50 ., The fourth and the fifth terms define native-based local interactions 32 ., The sixth term is non-local contact interaction for natively contacting pairs ., The last term is a generic excluded volume interaction ( See 32 for more details ) ., For the hydrophobic interaction , we take the function developed in40 , which is written in the form:, VHP=−∑i∈CαεA, ( i ) HPSHP ( ρi ), where εA, ( i ) HP is a parameter that reflects the hydrophobicity of amino acids for the amino acid type A, ( i ) ., SHP represents the buried-ness of the amino acid i and is defined as:, SHP ( ρ ) ={1ρ>1clinearρ+0 . 5 ( 1−clinear ) 1+cosπ ( 1−ρ ) 1−ρminρmin<ρ≤1clinearρρ<ρmin, where clinear and ρmin are constants and ρi represents local density and is calculated by:, ρi=∑j∈Cα , j≠inA, ( j ) uHP ( rij , rmin , A, ( i ) , A, ( j ) , rmax , A, ( i ) , A, ( j ) ) nmax , A, ( i ), where nA, ( i ) is the number of heavy atoms that defines the amino acid A, ( i ) represents and nmax , A, ( i ) is the maximum coordination number for particle type A, ( i ) ., The function uHP represents the degree of the contact between particle i and particle j and is defined as below a sigmoidal function:, uHP ( r , rmin , rmax ) ={1r<rmin12 ( 1+cosπr−rminrmax−rmin ) rmin<r<rmax0r>rmax, We note that the described hydrophobic interaction potential was first developed for a CG model that uses different resolution from the current work ., Thus we need to re-parameterize the function ., We estimated parameters rmin , A, ( i ) , A, ( j ) rmax , A, ( i ) , A, ( j ) , εA, ( i ) HP , nmax , A, ( i ) , and nA, ( i ) for each amino acid types in the following way ., Using Dunbrack’s culled PDB set 51 , we analyzed radius distributions of twenty types of amino acids ., For details , if a distance between heavy atoms of two amino acids is less than RvdW , i + RvdW , j + RvdW , H2O , where RvdW , i is the van der Walls radius of the atom i , we defined the distance between Cαs as an effective distance , obtaining a set of radial distribution of 20x20 amino acid combinations ., Then , we defined 95% confidence coefficient of their histograms as rmax , A, ( i ) , A, ( | Introduction, Results and Discussions, Materials and Methods | Co-translational folding ( CTF ) facilitates correct folding in vivo , but its precise mechanism remains elusive ., For the CTF of a three-domain protein SufI , it was reported that the translational attenuation is obligatory to acquire the functional state ., Here , to gain structural insights on the underlying mechanisms , we performed comparative molecular simulations of SufI that mimic CTF as well as refolding schemes ., A CTF scheme that relied on a codon-based prediction of translational rates exhibited folding probability markedly higher than that by the refolding scheme ., When the CTF schedule is speeded up , the success rate dropped ., These agree with experiments ., Structural investigation clarified that misfolding of the middle domain was much more frequent in the refolding scheme than that in the codon-based CTF scheme ., The middle domain is less stable and can fold via interactions with the folded N-terminal domain ., Folding pathway networks showed the codon-based CTF gives narrower pathways to the native state than the refolding scheme . | Proteins are synthesized in vivo by ribosome from their N-termini ., When N-terminal fragments of nascent proteins get out of the ribosome exit , they start folding , which is called co-translational folding ., It has been suggested that well-scheduled co-translational folding schemes would facilitate correct acquisition of their native structures for some multi-domain proteins ., In particular , an un-ambiguous experiment was recently reported for a model protein , SufI where pauses at certain positions in the translational elongation are obligatory for efficient folding ., Here , for the first time to our knowledge , we performed molecular dynamics simulations of SufI with co-translational folding as well as re-folding schemes ., We found a co-translational folding shceme with rare codon-based pauses indeed increased the success ratio of folding , which is consistent with recent experiments ., On top , molecular simulations provided much of structural insights on the folding routes and misfolding in the case of re-folding scheme ., This explains why pauses in the translational elongation rescue SufI from misfolding . | null | null |
journal.pcbi.1005627 | 2,017 | A data-driven modeling approach to identify disease-specific multi-organ networks driving physiological dysregulation | Complex disease conditions characterized by co-morbidities involve pathological dysregulation that evolves across multiple organ systems and over time ., Thus , a holistic approach is required to deconvolve the spatiotemporally distributed mechanisms of multifactorial disease pathogenesis at the tissue , cellular , and molecular levels of analysis ., From this systems perspective , time-series analyses of multiple organs are essential to determining the biological mechanisms of disease progression 1–4 ., New insights into complex disease mechanisms have been derived from analyses of gene expression across multiple human organs 5–7 ., The temporal dynamics of human multi-organ gene expression profiles have provided insight into the distributed mechanisms of diseases including hypertension 8 ., Such studies of animal models can be used to study disease pathogenesis by examining time points both before and after disease onset ., Existing studies have provided valuable information regarding the contributions of various organs to cardiovascular disease 9 , 10 , but the absence of global longitudinal studies precludes our understanding of the molecular mechanisms underlying disease pathogenesis ., Even when time-series data are available , complications with conventional analysis approaches often preclude new insights ., Common statistical methods that account for time as a categorical variable often fail to detect significant differences between the dynamics of phenotypes , necessitating an explicit consideration of time as a continuous variable in statistical analysis 11 ., Analytical techniques available to infer the network interactions underlying the gene expression dynamics require extensive experimental assessments of responses to targeted gene perturbations 12 ., The utility of combining time-series analysis with network-based approaches has been demonstrated extensively in developmental biology 13 , 14 , immunology 15–17 , neural systems 18 , 19 , and critical care medicine 20 , 21 ., Interactions between dynamics and structure have also been studied in mechanical , electrical , telecommunication , social , and economic networks 22 ., However , such approaches have been relatively underutilized in controlled studies of organismal physiology 23 ., It has been proposed that a triangular pattern of positive feedback—with vertices representing autonomic nervous system ( ANS ) activity , systemic inflammation ( INF ) , and renin-angiotensin system ( RAS ) signaling—underlies the pathogenesis of cardiovascular dysfunction in hypertension 24 ., Accordingly , cardiovascular function can be modulated by perturbations of peripheral T-cells 25 , bone marrow cells 26 , renal 27 , 28 and hepatic systems 29 , the adrenal gland 30 , as well as neurons and glial cells in the brain 31 , 32 ., Because of the positive feedback interactions amongst physiological systems involved in cardiovascular regulation 24 , it is difficult to determine causal mechanisms of disease pathogenesis ., The attribution of disease mechanisms can be facilitated by the temporal reconstruction of events underlying the multi-organ system’s evolution toward a pathological state 23 , 33 ., We performed such a temporal reconstruction by integrating experimental measurements with novel data-driven modeling and network analysis ., We profiled the temporal dynamics of ANS , INF , and RAS gene expression in the adrenal gland , brainstem , kidney , liver , and left ventricular muscle to characterize the multi-organ contributions to disease etiology ., We utilized a rat model of complex disease—involving cardiovascular , metabolic , and cognitive impairments—in which autonomic dysfunction is believed to be a key factor in controlling the development and persistence of the disease state 24 , 34 ., Hence , we refer to this model as an “autonomic dysfunction” phenotype ., Extensive evidence supports the relevance of this animal model to the pathogenesis of human hypertension ., For instance , pharmacological perturbations of ANS and RAS signaling , and surgical manipulations of ANS signaling , exert anti-hypertensive effects in both humans and the autonomic dysfunction model ., Other commonalities include elevated inflammation and elevated sympathetic activity that appears to precede hypertension in both humans and rats 24 , 35–38 ., Hence , it is plausible that the autonomic dysfunction animal model recapitulates key features of human disease pathogenesis ., We applied a robust technique for system identification to estimate the strength , direction , and sign of interactions amongst genes within and between organs ., We utilized a Hartley Modulating Function ( HMF ) -based system identification approach , which allowed us to estimate both continuous mathematical models of gene expression dynamics and corresponding network models of multi-organ gene regulatory interactions ., We analyzed the model structure and simulation results to test whether the temporal dynamics and gene regulatory interactions were globally affected during the pathogenesis of autonomic dysfunction ., We analyzed the model to identify disease-specific network motifs associated with aberrant temporal dynamics ., We were interested in identifying whether the gene expression dynamics and network interactions were prominently dysregulated in particular organs , suggesting an anatomical basis for disease development ., We further investigated whether single nucleotide variants were significantly associated with the transcription factor binding sites upstream of ANS , INF , and RAS genes ., Our analyses utilized a novel investigative framework to identify new candidate therapeutic targets for ANS-related diseases based on aberrant expression dynamics and network interactions involving genes in multiple organs ., All experimental work was performed according to protocols approved by the Thomas Jefferson University Institutional Animal Care and Use committee ., All protocols were approved by the Thomas Jefferson University ( TJU ) Institutional Animal Care and Use Committee ., Study subjects included male rats from the Spontaneously Hypertensive Rat ( SHR/NHsd ) and Wistar Kyoto ( WKY/NHsd ) strains , corresponding to autonomic dysfunction and control phenotypes , respectively ., Rats were purchased from Harlan Laboratories and experimental procedures were carried out one week following animal arrival at our facility ., All animals were housed socially in the TJU animal facility ., The facilities were maintained at constant temperature and humidity with 12/12 hour light cycles ( lights on at Zeitgeber time = 0 ) ., We harvested organ tissues at five time points: 4 , 6 , 8 , 12 , and 16 weeks of age ., Rats were humanely sacrificed via rapid decapitation ., CNS tissue was excised and the brainstem was isolated in ice-cold artificial cerebral spinal fluid ( 10mM HEPES; 140mM NaCl; 5mM KCl; 1mM MgCl2; 1mM CaCl2; 24mM D-glucose; pH = 7 . 4 ) ., We simultaneously harvested the adrenal gland , kidney , liver , and left ventricle of the heart ., Tissue samples were flash frozen and stored at -80°C ., Our original study was designed to include 50 animals ( 2 genotypes , 5 time points , 5 replicates ) ., One animal deceased prior to the designated time point for organ harvest ., Thirty-five animals were included in our study and 2–5 organ samples per strain were obtained at each time point for organs other than the brainstem ( S1A Fig ) ; 12 week brainstem tissues ( from n = 10 animals ) were not included in the present study as these samples were utilized for a parallel study that precluded the gene expression analysis employed here ., Five other animals were excluded from our study prior to performing qPCR analysis because either the respective RNA did not pass our quality criteria ( see below ) or because the tissue was used for other purposes ., S1A Fig shows the tissue samples included in our study for each animal ., Total RNA was extracted from 10–50 mg tissue samples using the Direct-Zol RNA extraction kit , which captures all RNA greater than 18 nucleotides in length ( ZYMO Research , Irvine , CA ) ., Samples were DNAse treated and stored at -80°C ., Concentration and integrity were assessed with a spectrophotometer ( ND-1000 from NanoDrop , Philadelphia , PA ) ., RNA samples with 260/280 ( nm/nm ) ratio <1 . 8 and 260/230 ratio 1 . 8–2 . 0 were purified with RNA Clean and Concentrator-100 ( ZYMO Research , Irvine , CA ) ., High-throughput PCR was implemented as described previously 39 , 40 ., Intron-spanning PCR primers were designed for 24 assays ( see Table 1 in S1 Text ) ., For each sample , 30 ng of total RNA was used ., The standard BioMark protocol ( Fluidigm , South San Francisco , CA ) was employed to reverse transcribe and pre-amplify cDNA samples for 12 cycles using TaqMan PreAmp Master Mix based on the manufacturer’s protocol ( Applied Biosystems , Foster City , CA ) ., The qPCR reactions were performed using a 192 . 24 BioMark Rx Dynamic Array for multiplex gene expression measurement ( Fluidigm , South San Francisco , CA ) ., Quantitative PCRs were implemented with 30 cycles ( 95°C for 15s , 70°C for 5s , 60°C for 60s ) ., We quantified qPCR products by determining threshold cycle ( Ct ) values ., We designed our study with Actb and Gapdh assays as potential reference genes for normalization ., To determine whether these assays were appropriate for normalization , we assessed the stability of these genes across samples , as in our previous studies 39 , based on a well established method for evaluation of putative reference genes 41 ., Our analysis revealed that Actb and Gapdh expression profiles were highly variable across samples , as has been shown previously ., No single gene showed consistent stability across all samples ., However , the median Ct across all genes was stable across samples , indicating superior utility of the median Ct as a ‘pseudo reference gene’ for data normalization ( S1B Fig ) ., Hence , we normalized the raw Ct data based on median expression levels , which were considered to represent reference gene expression levels ., For each sample ( s ) obtained from a specific organ, ( r ) at a specific time point ( t ) , we subtracted the median Ct computed across all genes ( g ) in that organ: Δ C t r g s ( t ) = C t r g s ( t ) - m e d ( C t r s ( t ) ) where med ( . ) is the median of the argument ., We next centered the data for comparison across genes based on the median expression level across all samples for each gene: Δ Δ C t r g s ( t ) = Δ C t r g s ( t ) - m e d ( Δ C t r g ( t ) ) ., We used −ΔΔCt values for analyses of gene expression ., We omitted Actb and Gapdh from all subsequent analysis due to ambiguity in the functional interpretation of the results ., Missing data based on our QC analysis were rare ( median = 1 . 8% , sd = 10% of samples per gene with NA values; median = 4 . 2% , sd = 9 . 1% of genes per sample with NA values ) ., Thus , missing data were imputed according to established approaches 6 , 42 by replacing missing values with the mean across 10 samples with most similar expression profiles according to Euclidean distance using the impute package in R 43 ., Note that we did not impute Brainstem data at age = 12 weeks because we did not obtain the gene expression data from the brainstem samples at this time point ., Both raw Ct and normalized data are available ( S1 and S2 Files ) ., To examine whether specific samples imparted systematic biases in our results , we implemented Principal Components Analysis in R using the princomp function ., To test whether the temporal dynamics of gene expression differed between autonomic dysfunction and control phenotypes , we applied the Optimal Discovery Procedure ( ODP ) using the EDGE package in R 44 ., Temporal profiles were modeled as natural cubic splines which connect a series of smooth polynomials between knots defined by the degrees of freedom for the spline fit 11 , 45 ( function ns with df = 3 in the splines package for R 46 ) ., The ODP analysis involved a comparison of a null model to an alternative model ., The null model was characterized by a single spline fit to the aggregated autonomic dysfunction and control time-series data for each gene ., The alternative model consisted of two splines fit to the respective phenotypes ., For each gene , errors between data points and fitted values were summed and squared for the null model ( SS0 ) and the alternative model ( SSA ) ., An analogue of the conventional F statistic was computed to evaluate the goodness of fit obtained for the null versus alternative model: F = ( SS0 − SSA ) /SSA ., The estimated distribution for this statistic was utilized to compute an estimate of the probability of the alternative model under the null hypothesis ( p value ) with correction for multiple testing ( q value ) ., Complete details can be found in 11 , 44 ., We scaled the data to the range ( 0 , 1 ) prior to implementing the HMF method ., To implement this scaling for expression profile E , we applied the following transformation: Escaled = ( E − min ( E ) ) / ( max ( E ) − min ( E ) ) ., We use E to represent Escaled in the context of our system identification studies ., The following description details the basic theory and procedure underlying the system identification approach using Hartley Modulating Functions ( HMF ) ., A signature attribute of this approach is that interaction coefficients can be estimated that jointly describe network dynamics and structure ., From a mathematical perspective , a principal advantage of the HMF-based system identification approach is that it obviates the need to compute temporal derivatives of the raw data 47 , 48 ., Instead , the interaction coefficients k ( see Fig 1D ) are determined by estimating inner products between both the expression data ( and the derivatives thereof ) and a set of basis functions—the carefully chosen Hartley modulating functions—and approximating these inner products using the Hartley transform to transform the data into the frequency domain ., This procedure facilitates the accurate and robust identification of continuous-time models from discretely sampled data , another principal advantage of the HMF method , as we have demonstrated previously 48 ., Furthermore , our approach entailed the use of powerful regularization techniques that mitigate against overfitting the interaction coefficients 45 ., Whereas frequency domain transformations of data have been previously been employed to implement systems identification , these approaches relied on optimization-based estimates of the interaction coefficients 49 ., In contrast , using HMF method , we directly estimated the interaction coefficients via regularized regression ., Thus , our approach , in principle , can overcome difficulties in parameter estimation that result from non-convex solution spaces characterized by local minima 50 ., The remainder of this section starts with a description of the mathematical underpinnings and implementation details underlying our use of the HMF method to identify multi-organ gene regulatory networks ., Following this description , we detail further analyses that demonstrate the robustness of our approach ., The expression level E of gene g in organ r at time t was modeled as follows for a data set with samples obtained between time t = 0 and time t = T:, d d t E r g ( t ) = ∑ i N r ∑ j N g k i j ( r g ) E i j ( t ) - γ r g E r g ( t ), where Nr is the number of organs and Ng is the number of genes ., The degradation coefficient for Erg is referred to as γrg ., For simplicity , we avoided using the degradation term explicitly such that degradation was implicitly incorporated in kij for i = r and j = g:, d d t E r g ( t ) = ∑ i N r ∑ j N g k i j ( r g ) E i j ( t ) ( 1 ), The full network can be compactly expressed in matrix form as, d d t E 11 d d t E 12 ⋮ d d t E N r N g = k 11 ( 11 ) k 12 ( 11 ) ⋯ k N r N g ( 11 ) k 11 ( 12 ) k 12 ( 12 ) ⋯ k N r N g ( 12 ) ⋮ ⋮ ⋱ ⋮ k 11 ( N r N g ) k 12 ( N r N g ) ⋯ k N r N g ( N r N g ) E 11 E 12 ⋮ E N r N g, with the following simplified representation:, d d t E ( t ) = K E ( t ) ( 2 ), Note that that parameter matrix K from Eq ( 2 ) is equivalent to the Jacobian matrix corresponding to this linear system: J = Jij where Jij = ∂fi/∂Ej , fi = dEi/dt , and ( i , j ) each refer to a particular gene-organ combination ., Thus , this matrix gives the influence of a gene in column j on a gene in row i ., We estimated the interaction coefficients k by applying the HMF method 47 , 48 ., This method entails the multiplication of Eq ( 1 ) by M different modulation functions ϕm ( m = 1 , 2 , … , M ) ., Integrating these products gives the following relation:, ∫ 0 T ϕ m d d t E r g d t = ∑ i N r ∑ j N g ∫ 0 T ϕ m E i j d t k i j ( r g ) ( 3 ), where ϕm = f ( t ) is chosen such that d d t ϕ m ( t ) = 0 for t = 0 and t = T . Note that the times t = 0 and t = T corresponding to sampled ages of 4 and 16 weeks , respectively , such that T = 12 in our computations ., The modulating functions ϕ are chosen as follows:, ϕm ( t ) =∑j=0n ( −1 ) j ( kn ) cas ( ( n+m−j ) ω0t ), where n is the order of the highest derivative of the system described by Eq ( 1 ) ( i . e . , n = 1 ) , cas ( x ) = sin ( x ) + cos ( x ) , and ω 0 = 2 π T . The integrals on the right and left hand sides of Eq ( 3 ) can be estimated using the Hartley transform 51 , the m-th HMF spectral component of gene expression profile E ( t ) , and the HMF spectra for the i-th derivative of E ( t ) 47 ., These computations are defined respectively as follows:, H r g ( ω ) = ∫ 0 T E r g ( t ) c a s ( ω t ) d t ( 4 ), H¯rg ( mω0 ) =∑j=0n ( −1 ) j ( jn ) Hrg ( ( n+m−j ) ω0 ) ( 5 ), H ¯ r g i ( m ω 0 ) = ∑ j = 0 n f 1 f 2 f 3 ( 6 ), where, f 1 = ( - 1 ) j n j d d t c a s i π 2, f 2 = ( n + m - j ) i ω 0 i, f 3 = H r g ( - 1 ) i ( n + m - j ) ω 0, Importantly , given a solution to Eq ( 4 ) , numerical solutions to Eqs ( 5 and 6 ) can be obtained ., The numerical solutions to Eqs ( 5 and 6 ) can be used to compute the interaction coefficients based on the following relations:, H ¯ r g ( m ω 0 ) = ∫ 0 T ϕ m E r g d t ( 7 ), H ¯ r g i ( m ω 0 ) = ∫ 0 T ϕ m d i d t i E r g d t ( 8 ), Then Eq ( 2 ) can be written as follows, H ¯ r g 1 ( m ω 0 ) = ∑ i N r ∑ j N g H ¯ r g ( m ω 0 ) k i j ( r g ) ( 9 ), where the only unknowns are the interaction coefficients , which can be determined by linear regression ., However , Eq ( 4 ) must be computed first ., Following our previous work 48 , we computed Eq ( 4 ) by linearly interpolating between average gene expression values at adjacent time points and analytically evaluating the integral:, ∫ t i t f E ( t ) c a s ( ω t ) d t = ∫ t i t f ( m t + b ) c a s ( ω t ) d t, m = E ( t f ) - E ( t i ) t f - t i , b = E ( t i ) - m t i, Then we computed the values in Eqs ( 5 and 6 ) given n = i = 1 48 ., Note that the analytical solution to the integral in Eq ( 4 ) evaluates to zero for n + m − j = 0 ( m = −1 , j = 0 and m = 0 , j = 1 ) ., However , allowing Eq ( 4 ) to be zero resulted in poor fits of the model to the data ., For these cases , we arbitrarily set n + m − j = ϵ with ϵ = 10−6 ., This choice of ϵ resulted in robust model fits as detailed below ., The set of interaction coefficients can be obtained for interactions regulating the expression dynamics of each organ/gene combination by solving Eq ( 9 ) using a range of m values:, H¯rg1 ( m1ω0 ) =∑iNr∑jNgH¯rg ( m1ω0 ) kij ( rg ) H¯rg1 ( m2ω0 ) =∑iNr∑jNgH¯rg ( m2ω0 ) kij ( rg ) ⋮H¯rg1 ( mMω0 ) =∑iNr∑jNgH¯rg ( mMω0 ) kij ( rg ), In matrix notation , this relation can be expressed as, y r g = X r g β ( r g ) ( 10 ), where y is a column vector of H ¯ r g 1 terms in which each row entry is computed with a different m value , X is a matrix of H ¯ r g terms with a row for each m value and a column for each term corresponding to a particular organ/gene combination , and β is a vector of interaction coefficients with the same length as the number of columns in X . Solving the linear system for β provides interaction coefficients that determine the dynamic regulation of a gene in a given organ , based on the expression profiles of all genes in all organs ., However , we wanted to perform system identification for gene regulatory networks spanning multiple organs and genes ., To globally estimate the entire set of interaction coefficients using the HMF method , a matrix of the form Eq ( 11 ) was utilized ., The full matrix was established using instances of Eq ( 10 ) for each organ/gene combination considered in the system:, y 11 y 12 ⋮ y N r N g = X 11 0 ⋯ 0 0 X 12 ⋯ 0 ⋮ ⋮ ⋱ ⋮ 0 0 ⋯ X N r N g β ( 11 ) β ( 12 ) ⋮ β ( N r N g ) ( 11 ), where 0 represents a matrix of zeros with the same dimensionality as X . In this formulation , y11 = X11 β ( 11 ) , y12 = X12 β ( 12 ) , and yNr Ng = XNr Ng β ( Nr Ng ) ., Thus , the solution to Eq ( 11 ) provides a global fit to the interaction network including all assayed genes in all sampled organs ., The regression problem can be described compactly as y = X β and solved using linear regression ., To circumvent overfitting of the model , we applied well established regularization techniques in which the regression coefficients were determined by solving an optimization problem with the following objective function:, J r e g = m i n β | | y - X β | | 2 + λ α | | β | | 1 + ( 1 - α ) | | β | | 2 2 ( 12 ), where ||x||1 = ∑|x| and | | x | | 2 = ∑ | x | 2 ., The regularization parameters α ∈ 0 , 1 and λ ∈ λmin , λmax ( see below ) impose sparsity on the network by forcing the interaction coefficients towards zero 45 ., This form of regularization is known as the ‘elastic net’ 52 , where the ||β||1 term represents the ‘lasso’ penalty 53 and ||β||2 term represents the ‘ridge’ penalty 54 ., Thus , α weights the lasso penalty and ( 1−α ) weights the ridge penalty ., The elastic net formalism exhibits positive attributes of both regularization techniques with respect to enhancement of network interpretation , based on sparsity of connectivity , and augmentation of prediction accuracy 45 ., We performed network identification as described above using a range of m value sets ., The regression problem represented by Eq ( 11 ) requires that the number of m values exceeds the number of interaction coefficients , Nr ⋅ Ng ., Each m value set had the form m = 0 , ±1 , ±2 , … , ± ( M − 1 ) , ±M , where M was varied between M m i n = N r · N g 2 and M m a x = N s a m p l e s 2 48 ., Initial simulations showed that the regression results were not sensitive to the range of the m values , given M ≥ Mmin ., We selected ten sets of m values that were evenly spaced within the M range ., For each of the ten m value sets , we performed the regression analysis for α = 0 , 0 . 2 , 0 . 4 , 0 . 6 , 0 . 8 , 1 , and we used 10 λ values for each α ., The λ values were also varied over a range ., The λ range was bounded by λmax , the minimal λ value associated with a particular α that resulted in all zero coefficients ., That is , for λ = λmax there was no network connectivity ., The λ values were incrementally varied from λmin = λmax × 10−4 to λmax on a logarithmic scale according to the default functionality of the glmnet package used for elastic net regression in R 46 , 55 ., Optimal solutions to the regularized regression problems ( 11 and 12 ) were determined based on simulation results ., We simulated the identified network models ( 2 ) using MATLAB’s ode45 and ode15s functions , or with R using the lsoda function from the deSolve package 56 ., Differences in simulation results , with respect to the choice of numerical integrator , were not visually detectable ., To select the optimal fits , we initially considered the sum of squared residuals as an objective function for minimization , ∑ ( y ^ - y ¯ ) 2 , where y ^ is the simulation result and y ¯ is the average for the corresponding gene expression data set ., However , according to this objective , the best fits could be those with all zero coefficients ., Hence , we revised our objective to penalize fits with low variability:, J s i m = ∑ ( y ^ - y ¯ ) 2 V a r ( y ^ ) + ϵ ( 13 ), and set ϵ = 10−10 ( the choice of ϵ did not make a detectable difference in the selection of the best fit ) ., Molecular networks underlying the physiology of blood pressure control have been shown to be distinct for autonomic dysfunction versus control 40 ., For all network reconstructions , we considered autonomic dysfunction and control phenotypes separately ., We initially performed our system identification analysis of a network with all organs ( Nr = 5 ) and genes ( Ng = 22 ) with ( Nr ⋅ Ng ) 2 = 12 , 100 possible connections ., We implemented 600 iterations of the system identification algorithm with distinct combinations of the m value range ( n = 10 ) , λ value ( n = 10 ) , and α value ( n = 6 ) ., In general , for both phenotypes , small λ values were associated for good fits ( i . e . , Jsim ∼ min ( Jsim ) ) irrespective of the α value for α > 0 ( S2 Fig ) ., For the control , Jsim = min ( Jsim ) for α = 0 . 2 at λ = 6 . 5⋅10−5 ( 221 m values ) ., For autonomic dysfunction , Jsim = min ( Jsim ) for α = 0 . 4 at λ = 8 . 3⋅10−7 ( 111 m values ) ., Based on these pilot studies , further analyses were implemented with ten λ values , ten sets of m values , and α = 0 . 2 ., To evaluate the robustness of the HMF Method , we compared the ‘best fit’ network with multiple subnetworks characterized by log ( Jsim ) < 10 ( see S2 Fig ) ., For these comparisons , we considered interaction coefficients that were larger in absolute magnitude than two standard deviations from the median ( median ∼ 0 in all cases ) ., Further , we only considered the coefficients for which | k | > | 2 σ ^ k - m e d ( k ) | in each phenotype-specific best fit network and comparison network , and analyses were completed for coefficients that met this criterion for both the best fit network and the comparison network ., Spearman rank correlation coefficients were determined and we investigated whether the coefficient sign was sensitive to the regularization parameters using the Fishers exact test ( FET ) ., For the FET analysis , we computed the sign ( i . e . , +1 or −1 ) of the coefficients considered in the correlation analysis ., The contingency table illustrated in S3 Fig was formulated for the FET ., Both Spearman rank and FET p-values were corrected for multiple testing according to the Benjamini-Hochberg method using the qvalue package 57 ., We also computed the odds ratio , based on the same data , to quantify the degree of agreement between networks ( S3 Fig ) ., To further evaluate robustness , we examined graph theoretic metrics ., A path through a network consists of the sequence of edges between two nodes or vertices ., The shortest path length refers to the minimal number of edges connecting two nodes , and the average path length < ℓ > is the mean of all shortest paths between every pair of nodes ., This measure is an index of the efficiency with which the network can be navigated ., The clustering coefficient Ci generally quantifies the connectivity among all nodes connected to node i ., The number of nodes with links to node i is ni and the number of links amongst the ni nodes is nc: Ci = 2nc/ ( ni ⋅ ( ni−1 ) ) 58 ., We computed a variant of this measure , the global clustering coefficient , which is the number of closed node triplets ( i . e . , ‘triangles’ ) divided by the total number of connected triplets ., The global clustering coefficient is also known as the ‘transitivity’ metric 59 ., Finally , we assessed the distribution of degrees kc; that is , the distribution for the number of edges connecting a node to its neighbors , which follows a power law of the following form for many biological networks: P ( k c ) ∼ k c - γ ., This distribution gives the probability that a given node has kc connections 58 , 60 ., To characterize the degree distribution , we fit a power law to the vector of degree frequencies and utilized a Kolmogorov-Smirnov ( K-S ) test of the null hypothesis that the data were distributed according to the power law ., The fit returned an estimate of γ and a p-value indicating the probability of the test statistic given the null hypothesis ., Thus , high p-values indicate the absence of evidence in support of the conclusion that the graph under consideration does not exhibit a power law degree distribution ., Network analyses were implemented using the igraph package for R 59 ., We assessed the patterns of gene expression dynamics across and within organs by temporally ordering the expression profiles identified using the HMF method ., We used temporal ordering schemes based on peak timing and valley timing , which were established as follows ., The dynamic profiles Ei of each gene i were scaled to Esc ∈ 0 , 1 ., We then determined all local extrema for the scaled waveform Esc ., We refer to local maxima as peaks and local minima as valleys , with associated times denoted as tp and tv ., Putative peaks and valleys ( t p ^ and t v ^ ) were considered as veritable extrema if their expression levels relative to the initial time point ( i . e . , t0 = 4 weeks ) exceeded a given threshold ( Eth ) :, i f | E ( t x ^ ) - E ( t 0 ) | > E t h ,, t h e n t x = t x ^ , x = p , v, For profiles with both peaks and valleys , we determined the first peak or valley that satisfied this condition ., We also classified monotonically decaying or increasing profiles as profiles without peak or valleys as defined above ., Further , monotonicity required the additional condition that, | E ( t x ^ ) - E ( t 0 ) | < E t h 0, for all x = ( p , v ) ., Monotonic increases versus decays were distinguished based on estimates of the profile’s first time derivative; tp = t0 and tv = t0 for monotonically decaying and increasing profiles , respectively ., For visualization , scaled profiles were sorted and plotted according to peak or valley time ., For this analysis , we set Eth = 0 . 5 and Eth0 = 0 . 1 ., We evaluated the differences in the multi-organ gene-gene interaction networks between autonomic dysfunction versus control phenotypes ., We describe our analyses based on conventional graph theoretic terminology , according to which the network is considered to be a graph G with vertices or nodes V that refer to genes which are connected by edges E: G = ( V , E ) ., The edges E are characterized by the interaction coefficients k described above ., In particular , we addressed whether edges were added , removed , or switched in sign from autonomic dysfunction to control ., We focused on the strongest network connections by considering edges for which the following condition was met:, | E i j | > m e d ( E ) + 2 s d ( E ), where Eij is an edge from node Vi to node Vj , med ( . ) is the median , and sd ( . ) is the standard deviation ., Median values computed across all edges were exactly zero ., Edges that did not meet our criteria were set to zero ., All edge values were scaled to the interval ( −1 , 1 ) by applying Esc = E/max ( |E| ) ., We evaluated differential network properties as follows ., For each edge in G , we first computed the difference between unscaled edge values for control ( WKY ) - versus autonomic dysfunction ( SHR ) -specific networks as, Δ E i j = E i j S H R - E i j W K Y, and we defined a threshold edge difference as follows:, E t h = m a x 2 s d ( E W K Y ) , 2 s d ( E S H R ), Then we evaluated each edge for a set of conditions ., Edges were considered to be added to the autonomic dysfunction network ( absent for control but present for the autonomic dysfunction phenotype ) if the following conditions were satisfied:, | E i j s c , S H R | > 0, | E i j s c , W K Y | = 0, | Δ E i j | > E t h, Note that the third condition ensured that an edge could not be considered to be added to the SHR network if that edge was slightly larger than Eth for SHR but slightly smaller than Eth for WKY , such that the actual difference was negligible . | Introduction, Materials and methods, Results, Discussion | Multiple physiological systems interact throughout the development of a complex disease ., Knowledge of the dynamics and connectivity of interactions across physiological systems could facilitate the prevention or mitigation of organ damage underlying complex diseases , many of which are currently refractory to available therapeutics ( e . g . , hypertension ) ., We studied the regulatory interactions operating within and across organs throughout disease development by integrating in vivo analysis of gene expression dynamics with a reverse engineering approach to infer data-driven dynamic network models of multi-organ gene regulatory influences ., We obtained experimental data on the expression of 22 genes across five organs , over a time span that encompassed the development of autonomic nervous system dysfunction and hypertension ., We pursued a unique approach for identification of continuous-time models that jointly described the dynamics and structure of multi-organ networks by estimating a sparse subset of ∼12 , 000 possible gene regulatory interactions ., Our analyses revealed that an autonomic dysfunction-specific multi-organ sequence of gene expression activation patterns was associated with a distinct gene regulatory network ., We analyzed the model structures for adaptation motifs , and identified disease-specific network motifs involving genes that exhibited aberrant temporal dynamics ., Bioinformatic analyses identified disease-specific single nucleotide variants within or near transcription factor binding sites upstream of key genes implicated in maintaining physiological homeostasis ., Our approach illustrates a novel framework for investigating the pathogenesis through model-based analysis of multi-organ system dynamics and network properties ., Our results yielded novel candidate molecular targets driving the development of cardiovascular disease , metabolic syndrome , and immune dysfunction . | Complex diseases such as hypertension often involve maladaptive autonomic nervous system control over the cardiovascular , renal , hepatic , immune , and endocrine systems ., We studied the pathogenesis of physiological homeostasis by examining the temporal dynamics of gene expression levels from multiple organs in an animal model of autonomic dysfunction characterized by cardiovascular disease , metabolic dysregulation , and immune system aberrations ., We employed a data-driven modeling approach to jointly predict continuous gene expression dynamics and gene regulatory interactions across organs in the disease and control phenotypes ., We combined our analyses of multi-organ gene regulatory network dynamics and connectivity with bioinformatic analyses of genetic mutations that could regulate gene expression ., Our multi-organ modeling approach to investigate the mechanisms of complex disease pathogenesis revealed novel candidates for therapeutic interventions against the development and progression of complex diseases involving autonomic nervous system dysfunction . | computer and information sciences, medicine and health sciences, genetic networks, anatomy, pathology and laboratory medicine, phenotypes, network analysis, neural networks, gene identification and analysis, gene expression, genetics, gene regulation, biology and life sciences, brain, pathogenesis, brainstem, neuroscience | null |
journal.pgen.1005759 | 2,016 | The Polymerase Activity of Mammalian DNA Pol ζ Is Specifically Required for Cell and Embryonic Viability | In eukaryotes , DNA polymerase ζ ( pol ζ ) is critical for the tolerance of many types of DNA replication blocks , by playing a central role in translesion DNA synthesis ( TLS ) ., Primary replicative DNA polymerases ( pol δ or pol ε ) are stalled when they encounter many types of template DNA adducts or DNA sequences forming stable secondary structures ., Such stalled replication forks are prone to formation of a dangerous DNA double-strand break ., The process of TLS helps avoid catastrophes by using a lower fidelity DNA polymerase ( such as pol ζ or pol η ) , to incorporate nucleotides across from a lesion ., TLS may occur either in S phase during primary DNA replication or in G2 phase during post-replication DNA synthesis ., In yeast and in mammalian cells , pol ζ is important for this process , but it leads to endogenous and DNA damage-induced point mutations because of errors introduced during TLS 1–5 ., Elimination of the pol ζ catalytic subunit Rev3l in mice leads to death during embryogenesis ( reviewed in 6 ) ., Primary cells in culture also cannot survive in the absence of Rev3l , because chromosomal DNA breaks quickly accumulate 7 , 8 ., Circumvention of damage-dependent checkpoints by SV40 large T antigen ( TAg ) immortalization of cells or by Tp53 knockout allows Rev3l-deficient cell lines to grow , but the cells continue to display gross chromosomal instability and DNA damage sensitivity 8–10 ., Mice conditionally deleting Rev3l in a fraction of hematopoietic cells or in basal skin keratinocytes are viable , but exhibit enhanced tumor incidence , as a consequence of the chromosomal instability of Rev3l-null cells 7 , 11 ., The hypersensitivity of REV3L-defective cells to some clinically-used DNA damaging agents indicates that REV3L is a possible target for enhancing the sensitivity of tumors to chemotherapeutic agents 12 ., Although the consequences of pol ζ disruption are dramatic , it is not clear that these arise from a specific DNA polymerase defect ., In mammalian cells , REV3L is a large protein ( >3000 amino acids ) , with multiple functional domains ., The DNA polymerase domain occupies only the last third of the protein ( Fig 1A ) ., The structural integrity of REV3L may be required in DNA processing complexes and for protein-protein interactions necessary to maintain cell viability and DNA integrity ., Indeed , REV3L serves as a multi-DNA polymerase scaffold ., The central region harbors two adjacent binding domains for REV7 ( gene name MAD2L2 ) ., REV7 is necessary for pol ζ activity in vitro and serves an important function as a bridge protein for interaction with the REV1 protein 13–15 ., REV1 in turn interacts with Y-family DNA polymerases that insert bases opposite sites of DNA damage and work in tandem with pol ζ 16–18 ., REV7 also has other cellular functions in chromatin assembly and structure 19–21 ., An N-terminal region of REV3 is conserved with yeast homologs 22 ., At the C-terminus of REV3L 23 , an Fe-S cluster is present that binds two other subunits of the pol ζ enzyme , POLD2 and POLD3 ., Both of these proteins also serve as subunits of the replicative DNA polymerase δ 23–26 ., More recently , a conserved positively charged domain in the central region has been recognized as necessary for the efficient polymerase function of the recombinant protein 24 ., Another domain in the central region has strong homology to the KIAA2022 gene ( S1 Fig ) ., A provocative hypothesis has been put forward to explain the severe genotoxic effects of Rev3l deletion 27 ., It was suggested that these are the consequence of the function of a second DNA polymerase , pol η ( gene Polh ) ., As in mammalian cells , chicken DT40 cells with a disruption of pol ζ exhibit growth defects , chromosomal aberrations and DNA damage sensitivity 27 ., Remarkably , it was reported that co-disruption of Polh and Rev3l corrects all of these phenotypes in DT40 ., The suggested interpretation was that pol η and pol ζ always work together in bypass of DNA damage , and that a toxic intermediate is formed by pol η that cannot be resolved in the absence of pol ζ ., It is clearly important to determine , in mammalian cells , whether the genome instability caused by pol ζ disruption is dependent on pol η ., Here we describe experiments with knockout cells and a specific knock-in mouse model to test whether the catalytic activity of pol ζ is responsible for the phenotypes observed in pol ζ knockout mutants ., We describe complementation of Rev3l-deficient mouse embryonic fibroblasts ( MEFs ) by expression of full-length human wild-type REV3L , and show that DNA polymerase-defective mutant REV3L cDNA is unable to complement cell survival or increased levels of DNA breaks ., Using a Rev3l polymerase-dead knock-in mouse model , we show that specific disruption of the polymerase activity prevents the completion of embryogenesis ., Finally , we tested whether pol ζ defects can be rescued by ablation of pol η function ., Rev3l deletion in mouse cell lines is associated with an elevated baseline level of DNA breaks and an increased sensitivity to DNA damaging agents such as cisplatin and UV radiation 3 , 8–10 ., We wanted to test definitively whether these phenotypes are caused by the disruption of Rev3l ., A pOZ expression vector harboring an IL2R selectable marker ( Fig 1A ) 28 was used to express human REV3L cDNA in Rev3l-deficient MEFs 8 ., Cells were selected for IL2R expression by repeated cycles of magnetic bead sorting and clonal populations were isolated ., The integrity of the expression vector was confirmed by PCR-based detection , and cells were assayed for expression of REV3L mRNA by real-time RT-PCR ., Human REV3L was expressed in the Rev3l-deficient MEFs at about one-half of the normal endogenous level ( Fig 1B ) ., Mouse cells expressing one or two alleles of Rev3l have indistinguishable low levels of spontaneous senescence , apoptosis , and chromosome aberrations 8 and there is no haploinsufficiency apparent regarding embryonic or adult viability in mice 7 ., We expressed both wild-type REV3L ( WT ) , as well as REV3L with a dual point mutation ( ASM: D2781A; D2783A ) in residues essential for divalent metal binding in conserved DNA polymerase motif I . Equivalent changes in all other tested DNA polymerases inactivate Mg2+ coordination in the active site , and eliminate enzymatic activity 29 , 30 ., We tested the growth of Rev3l-proficient and deficient cells expressing an empty vector ( EV ) , as well as deficient cells expressing WT and ASM REV3L cDNA ., Rev3l-deficient cells experienced S-phase associated delay and mitotic failure , leading to a population doubling time that was longer than Rev3l-proficient populations 8 ., REV3L re-expression in the deficient cell lines significantly decreased their doubling time to a level similar to Rev3l-/+ cells , whereas expression of the polymerase-inactive mutant had no effect ( Fig 1C ) ., Deletion of Rev3l causes sensitivity to DNA damaging agents 8–10 ., To determine whether REV3L expression could rescue this phenotype , cells were exposed to cisplatin or UVC radiation and cell survival was measured ., Rev3l-deficient cells displayed the expected sensitivity to these damaging agents when compared to the Rev3l-proficient cells ( Fig 1D and 1E ) ., Assays were repeated with multiple clones for each genotype ., Expression of wild-type REV3L rescued the sensitivity to all three DNA damaging agents , but expression of ASM REV3L did not ., Rev3l-deficient cells manifest an increased formation of DNA breaks in the absence of exogenous DNA damage ., We measured a 10 to 20-fold increase in cellular micronuclei ( Fig 2A and 2C ) in Rev3l-defective cells , with 30–40% of all cells displaying micronuclei ., The Rev3l defect was also accompanied by an increased frequency of DNA breaks as quantified by 53BP1 foci per cell , with a pronounced shift in distribution towards larger numbers of foci per cell ( Fig 2B and 2D ) ., Expression of wild-type REV3L in Rev3l-deficient MEFs rescued both of these phenotypes , but expression of ASM REV3L did not ., The frequency of sister chromatid exchange ( SCE ) was not decreased in Rev3l-deficient cells ( Table 1 ) , indicating that this mitotic recombination event is not impaired by a REV3L defect ., These experiments demonstrate that sensitivity to DNA damaging agents and the presence of DNA breaks in Rev3l-deficient cells is caused by the absence of the REV3L protein , and REV3L polymerase activity is required for prevention of these phenotypes ., We also investigated two reported human REV3L knockout lines designated 332 and 504 , derived from the Burkitt lymphoma cell line BL2 31 ., However , Rev3l mRNA is still transcribed in the 332 and 504 subclones , the subclones were no more sensitive to cisplatin than the parental BL2 , there was no significant increase in spontaneous double-strand break incidence in the subclones , and no complementation of the mild UV sensitivity was observed with Rev3L cDNA ( S2 Fig ) ., These results and uncertainties regarding the targeting strategy ( S3 Fig ) indicate that the BL2 subclones may not be pol ζ defective ., To determine the in vivo consequence of specifically inactivating the DNA polymerase function of Rev3l , a genetically engineered mouse was constructed to express an ASM knock-in allele from the endogenous promoter ( Fig 3A ) ., Variant lox sites 32 were used to control knock-in of the Rev3l allele ., The mice were crossed to CMV-Cre , producing a constitutive ASM allele ( abbreviated the “M” allele for the mice here ) , in a pure C57BL/6J background ., All steps of genomic engineering were extensively monitored by Southern blotting analysis ( Fig 3B ) , PCR analysis and DNA sequencing ., Heterozygote mutant Rev3l+/M mice were viable and fertile , demonstrating that the mutant allele does not have dominant-negative activity affecting viability ., Heterozygous mutant Rev3l+/M mice were bred and pups genotyped ., No homozygous mutant animals were identified at weaning ( Fig 3C ) ., In addition , 48 embryos from 6 pregnancies were isolated between 8 . 5 and 10 . 5 dpc ., Rev3lM/M embryos were rare at the earlier timepoints , and by 10 . 5 dpc only a few very small Rev3lM/M embryos were identifiable ., The severely impaired development of homozygous Rev3l ASM embryos mirrors the lethality of the Rev3l null allele on a C57BL/6 background 33 ., Due to the early embryonic lethality in Rev3lM/M embryos , we were never successful in deriving MEFs from them ., To circumvent this problem we crossed Rev3l+/M mice with Rev3l-/lox mice ., This mating produced embryos for derivation of viable Rev3lM/lox MEFs ., The floxed ( lox ) allele of Rev3l is functional , but becomes a knockout allele ( termed the Δ allele ) after action of the Cre recombinase ., We expressed Cre recombinase in the cells to yield Rev3lM/Δ MEFs ., The mice also harbored the mT/mG transgene to monitor Cre activity ., This mT/mG transgene constitutively expresses red fluorescent protein ( RFP ) ., When Cre is active , the RFP gene is removed and green fluorescent protein ( GFP ) is expressed 34 ., This allows GFP to be used for flow sorting and as a marker of cells in which Cre recombinase has been expressed ., Cre was introduced via an adenovirus vector into primary MEFs 8 to compare Rev3lM/Δ MEFs with Rev3lM/+ MEFs ( retaining a wild-type allele of Rev3l ) ., We measured cell growth , cisplatin sensitivity and DNA double-strand breaks in GFP-positive cells ., ASM MEFs had a growth defect compared to wild-type allele-containing MEFs ( Fig 3D ) and eventually failed to thrive ., ASM MEFs were hypersensitive to cisplatin , compared to control MEFs ( Fig 3E ) ., Additionally , there was a two to three-fold increase in the number of ASM MEFs containing 53BP1 and γ-H2AX foci ( a measure of DNA breaks ) compared to controls at 9 days after Cre recombinase expression ( Fig 3F ) ., These phenotypes are similar to those seen in Rev3l null primary MEFs 8 ( and compare Figs 3F and 4B ) ., This result demonstrates that the DNA polymerase activity of REV3L is specifically required to allow for cell proliferation , to protect genome stability and to moderate cisplatin sensitivity ., We wanted to determine in mammalian cells whether the DNA damage sensitivity and genome instability caused by pol ζ disruption is dependent on pol η , as has been reported for the DT40 cell line 27 ., We crossed parental mice with the genotypes Rev3l-/lox Polh-/- , and investigated the genotypes of the pups ., In the Polh-/- background , no Rev3l-/- mice were born ( Fig 4A ) , consistent with the complete lethality of the Rev3l-/- genotype in a Polh+/+ background 6 ., We attempted to produce Rev3l-/- Polh-/- MEFs from mouse embryos , but were unable to obtain sufficient material to produce viable MEFs because of the early death during embryogenesis ., Instead we derived primary MEFs from viable Rev3l-/lox Polh-/- embryos ., Following introduction of Cre via an adenovirus , Rev3l-/Δ Polh-/- cells were produced ., These Rev3l-defective primary MEFs had an elevated level of DNA breaks that was indistinguishable from Rev3l-/Δ Polh+/+ cells ( Fig 4B ) ., Consistent with published results 35 the pol η defect in Rev3l-/lox Polh-/- MEFs conferred enhanced sensitivity to cisplatin ( by comparison with Rev3l-/lox Polh+/+ MEFs ) ( Fig 4C ) ., A Rev3l defect independently enhanced cisplatin sensitivity , and the sensitivity of the Rev3l-/Δ Polh-/- and the Rev3l-/Δ Polh+/+ MEFs was similar ., Therefore , deletion of pol η does not rescue the cell and organismal defects caused by loss of pol ζ , showing that the absence of pol ζ does not create a pol η-dependent toxic intermediate in mouse cells ., A major objective of this study was to determine whether the catalytic activity of pol ζ is responsible for the severe consequences observed in pol ζ mutant mouse cells ., These include hypersensitivity to DNA damaging agents , a greatly increased generation of double-strand breaks in unchallenged cells , a slower growth rate , and a required role for pol ζ in embryonic viability ., The impetus for this question is the existence of numerous other functional domains within the catalytic subunit of REV3L ., These include a conserved N-terminal domain , two REV7 binding domains 14 , 19 , and a C-terminal Fe-S cluster that interacts with the POLD2 subunit and is necessary for in vitro activity ., In addition , the central region contains a conserved positively charged domain 24 that likely promotes protein-protein and protein-DNA interactions , and a KIAA2022 homology domain , described in detail here for the first time ( S1 Fig ) ., The presence of all of these domains introduces the possibility that the essential functions of REV3L could be structural , rather than directly related to the DNA polymerase activity itself ., A catalytically deficient but otherwise intact REV3L may have been able to specifically interact with protein partners and DNA substrates , allowing viability of cells and mice ., There is ample precedent for such a situation ., One example is the mammalian ERCC2/XPD gene ., Complete disruption of XPD is incompatible with viability 36 ., However , an amino acid substitution that inactivates the catalytic helicase activity of XPD specifically compromises nucleotide excision repair capacity , but allows cellular viability 37 ., This is because the presence of XPD as a subunit of transcription factor TFIIH is necessary for the integrity of that complex , even though XPD activity itself is unnecessary for transcription 38 , 39 ., Another example is provided by the REV1 protein ., REV1 has a DNA polymerase domain that can catalyze dCMP incorporation in DNA ., Cells lacking REV1 are hypersensitive to UV radiation , but this DNA damage tolerance activity does not require the polymerase catalytic domain of REV1 ., Instead , the damage tolerance activity is conferred by a protein-protein interaction domain at the C-terminus of REV1 that interacts with REV7 in pol ζ and with Y family DNA polymerases 40 ., Recently , a non-catalytic role has been reported for human DNA pol κ in protection against oxidative stresses 41 ., Here , we analyzed the consequence of a homozygous mutation of the Rev3l DNA polymerase active site ., No viable homozygous mice were produced , and the corresponding embryos died early in embryogenesis , as with a complete knockout allele ., To investigate cell-autonomous consequences of the specific polymerase alteration , we derived primary MEFs that carried one null Rev3l allele , and one active site mutant allele ., The growth defects , DNA break formation and cisplatin sensitivity of these cells were similar to cells harboring two null alleles 8 ., These results show that the DNA polymerase activity of REV3L is essential for all functions so far measured in mice and in cells ., Loss of Rev3l causes chromosomal instability in cells ., This complicates studies of the consequences of Rev3l deficiency , as genomic alterations may accumulate during each cell cycle and lead to new phenotypes ., A rigorous way to determine which phenotypes are directly caused by Rev3l loss is to complement the cells by expression of Rev3l cDNA ., Here we utilized a complementation system for REV3L in mammalian cells , allowing definitive testing of whether phenotypes seen in Rev3l-deleted cells are due to Rev3l-deletion 19 , 42 ., Our results with specific mutant cDNAs establish that the polymerase activity of REV3L is specifically essential for preserving genome integrity and protecting against DNA damage ., It is of course possible that other domains within REV3L also have critical functions for viability or genome integrity , and this complementation system will allow investigation of that possibility ., For example , we recently demonstrated that the REV7-binding domains of REV3L are essential for pol ζ function 19 ., We also attempted complementation of Rev3l-deficient phenotypes using human BL2 cell lines , reported to carry disruptions of REV3L 31 ., It is notable that there were no major differences in phenotypes between the wild-type BL2 cells and the nominal 332 and 504 REV3L mutants ., In contrast to the marked phenotypes found with Rev3l-deficient MEFs , the BL2 lines exhibited no statistically significant differences in cell doubling times , micronuclei formation or double-strand break formation as assessed by 53BP1 foci per cell ., A modest sensitivity of 332 and 504 cells to UVC radiation and cisplatin was not rescued by complementation with REV3L ., The limited sensitivity of 332 and 504 cells to a variety of DNA damaging agents has been noted 43–45 ., Others have also reported no significant differences in spontaneous DNA breaks in 332 and 504 cells compared to wild-type BL2 cells 43 ., In a study with wild-type BL2 cell extracts and extracts from the nominal REV3L-deficient cells 46 , it was concluded that REV3L does not contribute to acetylaminofluorine-induced frameshift mutagenesis ., This should probably be re-examined with a different REV3L-defective cell system ., It is possible that the modest increased sensitivity of the BL2 subclones to UV radiation and cisplatin 43 might be due to inadvertent disruption of an unrelated gene by the targeting strategy , as may have occurred with BL2 cells deleted for pol ι 47–49 ., Our data indicate that the 332 and 504 cell lines may not be truly ( or only ) REV3L-deficient , and are not well-suited for studies of REV3L function ., The DT40 chicken cell line has been widely used to examine the consequences of DNA repair defects , because it is amenable to genetic manipulation by homologous recombination ., Some characteristics of Rev3l-deficient DT40 cells are similar to Rev3l-deficient mouse cells , including elevated levels of spontaneous DNA breaks and sensitivity to DNA damaging agents ., Intriguingly , it was reported that deletion of polh ( pol η ) could rescue the severe phenotypes of Rev3l-deficient DT40 cells 27 ., This led to the model that the major defects in Rev3l –deficient cells are a consequence of a polh-dependent toxic intermediate ., To test this model in mammalian cells , we investigated whether Rev3l-/- Polh-/- mice could be generated ., We found that embryonic lethality of this double mutant was complete and similar in timing to Rev3l-/- Polh+/+ mice ., Moreover , Rev3l-/Δ Polh-/- MEFs showed levels of DNA breaks and cisplatin sensitivity analogous to that seen with Rev3l deletion in the presence of pol η ., The Rev3l-/lox Polh-/- MEFs were more sensitive to cisplatin than the Rev3l-/lox Polh+/+ MEFs , consistent with the cisplatin sensitivity of human polh-defective cells 35 ., Notably , the pol η defective and pol η pol ζ double mutant MEFs had similar sensitivities to cisplatin ., This epistatic interaction suggests that these two proteins act in the same pathway to mediate resistance to cisplatin ., In fact both polymerases can cooperate to bypass a cisplatin-DNA adduct 24 ., In summary , the severe phenotypes caused by Rev3l deletion cannot be rescued in murine cells by concurrent deletion of pol η ., This is consistent with results found in the yeast S . cerevisiae , where a Rev3 Rad30 ( pol ζ pol η ) mutant is more sensitive to ultraviolet radiation than a single Rev3 mutant 50 , 51 ., Although the absence of pol η causes sensitivity to some DNA damaging agents , it is not specifically toxic in the absence of pol ζ ., In the absence of pol ζ , it is possible that TLS does not occur at all , and that other modes of replication fork rescue are relied upon , which leads to a higher prevalence of DNA double-strand breaks 1 ., The genetic interaction between pol η and pol ζ reported for chicken DT40 cells might reflect a peculiarity of that cell line ., DT40 cells harbor mutations in TP53 , and no poli gene has been found in the chicken genome ., but it seems unlikely that either gene is relevant in this context ., Previously reported Tp53-/- Rev3l-/- MEFs are also pol i deficient ( an allele from the 129 ES cell background ) , and show major genome instability and DNA damage sensitivity 9 ., Polh poli double mutant mice are apparently normal with no deficits in development ., A poli defect does not exacerbate the UV radiation sensitivity of polh-defective mouse cells , indicating that pol ι does not have a significant backup function protecting against lethality in the absence of pol η 52 ., Our results with the Rev3l knock-in polymerase mutant mouse are relevant to development of REV3L as a target for chemotherapy ., Suppression of REV3L sensitizes cancer cells to cisplatin in mouse model systems , and can limit chemo-resistance 12 , 53 because loss of pol ζ diminishes point mutagenesis 2–5 ., These studies used siRNA knockdown of REV3L to demonstrate this effect , but future use of small molecule DNA polymerase inhibitors may be more clinically feasible ., Until now it has not been known whether inhibition of the catalytic activity of REV3L mimics the cytotoxic effects of a knockdown of the entire gene ., Our work demonstrates that loss of REV3L catalytic activity is equivalent , in the assays used here , to gene knockout ., This validates and encourages strategies to directly inhibit pol ζ DNA polymerase activity ., Rev3l-deficient TAg-immortalized MEFs were derived as in Lange et al 8 ., Briefly , MEFs were made from mouse embryos with the genotypes mT/mG+/- Rev3l-/lox or mT/mG+/- Rev3l+/lox , where “lox” represents a functional allele flanked by loxP sites ., The mT/mG transgene constitutively expresses RFP , until Cre recombinase activity removes the RFP and allows expression of GFP 34 ., The strain background of the mice used to derive these alleles was mixed C57BL/6 and 129 ., We genotyped DNA polymerase iota ( pol ι ) in these cell lines , because 129 mice carry a mutant allele of pol ι 54 ., All cell lines were heterozygous for this mutation , and so can be considered pol ι proficient ., These cell lines were immortalized with SV40 large T-antigen , and then treated with adenovirus Cre ( AdCre ) to delete the floxed allele of Rev3l ( and generate the knockout Δ allele ) ., The cells were subcloned and selected for GFP positivity and for complete deletion of the floxed Rev3l allele ., They were grown as in Lange et al 8 , in an atmosphere containing 2% O2 ., The primary MEFs were also derived and cultured in 2% O2 as in Lange et al 8 ., They were made from mouse embryos with the genotypes mT/mG+/- Rev3lM/lox or mT/mG+/- Rev3l+/lox , as well as from Rev3l-/lox Polh-/- or Rev3l-/lox Polh+/+ embryos ., The loxP-flanked allele of the Rev3l gene was deleted using AdCre adfection , and the deletion efficiency was measured as described 8 ., For all cell lines , cell number was counted at each passage , and was used to calculate population doublings and doubling time ., The BL2 parental cell line and subclones 332 and 504 31 were kindly provided by Claude-Agnés Reynaud ( Institut Gustave Roussy , Villejuif , France ) ., Genomic DNA samples from the three cell lines were compared using short tandem repeat ( STR ) fingerprinting by the Cell Line Identification Core at MD Anderson ., All yielded identical profiles of the 16 standard STR markers , confirming the relationship of the three cell lines ., The human REV3L full-length cDNA was acquired in the pUC19M1 vector from Zhigang Wang 55 ., The following modifications were made to the REV3L cDNA: a C-terminal Flag tag was added and the 5’-UTR was eliminated and replaced with an optimized mammalian Kozak sequence ., This cDNA was cloned into the pTSIGN vector , which contains an EF1α promoter and an internal ribosomal entry site ( IRES ) fused to a neomycin-eGFP reporter ., The active site mutation ( residues D2781A/D2783A in human REV3L ) was introduced into this pTSIGN-REV3L vector using PCR primers containing the REV3L mutations , and then the mutated PCR fragment was ligated into the REV3L-vector , replacing the wild-type sequence ., The full-length human REV3L gene was PCR amplified from the pTSIGN-REV3L and pTSIGN-REV3L-ASM vectors and was cloned into the pETDuet-1 vector ( Novagen ) ., The REV3L gene was removed from the REV3L-pETDuet-1 vectors using XhoI/NotI digestion , and the resulting fragments were inserted into the pOZN vector ( contains a Flag-HA tag on the N-terminal side of the inserted gene 28 ) ., For the pCDH vector , the XhoI/NotI fragment from the full-length REV3L-pETDuet-1 vector was inserted into the pCDH-EF1α-Flag-HA-MCS-IRES-Puro vector ( System Biosciences ) ., All vectors were completely sequenced to verify the integrity of the REV3L gene and the plasmid backbone ., Full-length Flag-HA tagged REV3L can be expressed from this cDNA 19 ., The pOZN-REV3L or pCDH-REV3L vectors were transfected into HEK-293T cells using lipofectamine 2000 ( Life Technologies ) , together with the retroviral packaging vectors psPAX2 ( plasmid 12260 , Addgene ) and pMD2 . G ( plasmid 12259 , Addgene ) ., 48 hr later , the media ( containing pOZ or pCDH lentivirus ) was collected ., It was filtered , and polybrene was added to 4 μg/mL ., This media was added to plates of immortalized MEFs ( pOZ ) or flasks of BL2 cells ( pCDH ) ., 48 hr later , the cells began selection for puromycin expression ( pCDH , 10 day incubation ) , or for IL2R expression ( pOZ ) ., The latter required incubation of the infected cells with IL2R-antibody conjugated magnetic beads followed by washing of the beads ( as in 56 , 57; IL2R antibody from Millipore , 05–170 ) ., This was repeated 5 times ., The population was then sorted for single-cells , and clones were selected and verified ., The cells were confirmed to contain both the N and C-terminal portions of the REV3L expression construct using the following PCR primers: NFwd: 5’ TAC ACA GTC CTG CTG ACC AC 3’ , NRev: 5’ GAG GTA AGG AAA GAT GCC ATG TAG 3’ , CFwd: 5’ ACC TAA CTC AGC ATG GCA TCT G 3’ , CRev: 5’ CGG AAT TGA TCC GCT AGA G 3’ ( at an annealing temperature of 50°C ) ., Expression of the recombinant human REV3L was confirmed using a human-specific Taqman assay ( Life Technologies ) at the exon 14–15 boundary: Ex14Fwd: 5’ CAC CTG GCC TTA GCC CAT TAT 3’ , Ex15Rev: 5’ CTC TTC TAA GAG TGT CAG TAT TAC TTC CTT TC 3’ Probe: FAM-MGB-5’ CAA CAG AAC CAA AAA CA 3’ ., In order to compare the recombinant expression to that of endogenous mouse Rev3l , we designed a set of primers and a probe that would recognize both human and mouse Rev3l , and would not amplify any knockout transcript ., The primers/probe were at the exon 26/27 boundary: Ex26Fwd: 5’ GTG AAT GAT ACC AAG AAA TGG GG 3’; Ex27Rev: 5’ GTG AAT GAT ACC AAG AAA TGG GG 3’; Probe: FAM-MGB-5’ TAC TGA CAG TAT GTT TGT 3’ ., An additional gene expression analysis was completed on the hREV3L-expressing BL2 cells in order to distinguish the endogenous REV3L transcript ( which was expressed at approximately equal levels in the REV3L knockout and wild-type BL2 cells ) from the exogenously expressed REV3L ., We used primers and a probe that crossed the FLAG tag on the exogenous gene: FlagFwd: 5’–GTCTTTGTTTCGTTTTCTGTTCTG C– 3’; FlagRev: 5’–GCTTGTCATCGTCGTCCTTG– 3’; Probe: FAM-MGB-5’–GCT GTG ACC GGC GCC TAC TCT AG– 3’ ., Gene expression ( with mouse or human GAPDH as an expression control ) was measured on an Applied Biosystems 7900HT Fast Real-Time PCR System ., To test sensitivity to chemical DNA damaging agents , the immortalized MEFs or BL2 cells were plated into white 96-well plates ( immortalized MEFs– 5 , 000 cells/well; BL2 cells– 10 , 000 cells/well ) ., The following day , various concentrations of cisplatin ( Sigma ) or bleomycin ( Sigma ) were added to the wells , and the cells were incubated for 48 hr ., Then the cells were lysed , a reagent was added that emits light in the presence of ATP ( ATPLite One Step , Perkin Elmer ) , and luminescence was measured using a plate reader ( Biotek Synergy II ) ., The luminescence measurement was normalized to undamaged control ., To test cisplatin sensitivity in Rev3l-deleting primary MEFs , 1 day after deleting the Rev3l floxed allele with AdCre , the cells were plated into white 96-well plates ( 10 , 000 cells/well ) ., On day 3 , cisplatin at various concentrations was added , and the cells were incubated for 5 days ., Then ATP content was measured by luminescence , as above ., To test sensitivity of immortalized MEFs or BL2 cells to UVC radiation , 3 x 105 cells were pelleted and resuspended in 300 μL of phosphate-buffered saline ., Three 100 μL drops were placed into the middle of a plastic dish and 10 μL aliquots from each were plated into 100 μL of growth media in a white 96-well plate after 0 , 2 . 5 , 5 , 7 . 5 , 10 , 15 or 20 J/m2 UVC radiation at a fluence of 0 . 4 J/m2 s-1 ., 48 hr after irradiation , ATP content was measured as above ., To measure the formation of DNA double-strand breaks , immortalized MEFs were plated in an 8-well chamber slide ., The following day they were fixed and stained for DAPI , 53BP1 and γ-H2AX , as in Lange et al 8 ., BL2 cells were applied to microscope slides using a Cytospin ( Thermo Scientific ) , and then fixed and stained as with the MEFs ., Immunofluorescence images were photographed through a Leica DMI6000B microscope ., Micronuclei were counted based on small , separate DAPI foci associated with DAPI-stained nuclei ., 53BP1 foci per cell were counted using the CellProfiler program 1 59 with a threshold correction factor of 1 . 7 ., To measure DNA double-strand breaks in the primary MEFs , cells were plated into 8-well chamber slides 7 days after deletion of the Rev3l floxed allele using AdCre ., 48 hr later they were fixed and stained for DAPI , 53BP1 and γ-H2AX as above ., Photographs of the immunofluorescence were taken on the Leica microscope , and cells containing double-strand breaks were scored as those with 3 or more 53BP1 + γ-H2AX foci ., Assessment of the Rev3L-/Δ and Rev3L+/Δ cell lines for sister chromatid exchanges ( SCEs ) was as described 57 ., BrdU ( 10 μM ) was added to growing TAg-immortalized MEF | Introduction, Results, Discussion, Materials and Methods | DNA polymerase ζ ( pol ζ ) is exceptionally important for maintaining genome stability ., Inactivation of the Rev3l gene encoding the polymerase catalytic subunit causes a high frequency of chromosomal breaks , followed by lethality in mouse embryos and in primary cells ., Yet it is not known whether the DNA polymerase activity of pol ζ is specifically essential , as the large REV3L protein also serves as a multiprotein scaffold for translesion DNA synthesis via multiple conserved structural domains ., We report that Rev3l cDNA rescues the genomic instability and DNA damage sensitivity of Rev3l-null immortalized mouse fibroblast cell lines ., A cDNA harboring mutations of conserved catalytic aspartate residues in the polymerase domain of REV3L could not rescue these phenotypes ., To investigate the role of REV3L DNA polymerase activity in vivo , a Rev3l knock-in mouse was constructed with this polymerase-inactivating alteration ., No homozygous mutant mice were produced , with lethality occurring during embryogenesis ., Primary fibroblasts from mutant embryos showed growth defects , elevated DNA double-strand breaks and cisplatin sensitivity similar to Rev3l-null fibroblasts ., We tested whether the severe Rev3l-/- phenotypes could be rescued by deletion of DNA polymerase η , as has been reported with chicken DT40 cells ., However , Rev3l-/- Polh-/- mice were inviable , and derived primary fibroblasts were as sensitive to DNA damage as Rev3l-/- Polh+/+ fibroblasts ., Therefore , the functions of REV3L in maintaining cell viability , embryonic viability and genomic stability are directly dependent on its polymerase activity , and cannot be ameliorated by an additional deletion of pol η ., These results validate and encourage the approach of targeting the DNA polymerase activity of pol ζ to sensitize tumors to DNA damaging agents . | Translesion synthesis allows DNA replication to occur in the presence of damaged DNA ., This process is mediated by low-fidelity DNA polymerases ( such as pol ζ or pol η ) that maintain genomic stability ., The action of these polymerases is crucial to limit cancer ., In mice , complete deletion of DNA pol ζ leads to embryonic lethality , and conditional deletion enhances tumorigenesis ., Pol ζ is a large protein with many domains that interact with other essential proteins and maintain the structural integrity of pol ζ ., It is not known if the polymerase activity of pol ζ mediates its essential activities ., Using a cell culture complementation system and in vivo knock-in mice , our work shows that pol ζ–mediated maintenance of genomic stability in the presence of DNA damage is absolutely dependent on its DNA polymerase activity ., Others have demonstrated in chicken cells that co-deletion of pol ζ and pol η rescues the pol ζ-dependent phenotypes , but our work in mice and in mouse cell culture does not support that conclusion ., These results demonstrate the physiological importance of pol ζ polymerase activity , and show that employing small-molecule inhibitors of the polymerase reaction is a valid strategy for sensitizing tumor cells to chemotherapeutic agents . | null | null |
journal.pntd.0005673 | 2,017 | Flavivirus and Filovirus EvoPrinters: New alignment tools for the comparative analysis of viral evolution | Flaviruses , including Dengue , Yellow Fever , Japanese Encephalitis and West Nile viruses , are significant public-health pathogens responsible for wide-spread epidemics ., Recently , another member of this genus , Zika virus ( ZIKV ) , has emerged as a global public health threat ( reviewed in 1 ., Two major ZIKV lineages have been recognized: an African lineage first detected in the Uganda Zika forest in 1947 , and an Asian lineage , first isolated in South East Asia during the 1950s , that has since spread to the Americas ( for review , 2 , 3 ) ., Phylogenetic analysis has revealed that both the African and Asian lineages can be further divided into distinct sublineages or groups 4 , 5 ., Recent studies have also shown that ongoing epidemics are accompanied by the continued diversification of viral sequences via accumulation of base substitutions and recombinant exchanges between related sub-groups 3 , 6 , 7 ., Members of the Flavivirus genus have been grouped based on their vectors ( reviewed in 8 ) ., Mosquito-borne human pathogens include ZIKV , Yellow Fever virus , four Dengue virus species , St . Louis and Japanese encephalitis viruses , and West Nile virus , along with other highly diverse less-characterized groups for review , 8 ., Although mosquitos are considered the primary vector for ZIKV transmission , recent studies have identified human to human transmission via sexual contact 9 ., Analysis of Filovirus human outbreaks during the last 49 years , from the initial 1967 Marburg virus outbreak in Germany through the most recent 2014–15 Ebola virus epidemic in West Africa and in the Congo , indicates that these pathogens will continue to pose serious public health risks ( reviewed in 10–12 ., Ebola virus species involved in these outbreaks and other non-human infections include the Zaire , Sudan , Taï Forest , Reston and Bundibugyo species , with the Zaire strains responsible for the most extensive human outbreak 13 , 14 ., Likewise , multiple Marburg outbreaks have occurred in Kenya , the Congo , Angola , Uganda and South Africa ( for review , 11 , 15 ., Studies indicate that each Filovirus genus may have its own particular transmission cycle that includes non-human primates , bats , rodents , domestic ruminants , mosquitoes and ticks ( reviewed in 16 ) ., While bats are considered the primary reservoir for many of these viruses 17 , 18 , studies on humans that survive acute Ebola/Zaire infections reveal the presence of persistent active virus within immune-privileged or tissue sanctuary sites 19 ., Phylogenetic analyses of both Ebola and Marburg strains responsible for human and non-human primate hemorrhagic fevers reveal that genetically identifiable strains from distinct lineages are associated with individual outbreaks; during these outbreaks , evolving sublineages have emerged 20–27 ., For example , sequence analysis of Ebola isolates collected during the 2014–2015 West African Zaire/Makona outbreak has revealed the presence of multiple distinct sublineages that can be temporally traced to an initial Guinea strain that diversified during its spread into Liberia and Sierra Leone 13 , 28–31 ., The availability of hundreds of Flavivirus and Filovirus genomic sequences is an important resource for acquiring insights into the evolution of these pathogens 32 , 33 ., Using current web-accessed alignment tools , when multiple viral genomes are compared , alignments are often difficult to visually assimilate given the large size of their readouts ., For example , a ClustalW alignment 34 of 14 ZIKV strains produces a 51-page readout ., In addition , web-accessed alignment programs restrict the number of viral isolates that can be compared in an individual alignment ., To circumvent these limitations , we have developed a multi-genome alignment method that can superimpose hundreds of one-on-one alignments to reveal sequence polymorphisms and conservation as they exist within a sequence of interest 35 , 36 ., Individual one-on-one input:database alignments can also be accessed directly from the input-centric readouts ., The combined EvoPrinter/Clustal alignment algorithms described here access databases of hundreds of Flavivirus or Filovirus genomes , allowing the user to input a full or partial viral sequence to initiate a comparative analysis ., EvoPrint readouts identify sequences shared by all selected strains , in addition to highlighting ( through color-coding ) unique base substitutions and those shared by subsets of database entries ., EvoPrinter databases currently contains 1 , 094 Flavivirus entries including 148 ZIKV strains and 460 Filovirus genomes with 393 Zaire isolates from the recent West African Ebola outbreak ., To demonstrate the utility of these comparative tools , we show how, 1 ) alignment readouts highlight unique bases in both the input and database sequences;, 2 ) multiple sublineages are identified within ongoing Florida , Dominican Republic , Puerto Rico , and Brazil ZIKV outbreaks;, 3 ) SNP analysis of other ZIKV strains also reveals different Central American , Caribbean and Chinese sublineages;, 4 ) novel Dengue2 and Zika recombinant viruses and their parental lineages were identified using differential SNP pattern screens;, 5 ) SNP patterns differentiate between Ebola/Zaire sublineages;, 6 ) host cell A-to-I hyper-editing within Ebola and Marburg genomes is identified by SNP profiling and, 7 ) inter-species multi-genome Ebola virus alignments can identify ultra-conserved sequences ., Genomic sequences were curated from the NCBI/Genbank database 32 , and additional information about virus strains was obtained from the Virus Pathogen Database 33 ., To ensure that duplicate genomes do not interfere in the identification of uniquely shared sequences among different strains , redundant entries ( detected by BLAST or Evoprinter alignments ) were excluded ., Database genome names contain the following information: species , NCBI designation , country of origin and year of isolation ., When available , additional information is included in the names , such as lineage assignments , group designations and/or serotypes 14 , 25 , 37–44 ., A lineage represents a set of genomes that differ from others within a species by a unique assemblage of sequence polymorphisms when compared to other species members ., Different lineages are often marked by greater than 50 unique lineage-specific base differences ., In addition to FASTA formatted sequences , each entry was formatted for enhanced-BLAT ( eBLAT ) alignments to speed initial database searches 36 , 45 ., For eBLAT alignments , each genome was indexed into non-overlapping 11-mers , 9-mers and 6-mers and used to generate independent BLAT alignments that are superimposed to produce an eBLAT readout 36 ., As of April 2017 , the Flavivirus EvoPrinter database contains 1 , 094 non-redundant genomes that include the following: 574 Dengue ( groups 1–4 ) ; 37 St . Louis Encephalitis; 115 West Nile; 110 Japanese Encephalitis; 70 Yellow Fever; 148 Zika; 8 Aroa-related; 7 Edgehill-related; 3 Entebbe-related; 3 Natya-related; 2 Spondweni-related; 12 Yaounde-related; 14 Insect-specific; 4 No Known Vector; 5 Seabird Tick-associated; and 8 Tick-borne genomes ., Flavivirus groupings correspond to those previously described 8 , 46 , 47 ., Databases will be updated when new genomes are submitted to NCBI ., The Filovirus database currently consists of 460 genomes that include 66 Marburg strains and 393 Ebola ( 371 Zaire , 10 Sudan , 7 Reston , 4 Bundybuygo , and 1 Taï Forest ) isolates ., Also included in the database is a single Cuevavirus genus strain , Lloviu Cuevavirus , isolated from European cave bats 48 , 49 ., To initiate the comparative analysis of a user-provided sequence , an eBLAT search is performed to identify database genomes that closely match the input sequence 36 ., User-supplied sequences can range from 100 bases to complete genomes ., Once the eBLAT search identifies the input species , one-on-one Clustal alignments using the alignment algorithms developed by 34 are generated between the input sequence and the intra-species database genomes ., Although BLAT alignments are significantly faster than Clustal comparisons , aligning bases at or near sequence ends are often missed due to insufficient K-mer alignment target lengths ., Pairwise alignments are then converted to distinguish between aligning bases ( upper case ) and non-aligning bases ( lower case ) within the input sequence for each comparison 45 ., This input-centric format allows for the superimposition of alignment data from an unlimited number of pairwise comparisons 35 , 36 ., In addition , holding one-on-one alignment data in memory instead of multi-genome alignments allows for user-customized comparisons ., To achieve higher throughput volumes and processing speeds , we wrote a Java-based program that employs multithread parallel processing 50 to generate pairwise alignments concurrently ., By random allocation of 144 computational threads , database search and alignment processing speeds are significantly enhanced using a Hewlett Packard 2 . 5GHz/512 GB RAM; 4 socket , 18-core processor server operating with the RedHat Enterprise Linux 6 operating system ., User-provided Flavivirus sequences ( including full genomes ) are automatically aligned to all intra-species database genomes and , to speed up processing times , alignments to the larger 18 kb Filovirus genomes are done incrementally , with the initial alignment round to the top ten eBLAT scoring Filovirus genomes ., Additional database genomes can then be added to include strains of interest ., From the genome selection tree , the user can select genomes for single or multi-genome comparisons with the input sequence ., The genome selection page orders the one-on-one alignments , based on the number of base mismatches with the input sequence ( least to most ) ., The selection page allows the user to, 1 ) view individual alignments with the input sequence by selecting the genome of interest;, 2 ) view multi-genome superimposed alignments of all or a selected subset of genomes in order to either identify shared or conserved sequences via an EvoPrint readout or to highlight sequence differences by generating an EvoDifference print readout , and, 3 ) initiate inter-species alignments ., By moving back and forth between the genome selection page and alignment readouts , the user can quickly add or remove viral strains from the comparative analysis ., Sequence differences in multi-genome EvoDifference print readouts are color-coded to highlight base differences that are, 1 ) unique to the input ,, 2 ) differ in only one of the database genomes , or, 3 ) differ in two or more of the database entries ( Fig 1 ) ., While sequence identity among the aligning genomes is indicated by gray-colored text in the EvoDifference print , conserved sequences within an EvoPrint readout are denoted by black text ( Fig, 2 ) and less conserved sequences are shown in gray font highlighted in green ., In addition , bases that are unique to the input sequence and not present in any of the database genomes included in the EvoPrint readout are highlighted in red ., The start and stop translation codons of open reading frames are highlighted when included in the alignment ., For Flaviviruses , protein boundaries for the processed polyprotein are annotated ( positions taken from the Virus Pathogen Resource 51 ) ., Sidebars to the right of the readouts delineate protein encoding ORFs ., Sequence lines in both EvoDifference and EvoPrint readouts can be expanded to view the alignment details for each of the database genomes and , by selecting a virus strain listed in the readout , the user can view its one-on-one alignment with the input sequence ., Amino acid alignments can also be viewed from one-on-one ORF alignments , to allow the user to assess whether nucleotide changes result in different encoded amino acids ., A tutorial that details these alignment steps is available at the Flavivirus or Filovirus EvoPrinter websites via the EvoPrinter homepage ( https://evoprinter . ninds . nih . gov/evoprintprogramHD/evphd . html ) ., As an alternative to a user-provided sequence analysis , a database genome can be selected as the input reference sequence for either individual or multi-genome alignments ., EvoPrinter keeps a library of one-on-one alignment data between all Flavivirus database entries and a separate library for Filovirus database alignments that can be accessed for rapid comparative analysis ., As with the user supplied input sequence search , database alignments are ordered on the genome selection page based on numbers of base mismatches compared to the input and individual alignments can be viewed by clicking on the database genomes ., To resolve different lineages and/or sublineages , the user should select ten or more genomes that have similar mismatch numbers with the input reference sequence and generate a multi-genome EvoDifference print ., On the genome selection page , the bracketed numbers after the database name represent the number of base mismatches with the input sequence ., In the readout , bases in black text indicate two or more database mismatches , and when these multi-genome differences are identical in two or more strains they frequently represent lineage or sublineage identity SNPs ., In other words , when multiple strains have the same base substitutions these SNPs can be considered markers of lineage progression ., By expanding readout lines that contain multiple mismatches , sublineages can be differentiated by their uniquely shared differences with the input ( see Figs 3 and 4 ) ., One-on-one EvoDifference print SNP patterns can be used to identify recombinant viruses and their parental lineages ., Virus strains that are closely related to the input sequence , as revealed by low mismatch numbers ( listed after the database genome name on the Genome Selection Tree ) , usually have randomly distributed base differences throughout their pairwise alignments with the input sequence ., Discontinuity in mismatch scores between related database genomes , as seen by a sudden jump in score values , are often due to one of two reasons ., First , a higher score can indicate a sublineage difference and in this case , the increased SNPs are randomly distributed throughout the alignment ., Second , the higher score could indicate a recombinant exchange , and in this case , a cluster of high-density SNPs ( a recombinant fragment from a more divergent minor parent ) would be flanked by regions of lower SNP densities ( from the major parent ) ., Alternatively , if the recombinant is aligned with a member of the minor parental lineage , a significantly reduced low-SNP density region ( corresponding to the above high SNP density cluster ) is flanked by regions of higher SNP densities ( from the major parent ) ., To identify members of the minor parental lineage , the database search is repeated using the region of the input sequence that generates the high SNP density cluster along with flanking sequences of the putative recombinant strain ., If members of the minor parental lineage are present in the database , they will likely have the lowest mismatch numbers when compared to the other database genomes ., By repeating the initial search using the complete or nearly complete recombinant genome and then comparing one-on-one alignments with members of both parental lineages , the genomic region that generates the high SNP density when aligned to a major parental lineage strain ( Figs 5A and 6A ) will show near identity within the corresponding region when aligned to a minor parental strain ( Figs 5B and 6B ) ., If a member of the minor parental strain is detected first , the members of the major parental lineage can be identified in database searches by using the low SNP density region plus its flanking higher SNP density regions and examining high mismatch scoring strains ., Differential SNP patterning can also be used to identify recombinant strains that are decedents of multiple rounds of recombinant exchanges with different partners ., For example , if not all of the high SNP density clusters observed in the recombinant / major parental lineage alignment have corresponding “SNP clearings” when aligned to a member of the minor parental lineage , then the recombinant strain most likely is a mosaic of different recombination events involving multiple partners ., To confirm putative recombinants , we recommend that additional recombinant detection programs be employed such as the Recombination Detection Program 52 ., Both one-on-one and multi-genome EvoDifference prints of related Ebola or Marburg strains can be used to identify genomic sequences that have undergone A-to-I editing by host cell adenosine deaminases ., When the conversion occurs within the replicative template of Filoviruses , the inosines are read as guanine residues resulting in T/U -> C substitutions in the negative stranded RNA genome ( for review , 53 ) ., In both one-on-one and multi-genome EvoDifference prints , hyper-editing appears as clusters of T or C unique substitutions depending on whether the editing occurred in the input sequence or database genome ., Resolving sublineages during a viral outbreak or epidemic facilitates the identification of the genetic heterogeneity among viral isolates , identifies the spread of related strains to different countries , and allows for the detection of recombinant variants ., Based on phylogenetic analysis , previous studies have identified major ZIKV groups: two African groups , consisting of West and East African sublineages 3 and a diverse Asian/Western hemisphere lineage ( for review , 54 ) ., The West African group contains isolates primarily from Senegal and Cote-d’Ivoire , while the East African sublineages can be further resolved into isolates from Uganda and the Central African Republic ., EvoDifference print readouts can be used to highlight sequence differences among related and evolutionary distant ZIKV strains ( Fig 1 ) ., Using the capsid , pre-membrane and envelope encoding region from the Zika_KU321639 . 1_Brazil_2015 strain as the reference input sequence , one-on-one alignments with twenty-nine ZIKV database genomes were selected to identify, 1 ) bases that are unique to the input ,, 2 ) bases that differed in only one of the database genomes ,, 3 ) sequences that differed in two or more database genomes , and, 4 ) sequences shared by the input and all selected database genomes ( Fig 1A ) ., Alignment details and the color-coded names of the database isolates included in the comparative analysis can be viewed by selecting line numbers ( Fig 1B ) ., In this example , sequence line number 975 was expanded to highlight SNPs that are unique to the input or database genomes and shared SNPs ., The expanded sequence line also highlights the greater SNP density of the more divergent African isolates ( located below the horizontal line ) when compared to the Asian isolates ( above the line ) ., Database genomes are ordered by their total number of base differences when aligned to the input sequence ( least to most ) ., The genome ranking and base differences are also part of the database selection page ., Differentially shared SNP patterns among multiple ZIKV isolates can be used to resolve individual sublineages ., For example , when 525 bases of the Zika_KF383118 . 1_Senegal_2001 NS5 coding region are used to generate an EvoDifference print with database genomes from different African sublineages , their base differences with the input Senegal isolate or SNP profiles resolve different sub-groups ( S1 Fig ) , that correspond to previously described sublineages 3 , 4 , 55–57 ., Phylogenetic footprinting , identifying evolutionary conserved sequence elements using multi-genome alignment protocols , has become an important tool for resolving essential genomic information 35 , 58 , 59 ., A significant advantage of EvoPrinter is the ability to rapidly change the cumulative evolutionary divergence stringency of a multi-genome comparison ., By moving between the genome selection page and the EvoPrint readout , one can quickly add or remove viral strains from the analysis to reveal different levels of conservation of essential elements , as they exist within genomes of interest ., For example , to identify previous characterized Ebola virus conserved transcriptional start and stop regulatory elements ( for review 60 , 61 ) , we generated a multi-genome EvoPrint of the Zaire_lin6_Kissidougou_GIN_C15_KJ660346 . 2_2014 strain that included 271 non-redundant genomes from 3 Ebola species ( 269 Zaire , 1 Bundibugyo and 1 Taï Forest ) ( Fig 2 ) ., In addition to resolving transcriptional regulatory elements that flank each of the seven Ebola virus genes , the divergence stringency of the EvoPrint is sufficient to highlight essential amino acid codons by revealing their less-conserved wobble positions and identify the transcription editing site within the GP gene ( Fig 2 ) ., The EvoPrint also delineates less-conserved intergenic regions and the evolutionarily variable GP mucin-like domain encoding region 62 ( Fig 2 ) ., As with the Filoviruses , near-base resolution of essential information is obtained with Flaviviruses ., A multi-genome EvoPrint was generated using the YellowFever_GQ379162 . 1_Peru_2007 NS3 encoding region as the input reference sequence , comparing it with 15 South American and African Yellow Fever strains selected from the Yellow Fever database ( S2 Fig ) ., Together the 15 strains provide a cumulative evolutionary divergence sufficient to resolve essential bases , as evident from the less conserved codon wobble positions ( S2A Fig ) ., Flavivirus SNP differences can also be accessed by expanding readout lines of multi-genome EvoPrints ., The shared SNP profiles of different Yellow Fever Virus sub-groups ( S2B Fig ) ., correspond to previously identified phylogenetic tree groupings 63 ., Shared SNPs that highlight differences between groups of viruses serve as ancestry informative markers for identifying sublineages ( for review , 64 ) ., We call these identity SNPs ( ID-SNPs ) , since they represent lineage markers for descendants of an earlier parental strain and multiple shared ID-SNPs , or profiles can be used to resolve different sublineages and illuminate ancestral relationships among ZIKV strains during spreading epidemics ., Most ID-SNPs highlight differences between a sublineage and all other strains outside of the sublineage that have maintained the same ancestral base at those nucleotide positions ( Figs 3 and 4 ) ., Phylogenetic tree comparisons of Asian/Oceania strains have revealed that the South American epidemic ( first identified in Brazil ) derives from a distinct sublineage that arose from an outbreak in French Polynesia in 2013 4 , 55–57 ( for review 4 , 65 , 66 ) ., Our SNP profiles of Brazilian isolates reveal that they can be further divided into at least four different subgroups based on non-overlapping ID-SNP patterns shared among 20 isolates ( Fig 3 and S3 Fig ) ., For example , when the Zika_KX447510 . 1_FrenchPolynesia_2014 strain is used as the input reference genome and aligned to 13 Brazilian isolates , 3 subgroups ( Br1-3 ) ( each represented by multiple isolates ) were distinguished by 22 ID-SNPs that are positioned throughout the genome ( Fig 3 ) ., When isolates from China , Ecuador , Florida , Dominican Republic , Puerto Rico , Suriname and French Guiana are included in the analysis , all five of the Florida isolates , all of the Ecuador , and two of three Dominican Republic strains share ID-SNPs with the first Brazilian subgroup ( Br1 ) but not with the Br3 subgroup ( Fig 3 ) ., The second Brazil sublineage ( Br2 ) shares ID-SNPs with Florida isolates and with the Puerto Rico strains but not with Br1 or Br3 ( Fig 3 ) ., The alignment also reveals that the Puerto Rico , Suriname , French Guiana and a single Dominican Republic isolate share ID-SNPs with the third Br3 Brazil subgroup but not with the Br1 sublineage ., In addition , while isolates from Florida and Puerto Rico represent two distinct subgroups , the ID-SNP patterns of isolates from the Dominican Republic reveal that one isolate is related to the Puerto Rico subgroup while the other two share ID-SNPs with the Florida subgroup ( Fig 3 ) ., Interestingly , pairwise alignments between the Dominican Republic isolate that is related to the Puerto Rico subgroup , the Zika_KX766028 . 1_DominicanRepublic_2016 strain , and any of the China Ch2 sublineage members reveal near identity , suggesting that the Ch2 sublineage may have originated from the Caribbean ( Fig 3 and S4 Fig ) ., This possibility is further strengthened by the observation the China Ch2 strains share many ID-SNPs with isolates from Puerto Rico , Dominican Republic , Suriname , French Guinea , and members of the Brazil Br3 sublineage ., In addition , these observations are in agreement with Zhang et . al . , who report the presence of highly diversified ZIKVs that have been most likely imported into China 67 ., Comparative analysis of isolates from the recent southern Florida outbreak identify ancestral ID-SNPs that together suggest a progressive evolutionary divergence away from other related strains and other members of the Asian lineage ., For example , an EvoDifference print of the Zika_KX832731 . 1_Florida_2016 isolate with 71 other Asian/Oceanian/Western hemisphere strains ( both related and evolutionarily distant ) revealed ID-SNPs that are shared among Florida and Dominican Republic isolates while all other strains have maintained the same ancestral base at those positions ( Fig 4 ) ., Our analysis also identified ancestral ID-SNPs that are restricted to just a subset Florida and Dominican Republic strains and ID-SNPs that only distinguish a subset of Florida isolates from all other Asian lineage strains ., Taken together , the different subgroups indicate that progressive , multi-generation base substitutions at different genomic positions are playing a significant role in ZIKV divergence ., In addition , the multi-genome analysis demonstrated that the KX832731_Florida strain has recently acquired three unique SNPs that are not shared by any of the other Asian/Oceanian/South American strains ( two of the three unique SNPs are red highlighted in Fig 4 ) ., We have also used ID-SNP profiles to search for additional Western Hemisphere sublineages by examining pair-wise alignments of South/Central American and Caribbean isolates ., Our screen identified two Central American sublineages , differentiated from the Brazil Br1-4 subgroups by combinations of 15 ID-SNPs ( S3 Fig ) ., These subgroups contained isolates from Mexico , Guatemala , Honduras , Panama and Columbia ., Strains from Mexico fall into either the first or second central American group ., Our comparative analysis also revealed that the single Martinique isolate , Zika_KU647676 . 1_Martinique_2015 , most likely originated from a Mexican strain as it differs from the Zika_KU922960 . 1_Mexico_2016 isolate by only 4 bases ., To examine sublineage heterogeneity among Asian and Southeast Asian ZIKV strains , we searched for ID-SNPs that group isolates from different locations ., As indicated above , our SNP pattern screen revealed two Chinese subgroups that are differentiated by 31 ID-SNPs ( Fig 3 and S4 Fig ) ., Using Zika_KU955589 . 1_China_2016 as the input reference genome , our multi-genome analysis revealed that the Chinese Ch2 subgroup shares many ID-SNPs with Western hemisphere isolates , while the first China subgroup ( Ch1 ) constitutes a distinct ( perhaps older ) Asian sublineage ( Fig 3 and S4 Fig ) ., The French Polynesian strains share six ID-SNPs with the Ch1 subgroup and the Tonga strain shares eight ID-SNPs , suggesting that strains from Tonga and French Polynesia may be evolutionarily positioned between the Chinese Ch1 sublineage and Western hemisphere isolates ( S4 Fig ) ., Genomic diversity among Flaviviruses is driven in part by homologous recombination between related strains , with their recombinant exchanges occurring in both protein encoding and noncoding sequences 3 , 7 , 68 , 69 ., Alignment programs that scan for changes in sequence homology within multiple genomes and methods that examine differential phylogenetic clustering using genomic sub-regions have been used to identify recombinants and locate approximate recombinant fragment boundaries 70 , 71 ., Evoprinter screens can also identify recombinants and resolve the approximate boundaries of their recombining fragments within parental lineages ., By examining a previously characterized Dengue2 recombinant , we show how SNP profiling can be used to identify recombinant strains and their parental sublineages ( S5 Fig ) ., Phylogenetic tree clustering analysis of the Dengue2_AF100466 . 2_Venezuela_ 1990 ( Mara4 ) strain with other Dengue2 genomes revealed that Mara4 is the recombinant progeny of two distinct Dengue2 sublineages 71 ., Differential phylogenetic clustering analysis revealed that the first ~500 bases of Mara4 are nearly identical to Dengue2 strains from Thailand , while the remaining genome is related to American strains 71 ., Side-by-side EvoDifference SNP profile comparisons of the Mara4 recombinant with members of the parental sublineages ( from Thailand and Jamaica; S5A and S5B Fig , respectively ) demonstrate that the 5’ recombinant fragment originated from the minor parental Thailand sub-group ( boxed region in S5 Fig ) ., Note that , by convention , the strain that produces the highest SNP density within the recombinant region when aligned to the recombinant strain is designated as the major parental lineage , while the minor parental sub-group shares identity or near identity with the recombinant within the boundaries of the recombinant fragment ., In this example , the differing parental SNP pattern boundaries are located at positions 594 ( major parent ) and 600 ( minor parent ) , indicating that the recombinant exchange most likely occurred between bases 595 and 599 ( S5 Fig ) ., One advantage of the genome SNP profiling is that recombinants and their parental lineages can be identified by differental SNP patterning ., For example , Fig 5 identifies a novel Dengue2 recombinant strain ., Side-by-side comparisons of SNP profiles generated from one-on-one EvoDifference prints of the Dengue2_GQ398269 . 1_PuertoRico_1994 strain with another Puerto Rico strain ( Dengue2_KF955363 . 1_PuertoRico_1986 ) and with a New Guinea isolate ( Dengue2_AF038403 . 1_NewGuinea_1988 ) –Fig 5A and 5B , respectively—revealed that the Puerto Rico_GQ398269 . 1 strain is the resultant progeny of a recombinant exchange between a member of a Puerto Rican subgroup ( major parental sublineage ) and a New Guinea sub-group member ( minor parental sublineage ) ( Fig 5 ) ., The abrupt SNP density pattern change within the recombinant Puerto Rico/New Guinea strain alignment delineates an ~2 , 100 base region ( spanning the NS2B and NS3 protein encoding sequences ) that is identical in both the recombinant and New Guinea genomes ( Fig 5B ) ., Note that the higher density SNP cluster in the Puerto Rico ( major parent ) –recombinant strain SNP profile alignment corresponds to the region of identity shared between the recombinant and the minor parental strain ( Fig 5A and 5B ) ., Using SNP profiling , we have sought evidence of recombination within Asian and African ZIKV lineages ., Our initial screen of the China Ch-1 sublineage isolates revealed that many are nearly identical , however , the SNP profile generated when the Zika_KU963796 . 1_China_2016 strain was aligned to Zika_KU866423 . 1_China_2016 identified two genomic regions that have significantly higher SNP densities when compared to flanking sequences ( Fig 6A ) ., Further analysis that included other Asian strains revealed that when the KU866423 . 1 strain was aligned to a Cambodian isolate , Zika_JN860885 . 1_Cambodia_2010 , their genomes are identical in the same two regions that displayed higher density SNP clustering in the above KU866423 . 1—KU963 | Introduction, Materials and methods, Results and discussion | Flavivirus and Filovirus infections are serious epidemic threats to human populations ., Multi-genome comparative analysis of these evolving pathogens affords a view of their essential , conserved sequence elements as well as progressive evolutionary changes ., While phylogenetic analysis has yielded important insights , the growing number of available genomic sequences makes comparisons between hundreds of viral strains challenging ., We report here a new approach for the comparative analysis of these hemorrhagic fever viruses that can superimpose an unlimited number of one-on-one alignments to identify important features within genomes of interest ., We have adapted EvoPrinter alignment algorithms for the rapid comparative analysis of Flavivirus or Filovirus sequences including Zika and Ebola strains ., The user can input a full genome or partial viral sequence and then view either individual comparisons or generate color-coded readouts that superimpose hundreds of one-on-one alignments to identify unique or shared identity SNPs that reveal ancestral relationships between strains ., The user can also opt to select a database genome in order to access a library of pre-aligned genomes of either 1 , 094 Flaviviruses or 460 Filoviruses for rapid comparative analysis with all database entries or a select subset ., Using EvoPrinter search and alignment programs , we show the following:, 1 ) superimposing alignment data from many related strains identifies lineage identity SNPs , which enable the assessment of sublineage complexity within viral outbreaks;, 2 ) whole-genome SNP profile screens uncover novel Dengue2 and Zika recombinant strains and their parental lineages;, 3 ) differential SNP profiling identifies host cell A-to-I hyper-editing within Ebola and Marburg viruses , and, 4 ) hundreds of superimposed one-on-one Ebola genome alignments highlight ultra-conserved regulatory sequences , invariant amino acid codons and evolutionarily variable protein-encoding domains within a single genome ., EvoPrinter allows for the assessment of lineage complexity within Flavivirus or Filovirus outbreaks , identification of recombinant strains , highlights sequences that have undergone host cell A-to-I editing , and identifies unique input and database SNPs within highly conserved sequences ., EvoPrinter’s ability to superimpose alignment data from hundreds of strains onto a single genome has allowed us to identify unique Zika virus sublineages that are currently spreading in South , Central and North America , the Caribbean , and in China ., This new set of integrated alignment programs should serve as a useful addition to existing tools for the comparative analysis of these viruses . | Flaviviruses , including Zika and Dengue viruses , and Filoviruses , including Ebola and Marburg viruses , are significant global public health threats ., Genetic surveillance of viral isolates provides important insights into the origin of outbreaks , reveals lineage heterogeneity and diversification , and facilitates identification of novel recombinant strains and host cell modified viral genomes ., We report the development of EvoPrinter , a web-accessed alignment tool for the rapid comparative analysis of viral genomes ., EvoPrinter superimposes alignment data from multiple pairwise comparisons onto a single reference sequence of interest , to reveal both similarities and differences detected in hundreds of selected viral isolates ., Evoprinter databases provide easy access to hundreds of non-redundant Flavivirus and Filovirus genomes ., allowing the user to distinguish between sublineage identity SNPs and unique strain-specific SNPs , thus facilitating analysis of the history of viral diversification during an epidemic ., EvoPrinter also proves useful in identifying recombinant strains and their parental lineages and detecting host-cell genomic editing ., EvoPrinter should serve as a useful addition to existing tools for the comparative analysis of these viruses . | medicine and health sciences, pathology and laboratory medicine, pathogens, microbiology, viruses, genomic databases, filoviruses, rna viruses, genome analysis, microbial genomics, research and analysis methods, sequence analysis, viral genomics, sequence alignment, bioinformatics, medical microbiology, microbial pathogens, biological databases, comparative genomics, sequence databases, ebola virus, flaviviruses, virology, database and informatics methods, viral pathogens, genetics, biology and life sciences, genomics, computational biology, hemorrhagic fever viruses, organisms, zika virus | null |
journal.pcbi.1006611 | 2,019 | Top-down inputs drive neuronal network rewiring and context-enhanced sensory processing in olfaction | A key property of the mammalian brain that is essential for its vast computational power is the pervasiveness of centrifugal , top-down feedback from higher to lower brain areas 1–5 ., The top-down projections can provide the receiving brain area with information that is not available in the feedforward stream 6 and can switch the processing by the lower brain area between different modes , as demonstrated in the visual system 7 ., From a theoretical perspective , it has been posited that top-down signals can direct the lower brain area to suppress response to expected inputs or to inputs that have already been recognized by the higher brain area 8–11 and to transmit predominantly information about unexpected or unexplained inputs and the error in the prediction of the inputs 12 , focusing on task-relevant information ., Such specificity in the processing requires that the top-down projections be precisely targeted 10 , 13–16 ., The mechanisms that are at work in the formation of these specific connectivities are , however , not well understood ., Experiments in the visual system suggest that they are activity-dependent 17 ., Here we address this issue in the context of the extensive structural plasticity of the adult olfactory system ., In the olfactory bulb , which is the first brain area receiving sensory input from the nose , structural plasticity is not restricted to early development , but is also pronounced in adult animals ., At that point its key players are the granule cells ( GCs ) ., They receive sensory input from the bulb’s principal neurons—mitral and tufted cells ( MCs ) —and are the target of massive top-down projections from the olfactory cortex 2–5 ., Not only do the GC dendritic spines exhibit strong and persistent fluctuations 18 , 19 , but these interneurons themselves , which constitute the dominant neuronal population of the bulb , undergo persistent turnover through adult neurogenesis 20 ., Both , spine fluctuations and the survival of the granule cells depend on the sensory environment 18 , 21 ., Functionally , adult neurogenesis is observed to enable and improve various aspects of learning and memory 22 , 23 ., A particularly clear example is the perceptual learning of a spontaneous odor discrimination task , which is substantially compromised if adult neurogenesis is suppressed 24 ., The survival of GCs seems also to play an important role in retaining the memory of an odor task , with the survival of odor-specific GCs being contingent on the continued relevance of that odor memory 25 ., This suggests that GCs receive non-sensory , task-related information via the top-down projections ., These projections likely also underlie the GC activation that can be evoked by context alone , without the presence of any odor , if that context has previously been associated with an odor 26 ., Importantly , the context-evoked GC activation patterns reflect specifically the GC activation pattern that would be induced by the associated odor ., Taken together , the experiments suggest that sensory input together with non-olfactory information like context or valence may shape the connectivity within the olfactory bulb as well as that of the top-down projections onto the GCs ., It is , however , poorly understood how the bottom-up and top-down inputs into the GCs jointly shape the network connectivity , what mechanisms are at work , what kind of connectivities arise , and what kind of functionalities these connectivities support ., To address these issues we have employed computational modeling using a biophysically supported framework ., This model makes a number of predictions that can be tested with physiological and behavioral experiments and that can guide future experiments aimed at unraveling the cortico-bulbar connectivity ., Behavioral experiments on spontaneous odor discrimination using a habituation protocol have shown that exposure to an odor related to the odors used in the discrimination task can induce a perceptual learning of that task; however , this learning was compromised when adult neurogenesis was suppressed 24 ., A parsimonious interpretation of this finding is that the restructuring of the bulbar network by adult neurogenesis enhances differences in the bulbar representations of the similar odors rendering them more discriminable ., To assess the impact of the network structure on odor discriminability we envisioned a read-out of the bulbar output that consists of the sum of the suitably weighted outputs of all MCs ., Discriminability can then be characterized by Fisher’s linear discriminant F given by the square of the difference between the trial-averaged read-outs corresponding to the two odors divided by the trial-to-trial variability of the read-outs ., Our firing-rate framework did not include any trial-to-trial variability ., We therefore took as a proxy for it the firing rate , which would be proportional to the variability if the rates arose from Poisson-like spike trains ., We considered here the optimal value F o p t that is obtained if the weights of the outputs to the read-out are chosen to maximize F for the stimuli in question ., Such optimal weights could be the result of the animal learning the task ., For similar odors F o p t typically increased in our model as the network structure evolved in response to these odors , typically in parallel with a reduction in the Pearson correlation of the MC activity patterns , capturing the observed perceptive learning 24 ( cf . our previous results for a purely bulbar model 39 ) ., We assessed network performance based on the MC activity patterns for two reasons ., As mentioned above , through the MC projections the information flows from the bulb to a variety of other brain areas 27–29 , each of which making use of that information for different purposes ., The MC activity patterns constitute therefore the central basis for multiple types of odor processing , and enhancements of the bulbar odor representations will affect all of them ., Moreover , the CCs included in our model represented only a single aspect of such processing: associating odors with contexts 26 and associative pattern completion 30–32 ., Thus , the CC-activity had a very strong contextual excitatory component , which was independent of the odor stimuli ., The impact of contexts on CCs was therefore opposite to that on MCs and reduced the discriminability of these CC-patterns ( cf . Supporting Information S10 Fig ) ., These CCs were therefore not suitable as a read-out that aims to discriminate very similar odors ., We envision that the modeled CCs effectively represent only a subpopulation of cortical cells , while other cortical cells , which were not included in our model , are engaged in tasks like fine discrimination of odors ., This interpretation is in part motivated by the fact that the balance between feedforward sensory input and recurrent associative input varies along the anterior-posterior axis of piriform cortex 42 and between its types of principal cells 43 , 44 as well as by recent observations indicating that different populations of cells in anterior piriform cortex encode odors differently 45 ., The survival of the model GCs depended on the sensory input they received from the MCs as well as on the top-down input from the CCs ., The top-down input was enhanced if the presented odor had previously been ‘memorized’ , i . e . if the corresponding recurrent excitatory connections among the CCs had been strengthened ., What happens if the cortical memory of one of the learned odors is erased , i . e . if the recurrent connections of the CCs representing that odor are removed ?, In our simulations removal of the memory of odor pair 1 ( Fig 3B ) significantly enhanced cell death and quickly reduced the total number of GCs ( Fig 3C ) ., The GCs that were removed during this phase were predominantly GCs that had previously responded to the odors in that pair and whose top-down inputs had been enhanced by the cortical memory of that odor pair ( Fig 3D and Supporting Information S1 Fig ) ., In parallel , the network’s ability to discriminate between the odors in that pair was substantially reduced ( Fig 3E ) ., This degradation of the performance did not occur if the removal of GCs was blocked in the simulation ., These results capture essential features of experiments in mice in which the extinction of an odor memory enhanced the apoptosis of GCs , particularly of those GCs that had been responsive to that odor ., However , fewer of the GCs died and the mice did not forget the task when apoptosis was blocked during the extinction of the odor memory 25 ., Experimentally it has been found that specific GC activity patterns could be evoked even in the absence of odors , if the animal was placed in an environment that previously had been associated with an odor 26 ., In fact , the GC activation pattern that was induced by this environment had substantial overlap with the pattern evoked by the associated odor ., This was not the case for a different environment ., A natural interpretation of these observations is that the odorless GC activation patterns were driven by top-down projections onto the GCs from higher brain areas that have access to non-olfactory , e . g . visual , information 26 ., To capture this aspect we extended our cortical model to include CCs that did not receive direct input from the olfactory bulb but were driven by non-olfactory , contextual information ., This could , for instance , represent information from other sensory modalities , information about the task the animal is to perform , or an expectation by the animal ., We introduced excitatory associational connections with Hebbian plasticity between these cells and the CCs that received MC input and extended the global inhibition to those cells ., Presenting two odors to the network , each in the presence of a different context , established associational connections in the cortical network between the CCs that received odor input ( cell indices 1 to 110 ) and the CCs that received the corresponding contextual input ( marked ‘context 1’ and ‘context 2’ in Fig 4B ) This enabled the contextual input to excite GCs via top-down projections ( Supporting Information S2 Fig ) even in the absence of any odor stimulation ., Due to the specificity of the cortico-bulbar network the resulting odorless GC activity patterns were highly correlated with the patterns induced by the associated odor , but not with those of the other odor ( Fig 4C and 4D ) , recapitulating the experimental observation ., In contrast , the context-evoked , odorless MC patterns were anti-correlated with the odor-evoked patterns , akin to representing a novel ‘negative’ odor , which may evoke a very different percept than the odor itself ., This may account for the enhanced response of the animals observed in the familiar , but odorless context , but not in the unfamiliar context 26 ., What functionality is enabled by the learned network structure , which allows CCs to inhibit specific MC ?, To assess this question we considered two scenarios:, i ) the detection and discrimination of odors in a cluttered environment and, ii ) rapid switching between different odor tasks ., We considered cluttered environments in which the presence of additional odors may or may not occlude ( mask ) the odors of interest , an olfactory analog of the ‘cocktail party’ problem 46 ., As an illustrative example we considered the detection of a weak target odor ( Fig 5 ) ., In the presence of a strong odor that activated a large number of MCs and occluded the target odor this required the discrimination between the occluder alone and the occluder with the target ( Fig 5E ) ., This was difficult because the MCs carrying the information about the target were also driven by the occluder , rendering the relative difference between the overall pattern with target ( red , solid symbols ) and that without target ( black , open symbols ) small ., If the activation by an occluding odor could be reduced without suppressing the contribution from the target odor , detection of the target should be significantly enhanced ., Indeed , in our model such an odor-specific inhibition was possible if the occluding odor was familiar , i . e . if it had been one of the training stimuli ., In these simulation experiments we associated the occluding odor during the training with context 1 ( Fig 5D ) ., The odor-specific inhibition then had two contributions ., The intra-bulbar component did not require cortical activation and suppressed the familiar occluder on its own ( Fig 5F left panel ) ., In the presence of the context the detectability of the target was even further enhanced by the cortical excitation of the GCs ( Fig 5G left panel ) , increasing the Fisher discriminant F o p t ( Fig 5H ) ., However , the same context was detrimental for the detection of the target odor if the occluding odor was not present: the strong context-enhanced inhibition almost eliminated the response to the target odor ( Fig 5F and 5G right panels ) ., The flexibility in the control afforded by top-down inputs therefore substantially enhanced performance ., The activity of the CCs and their inputs to the bulb play dual roles: they shape the cortical-bulbar connectivity during the network development and they provide—via that connectivity—input to the bulb during the task performance ., To illustrate the impact of the established connections from the CCs to the GCs , we include in Fig 5H the outcome when the top-down input was blocked ( wGC = 0 ) after the network had been established ., Even in the absence of the context signal the CCs representing the familiar , occluding odor were excited , albeit less strongly ., Blocking their input to the bulb therefore disinhibited the MCs representing the occluding odor and reduced the Fisher discriminant ., Blocking the top-down input in the absence of the occluder avoided the excessive suppression of the target odor by the context , leading to a larger Fisher discriminant ., However , the discrimination turned out not as good as in the presence of top-down input but without the context signal , since that top-down input also reduced the background activity of the MCs ., If the odors of interest are not occluded by any other odors in the environment , read-out cells can , in principle , adapt the weights of their input synapses so as to focus only on the relevant odors and ‘ignore’ the cluttered environment ., However , this weight optimization is often not possible , since the animal may not know yet which odors are to be discriminated ., This is , for instance , likely the case in the early phase of learning a new discrimination task 35 ., During this phase it is reasonable to envision that animals rely on a large number of read-out cells each of which receives inputs from different combinations of MCs and is therefore sensitive to different aspects of the MC activity patterns ., The activity of many of these read-out cells will be dominated by the uninformative components of the odor environment ( ‘distractor’ ) , making it difficult to discriminate between two weak target odors , even if they are not occluded by the environment ( Fig 6E , right panel illustrating optimal and random read-out ) ., To analyze such a situation we employed a non-optimal Fisher discriminant F n o n o p t based on a large number of random read-outs of the MCs ( cf . 10 in Methods ) ., We considered the discrimination between two similar , novel target odors in the presence of a strong , familiar odor that did not occlude the targets but served as a distractor ( Fig 6E ) ., It was associated with context 2 ., In addition , the network was familiarized with odors 1 and 2 , which were associated with context 1 and partially overlapped with the novel target odors ( Fig 6A–6D ) ., Even in the absence of any contextual signal the learned intra-bulbar connectivity was able to suppress to some extent the distracting , familiar odor relative to the novel odors ( Fig 6F ) ., In the presence of context 2 , which was associated with the distractor ( ‘correct’ context ) , this suppression was substantially enhanced through the cortical feedback driven by that context , leading to much better discriminability of the two novel odors ( Fig 6G and 6H ) ., As in the case of discrimination via an optimal read-out ( Fig 5 ) it was not beneficial to have indiscriminate strong cortical feedback for all familiar odors , even if these odors were not present ., For instance , the feedback driven by context 1 ( ‘incorrect’ context ) was detrimental to the discrimination of the target odors if neither of the familiar odors 1 and 2 , which were associated with context 1 , were part of the odor scene ., While the target odors were not occluded by these familiar odors , they had significant overlap with them ., Therefore the connectivity that was learned through the training included a sizable number of inhibitory projections—via the GCs—from the CCs representing context 1 to the MCs representing the target odors ., This lead to a strong suppression of the MC response to the target odors , resulting in poor discriminability ( Fig 6G and 6H ) ., If the top-down input was blocked ( wGC = 0 ) , neither the beneficial nor the detrimental impact of the two contexts arose ., Thus , the ability of top-down inputs to induce specific inhibition in a flexible manner substantially enhanced olfactory processing ., The neurogenic evolution of the structure of the cortico-bulbar network occurs on a time scale of days , particularly since the apical reciprocal MC-GC synapses form only days after the proximal synapses of the top-down projections 47 ., Synaptic plasticity based on changes in the synaptic weights can act on much shorter time scales , allowing cortical associational connections to adapt faster to changes in the tasks that the animal needs to perform ., This could modify the top-down signals to the bulb , altering bulbar processing ., As an example we considered a situation in which the network was trained on two pairs of odors , O1 , 2 and O3 , 4 ., The two odors in each pair were very similar , but the pairs were dissimilar from each other ( Fig 7A ) ., As expected , the training increased the discriminability of the odors within a pair ., However , for the discrimination of the mixture M1 = 0 . 55O1 + 0 . 45O3 from the mixture M2 = 0 . 45O1 + 0 . 55O3 , obtained by combining odors from the two pairs , the training on the individual components O1 , 2 and O3 , 4 was detrimental ., While it established mutual inhibition of MCs that were activated by the same mixture component ( O1 or O3 ) , it provided only little mutual inhibition between MCs that were activated by different mixture components: the number of connections between MCs representing odor O1 ( MCs with index near 30 in Fig 7B ) and MCs representing odor O3 ( MC index near 80 ) was small ., However , these MCs are activated simultaneously in the mixtures and mutual inhibition of these MCs is needed to enhance the discrimination between the mixtures 19 , 39 , 48 , 49 ., As a result the inhibition significantly reduced the relative difference between the two mixtures and with it their Fisher discriminant ( Fig 7F , left panel , and Fig 7G ) ., Inhibition between the MCs representing O1 and those representing O3 can be effected by establishing associational excitatory connections between the CCs that disynaptically inhibit these two groups ., To do so we exploited the Hebbian plasticity of the cortical synapses and trained the cortical network briefly on the mixture 0 . 5O1 + 0 . 5O3 ( Fig 7E , right panel ) ., This enhanced the discriminability of the mixtures substantially ( Fig 7F , right panel , and Fig 7G ) ., The role of the top-down inputs before and after learning are qualitatively different ., Before learning , the intra-bulbar as well as the cortically driven inhibition of mixture component O1 increases with O1-activation , which reduces the difference in the O1-activity in the two mixtures , and analogously for component O3 ., This reduces the discriminability of the mixtures ., After learning , the cortical contribution to the inhibition of each component is driven by the combined activity of both components and is therefore identical for both mixtures ., This preserves their difference ., In S11 Fig ( Supporting Information ) we show that this improvement is not due to the overall reduction in MC-pattern amplitude ., Thus , cortical projections can exploit the learned , odor-specific subnetwork structure ( Fig 1C ) to switch between different cortico-bulbar processing modes , adapting to the odor objects at hand ., Since the top-down input is at the core of the bidirectional mapping between the bulbar and cortical odor representations , we considered the influence of its strength wGC on the connectivity and function of the bulbar-cortical network in some detail ., Here we focus on the context-enhanced processing of a stimulus in the presence of a distractor ., The results for an occluded stimulus are presented in S9 Fig in Supporting Information ., The top-down inputs are instrumental during the network development as well as during the processing of stimuli ., The survival of GCs depends on their overall activity , which results from bulbar and from cortical inputs ., Therefore , increasing the strength wGC of the top-down projections shifts the balance from the bulbar inputs determining the survival and with it the network connectivity to the cortical inputs dominating the development ., This is demonstrated in Fig 8 for the cluttered environment investigated in Fig 6 ., For very small wGC training with the stimuli shown in Fig 6A resulted in a highly selective intra-bulbar connectivity in which the mutual inhibition was essentially restricted to MCs that were co-active for one of the training stimuli ., The MCs received , however , disynaptic top-down inhibition that only slightly reflected the receptive fields of the MCs and the CCs , and most CCs induced inhibition on most MCs that responded to any of the training stimuli ( Fig 8B ) ., In the opposite limit of large wGC only few CCs had projections to the bulb , but the inhibition they induced was only weakly targeted to specific MCs ., Moreover , the intra-bulbar connectivity was essentially homogeneous ., As a result , the inhibition induced by the top-down projections did not allow different CCs to inhibit specific different sets of MCs ., Thus , to obtain a subnetwork structure of the type sketched in Fig 1C , wGC had to be in an intermediate range in which the bulbar and cortical inputs to the GCs were of similar magnitude ( 1 ≲ wGC ≲ 3 . 5 in Fig 8 ) ., We probed the performance of the resulting networks with two tasks ., In both , the stimuli involved a distractor that is very different from the target odors ( Fig 8D ) and we considered the non-optimal read-out discussed in Fig 6 ., For the context associated with the distractor to enhance the processing , the top-down input has to suppress the MC-activity driven by the distractor but not that due to the target odors ., This was the case for intermediate values of wGC and resulted for both types of target odors in an increase in F n o n o p t in the presence of the correct context , but in a decrease for the incorrect context ( Fig 8C and 8E ) ., For larger wGC , however , the performance deteriorated , both with the correct context and without any context ., This was caused by the decrease in the specificity of the connectivity , which is quantified in S7 Fig . As a result even the correct context suppressed the activity of the MCs responding to the target odors , reducing F n o n o p t ., This is shown explicitly in S8 Fig in Supporting Information ., A similar dependence on wGC was found for the optimal detection of an occluded target ( S9 Fig in Supporting Information , cf . Fig 5 ) ., The key anatomical feature of the model network resulting from learning is its connectivity ., Specifically , the projections that the GCs receive from the MCs and from the CCs are predicted to be matched: a given GC receives cortical inputs predominantly from those CCs that respond to the same odors as the MCs projecting to that GC ., This matching of the GCs’ receptive fields can be tested experimentally ., One possibility is to express—after suitable training—ChR2 conditionally ( e . g . , via c-Fos 50 ) in those principal cells of piriform cortex that are activated by the training odor , combined with expression of a calcium-indicator in the GCs ., Note that the training needs to activate the neurogenic plasticity of the bulb 24 , which is not the case if the odor exposure is only passive 51 ., The model predicts that optical stimulation of cortical cells in the absence of an odor will then lead to excitation patterns of the GCs that are strongly correlated with the patterns excited by the training odor ( Fig 4C and 4D ) ., Previous experiments in which odor-evoked and context-evoked GC activation patterns were found to be correlated in different animals are suggestive of this outcome 26 ., On a behavioral level the model makes specific predictions for the learning of odor discrimination or detection in cluttered environments ., Recent experiments have shown that the detection of an odor in a go/no-go task is particularly difficult if that odor is masked or occluded by an odor 46 ., Our model predicts that the performance in such a detection task would be enhanced if the animal is first familiarized with the occluding odor over an extended period of time 24 ., The resulting restructuring of the bulbar network would lead to a reduction in the response to that familiar odor , partially unmasking the task-relevant odor ., If , in addition , the occluding familiar odor has been associated with a non-olfactory context 26 , the model predicts that the performance is further enhanced if the task is performed in that context , but not in a different , novel context ( Fig 5A ) ., Even if the cluttered odor environment does not occlude the task-relevant odors , it is expected that the learning speed in an odor-discrimination task 35 is reduced by strong distracting odors ., The model predicts that sufficient familiarization 24 with the distracting odors will reduce their uninformative contributions to the overall MC activation pattern ., This is expected to increase the signal-to-noise ratio and with it the learning speed by reducing the contributions from the uninformative MCs to the variability of the read-out ., If , in addition , the distracting , uninformative odors are associated with a context , the learning speed is predicted to increase further in the presence of that context ( Fig 6B ) ., If the cluttered environment makes the task too hard to learn for naive animals , our model suggests that prior familiarization with the distracting odors , preferably in a specific context , may enable the animals to master this difficult task ., Adult neurogenesis is a striking mechanism of structural plasticity that has the potential to rewire a network extensively ., In mammals it arises predominantly in two brain areas ., In the dentate gyrus it involves excitatory granule cells; their role in the network has been studied in detail 52 , 53 , also in terms of computational models 54 ., In the olfactory system , where adult neurogenesis involves inhibitory rather than excitatory granule cells , there is also a substantial body of experimental work 20 , but only few modeling studies are available 39 , 55 ., Here we have developed a computational model for the neurogenic evolution of the network connectivity with an emphasis on the possible role of the pervasive top-down projections from cortical areas ., The model is based on a number of experimental observations: GC survival depends on GC activity 21 , 38 , GC activity can be induced in the absence of odor stimulation 26 , and piriform cortex exhibits extensive recurrent excitation , which can support associational memory 30 , 31 ., The model captures qualitatively the experimentally observed perceptual learning afforded by neurogenesis 24 as well as the enhanced apoptosis of specific GCs and the reduced odor discriminability after the extinction of memories 25 ., Without theoretical guidance , it is difficult in experiments to identify the functional structure of the cortico-bulbar connectivity , in particular , because odor representations in the olfactory system do not reflect detailed spatial maps like those of other sensory systems 56 , 57 ., An important contribution of the model is therefore its prediction that through the structural plasticity the network develops a structure that reflects the learned odors and provides enhanced inhibition that is specific to these odors ., This inhibition is in part intra-bulbar and in part driven by top-down ( cortical ) inputs ., The latter reflects the formation of a bidirectional connection between the bulbar and the cortical representation of the familiar odors , which allows the cortical cells associated with such an odor to inhibit specifically those MCs that are excited by that odor ., This inhibition is mediated by GCs ., For this connectivity to arise the reciprocal nature of the MC-GC synapses is essential ., Our results therefore suggest that a key function of the reciprocity of these synapses may be to guide the wiring of the cortico-bulbar network connectivity ., The predicted matching of the GC receptive fields for olfactory and for cortical input can be tested experimentally ., Moreover , the model can guide future experiments aimed at elucidating the cortico-bulbar connectivity ., Functionally , the model predicts , in particular , that the learned connectivity improves the detection and discrimination of novel odors in cluttered environments , if the occluding or distracting odors are familiar ., This is achieved by a reorganization of the cortical-bulbar network so as to enhance the inhibition of familiar odors ., Processing of odors in cluttered environments can also be enhanced by adaptation of sensory neurons or adaptation further downstream 58 ., For those mechanisms to be successful , the occluding/distracting stimuli need to be present before the novel stimulus ., In contrast , the neurogenically formed structured network suppresses familiar distractors and occluders independent of their relative onset times , even if the novel odor precedes the occluding or distracting odor ., Moreover , the processing of cluttered environments can be enhanced by top-down input when the occluding or distracting odor activates cortical memory ., This may indicate that a higher brain area has recognized parts of the odor scene and may allow something akin to the ‘explaining away’ of components of a complex odor mixture that is theoretically predicted for the optimal processing of stimuli 8–10 , 16 ., Recent work has identified networks that demix familiar odors employing approximate optimal Bayesian inference; the anatomical structure of these networks is very close to that emerging naturally in our neurogenic model 10 , 11 ., By providing a biophysically supported mechanism through which the system can learn the required network structure our model complements this abstract normative approach ., Going beyond the purely olfactory aspect , the top-down input could encode task-related expectations or contextual information originating from other sensory modalities ., Thus , it may implement a predictive coding in the bulb that reflects the context or task at hand 59 ., Our model demonstrates how such contextual information can enhance performance ., The formation of the bulbar and cortico-bulbar network via structural plasticity is a relatively slow process ., However , our model predicts that the network structure emerging from it can be exploited by faster synaptic learning processes in cortex , which allow the system to switch relatively quickly between different discrimination tasks ., This is reminiscent of the task-dependent switching of neuronal responses | Introduction, Results, Discussion, Methods | Much of the computational power of the mammalian brain arises from its extensive top-down projections ., To enable neuron-specific information processing these projections have to be precisely targeted ., How such a specific connectivity emerges and what functions it supports is still poorly understood ., We addressed these questions in silico in the context of the profound structural plasticity of the olfactory system ., At the core of this plasticity are the granule cells of the olfactory bulb , which integrate bottom-up sensory inputs and top-down inputs delivered by vast top-down projections from cortical and other brain areas ., We developed a biophysically supported computational model for the rewiring of the top-down projections and the intra-bulbar network via adult neurogenesis ., The model captures various previous physiological and behavioral observations and makes specific predictions for the cortico-bulbar network connectivity that is learned by odor exposure and environmental contexts ., Specifically , it predicts that—after learning—the granule-cell receptive fields with respect to sensory and with respect to cortical inputs are highly correlated ., This enables cortical cells that respond to a learned odor to enact disynaptic inhibitory control specifically of bulbar principal cells that respond to that odor ., For this the reciprocal nature of the granule cell synapses with the principal cells is essential ., Functionally , the model predicts context-enhanced stimulus discrimination in cluttered environments ( ‘olfactory cocktail parties’ ) and the ability of the system to adapt to its tasks by rapidly switching between different odor-processing modes ., These predictions are experimentally testable ., At the same time they provide guidance for future experiments aimed at unraveling the cortico-bulbar connectivity . | In mammalian sensory processing , extensive top-down feedback from higher brain areas reshapes the feedforward , bottom-up information processing ., The structure of the top-down connectivity , the mechanisms leading to its specificity , and the functions it supports are still poorly understood ., Using computational modeling , we investigated these issues in the olfactory system ., There , the granule cells of the olfactory bulb , which is the first brain area to receive sensory input from the nose , are the key players of extensive structural changes to the network through the addition and also the removal of granule cells as well as through the formation and removal of their connections ., This structural plasticity allows the system to learn and to adapt its sensory processing to its odor environment ., Crucially , the granule cells combine bottom-up sensory input from the nose with top-down input from higher brain areas , including cortex ., Our biophysically supported computational model predicts that , after learning , the granule cells enable cortical neurons that respond to a learned odor to gain inhibitory control of principal neurons of the olfactory bulb , specifically of those that respond to the learned odor ., Functionally , this allows top-down input to enhance odor discrimination in cluttered environments and to quickly switch between odor tasks . | learning, medicine and health sciences, neurogenesis, neural networks, nervous system, brain, social sciences, electrophysiology, neuroscience, learning and memory, cognitive psychology, cognition, network analysis, memory, granule cells, computer and information sciences, developmental neuroscience, adult neurogenesis, psychology, cellular neuroscience, network reciprocity, olfactory bulb, anatomy, synapses, cell biology, physiology, biology and life sciences, cellular types, cognitive science, neurophysiology | null |
journal.pntd.0002825 | 2,014 | A New Mouse Model for Female Genital Schistosomiasis | An estimated 240 million humans worldwide have schistosomiasis , an infection by Schistosoma worms of various species 1 ., Human infection begins when aquatic cercariae found in contaminated water penetrate intact skin ., Once in the human host , these cercariae migrate into the circulation as schistosomula where in the portal vein they mature into adult worm mating pairs and then migrate to various venous plexi 2 ., Three species of Schistosoma are primarily responsible for human disease , and Schistosoma haematobium contributes to over half of all cases of schistosomiasis 3 ., With S . haematobium infection , worms can live and lay eggs for an average of 3 . 4 years 4 ., When S . haematobium eggs deposit along the female genitourinary tract such as the urinary bladder , lower ureters , cervix and vagina , girls and women can experience hematuria , dysuria , urinary frequency , and an increased risk of bladder cancer 5 ., However , S . haematobium infection is postulated to also cause dyspareunia , vaginal bleeding , pruritis , and giant granulomata that appear as tumors in the female genital tract 6 ., Collectively , these signs and symptoms are termed female genital schistosomiasis ( FGS ) 7 ., Recent studies suggest that FGS may cause women to be more susceptible to human immunodeficiency virus ( HIV ) infection 8–10 and those girls and women with FGS may have a 3-fold increased risk of contracting HIV 11 ., Unfortunately , the pathophysiology of this co-infection is not well understood ., Several studies have indicated , however , that other female genital infections , such as syphilis , human papilloma virus , and chlamydia , may increase the risk of HIV transmission 12 , 13 ., Genital infections that produce ulcers or vaginal discharge likely have the greatest impact on HIV shedding ., This may be due to high concentrations of leukocytes in the genital tract , for example , during gonorrheal or chlamydial infections , that thereby lead to greater HIV shedding 14 ., Syphilis is also associated with increased HIV shedding in the blood as well as genital tract 15 ., Clinical features of FGS , including vascularized , “sandy patches” of disrupted vaginal mucosa which are susceptible to contact bleeding , likely promote viral transmission through sexual contact 9 , 16 , 17 ., These lesions arise from an inflammatory response to deposited S . haematobium eggs , and contain inflammatory infiltrates , which may provide the optimal milieu for HIV transmission 18 ., Eggs can trigger a significant immune response , including primarily Th2-skewed systemic immune deviation 19 , 20 as well as the formation of egg-based granulomata ., 19 , 20 ., Accordingly , an additional hypothesis for the increased HIV susceptibility of S . haematobium-infected girls and women ( besides contact bleeding of genital lesions ) postulates that S . haematobium infection results in systemic immune deviation which renders affected individuals more vulnerable to HIV infection ., A third hypothesis for the enhanced HIV susceptibility of girls and women with FGS is that the close proximity of large numbers of granuloma-associated CD4+ T cells , macrophages , and dendritic cells ( so-called HIV target cells ) to infected genital tissues creates a convenient portal for HIV entry 21 ., Currently there are no relevant animal models to study FGS-related pathology ., Many questions remain regarding the mechanisms responsible for the genitourinary symptoms and possible increased rates of HIV transmission associated with FGS ., Knowing the kinetics of how rapidly HIV target cells accumulate in FGS lesions is directly relevant to HIV prevention strategies for women and girls at risk of HIV ., This in turn may drive the development of therapeutic interventions capable of limiting the immune and tissue pathology responsible for FGS-related sequelae ., Unfortunately , natural transdermal infection of mice with S . haematobium cercariae results in hepatoenteric disease and very little if any pelvic organ pathology 22 , 23 ., Since the immune response is primarily directed against S . haematobium eggs , and not as prominently to other stages of the parasite lifecycle , we previously developed a mouse model of S . haematobium egg-induced bladder disease by direct injection of S . haematobium eggs into the mouse bladder wall 24 ., This model recapitulates multiple aspects of human urinary schistosomiasis-associated bladder disease , including urinary frequency , hematuria , granuloma formation , and systemic immune responses ., Although , akin to oviposition in the bladder wall , the morbidity associated with FGS infection is strongly associated with egg deposition into the vagina and cervix , it is currently unclear whether oviposition alone , in the absence of adult worms , is sufficient to induce vaginal pathology ., This is relevant to girls and women who have cleared S . haematobium infections through drug therapy or natural immunity and yet still have parasite eggs in their reproductive tracts ., To address this issue , we directly microinjected viable S . haematobium eggs into the vaginal walls of female BALB/c mice ., Our overall aim was to create a mouse model to study FGS ., All animal work was conducted according to relevant U . S . and international guidelines ., Specifically , all experimental procedures were carried out in accordance with the Administrative Panel on Laboratory Animal Care ( APLAC ) protocol and the institutional guidelines set by the Veterinary Service Center at Stanford University ( Animal Welfare Assurance A3213-01 and SDA License 93-4R-00 ) ., Stanford APLAC and institutional guidelines are in compliance with the U . S . Public Health Service Policy on Humane Care and Use of Laboratory Animals ., The Stanford APLAC approved the animal protocol associated with the work described in this publication ., A total of 50 mice were used for the experiment ( 30 egg- and 20 vehicle-injected controls ) ., Seven to eight week old female BALB/c mice were purchased from Jackson Laboratories and housed in the Veterinary Service Center at Stanford University ., S . haematobium-infected LVG hamsters were obtained from the National Institute of Allergy and Infectious Diseases Schistosomiasis Resource Center of the National Institutes of Health ., The hamsters were sacrificed at the point of maximal liver and intestinal Schistosoma egg levels ( 18 weeks post-egg injection 25 , at which time livers and intestines were minced , homogenized in a Waring blender , resuspended in 1 . 2% NaCl containing antibiotic-antimycotic solution ( 100 units penicillin , 100 µg/mL streptomycin and 0 . 25 µg/mL amphotericin B , Sigma-Aldrich ) , passed through a series of stainless steel sieves with sequentially decreasing pore sizes ( 450 µm , 180 µm , and 100 µm ) , and finally retained on a 45 µm sieve ., Control injections were performed using similarly prepared liver and intestine lysates from age-matched , uninfected LVG hamsters ( Charles River Laboratories ) ., Seven to eight week old female BALB/c mice were anesthetized with isoflurane ., Freshly prepared S . haematobium eggs ( 1 , 000 eggs in 50 µl of phosphate-buffered saline , experimental group ) or uninfected hamster liver and intestinal extract ( in 50 µl of phosphate-buffered saline , control group ) was injected submucosally into the mouse posterior vaginal wall at 6 oclock over 5 seconds ( Figure 1 ) ., For mice undergoing sacrifice for flow cytometry and Luminex experiments , additional eggs ( 1 , 000 eggs in 50 µl of phosphate-buffered saline , experimental group ) were injected at the 3 , 6 , and 9 oclock positions into the mouse posterior vaginal wall ., By both observation of miracidial activity and hatch testing a high proportion of viable eggs was confirmed for each batch of eggs injected ., All egg injections were performed within 8–10 hours of egg isolation ., ( >70% of eggs from by each batch were confirmed to be viable through observation of motile miracidia within eggs and successful hatch tests ) ., Voided spot on paper analysis was performed as previously described 26 ., In brief , mice underwent vaginal injection with either eggs ( n\u200a=\u200a15 ) or control vehicle ( n\u200a=\u200a5 ) ., One week later , mice were housed singly and acclimated for one hour in cages lined with filter paper laid underneath a wire floor bottom ., Animals were given ad libitum access to food and water-soaked sponges placed on wire cage covers ., After 8 hours , each piece of filter paper was photographed under ultraviolet light to localize voided urine spots ., Total spots were counted for each mouse and the average number of voids were compared between the egg- and vehicle-injected mice using two-tailed T-tests ., Mice were sacrificed at serial time points 2 , 4 , 6 and 8 weeks after vaginal injection and the vaginas , cervices , and bladders processed for routine histology ., Step sectioning was performed by alternating between discarding and analyzing 10 sequential 5 micron sections ., Morphologic analyses were conducted on hematoxylin and eosin- stained sections ., The entire vagina , bladder , and cervix of each mouse was processed , sectioned , and examined for pathology ., Each mouse vagina was dissected in its entirety from the introitus up to the level of the cervix ., The posterior cul de sac was separated from adjacent adipose tissue and skin with sharp dissection ., The pubic bone was split and the vagina was gently removed from the pelvis by transecting it 5 mm below the cervix ., Freshly excised vaginas were minced and incubated with agitation in 0 . 5% heat-inactivated FBS ( Thermo Scientific Hyclone , IL ) , 20 mM HEPES pH 7 , 125 U/ml ( 1 mg/mL ) collagenase VIII ( Sigma-Aldrich , Saint Louis , MO ) in RPMI 1640 medium for 1 hr at 37°C 27 ., The tissue was then passed through a 70 µm nylon cell strainer to remove undigested tissue and macrocellular debris ., A total of 106 cells/sample were treated with mouse anti-CD16/CD32 ( clone 2 . 4G2 , BioLegend , San Diego , CA ) for 20 min and stained with surface markers of lymphocyte lineages mouse anti–CD3-APC-Cy7 ( clone 17A2 , BD Pharmingen , San Diego , CA ) , anti–CD4-Pacific Blue ( clone RM4-4 , BioLegend ) , anti–CD8a-Alexa Fluor 647 ( clone 53-6 . 7 , BioLegend ) , anti-CCR5-PE ( clone HM-CCR5 , BioLegend ) and anti-CXCR4-PerCP efluor 710 ( clone 2B11 , eBioscience , San Diego , CA ) ; or surface markers of myeloid lineage anti-F4/80-FITC ( clone BM8 , Biolegend ) , anti–CD11b-APC-Cy7 ( clone M1/70 , BioLegend ) , anti–CD11c- Pacific Blue ( clone N418 , BioLegend ) , anti-CD64 ( clone X54-5/7 . 1 , BioLegend ) , anti-CXCR4-PerCP efluor 710 and anti-MerTK ( clone AF591 , R&D Systems , Minneapolis , MN ) with anti-goat-IgG-APC ( R&D systems ) for 30 minutes at 4°C ., Flow cytometry was performed using a BD LSRII flow cytometer and BD FACS Diva software ., To ascertain whether the presence of S . haematobium eggs would induce a systemic immune response we performed serum cytokine assays ., Serum samples were assayed using a mouse 26-plex cytokine kit ( Affymetrix , Santa Clara , CA ) according to the manufacturers instructions ., Samples were read using a Luminex 200 ( Luminex , Austin , TX ) with a lower cut off of 100 beads per sample ( Human Immune Monitoring Core , Stanford University ) ., Assayed proteins analyzed included: IL-1α , IL-1 β , IL-2 , IL-3 , IP10 , IL-4 , IL-5 , IL-6 , IL-10 , TGF-β , IL-12p40 , IL-12p70 , IL-17 , IL-13 , KC , IL-23 , RANTES , IFN-γ , GM-CSF , TNF-α , G-CSF , MIP-1α , MCP-3 , eotaxin , MCP-1 , and VEGF ., Flow cytometric data were analyzed using FlowJo v7 . 2 . 4 ( Tree Star , Ashland , OR ) ., An unpaired Mann-Whitney U test was used to analyze flow cytometric data and Luminex analysis between control- and egg-injected mice at each time point ., Data were expressed as medians ., A p value of <0 . 05 was considered statistically significant ., Over 8 weeks , the egg-associated mixed inflammatory infiltrate expanded and organized into well-defined granulomata surrounded by peripheral eosinophils and neutrophils and containing a diffuse , peripheral lymphocytic infiltrate ( Figures 2–5 ) ., This is consistent with our flow cytometry data , which demonstrated an initial increase in numbers of T-cells followed by a later expansion of the macrophage pool ., Intact granulomata were still present 8 weeks after egg injection ., Interestingly , disruption of the vaginal mucosa , was not observed in our model ., We also did not appreciate any pathology in the mouse cervix on H&E ( data not shown ) ., Accordingly , levels of keratinization and the thickness and integrity of the vaginal mucosa showed no difference in egg-injected mice compared to controls ( Figure 6 ) ., Histologically we have identified intact miracidia within eggs at least two weeks after injection into mouse tissue ( data not shown ) ., This suggests that eggs remain viable for a period after injection into mouse vaginal submucosal tissues ., Vaginal submucosal S . haematobium egg injection induced urinary frequency with an increase in the number of urinary voids ( median number\u200a=\u200a5 ) relative to vehicle-injected animals ( median number\u200a=\u200a2; p\u200a=\u200a0 . 0423 ) ( Figure 7 ) ., Given the association between FGS and HIV transmission we sought to characterize potential HIV target cell populations and their HIV co-receptor ( CCR5 and CXCR4 ) surface expression in vaginal tissue from S . haematobium egg-injected mice ., Total T-cell subsets were defined by surface expression of CD3 , and then further categorized by the surface expression of CD4 , CD8 , CCR5 , and CXCR4 ., Macrophage populations were defined by the surface markers CD11b , F4/80 , MerTK , and CD64 , and further characterized by CXCR4 expression ., Egg-injected vaginal tissue contained significantly higher numbers of both CD4+CCR5+ T cells ( p\u200a=\u200a0 . 0079 ) and CD4+CXCR4+ T cell ( p\u200a=\u200a0 . 0079 ) populations by week two post-egg injection ( Figure 8A ) ., Egg-injected vaginal tissue also had greater numbers of T cells , CD4+CXCR4+ T cells , and CD4+CCR5+ T cells throughout the 8 week time course , though these trends were not statistically significant ., An increased number of macrophages expressed the HIV co-receptor CXCR4 in egg-injected mice at week 6 post-egg injection ( p\u200a=\u200a0 . 0173 ) , compared to vehicle-injected mice ( Figure 8B ) ., Macrophage numbers were increased in egg-injected vaginal tissue at week 4 ( p\u200a=\u200a0 . 043 ) and 6 ( p\u200a=\u200a0 . 0303 ) compared to vehicle-injected tissue ( Figure 8C ) ., RANTES protein levels were increased at 2 weeks post-egg injection ( median 23 . 01 pg/ml ) compared to vehicle-injected ( median 11 . 23 pg/ml , p\u200a=\u200aNS ) ., RANTES protein levels decreased at 4 weeks post egg-injection ( median 13 . 08 pg/ml ) compared to vehicle controls ( median 10 . 5 pg/ml ) p\u200a=\u200aNS ., There were no differences in any other assayed cytokine in egg- versus control-injected mice at 2 and 4 weeks post-egg injectionnjection ( data not shown ) ., We present a mouse model of female genital schistosomiasis amenable to the study of immune modulation and genitourinary changes that occur with S . haematobium egg exposure ., This model did result in an increase in numbers of potential HIV target cells in egg-injected mice ., The presence of S . haematobium eggs in the vagina did not induce significant shifts in the overall systemic immune response ., We also detected an increase in urinary frequency in S . haematobium egg-injected mice ., Besides identifying increased urinary frequency , we also found vaginal granuloma formation in mice after S . haematobium egg injection as early as 2 weeks post-egg injection ., The vaginal lesions we describe herein feature cellular infiltrates that differ in composition from those seen in the mouse bladder wall egg injection mouse model ., In our model there is an increase in numbers of T cells at 2 weeks whereas in the bladder model there is an increase in numbers of eosinophils and B cells 24 ., At 4 weeks , our model demonstrated an increase in numbers of macrophages whereas the bladder model found an increase in numbers of T cells , B cells and neutrophils 24 ., We speculate that these differences exist because the resident leukocyte populations and lymphatic tissue organization of the vaginal submucosa is distinct from that of the bladder lamina propria ., These differences likely guide any resulting leukocyte responses to S . haematobium egg exposure ., Natural infection of experimental animals with S . haematobium cercariae can be inefficient and slow to evolve , often taking greater than 15 weeks and yielding low worm returns 28 ., Bladder pathology in mice is infrequent and often is not seen until 20 weeks post-egg injection 29 ., In contrast , non-human primate models of S . haematobium worm-based oviposition in the pelvic organs are more consistent ., One study of S . haematobium-infected African baboons reported that their internal genitalia possessed tan , firm polypoid patches with diffuse infiltrate of eosinophils , macrophages , plasma cells and lymphocytes seen after infections of greater than 30 weeks of duration 30 ., However , compared to use of experimental mice , the utilization of non-human primates in research is more expensive , fraught with more ethical concerns , and suffers from a lack of species-specific tools ., To our knowledge , the work presented herein is the first mouse model to describe vaginal immune modulation by the presence of S . haematobium eggs ., The granulomas we report are similar to those seen in human immunopathology , with recruitment of lymphocytes , macrophages , and eosinophils to egg-containing sites 31 , 32 ., These inflammatory cells include CD4+ T-cells , which are the primary cellular targets for HIV ., Given that HIV primarily infects CD4+ T cells and macrophages bearing the co-receptors CCR5 33 and CXCR4 34 , we sought to characterize potential HIV target cell populations in vaginal tissue from S . haematobium egg-injected mice by studying these specific co-receptors ., Indeed , it has been previously demonstrated that schistosomal infection elevates expression levels of CCR5 and CXCR4 on peripheral CD4+ T-cells in Schistosoma mansoni-infected individuals , and biopsies of FGS lesions demonstrate increased numbers of both CD4+ T cells and macrophages 18 , 35 ., Relative to controls , egg-injected vaginal tissue featured increased numbers of CD4+CCR5+ T cells and CD4+CXCR4+ T-cells out to 8 weeks ., This could represent a shift from acute to chronic inflammation , induced in this synchronous model; however it is difficult to say with certainty that this is a chronic phenomenon ., Nevertheless , the purported causal link between FGS and increased susceptibility to HIV transmission is a hypothesis and believed to be mechanistically multifactorial ., Studies of other co-infections with HIV suggest other mechanisms for an increased susceptibility to HIV transmission 36–39 ., One study in humans co-infected with chlamydia and HIV-1 reported that HIV replication increases in association with granulocyte generation of reactive oxygen species and increases in cytokine production ( based on in vitro assays ) may impact numbers of HIV-receptive cells 36 , 37 ., Another study found that both native lipoprotein and synthetic lipopeptides derived from Treponema pallidum induced the production of HIV in a chronically infected cell line 38 ., HIV-1 has also been found to utilize the host transcription factor NF-κB to drive viral gene expression in T . pallidum infected cells 39 ., It is likely that schistosome-HIV co-infection may induce similar host inflammatory signaling cascades , and these additional mechanisms of enhanced viral replication and transmission warrant future exploration ., In addition to co-infection associations , urogenital schistosomiasis is well-known to induce genitourinary symptoms ., A recent study in an S . haematobium endemic area of South Africa found 35% of young girls between the ages of 10–12 reported urogenital symptoms associated with urinary schistosomiasis 40 ., Symptoms included increased dysuria , burning sensation in the genitals , as well as stress and urge urinary incontinence 40 ., While not all symptoms were statistically significant compared to girls without urinary schistosomiasis , infected girls reported increased episodes overall 40 ., In our model , S . haematobium-injected mice were found to show signs of urinary frequency more often than control-injected mice ., Step sectioning of pelvic organs by alternating between discarding and H&E staining 10 sequential 5 micron sections demonstrated that granulomas were restricted to the vaginal submucosa and did not migrate beyond to perivesical tissues ., To our knowledge , this is the first report of FGS inducing urinary frequency in the absence of S . haematobium eggs in the bladder ., Although egg injections were administered to the posterior vaginal wall ( 6 oclock ) of infected mice , they were found to have an increase in the number of voids compared to controls ., Several animal models have confirmed cross-organ sensitization among the lower urinary tract and gynecologic structures 41 , 42 ., In an induced model of endometriosis , female rats were found to have bladder inflammation and urinary bladder hypersensitivity , reflected as a decrease in micturition thresholds 41 ., A different study reported that uterine inflammation in female rats causes plasma extravasation , suggesting the existence of cross-organ inflammation 42 ., Viscero-visceral referral and sensitization ( termed cross-organ sensitization ) has recently been described to include peripheral mechanisms 43 ., This is likely due to neurons from the peripheral nervous system ( PNS ) that converge centrally in the spinal cord with input from the viscera , skin , muscles and blood vessels ., A large number of spinal neurons are receptive to visceral afferents ., There are no second order spinal neurons that specifically transmit visceral signals , thus leading to convergence of both somatic and visceral inputs into the same second order neurons 44 ., Besides inducing genitourinary symptoms , FGS is widely regarded as an immunomodulating infection ., We assessed a large cytokine panel and did not find that FGS induced broad , systemic immunomodulation in this mouse model ., RANTES was the only chemokine found to be increased in S . haematobium egg-injected versus control- injected vaginal tissue , however this was a non-significant trend ., We believe that RANTES could possible be elevated as an acute response to S . haematobium eggs ., RANTES has previously been described to aid in immunity against HIV-1 by competing to bind to CCR5 ., Sustained RANTES binding has been reported to chronically reduce cell surface levels of CCR5 45 ., A recent meta-analysis suggested that Asians with the RANTES -28G allele may have decreased susceptibility to HIV-1 infection 46 ., Few studies have described the role of RANTES in schistosomiasis infection ., One study reported a classification tree created from both factor analysis and risk analysis that showed high levels of TNF-α and low levels of RANTES in men were associated with a high risk of schistosomal liver fibrosis 47 ., In our mouse FGS model , RANTES was found to have fallen by week 4 , which coincides with the post-granuloma formation period ., A decrease in RANTES could potentially cause an increase risk of HIV transmission due to weakened immunity ., To our knowledge , no study to date has reported on the relationship between RANTES and FGS ., Although there are few animal models for S . haematobium infection in general , most existing models are of urinary schistosomiasis 30 , 48 ., Given the large number of available mouse-specific tools , our model may aid the further investigation of FGS ., FGS results from a very complex natural history that is challenging to replicate via transdermal infection of experimental animals with cercariae , the route of infection for humans ., Instead , we have injected live eggs into the vagina and have confirmed similar granulomatous pathology seen in humans ., We recognize there are limitations to this model , as it does not reproduce true disease in which ova migrate from the lumens of host blood vessels to the epithelial surface ., Instead , our model generated oviposition in the vaginal submucosa , below the vaginal epithelium ., Eggs were injected below the epithelial surface and did not migrate as seen in natural infection ., We therefore did not find any vaginal mucosal lesions ., Sandy patches on the cervix or vaginal mucosa are pathognomonic lesions associated with human FGS and are indicative of mucosal abnormalities 16 ., Thus , our model is unsuitable for examining the pathobiology of sandy patches and contact bleeding associated with FGS ., Another consideration is that much of human FGS pathology is seen in the human cervix 16 ., Due to the technical challenge of injecting eggs into the mouse cervix we were not able to incorporate this into our model ., All mice were injected at one time point and to the same depth likely because a consistent vaginal submucosal tissue plane naturally developed during the injections ., We also believe this to be a limitation of our model as we were not able to study migrating eggs at various depths within the tissue ., Given that our model features synchronous progression of egg-based inflammatory lesions by virtue of a single bolus injection , all lesions evolve at the same rate and as a result appear similar to each other ., In this important sense the lesions that result in our model do not resemble human FGS , in which lesions are of varying chronicity depending on when oviposition has occurred ., We have also injected Sepharose beads into mouse tissues and this foreign body control also does not result in epithelial alterations ( data not shown ) ., Ultimately , a humanized animal model of FGS ( including HIV co-infection ) may be informative ., The exact natural history of the local immune reactions to S . haematobium eggs in different phases of FGS is not known , but should be explored because it will have implications for treatment schedules and in choosing the best target populations ( i . e . , schoolgirls versus women ) ., The results presented herein suggest that our novel model of FGS may give insights regarding the evolution of FGS lesions ., Finally , it may be amenable to the study of S . haematobium-induced female reproductive tract inflammation and HIV susceptibility . | Introduction, Methods, Results, Discussion | Over 112 million people worldwide are infected with Schistosoma haematobium , one of the most prevalent schistosome species affecting humans ., Female genital schistosomiasis ( FGS ) occurs when S . haematobium eggs are deposited into the female reproductive tract by adult worms , which can lead to pelvic pain , vaginal bleeding , genital disfigurement and infertility ., Recent evidence suggests co-infection with S . haematobium increases the risks of contracting sexually transmitted diseases such as HIV ., The associated mechanisms remain unclear due to the lack of a tractable animal model ., We sought to create a mouse model conducive to the study of immune modulation and genitourinary changes that occur with FGS ., To model FGS in mice , we injected S . haematobium eggs into the posterior vaginal walls of 30 female BALB/c mice ., A control group of 20 female BALB/c mice were injected with uninfected LVG hamster tissue extract ., Histology , flow cytometry and serum cytokine levels were assessed at 2 , 4 , 6 , and 8 weeks post egg injection ., Voiding studies were performed at 1 week post egg injection ., Vaginal wall injection with S . haematobium eggs resulted in synchronous vaginal granuloma development within 2 weeks post-egg injection that persisted for at least 6 additional weeks ., Flow cytometric analysis of vaginal granulomata revealed infiltration by CD4+ T cells with variable expression of the HIV co-receptors CXCR4 and CCR5 ., Granulomata also contained CD11b+F4/80+ cells ( macrophages and eosinophils ) as well as CXCR4+MerTK+ macrophages ., Strikingly , vaginal wall-injected mice featured significant urinary frequency despite the posterior vagina being anatomically distant from the bladder ., This may represent a previously unrecognized overactive bladder response to deposition of schistosome eggs in the vagina ., We have established a new mouse model that could potentially enable novel studies of genital schistosomiasis in females ., Ongoing studies will further explore the mechanisms by which HIV target cells may be drawn into FGS-associated vaginal granulomata . | Over 112 million people worldwide are infected with Schistosoma haematobium worms ., S . haematobium eggs tend to be deposited in the tissue of pelvic organs such as the urinary bladder , lower ureters , cervix and vagina ., Key sequelae include hematuria , dysuria , urinary frequency , and an increased risk of bladder cancer ., This form of schistosomiasis can also cause dyspareunia , vaginal bleeding , pruritis , and granulomata that appear as tumors in the female genital tract ., Collectively , these signs and symptoms are termed female genital schistosomiasis ( FGS ) ., Recent studies suggest that FGS occurs more commonly in girls and women with HIV , suggesting that it may be a risk factor for becoming HIV-infected ., Unfortunately , the pathophysiology of this co-infection is not well understood ., A lack of an experimentally manipulable model has contributed to the paucity of research focusing on this parasite ., We have circumvented the barriers to natural S . haematobium oviposition in the mouse by directly microinjecting parasite eggs into the vaginal mucosa ., The injection of S . haematobium ova appears to trigger vaginal inflammation and scarring infiltration by leukocytes expressing HIV co-receptors , and increased urinary frequency ., Our approach may provide a representative animal model that could contribute to new opportunities to better understand the basic molecular and cellular immunology of female genital schistosomiasis . | infectious diseases, medicine and health sciences, womens health, obstetrics and gynecology, gynecologic infections, parasitic diseases | null |
journal.pcbi.1005679 | 2,017 | Growth of bacteria in 3-d colonies | In 1942 , Jacques Monod developed a mathematical model of bacterial growth in a liquid culture , where the bacterial cells and nutrient molecules were homogeneously distributed 1 , 2 ., A simple ordinary differential equation was accurate enough to account for the exponential growth of bacteria and their ascent into stationary phase following the exhaustion of the limiting resource ., The model has proven to be long-lived since most experimental studies of the population dynamics of bacteria are in liquid culture 3 , 4 ., In contrast , outside the tubes , flasks , and chemostats of laboratory culture , bacteria most commonly live in physically structured habitats as colonies or microcolonies ., Such colonies are heterogeneous; at a minimum , cells vary in their access to nutrients depending on their position within the colony and thereby divide at different rates ., The majority of research directed at understanding structured bacterial population growth has been confined to two dimensional ( 2-d ) surfaces 5–10 , including studying the interplay of evolution and the physical structure 11 , 12 , or analyzing effects of mechanical interactions in an expanding colony 13–15 ., However , diffusion in two dimensions is different from three , making it easier to form diffusion-limited instabilities 5 , 16 , 17 ., In 3-d , work has been done to understand nutrient shielding of the interior of a colony by the microbes on the surface , treating them as individual agents 18 ., In addition , in the context of modeling biofilms , there exist many complex models that account for mechanical stresses , adhesion properties , fluid flows , fluxes of multiple metabolites and waste product , and so on ( see Refs . 19–26 for just a few examples ) ., These models typically involve many free parameters , not all of them experimentally constrained ., Their complexity prevents analytical treatments , so that most such models are formulated in terms of agent-based or cellular automata approaches ., As a result of this complexity , very few of these models have provided analytical insights , or have been compared to experiments quantitatively ., In summary , we are not aware of 3-d models of colony growth that account for the spatially varying density of nutrients and bacteria , explain the observed experimental phenomenology of bacterial growth in such colonies , and do so in a relatively simple coarse-grained ( PDE ) Monod-style manner , rather than relying on complex agent-based simulations of individual bacteria ., Here we develop such a model that treats the growth rate heterogeneity due to the non-uniform nutrient distribution produced self-consistently by consumption of a nutrient by the bacteria ., We explore its fit experimentally with the growth of E . coli maintained and growing as colonies embedded in 3-d matrix of soft agar with an initially uniform spatial distribution of a limiting carbon source , glucose ( Fig 1 ) ., We compare dynamics of growth of bacteria in colonies with that of planktonic cells in liquid culture with the same concentration of limiting glucose ., In our model , we assume that the colony is essentially unconstrained by the soft agar and is free to expand , and the bacteria within it are non-motile ., This combined theoretical-experimental study reveals two surprising features of bacterial populations growing as colonies:, ( i ) the bacteria within these structures exist as loosely packed viable cells , and, ( ii ) the viable cell densities of bacteria produced in colonies is more than two-fold greater than that in liquid cultures with the same concentration of the limiting glucose ., We use population growth of E . coli in minimal medium as the basis for developing the model ., To control the amount of nutrients available to the bacteria , we use glucose at the initial concentration 0 . 2 mg/ml , at which it limits the stationary phase density of E . coli produced as planktonic cells and as colonies in soft agar ., We grow bacteria either in liquid cultures or as three-dimensional colonies embedded in soft agar ( Fig 1 ) ., Unless otherwise noted , 3-d colonies are grown in 3 ml of soft agar , inoculated with approximately 50 bacteria/ml ., Under these conditions , each colony has an access to a nutrient subvolume of v ∼ 1/50 ml , or , on average , a nutrient sphere of radius R = ( 3v/4π ) 1/3 ≈ 1 . 7 mm ., For the liquid and the 3-d growth , we estimate the density of viable cells , N ( t ) , at different times diluting and plating and then counting the number of resulting colonies ( colony-forming units , or CFU ) , see Methods for details ., For each time point , we obtain 6 independent replicates of CFU density estimates , and each experiment was replicated at least 3 times ., The results of these population growth experiments are shown in Fig 2 ( data points ) ., In liquid , the density of the population increases exponentially , and then abruptly stops and begins to decline at a low rate , presumably because the bacteria consume the available glucose and enter the stationary phase , at which time the rate of cell mortality exceeds that of division ., In contrast , in 3-d colonies , the exponential growth and the stationary phase are separated by a gradual decline in the net rate of growth ., We expect that this is because the growth of the population here is limited by the speed with which diffusion brings glucose from the periphery of the available nutrient volume to the colony , where it is consumed by the bacteria ., Surprisingly , the maximum density of bacteria growing as colonies is substantially greater than that in liquid , despite the concentration of the limiting glucose being equal for liquid and the soft agar ., To understand these findings quantitatively , we now develop a simple ( minimalist ) mathematical model of resource-limited bacterial growth in liquid and as spatially structured colonies ., Our liquid culture model of bacterial growth is a variant of that of Monod 2 ., In this model , all bacteria have the same resource ( glucose ) concentration dependent growth rate g ( ρ ) = gmax ρ/ ( ρ + K ) , where ρ is the concentration of glucose , gmax is the maximum growth rate , and K , the Monod constant , is the concentration of the resource when the growth rate is half its maximum value gmax/2 ., With these parameters , the rate of change of the density of the bacterial population n = N/v is given by, d n d t = n Θ ( t - τ lag ) g max ρ ρ + K - n m , ( 1 ), d ρ d t = - 1 a l n Θ ( t - τ lag ) g max ρ ρ + K ., ( 2 ), Here v is the volume of the liquid where the culture grows , and al is the liquid yield , which measures the number of bacteria produced by a microgram of the nutrient ., Further , Θ ( t − τlag ) is the Heaviside Θ-function , which is equal to zero for t < τlag , and to unity otherwise ., It represents the lag phase before the growth starts after a transfer to a new environment ., Note that Eqs ( 1 ) and ( 2 ) differ slightly from the standard Monod model ., Specifically , we added a small constant death rate m to account for the decrease of the population in the liquid culture after the saturation ( Fig 2 ) ., Thus the population has a zero net growth at a critical nutrient density of ρm = mK/ ( gmax − m ) , which represents the minimum nutrient concentration needed to sustain life without growth 27 ., We fit the five growth parameters ( gmax , K , al , τlag , and m ) to the experimental data using nonlinear least squares fitting , and estimate uncertainties of the fit using bootstrapping ( see Materials and methods , and also Table 1 ) ., As seen in Fig 2 ( blue curve ) , after the lag phase , the population increases exponentially before it saturates abruptly when all the cells in the colony run out of food at the same time ., The excellent agreement between the experiments and the model is encouraging ., It allows us to use the Monod model with death as the basis for 3-d studies ., To develop the minimal model of 3-d growth , we assume that the bacteria within the colony are physiologically identical , but depending on their position , vary in their access to the diffusing carbon source ., Thus all cells grow according to the Monod model , differing only by the local availability of the limiting nutrient , glucose , ρ ( x , y , z ) ., For 3-d colonies , the dynamics of the bacterial density n ( x , y , z ) is a result of a complex interplay between mechanical properties of the extracellular colony matrix and the substrate , in which the colony grows , the stiffness of the bacterial wall , and the growth properties of bacteria themselves ., Mathematical models that account for these complexities are typically over-parameterized , making it hard to make precise quantitative predictions 23 , 24 ., In contrast , here we aim at building the simplest possible model consistent with data ., We assume that soft agar is too soft to provide mechanical resistance to the colony , but sufficiently dense to keep cells from moving ., Thus the colony would grow in density until bacteria get as close-packed , on average , as possible given the amount of the extracellular matrix they secrete ., We denote this maximum cell number density as μ ., Having reached the maximum density , any new cellular divisions must expand the colony into the soft agar , with the same fixed maximum density in the volume occupied by the colony , save for possibly smaller density at the colony edge ., Further , as seen in Fig 1 , the colony spreads spherically-symmetrically , so that the density of the cells and the nutrient concentration are functions of the radius and the time only , n ( r , t ) and ρ ( r , t ) ., Thus we have, ∂ n ( r , t ) ∂ t = n ( r , t ) Θ ( t - τ lag ) g max ρ ( r , t ) ρ ( r , t ) + K - m , ( 3 ) ∂ ρ ( r , t ) ∂ t = D ∇ 2 ρ - 1 a c n ( r , t ) Θ ( t - τ lag ) g max ρ ( r , t ) ρ ( r , t ) + K , ( 4 ), with the initial uniform spatial concentration of the nutrient ρ ( r , 0 ) = ρ0 at time t = 0 , and a single bacterium starting the colony at r = 0 ., In these equations , D is the nutrient ( glucose ) diffusion coefficient ., Further , we allow for the yield in the colony ac to be different from the liquid yield al to account for the different saturation values in Fig 2 , as further discussed below ., Importantly , since the agar is more than 99% liquid , the four other growth parameters gmax , K , τlag , and m are taken to be the same in both media ., Bacterial density in Eq ( 3 ) would grow to infinity with time , which is physically unrealistic ., Thus we need to bound the cell density from above by a maximum value , corresponding to maximally packed cells ( and extracellular matrix ) that can be compressed no further ., To do this , and to establish such maximum cellular packing density μ in Eq ( 3 ) , we now impose that the overall increase in cell number beyond the maximum density leads to the proportionate growth of the colony radius rc , so that N ≡ 4 π ∫ d r r 2 n ( r , t ) = ( 4 / 3 ) π r c 3 μ ., In other words , at each point in time , we impose the condition that, n ( r , t ) = { μ , 0 < r ≤ r c = ( 3 N / 4 π μ ) 1 / 3 , 0 , r c < r ≤ R , ( 5 ), where R = ( 3v/4π ) 1/3 is the radius of the nutrient subvolume accessible to the colony ., To reconcile Eqs ( 3 ) and ( 5 ) , we say that all new growth is accounted for by the expansion of the colony edge , rc ( t ) , while the death results in a decrease in the cell density locally ( see Materials and methods for description of the algorithm for simulating this growth model ) ., This is reminiscent of earlier hybrid differential-discrete simulations 28 , 29 ., However , we note that , in our model , the biomass changes differentially: the cell density in each spherical shell is a real number , and it is not necessarily equal to μ in either the largest shell ( due to growth ) or in all other shells ( due to cell death ) ., In fact , we emphasize that one should not view the colony growth as biomass transfer , but rather as growth in the inner shells creating pressure that expands the colony radius continuously ., It might be possible to write this mass dynamics as an integro-differential equation ., However , for the purpose of solving the model and comparing to experiments , an integro-differential equation will not be more useful than the explicit dynamical rules that we have provided ., Although the assumption of redistribution of new growth into a spherically symmetric front of a growing colony will likely be violated for a generic colony or biofilm , it is clearly satisfied for colonies in our experiments ( cf . Fig 1 ) ., One would expect violations of the symmetry if there are nearby colonies competing for the same nutrients , and thus partially shielding each other ., However , in our experiments , colonies are either well separated and hence weakly interacting ( small initial bacterial densities ) , or there are many colonies in arbitrary directions from each other ( large bacterial densities ) ; in both cases , the approximate spherical symmetry is restored ( especially since we average over multiple colonies before counting CFUs ) ., Thus there are no obvious reasons to go beyond the spherical symmetry assumption ., In fact , we will see that predictions of this simple model will be verified against new experimental data , further confirming the spherical symmetry assumption a posteriori ., As mentioned above , there are a lot of models in the literature describing growth of bacterial communities in spatially structured environments ., 19 , 23 , 24 ., However , we have not found any readily available 3-d spherically symmetric soft agar colony models , where every included biological process or physical feature is essential to the population biology of the colony ., This necessitated development of our model in this section ., To illustrate the behavior of the 3-d model of bacterial growth as colonies , we plot numerical solutions of Eqs ( 3 ) – ( 5 ) for different values of the nutrient diffusion coefficient in Fig 3 ( A ) ., Especially at small D , two different growth regimes are clearly visible after the lag but before the ultimate saturation and the slow cell death ., The first is the fast exponential growth based on local , immediately accessible resources ., This regime is indistinguishable from the growth in liquid ., When the local nutrients are depleted at a certain time τ1 following the start of the growth at τlag , new nutrients must be brought from afar by diffusion ., This is slow , resulting in a slower diffusion-limited growth regime ., Here the overall colony growth rate is an average over cells growing at different rates due to different concentrations of the locally accessible nutrient ., Our numerical solutions suggest that , in this regime , the nutrient concentration at the colony edge decays exponentially fast , in agreement with Ref ., 18 , cf ., Fig 3 ( B ) ., The nutrient penetration depth is only a few μm , or a few cell layers ., Therefore , in the diffusion-limited regime , there are , essentially , no nutrients deep inside a colony , and only cells at the periphery can grow ., In the absence of resource storage 30 , nutrient sharing from the outer cells , or cannibalism ( we model none of these ) , interior cells would not grow at all and will eventually die ., The diffusion-limited growth regime finally ends with saturation and slow death when most of the nutrients in the accessible subvolume are depleted at time τ2 after τlag ., The onset of the saturation takes longer than in liquid since small ( but larger than ρm ) amounts of the nutrient linger at the far edges of the nutrient subvolume for a long time ., Analytical expressions for τ1 , τ2 , and the growth dynamics can be obtained from the following arguments ., First , in the exponential growth regime , the population grows as N ∼ e g max t ., This requires e g max t / a c of the nutrient mass , which must come from the volume immediately accessible by diffusion , equal to ∼ ρ 0 ( D t ) 3 ., Equating the two expressions gives , to the leading order , τ 1 ∼ g max - 1 log ρ 0 a c ( D / g max ) 3 / 2 ., When local resources are exhausted , growth is limited by nutrients diffusing in from the volume ∼ ( D t ) 3 ., However , because the encounter probability for a 3-d random walk is less than one 31 , most of the nutrient molecules coming from afar will not be immediately absorbed ., In fact , since the box-counting dimension of a diffusive process is two , only ∼ ρ 0 ( D t ) 2 r c nutrient molecules will be captured in time t , resulting in N ∼ ρ0 Dtrc ac ., On the other hand , the radius of the colony grows as rc = ( 3/4π ) 1/3 ( N/μ ) 1/3 ., Combining these expressions gives N ∼ ( ac ρ0 D ) 3/μ1/2 t3/2 in the diffusion-limited regime ., Finally , the total amount of nutrients available to the colony is ∼ρ0 R3 , and so the diffusion-limited growth will saturate , and the cells will start dying with the rate of m when the colony grows to N ∼ ac ρ0 R3 , which occurs at τ2 ∼ ( μ/ac ρ0 ) 1/3 R2/D ., Altogether , we find, N ∼ { const , t < τ lag , e g max t , t − τ lag ≪ τ 1 ∼ log ρ 0 a c ( D g max ) 3 / 2 g max , ( a c ρ 0 D ) 3 μ 1 2 t 3 / 2 , τ 1 ≪ t − τ lag ≪ τ 2 ∼ ( μ a c ρ 0 ) 1 3 R 2 D , a c ρ 0 R 3 e − m t , τ 2 ≪ t − τ lag , ( 6 ), These analytical estimates are supported by the numerical solutions in Fig 3 ( A ) ., We note that in one or two dimensions , the diffusion limited growth would scale as N ∝ td/2 for dimension d , independently of the ( small ) colony radius , or even for a point colony , since the random walk encounter probability there is one 31 ., In contrast , our three-dimensional result depends critically on knowing how the radius of the colony scales with the number of growing bacteria ., In particular , here we cannot model the colony as a point-like object ., Thus the exponent of the power law scaling is not universal in 3-d , and it may change for heterogeneous colonies with varying cell size and cell density ., To determine the extent to which our minimal model accounts for the dynamics of growth of bacteria in colonies , we fit the model to data using nonlinear least squares fitting , similar to the liquid case ., We keep the parameters al , K , gmax , m , and τlag equal to the values inferred for liquid , and only optimize D , μ , and ac for the 3-d culture data ., See Materials and methods for the details of the fits , including estimation of the prediction uncertainty using bootstrapping ., Table 1 shows fitted parameter values with the corresponding nominal values from the literature ., The fitted parameters are consistent with the nominal values where the latter are known ., A possible exception is the value of the glucose diffusion coefficient D , which is lower than those reported in previous publications ( though the confidence interval on our fits is rather large ) ., This could be a result of the previous measurements done in hydrogels , rather than in 0 . 35% agar preparation used in this study ., Further , the best fit curve shows an excellent agreement with data ( cf . Fig 2 , red ) , and the prediction confidence bands are very narrow ( cf . Fig 4 ) ., This suggests that nutrient diffusion and the ensuing geometric heterogeneity of growth are sufficient to explain the population dynamics of these E . coli colonies in 3-d at our experimental precision , and consideration of additional phenotypic inhomogeneities is not needed ., Our analysis also provides estimates of two previously unknown parameters , μ ( the maximum packing density of viable cells ) and ac ( yield in 3-d colonies ) ., The inferred packing density is μ = 3 . 0 ⋅ 10−2 CFU/μm3 , with the 80% confidence interval of 1 . 7 , 4 . 2 ⋅ 10−2 CFU/μm3 ., Since an E . coli cell has a volume of between 0 . 5 and 2 μm3 32 , 33 , this suggests that only about ∼3% of all space in a colony is occupied by viable cells ., This is a surprising finding , and it requires an independent corroboration ., Towards this end , we measure radii of large colonies and calculate their packing densities by diving colony volumes by the average CFUs per colony ., This gives μ = 1 . 5 ± 0 . 08 × 10−2 CFU/μm3 , consistent with our estimation of μ from the fitted growth model ., In other words , in our experiments , viable E . coli cells are sparsely packed ., Notice that here we only say that , in large colonies , there is a small density of viable cells that grow into visible , countable colonies when plated ., There could be many other cells , which , for whatever reason , do not grow into large colonies after plating ., Without additional investigations , we cannot make the distinction ., Also notice that we can only make this claim for large , old colonies which is when most cells in growing colonies emerge ., In particular , the low density claim is not valid during early growth , when each colony starts with an individual cell , which by definition takes 100% of the colony volume ., The second inferred parameter is ac ., We find that the yield as measured by the ratio of the CFU estimated stationary phase density and the quantity of glucose in 3-d is 2 to 3 times higher than that in liquid culture , ac > al ( cf . Table 1 ) ., This implies that , at saturation , colonies produce more CFUs than liquid cultures , which is directly apparent from Fig 2 ., This is a surprising result , since in the colony the bacteria grow more slowly and there is more time for cell death ., Nonetheless , similar results have been reported for colonies growing on surfaces 34 ., Here this effect is likely a direct consequence of the growth dynamics during the diffusion-limited regime ., Indeed , E . coli cells growing at a rate of >1 hr−1 grow to be 2 to 3 times larger than cells growing at a rate of <0 . 1 hr−1 35 ., While the diffusion limited regime lasts only for a few hours ( cf . Fig 2 ) , more than 90% of all cells emerge at that time , so that the majority of cells in the colony are smaller than in liquid , yielding more cells from the same nutrient amount ., As an independent test of the developed 3-d growth model , we use it to predict results of experiments distinct from those used for fitting the model ., Specifically , we investigate how the population size depends on the density of bacteria used to inoculate the soft agar ., At a long measurement time ( 72 hrs ) , our model predicts a non-monotonic dependence of the population size on the inoculation density ( cf . Fig 5 , dashed line ) ., This is because , at very low densities , each colony has access to a large nutrient subvolume , and the colony cannot clear this subvolume by diffusion in just 72 hrs ., As a result , at the end of the experiment , there are still nutrients in the media , and the colony does not reach its maximum size ., In contrast , at very high inoculating densities , colonies rapidly exhaust their small available nutrient subvolumes , the cell death becomes important throughout much of the experiment duration , and the population is smaller again ., Thus the population reaches its maximum at intermediate densities , where these two effects balance ., We test this prediction by experimentally measuring population sizes at 72 hrs for E . coli growing in soft agar at inoculums varying from 101 to 105 cells/ml As seen in Fig 5 , the experimental data agree with the prediction within errors and , in particular , exhibit the expected non-monotonicity ., We emphasize that no additional fitting was done for this figure , and yet the agreement between the experiment and the theory is very good ., Our simplest 3-d spherically symmetric colony growth model has been able to fit bacterial population dynamics data remarkably well ., To further challenge the model and to suggest its possible future improvements , we now use microscopy ( see Materials and methods ) to collect additional data that is of a very different nature compared to the data used for building the model ., We use these new experiments to further investigate the most salient prediction of the model , namely the difference in the liquid vs . colony yield ., According to our fits , the bacterial yield per microgram of glucose in 3-d colonies is 2 to 3 times higher than that in the liquid culture ., We proposed that this is because the cells in liquid grow faster ( and hence are larger ) than those in colonies ., To verify this directly , we measured the cell size ( length ) in these different growth conditions as a function of the time since inoculation ( see Materials and methods ) ., For liquid , the mean cell length at 6 hours post-inoculation was 1 . 9 ± 0 . 7 μm ., For older cultures , we have no way of distinguishing young and old cells , and so the distribution of cell sizes includes both cells that were born in earlier stages of the experiment , as well as recently ., Crucially , as the cultures grew older , long , filamentous cells emerged ( for old cultures , the longest cells were > 100 μm ) ., While the fraction of such extremely long cells was small , this tail of the cell size probability distribution 42 , 43 had a pronounced effect on the mean cell length , increasing it to ∼5 μm for the oldest cultures ( Fig 6 ) ., To account for this long tail , we report both the mean and the median cell sizes , as well as the fraction of cells that remained non-filamentous ( defined as < 5 μm in length ) ., As seen in Fig 6 , the number of short cells stabilizes near 60 − 70% for the oldest cultures ., At the same time , the median cell size in liquid does not depend on the culture age , hovering around 2 μm ., In contrast , the distribution of cell sizes in colonies is much less skewed ., Less than 1% of the sampled cells become filamentous at long times ( Fig 6 ) , so that the mean and the median cell lengths are nearly equal ., The average cell size drops when the diffusion-limited growth starts , and it saturates near 1 . 5 μm for very old colonies ., Combining these measurements , the size ratio of cells grown in the liquid and in colonies is between ∼1 . 6 ( for the median length ) and ∼3 . 4 ( for the mean length ) , in agreement with the population biology estimate above , again validating our model ., Crucially , these experiments also suggest that the immediate next modification of our growth model should not be inclusion of mechanical stresses and more complicated nutrient and waste product fluxes , but rather the growth-speed dependence of the yield , a = a ( g ( ρ ) ) , which would replace the two parameters al and ac with a single function and unite the two growth models ., To our knowledge , the model developed here is the first continuous , rather than agent-based , model to explicitly study bacterial growth as colonies ., We consider this the minimal model because it assumes spherical symmetry and that the availability of nutrients ( a carbon source ) is the sole factor determining the rate of cell division within colonies ., In reality , the cellular growth , division , and death rates would also depend on cell-to-cell interactions of various sorts , on the enrichment and deterioration of the environment due to the buildup of secondary metabolites and waste , on cell-environment mechanical interactions , and on diverse cellular phenotypic commitments ., The model we developed and experimentally tested here only accounts for the spatial heterogeneity in access to the diffusing nutrient and assumes no such additional effects 30 , 44 , 45 ., Nevertheless , despite these limitations , with five parameters describing the growth in liquid , and three additional parameters specific to the 3-d spherically symmetric colony growth , this model provides an impressively accurate description of growth of populations of E . coli as colonies in soft agar as well as planktonic cells in liquid ., Unlike the anticipated and observed nearly precipitous termination of growth in liquid culture as nutrients become depleted , our 3-d model accounts for the experimentally observed gradual reduction in net rate of replication as diffusion of the resource increasingly limits colony growth with time ., With no additional fitting , the model also correctly predicted the non-monotonic , upside-down U shaped dependence of the population size on the inoculating bacterial density ., Moreover , all of the best-fit parameters inferred from the data agreed with prior estimates in the literature , where these are available ( see Table 1 and references therein ) , indicating high-quality fits without overfitting ., Our study has revealed and/or confirmed several intriguing observations about bacteria growing in colonies ., First , the growth in colonies yielded substantially greater viable cell densities than obtained in liquid culture with the same concentrations of limiting carbon source ., We proposed that this was a direct consequence of the diffusion-limited growth , which happens at a slower division rate ., In turn , slow division is correlated with smaller size of bacterial cells 35 , resulting in more bacteria for the same nutrient amount ., This slowing down is very important phenotypically—according to our model , over 90% of all bacteria in the colony are formed at such decreased growth rate , and the yield ac is an average over yields at different stages of the slowing ., We have directly verified this hypothesis by measuring cell sizes in liquid and colony cultures as a function of the time since inoculation ., These new data provided a confirmation of our population biology estimates of the yields and of the hypothesis behind their difference ., The experiments also suggest a natural future extension of our model , which would come from measuring the dependence of the cell size and the yield on the growth rate and then verifying if both the liquid and the colony growth can be described by the same dependence a = a ( g ( ρ ) ) ., Our second intriguing observation , which is supported by two independent sets of measurements , is that the packing density inside colonies is very low , μ ∼ 0 . 03 CFU/μm3 , so that the vast majority of a volume of a colony is not occupied by viable cells ., The accuracy of this observation depends strongly on whether , when the 3-d cultures are liquified and plated for counting , cells get perfectly separated from each other , and the counted colonies start from individual cells ., We notice that , when we put liquified cultures under a microscope to produce data for Fig 6 , we observe that the cells are well-separated ., Further , if low μ was a result of us underestimating the number of cells in the colony , it would mean that ac must be even larger than our current estimate ., Thus one would be able to make even more cells from the same nutrient amount in 3-d cultures ., This is extremely unlikely since we additionally measured the cell lengths in liquid and in 3-d colonies ( Fig 6 ) , with the ratio of cell lengths agreeing with the independently estimated ratio al/ac ., Thus we are confident in our estimate of μ ., What could be the reasons for its small value ?, It is possible that the colonies are , indeed , largely void of viable cells , with extracellular fluids and matrix fibers filling in the gaps ., Another possibility is that cells deep inside the colony are dead or dormant due to the absence of nutrients , or due to other effects , such as mechanical stresses , so that the viable cells that we measure are a minority of all the bacterial cells that existed ., Our experiments show no evidence for such deviations from the minimal growth model , but it is clear that additional studies , including direct imaging of the colony structure , must be done in the future ., One interpretation of the close fit between the predictions of this minimal model and the results of our soft-agar experiments is that heterogeneities beyond nutrient access contribute little to the growth dynamics of bacteria in colonies ., It remains to be tested how general this result is ., Is the E . coli in glucose-limited minimal medium used in this experiment exceptional ?, Is the spherically symmetric growth special ?, Will the results hold for other bacterial species and for c | Introduction, Results, Discussion, Materials and methods | The dynamics of growth of bacterial populations has been extensively studied for planktonic cells in well-agitated liquid culture , in which all cells have equal access to nutrients ., In the real world , bacteria are more likely to live in physically structured habitats as colonies , within which individual cells vary in their access to nutrients ., The dynamics of bacterial growth in such conditions is poorly understood , and , unlike that for liquid culture , there is not a standard broadly used mathematical model for bacterial populations growing in colonies in three dimensions ( 3-d ) ., By extending the classic Monod model of resource-limited population growth to allow for spatial heterogeneity in the bacterial access to nutrients , we develop a 3-d model of colonies , in which bacteria consume diffusing nutrients in their vicinity ., By following the changes in density of E . coli in liquid and embedded in glucose-limited soft agar , we evaluate the fit of this model to experimental data ., The model accounts for the experimentally observed presence of a sub-exponential , diffusion-limited growth regime in colonies , which is absent in liquid cultures ., The model predicts and our experiments confirm that , as a consequence of inter-colony competition for the diffusing nutrients and of cell death , there is a non-monotonic relationship between total number of colonies within the habitat and the total number of individual cells in all of these colonies ., This combined theoretical-experimental study reveals that , within 3-d colonies , E . coli cells are loosely packed , and colonies produce about 2 . 5 times as many cells as the liquid culture from the same amount of nutrients ., We verify that this is because cells in liquid culture are larger than in colonies ., Our model provides a baseline description of bacterial growth in 3-d , deviations from which can be used to identify phenotypic heterogeneities and inter-cellular interactions that further contribute to the structure of bacterial communities . | The vast majority of theoretical and experimental studies assume that bacteria exist as planktonic cells in well-mixed liquid cultures , all with equal access to nutrients , wastes , toxins , antibiotics , bacterial viruses , and each other ., However , in the real world , bacteria are more often found in physically structured , spatially heterogeneous habitats as colonies and micro-colonies ., While one can experimentally explore the population and evolutionary dynamics of bacteria in such physically structured habitats , there is dearth of mathematical models to generate hypotheses for and to interpret results of these experiments ., As a step towards the construction of a theory of the population dynamics of bacteria in physically structured habitats , we develop and experientially explore the simplest such model of the dynamics of bacterial growth in 3-d structured colonies . | death rates, classical mechanics, chemical compounds, fluid mechanics, demography, microbiology, carbohydrates, geometry, organic compounds, glucose, mathematical models, developmental biology, mathematics, statistics (mathematics), population biology, microbial growth and development, research and analysis methods, packing density, microbial physiology, mathematical and statistical techniques, chemistry, fluid dynamics, bacterial growth, continuum mechanics, physics, people and places, population metrics, organic chemistry, confidence intervals, monosaccharides, biology and life sciences, physical sciences, population density | null |
journal.pgen.1000164 | 2,008 | Identification and Functional Analysis of Light-Responsive Unique Genes and Gene Family Members in Rice | Gene inactivation by the insertion of T-DNA is a common tool used for functional studies of genes in model plants such as Arabidopsis or rice ., T-DNA insertional mutants have been generated for virtually all of the annotated genes in Arabidopsis thaliana ., Recently , researchers working with Arabidopsis have combined the use of microarray technology with screens of T-DNA insertional mutant libraries to identify and characterize genes of unknown function ., This strategy has successfully been used to characterize genes related to the formation of the secondary cell wall , those differentially regulated in phytochrome-mediated light signals , and those involved in host responses to pathogens 1 , 2 , 3 ., These studies did not address the issue of multi-gene families , which limits functional analysis of genes in many plant species ., Rice is a staple crop for more than half of the worlds population and a model for other cereal crops ., Therefore , studies of the rice genome are expected to help elucidate the function of genes in other major cereal crops , all of which have much larger genomes than rice ., So far the functions of only a handful of rice genes have been characterized ., The rice research community has recently generated a large collection of rice insertional mutant lines and thus far 27 , 551 gene loci with knockout mutations have been collected and 172 , 500 flanking sequences tagging those mutations have been sequenced ( http://signal . salk . edu/RiceGE/RiceGE_Data_Source . html ) 4 ., Four microarray platforms covering nearly the entire rice transcriptome have also been developed 4 , 5 , 6 , 7 ., Accordingly , studies similar to those carried out in Arabidopsis , combining microarray-derived expression data with reverse genetics to address gene function , can now be carried out in rice ., Plants are continuously exposed to biotic and abiotic factors , light being one of the most important ., In addition to providing the source of energy , light is also involved in regulating growth and development throughout the plant life cycle 8 ., Genome-wide gene expression profiles of various light responses can help reveal the complicated physiological networks with which plants adapt to environmental changes ., Toward that end , plant researchers have carried out microarray experiments with various plant species in efforts to identify light-regulated genes 9 , 10 , 11 ., One of the major challenges facing scientists in the field of functional genomics , even those working with relatively simple model organisms like Arabidopsis , is the prevalence of gene families 12 , 13 ., Such family members often encode redundant functions 14 , 15 ., Because of the presence of gene families with functional redundancy , it is often the case that a gene , when mutated , will display no detectable phenotype ., For example , an Arabidopsis line carrying a T-DNA insertion in a gene encoding a protein with a putative R3H domain , At5g05100 , hypothesized to be of importance in seed organ development , was reported to have no observable mutant phenotype 13 ., Because Arabidopsis also has three genes , At2g40960 , At3g10770 , and At3g56680 , the lack of a mutant phenotype in this line is probably due to one or more of these genes encoding a function redundant of At5g05100 13 ., Efforts to overcome such functional redundancy have mostly focused on generating mutants that disrupt more than one gene family member simultaneously 16 , 17 ., For example , RNA interference ( RNAi ) technology has also been used to simultaneously silence multiple members of a gene family 18 ., Rice researchers encounter even larger gene families , more difficulty in scaling up experiments and greater time constraints associated with rices longer life cycle than do researchers working with Arabidopsis ., For example , whereas Arabidopsis has more than 600 genes in its receptor-like kinase ( RLK ) family , 80 in the bric-a-brac/tramtrack/broad ( BTB ) protein family , and 37 genes in the GARS family , rice has more than 1131 , 149 , and 52 members in those families , respectively 19 , 20 , 21 ., Accordingly , a strategy for overcoming these difficulties and accelerating functional genomics analysis is especially helpful in crops like rice ., Interestingly , the presence of multiple gene family members does not always mask phenotypic changes in gene “knockout” mutants ., For example , 43 . 9% of the Arabidopsis seedling-lethal mutants and 45 . 2% of the Arabidopsis embryo-defective mutants phenotypically identified in two Arabidopsis studies carried defects in genes that were members of gene families 22 , 23 ., These results suggest that all members of gene families are not necessarily functionally equivalent and individuals among them may play predominant roles 24 ., Here we report the use of a microarray that covers nearly the entire rice transcriptome , the National Science Foundation-supported 45K microarray ( NSF45K , http://www . ricearray . org/ ) , to identify light-responsive genes in rice and subsequent functional analysis of a subset of those genes through screening gene-indexed mutant lines of rice ., We describe the advantages of utilizing candidate genes derived from rice gene expression profiles to screen rice insertional mutant collections and discuss the biological significance of our findings ., We performed expression-profiling experiments with a newly-developed NSF45K microarray ( http://www . ricearray . org/ ) using two weeks-old rice leaf tissues harvested from plants grown under light and dark conditions ( See Materials and Methods ) ., Because genetic backgrounds can affect expression profiles 25 and we wanted to select light-responsive genes that behaved similarly in different genetic backgrounds , identical experiments were carried out with four different rice varieties: Kitaake , Nipponbare , Tapei309 and IR24 ., We used the LMGene Package developed by Rocke ( 2004 ) to identify 10361 , 4962 , 1933 , and 453 genes differentially expressed in response to light versus dark treatment at FDR p-values of 0 . 01 , 10−4 , 10−6 , and 10−8 , respectively ( Table S1 ) ., To assess the difference in gene expression profiles among the varieties , we calculated correlation coefficients ., Correlation coefficient values were 0 . 89–0 . 92 between the three japonica varieties ( i . e . Kitaake , Nipponbare , and Tapei309 ) whereas correlation coefficients values between subspecies ( i . e . japonica and indica ) were 0 . 80–0 . 82 ., This result indicates that differences in genetic background clearly affect the expression patterns ., Accordingly , by focusing on genes whose expression pattern is similar between varieties , we can avoid genes that show cultivar-specific light responses ., To assess the usefulness of our microarray data we surveyed the expression patterns we obtained through microarray analysis for seventeen genes , including gene family members , which encode proteins for the seven steps in the well-characterized chlorophyll biosynthetic pathway ( Figure 1 ) ., The expression patterns of all of the candidate genes in the seven steps of this pathway were validated using reverse transcriptase- ( RT- ) PCR ( Figure 1 ) ., Based on this analysis we identified four unique genes ( Figure 1B , 1a , 1b , 1c , and 3 ) , and two predominantly expressed light-responsive candidate genes ( Figure 1B , 2-1 and 7-1 ) as good targets for studying the biological functions of genes involved in rice chlorophyll biosynthesis ., A predominantly expressed light-responsive candidate gene is the one that is most significantly induced by light ( compared to other members of the multi-gene family ) ., Previous studies in rice showed that the knockout mutants of three of the unique genes ( encoding the subunits comprising magnesium chelatase complexes in rice: magnesium chelatase subunit H ( CHLH , 1a ) , magnesium-chelatase subunit I family protein ( CHLI , 1b ) , and magnesium chelatase ATPase subunit D protein ( CHLD , 1c ) exhibited chlorina phenotypes 26 , 27 ) ( Figure S1 ) ., The remaining unique gene ( magnesium-protoporphyrin IX monomethyl ester cyclase , MPE ) at step 3 is predicted to have a function similar to that of its Arabidopsis ortholog , CHL27 28 ., Knockout lines of the two predominantly expressed light-responsive candidate genes ( rice magnesium-protoporphyrin O-methyltransferase ( ChlM;Os06g04150; and gene 2-1 ) and rice chlorophyll a oxygenase ( Cao1; Os10g41780;and gene 7-1 ) ) had been previously identified and shown to display with light response-related phenotypes 29 , 30 ( Figure S1 ) ., These results indicate that identification of unique genes or the predominantly expressed member of a gene family in the light is an effective method to target genes for further functional analysis ., More descriptions on this strategy are shown in Figure S2 and Text S1 ., ( RiceGE , http://signal . salk . edu/RiceGE/RiceGE_Data_Source . html ) 4 ( Table 1 ) ., From our NSF45K light vs . dark experiment , we selected 365 candidate genes showing at least an 8-fold induction at the 10−4 FDR p-value ., We then utilized collections of rice insertional mutants to identify lines carrying mutations in the light-responsive genes identified through expression analysis ., To date , some 172 , 500 sequences have been generated from regions flanking insertional mutants in rice and they are publicly available at the Rice Functional Genomic Express Database ( http://signal . salk . edu/cgi-bin/RiceGE ) ., Because we wanted to include 2 independently derived mutant alleles in our analysis of each candidate gene so as to help discriminate between phenotypic changes generated by somaclonal variation versus those resulting from the insertional mutations themselves , we limited our phenotypic analysis to 74 mutant lines with T-DNA insertions in a total of 37 candidate genes ., The overall scheme we used for functional analysis based on our microarray experiment is presented in Figure 2A and Figure S2 ) ., We classified the 37 candidate genes for which we had corresponding mutants into two groups according to whether the candidate gene belonged to a gene family or not ., There were 12 unique genes ( those without gene family members ) ( Figure 2B and Figure S3A ) and 25 belonging to gene families ( see Materials and Methods ) ., The latter class was further divided into two subgroups by considering the predominance of each genes expression in the light based on the NSF45K light vs . dark array dataset ., As a result there were 13 predominantly expressed-light-induced gene family members ( referred to as “P” in Figure 2B and as “Predominant” , marked with asterisks in Figure S3B ) and 12 gene family members that were not the predominantly expressed in the light ( referred to as “NP” marked in Figure 2B and as “Non predominant” , marked with sharps in Figure S3C ) ., Because other more predominantly or equally expressed gene family members might compensate for a defective gene family member in the light , the non-predominantly expressed gene family members were not considered good candidates for functional analysis and were initially excluded from the functional analysis ( Figure S3C ) ., Next , we identified 5 genes , Os03g48030 , Os11g05050 , Os02g58790 , Os09g37620 , and Os09g16950 , for which light responses between the NSF45K and BGI/Yale light vs . dark array datasets were significantly inconsistent and deleted them from our primary list of candidate genes for the initial round of functional analysis ( Figure 2 ) ., Of these , Os03g48030 was a unique gene and Os11g05050 , Os02g58790 , Os09g37620 , and Os09g16950 were the members of their respective gene families the most predominantly expressed in the light in the NSF45K array data set ., We also included one unique gene ( Os07g46460 ) and one predominantly light-induced gene family member ( Os03g37830 ) in our candidate gene list based only on our own data because information on their light responses was not available among the BGI/Yale data ( Figure 2 ) ., We then screened the remaining mutants , those associated with 11 unique genes and 9 predominantly light-induced gene family members , to determine their phenotypes ( Figure 3 ) ., We also assayed the phenotypes of the knockout lines associated with the 17 genes we had eliminated from our primary list of candidate genes to check the efficiency of identifying mutant phenotypes for the not predominantly light-induced genes in a gene family and/or genes with expression patterns that werent consistent between the NSF45K and BGI/Yale light vs . dark array datasets ., ( Detailed data regarding all 37 of the genes that were functionally analyzed are presented in Table S2 . ), Our criteria for selecting candidate genes that respond to light might have inadvertently eliminated from consideration genes among the BGI/Yale data that exhibit condition-dependent light responsiveness ., We initially carried out functional analysis for 20 selected candidate genes , 11 unique genes and 9 predominantly light-induced gene family members ( Figure 3A and 3B ) ., First , we identified defective phenotypes associated with six of the 11 unique genes that we analyzed for function from our list of top candidate candidates ., Of these , the phenotypes of knockout lines associated with Os01g01710 ( 1-deoxy-D-xylulose 5-phosphate reductoisomerase , Dxr ) and Os03g04470 ( Expressed protein ) were albino and displayed chlorotic leaves , respectively , 2 weeks post-sowing ( Figure 3A ) ., Knockouts of Os01g71190 ( Photosystem II subunit 28 , Psb28 ) , Os02g57030 ( Expressed protein ) , and Os07g46460 ( ferredoxin-dependent glutamine:2-oxoglutarate aminotransferase , Fd-GOGAT ) displayed pale green phenotypes ( Figure 3A ) ., Of these , a mutation in the Fd-GOGAT gene displayed photo-bleached leaves two weeks later after revealing pale green leaves ( Data not shown ) ., Knockouts of Os04g37619 ( Zeaxanthin epoxiydase , Aba1 ) produced dwarf mutants ( Figure 3A ) ., We noted that the mutant phenotypes associated with three of these 6 genes , those encoding DXR , Fd-GOGAT , and ABA1 , were similar to those associated with their Arabidopsis orthologs 31 , 32 , 33 ., The Arabidopsis ortholog of Os01g71190 encodes photosystem II ( PSII ) reaction center Psb28 protein ( Psb28 ) that was first identified from PSII of Synechocystis 6803 with Psb27 34 ., The function of this PSB28 has not yet been well- characterized , although it is predicted to serve a role as a regulatory protein based on its substoichiometric amount 34 , 35 ., Similarly , Arabidopsis lines carrying a mutation in Psb27 did not display a severe phenotype ., Recovery of PSII activity after photoinhibition was delayed in the Arabidopsis psb27 mutant supporting a role in PSII for this gene 36 ., The mild phenotype displayed by the Psb28 T-DNA insertional rice plants in Psb28 gene suggests that this gene product might serve as a regulatory protein to stabilize PSII activity ., The other two unique genes for which we identified corresponding mutant phenotypes as a result of our analysis encode as yet unidentified proteins ., The endogenous retrotransposon Tos17 has been shown to be an efficient insertional mutagen in rice and phenotypes of some 50 , 000 M2 generation insertion lines carrying Tos17 insertions have been reported 21 ., We used the available Tos17 phenotypic data to assess the functions of some of our candidate genes ., Phenotypes similar to those identified in the T-DNA insertional mutants for Os01g71190 ( Psb28 ) , Os03g47610 ( Dxr ) , and Os04g37619 ( Aba1 ) have also been observed for the corresponding Tos17 insertional mutant lines ( Table 2; 21 , 37 ) ., We could not identify mutant phenotypes for the other unique genes containing T-DNA insertional mutations that we analyzed ., We assume that phenotypes associated with two genes , Os01g40710 and Os08g40160 , could not be determined due to confounding somaclonal variations ( Figure S4 ) in the one mutant line available for each of these genes for analysis ., We could not analyze the second mutant line corresponding to each of these genes due to poor seed set ., Lines with T-DNA insertions in Os03g47610 , lines 1A-08723 and 1A-17021 , did not produce any homozygous progenies and as a result the phenotypes associated with these mutations were also not determined ( Table 2; Table S2 ) ., On the other hand , while we did identify homozygous progenies and their siblings among mutants with insertions in Os03g06230 and Os03g17960 , we did not observe phenotypic differences between them ( Table 2 ) ., Targeting functional analysis to unique genes is an effective way to significantly increase the efficiency of identifying genes corresponding to defective phenotypes ., Utilizing microarray-derived expression profiles of unique genes can increase the efficiency of functional analyses 1 , 22 , 38 ., The efficiency ( 6/11 ) with which we were able to identify phenotypes associated with mutations in unique genes demonstrates the power of combining knowledge of gene copy number and gene expression patterns ., Of the nine predominantly light-induced gene family members that were consistently induced in the light , we found four for which mutant phenotypes co-segregated with a T-DNA insertion in the gene ( Figure 3B ) ., These four genes encode carbonic anhydrase 1 ( CA1 ) , serine hydroxymethyltransferase 1 ( SHMT1 ) , nitrate reductase 1 ( NR1 ) , and Vitamin C defective 2 ( VTC2 ) , respectively ., The rice T-DNA insertional line carrying a mutation in the Shmt1 gene ( Os03g52840 ) , line 1D-03944 , displayed variegated chlorina leaves ( Figure 3B ) ., The line 2B-60065 with a T-DNA insertion in Ca1 ( Os01g45274 ) showed an oxidative stress-related phenotype , necrosis in the middle of the leaf , and a little growth retardation ( Figure 3B ) ., Line 4A-50280 , with a T-DNA insertion in Nr1 ( Os08g36480 ) , displayed dwarfism , and line 3A-14221 , with a T-DNA insertion in the Vtc2 gene ( Os12g08810 ) , exhibited a pale green and later photo-bleached leaves phenotype ( Figure 2B ) ., The Tos17 line with an insertion in the rice Shmt1 gene exhibited the same phenotype as the corresponding T-DNA mutant , line 1D-03944 ( Table 2 ) ., The phenotypes associated with mutations in Shmt1 ( Os03g52840 ) , Nr1 ( Os08g36480 ) , and Vtc2 ( Os12g08810 ) were also reminiscent of the phenotypes associated with mutations in the orthologous genes in Arabidopsis 39 , 40 , 41 , 42 ., Mutant phenotypes associated with T-DNA insertions in Os01g08460 , and Os05g47540 were not determined due to confounding somaclonal variations ( Figure S4 ) ., The homozygous progenies of mutant lines with insertions in two other genes , Os03g37830 and Os06g04510 , did not show visible phenotypic changes ( Table 2 ) ., Our success in conducting functional analyses of gene family members that are the predominantly expressed in the light , and consistently so from one microarray experiment to the next , suggests that the functions of gene family members can be successfully analyzed by utilizing data obtained using microarrays representing nearly complete plant transcriptomes ., Analyses that consider both sequence similarity and predominance of gene expression has also been reported to be quite effective in functional analysis of yeast genes 14 ., We also carried out functional analysis of 5 genes that showed inconsistent gene expression patterns when we compared our NSF45K light-response experiments and the BGI/Yale light vs . dark experiments ., One of them , Os03g48030 ( designated U9 in Figure 2 and Figure 3 ) , is a unique gene and the other four genes , encoding lectin protein kinase ( Os09g16950 ) , stem-specific protein TSJT1 ( Os11g05050 ) , flavin-containing monooxygenase family protein ( Os09g37620 ) , and an expressed protein ( Os02g58790 ) , were the predominantly light-induced members of their respective families among the NSF45K array-derived data but not in the BGI/Yale light vs . dark dataset ., No defective phenotypes were observed among the mutant lines with T-DNA insertions in any of these genes ( Table S3 ) ., One possible reason for this result is the presence of redundant metabolic networks or the absence of appropriate screening conditions 15 , 43 , 44 ., Mutants carrying insertions in 12 genes that were not the predominantly expressed light-induced member of their respective gene family were also examined in this study ., A visible phenotype was observed in the homozygous mutant progenies associated with only one of these genes , Os07g05000 ( R8-2 ) ( Figure S3C and Figure S5 ) , which belongs to the family of genes encoding aldo/keto reductases ., Phenotypic changes were not observed in the homozygous segregants of lines with mutations in any of the other genes ( Table S3 ) ., The absence of detectable abnormal phenotypes associated with these other genes is generally believed to be due to one ( or more ) family members compensating for the function of the mutated gene 14 , 15 ., Line 3A-03008 , which carries a T-DNA insertion in Os07g05000 , showed a weakly pale green phenotype and slight growth retardation ( Figure S5 ) ., Identification of phenotypes associated with the other mutations may require specific conditions under which there will be no compensatory gene expression from other family members ., In cases of gene families without a predominantly expressed member under specific experimental conditions , microarray data can still be used to identify the multiple significantly expressed genes in a family so that they can be subjected to RNA-silencing techniques as has been carried out by Miki et al . 18 , 45 for the rice genes encoding homologs of mammalian Rac GTPase , OsRac1 and OsRac5 ., Of our list of non-predominantly light-induced genes , there were two genes from same family expressed under light conditions ( Os12g03070 and Os11g03390 ) ., Both encode an FHA domain , which is a putative nuclear signaling domain found in protein kinases and transcription factors ., However , we did not observe a phenotype in knockout lines of Os11g03390 ( R12-1 ) 46 ( Table S3 ) ., Similarly , we did not observe phenotypic changes associated with T-DNA insertions in members of the gene families encoding ABC1 proteins ( Os02g57160 , R3-1 and Os04g54790 , R5-1 ) , S1-RNA binding domain proteins ( Os04g54790 , R4-1 ) , and glycine dehydrogenases ( Os06g40940 , R7-1 ) ., Further experiments to generate double mutants for these family members and their light-induced relatives will be required to elucidate their functions ., In Arabidopsis , the light-responsive functions of genes in gene families such as POR and PHOT have been clarified by using double mutants of two family members , porbporc and phot1phot2 , respectively 47 , 48 ., When we consider severity of phenotypes , two of six lines carrying defects in unique genes and one of four lines carrying defects in the gene family member predominantly expressed in the light died at early seedling stage ( Figure 3 ) ., Therefore , these three genes are essential for survival ., Overall , phenotypes of mutants with defects in unique genes were more frequently observed than those of mutants carrying defects in predominantly light-induced gene family members ( compare Figure 3A with Figure 3B ) ., We suspect , therefore , that other members in a gene family may carry out a somewhat compensating role 14 , 15 ., Similarly , we did not find phenotypic changes associated with mutations in genes that were not the predominantly light-induced members of their respective families , with the exception of mutations in a gene encoding an aldo/keto reductase protein ., This result also supports our hypothesis that other gene family members can compensate for the mutation ., Despites these observations , we can not rule out the possibility that the absence of phenotype is due to non-optimal environmental conditions 44 ., In summary , we identified phenotypic changes in rice lines carrying mutations in 10 out of 20 unique genes or genes that were the most predominantly light-induced members of their respective families ., In contrast , we discovered only one phenotypic change among the lines carrying mutations in the 17 other genes that either showed inconsistently light-induced expression among different microarray data sets or were not the predominantly light-induced members of their respective gene families ( Table S3 ) ., Microarray data were very useful as criteria for prioritizing candidate genes for functional analyses ., Consideration of the expression patterns of all the genes within a gene family is an effective way to approach study of the functions of gene family members 14 ., Functional profiling of genes related to the phytochrome-mediated signaling pathway in Arabidopsis was recently carried out 3 ., We used this data set to further test the usefulness of our method ( see Materials and Methods ) ., Thirty two genes were selected for this functional profiling analysis ., Of these , mutants in seven genes displayed statistically significant photomorphogenic phenotypes ., Except for one gene ( At2g46970 ) whose gene expression profiling data was not available , we found that six genes were either unique sequences or the predominantly expressed gene family member in the red light ., In contrast , mutations in the remaining 25 genes showed less significant photomorphogenic phenotypes; thirteen of them displayed mild or severe defects in photoresponsiveness and 12 did not showed distinguishable phenotypes ( Figure S6 ) 3 ., Of these , 10 genes were not the predominantly expressed gene family member in the red light whereas 13 genes ( except two genes; At3g21550 and At3g21330 whose gene expression profiling data was not available ) were either unique sequences or the predominantly expressed gene family member in the light ( Figure S6 ) ., Thus , this functional profiling analysis in Arabidopsis also indicates that predominantly expressed gene family members as well as unique genes are good targets for functional validation ., Mutant phenotypes clearly suggest functions for targeted genes and also for the pathways those genes are associated with ., Therefore , understanding relationships among multiple pathways containing mutants defective in the plants response to light will help us elucidate the light response ., To do this , we identified genes among various pathways that were co-expressed with the 10 genes for which we had identified mutant phenotypes in this study ., We found that ten pathways ( http://www . gramene . org/pathway/ ) involved 7 of the genes for which we had identified mutants ( Figure 4; Table S4 ) ., Sixty nine genes in these 10 pathways were selected as described in Materials and Methods ., Additionally , three single-step reactions unlinked to any of these pathways but involving the other three genes for which we identified mutant phenotypes in this study , U4 ( Os02g57030 ) , U5 ( Os03g04470 ) , and Ca1 ( P2-1 , Os1g45274 ) were also included in this analysis ( see Table 2 ) ., All together , 72 genes involved in the 10 pathways and three reactions were selected for hierarchical clustering analysis ( Table S4 ) ., Then , we selected 10 datasets with which to carry out the analysis ( Table S4 ) : log2 fold change values of NSF45K light vs . dark and four different types of light vs . dark datasets generated by BGI/Yale array 49 , and log2 fold change values of five different tissues vs . cultured cells 50 ., We selected candidate genes in each pathway that are unique or are the predominantly light-induced gene family member ( except several steps not represented by predominantly light-induced gene family members ) ., We found that gene expression patterns of 67 out of the 72 genes are light-inducible in at least two of the light treatments ( Figure 4 ) ., Fifty-five genes have GO terms in the cellular component category and 46 of them have a chloroplast GO term in the cellular component category ., Seven genes are predicted to have role in the mitochondrion ( Figure 4 ) ., Most of the genes used for the clustering analysis are predicted to perform their light response-related function in chloroplasts or mitochondria ( Figure 4 ) ., As a result of the hierarchical clustering analysis we identified 10 gene clusters ( Figure 4 ) ., As has been previously reported 50 , 51 , 52 , co-expression analysis is useful for revealing functionally coherent groups of genes ., We next looked for relationships among different pathways by utilizing Cytoscape software to analyze the results we obtained from our co-expression analysis ( see Materials and Methods ) ., Cytoscape is an open source software for integrating biomolecular interaction networks with high-throughput expression data 51 ., The results of this analysis are shown in Figure S7 ., First , cluster III ( purple lines in Figure 4 and Figure S7 ) contained 3 components of PSI , one of PS II , three from the photorespiratory pathway , one from the ammonia assimilation pathway , and one from the chlorophyll biosynthetic pathway ., Of these , ferredoxin-dependent glutamine:2-oxoglutarate aminotransferase ( Fd-GOGAT , U11 ) couples with glutamine synthetase 2 ( GS2 ) for assimilating ammonium produced by photorespiration 33 , 52 ., In Figure 4 and Figure 5 , the Fd-GOGAT gene ( U11 ) , which generates glutamate at step 2 of the ammonia assimilation pathway is co-expressed with the PS I and PS II components Os08g44680 , Os12g23200 , Os07g25430 , and Os01g64960 ., GS2 ( OS04g56400 ) at step 1 of this pathway is co-expressed with other PS I and PS II components ( Os09g30340 and Os08g10020 ) ., PS I and PS II supplies ATP for the reaction of GS2 and reduced ferredoxin ( Fdrd ) for Fd-GOGAT 53 ., This dependency of the ammonia assimilation pathway on photosynthesis is supported by the co-regulation of several PS I and PS II components with two genes in the ammonia assimilation pathway ( Figure 4 , Figure 5 , and Figure S7 ) ., The photorespiratory pathway is known to be tightly correlated with nitrogen cycles 54 ., Co-expression patterns of genes encoding three early steps of the photorespiratory pathway with the Fd-GOGAT gene suggest a close relationship of photorespiration with the ammonia assimilation pathway ( purple lines in Figure 4 , Figure 5 and Figure S7 ) ., This makes sense because the ammonia assimilation pathway plays a role in preventing loss of ammonia , which is generated from step 6 of the photorespiratory pathway by refixing it to glutamate 53 , 55 , 56 ., Furthermore , the resulting amino acid ( i . e . glutamate ) reacts with glycine to synthesize glutathione in the peroxisome , suggesting that this pathway is important for protecting photosynthetic apparatus ( e . g . PS II ) from toxins such as free radicals 55 ., Also , co-regulation of genes in the three early steps of the photorespiratory pathway with several PS I and II components suggests that oxygen generated by photosynthesis triggers the photorespiration pathway 57 , 58 ., Co-regulation of genes involved in photosynthesis , ammonia assimilation and photorespiration can explain their coordinated functions 56 ., Of the genes in cluster III , the phenotypes of a mutant line with a T-DNA insertion in Fd-GOGAT ( U11 ) and a mutant lines under-expressing the Rca1 gene ( Os11g47970 ) were characterized ( Figure 4 , Figure 5 and Figure S7 ) 59 ., The Rca1 under-expressed mutant displays chlorotic leaves 59 ., The mutant ( 3A-01082 , this study ) with a T-DNA insertion in Fd-GOGAT gene displayed pale green leaves shortly after germination ( Figure 3 ) ., The same mutant displayed chlorotic leaves four weeks after germination and is similar to the phenotype of one of the Rca1 under-expressed mutants ( data not shown ) 59 ., These results indicate that co-expressed groups of genes carry out closely related functions as reported in other species 1 , 33 , 60 ., Therefore we can predict that mutations in other genes in this cluster will display similar phenotypes to those observed in the Fd-GOGAT and Rca1 mutant lines ., Those predi | Introduction, Results/Discussion, Materials and Methods | Functional redundancy limits detailed analysis of genes in many organisms ., Here , we report a method to efficiently overcome this obstacle by combining gene expression data with analysis of gene-indexed mutants ., Using a rice NSF45K oligo-microarray to compare 2-week-old light- and dark-grown rice leaf tissue , we identified 365 genes that showed significant 8-fold or greater induction in the light relative to dark conditions ., We then screened collections of rice T-DNA insertional mutants to identify rice lines with mutations in the strongly light-induced genes ., From this analysis , we identified 74 different lines comprising two independent mutant lines for each of 37 light-induced genes ., This list was further refined by mining gene expression data to exclude genes that had potential functional redundancy due to co-expressed family members ( 12 genes ) and genes that had inconsistent light responses across other publicly available microarray datasets ( five genes ) ., We next characterized the phenotypes of rice lines carrying mutations in ten of the remaining candidate genes and then carried out co-expression analysis associated with these genes ., This analysis effectively provided candidate functions for two genes of previously unknown function and for one gene not directly linked to the tested biochemical pathways ., These data demonstrate the efficiency of combining gene family-based expression profiles with analyses of insertional mutants to identify novel genes and their functions , even among members of multi-gene families . | Rice , a model monocot , is the first crop plant to have its entire genome sequenced ., Although genome-wide transcriptome analysis tools and genome-wide , gene-indexed mutant collections have been generated for rice , the functions of only a handful of rice genes have been revealed thus far ., Functional genomics approaches to studying crop plants like rice are much more labor-intensive and difficult in terms of maintaining the plants than when studying Arabidopsis , a model dicot ., Here , we describe an efficient method for dissecting gene function in rice and other crop plants ., We identified light response-related phenotypes for ten genes , the functions for which were previously unknown in rice ., We also carried out co-expression analysis of 72 genes involved in specific biochemical pathways connected in lines carrying mutations in these ten genes ., This analysis led to the identification of a novel set of genes likely involved in these pathways ., The rapid progress of functional genomics in crops will significantly contribute to overcoming a food crisis in the near future . | genetics and genomics/genomics, genetics and genomics/gene discovery, genetics and genomics/gene expression, genetics and genomics/functional genomics, plant biology, plant biology/plant-environment interactions, genetics and genomics/gene function, plant biology/plant growth and development, plant biology/agricultural biotechnology, plant biology/plant genetics and gene expression, genetics and genomics/plant genetics and gene expression | null |
journal.pntd.0002644 | 2,014 | Characterization of a Gene Family Encoding SEA (Sea-urchin Sperm Protein, Enterokinase and Agrin)-Domain Proteins with Lectin-Like and Heme-Binding Properties from Schistosoma japonicum | Schistosomiasis still ranks as the most important helminthic infection; second only to malaria in its socioeconomic burden in the resource constrained tropics and subtropics ., It affects over 200 million people worldwide with more than 700 million people at risk of getting infected 1 ., Although an effective treatment is available ( praziquantel ) , the fact that reinfection occurs very rapidly after mass treatment renders chemotherapy alone inadequate for disease control ., It is opined that a prophylactic alternative applied singly or in combination with other interventions , even with limited efficacy in limiting transmission is the optimum approach 2 ., This intervention is especially needed in S . japonicum endemic areas , where non-human mammalian hosts are complicating control efforts ., Schistosomes inhabit host vasculature , where they ingest erythrocytes and catabolize the host hemoglobin as a source of amino acids for their growth , development and reproduction 3 ., However , large quantities of potentially toxic heme ( Fe-protoporphyrin IX ) are released as ‘byproducts’ of hemoglobinolysis 3–6 ., The parasite is thus faced with the challenge of maintaining heme homeostasis by evolving strategies to sequester and detoxify heme 3 , 5–9 , and at the same time maintaining a heme acquisition mechanism to harness the needed iron from the heme molecules 4 , 10 ., Indeed , effective mechanisms for detoxification of toxic heme and controlled acquisition of heme iron are paramount for parasite survival and establishment ., Such mechanisms are major targets of effective drugs against hemoparasites , including malaria and schistosomiasis 11–13 ., However , information on the exact mechanisms and molecules involved in this ‘weak link’ is either lacking or equivocal 3 ., Such molecular targets should be localized at the host-parasite interfaces in contact with the host erythrocytes ., The tegument and gastrodermis are syncytial layers lining the entire parasite surface and the parasite gut , respectively 14–16 ., Heme liberated during hemoglobinolysis is sequestered in the parasite gastrodermis lining the gut lumen 4 , 17 , and subsequently detoxified to non-toxic crystalline aggregates called hemozoin 8 , 9 , 17 , 18 and regurgitated ., The exact mechanism is not fully understood but it is thought that heme-binding proteins initiate the nucleation step of the crystallization , while lipids mediate the elongation step in an amphiphilic interface created by lipid droplets in the gastrodermis and gut lumen 17 , 19 ., Equally , schistosomes like other obligate parasites scavenge molecules from the host , including heme as the major source of iron needed for development and reproduction 4 , 10 ., Also , newly penetrated schistosomulae obtain iron via heme-binding proteins on their teguments before their guts are developed 20 ., Thus , heme-binding proteins that are localized at these interfaces are most likely involved in the parasite heme acquisition and detoxification ., Over the years , enormous resources and technologies have been channeled towards identifying molecular targets involved in several biological mechanisms utilized by parasites for effective parasitism ., The recently completed genome 21 , transcriptome sequences 22 and proteomic studies 23 of this parasite represent invaluable feats towards identifying such targets ., Although the functions of many sequenced genes are readily known or inferred from their amino acid sequences , many of the genes that are potential determinants of successful parasitism sometimes do not have readily identifiable sequence homologs ., This is a major challenge for placing the vast amount of ‘omics’ data into functional contexts for identifying genes of interest 24 , 25 ., As a matter of fact , several of such proteins presently annotated as ‘hypothetical proteins’ may well represent the missing link to filling the gene ‘gaps’ in our understanding of host-parasite interactions ., Indeed , over 30% of S . japonicum proteins are yet of unknown functions 21 ., Therefore , adopting novel strategies for the characterization of otherwise ‘hypothetical proteins’ is highly needed and can provide valuable functional clues that may not be readily identifiable from sequence data alone 24 , 25 ., Our group had utilized a signal sequence trap ( SST ) to isolate secreted and membrane binding antigens from S . japonicum with appreciable success 26 ., Among the SST isolated candidates , we identified a novel gene family which we found to have originated through a repetitive element mediated DNA-level gene duplication mechanism 27 ., Although several transcripts from ∼27 duplicons were identified , no sequence homolog was readily identifiable in other organisms ., We here utilized an integrated strategy combining comparative structural homology modeling and biochemical analyses to identify remote structural homologs , and characterize an extracellular domain in this family as SEA ( sea urchin sperm protein , enterokinase and agrin ) -domain ., Similar approach was used to further identify and characterize a functional heme-binding site on the SEA-domain ., SEA-module is an extracellular structural domain originally identified in sea urchin sperm protein , enterokinase and agrin , the basis for the nomenclature 28–30 ., The domain is found in several functionally diverse proteins , and is known to assist or regulate binding to carbohydrate moieties ., SEA-domain evolved from the ancestral ferredoxin-like fold , which is able to acquire various active sites including heme-binding sites 24 ., The identification of a functional heme-binding protein in this hemophagous trematode is a significant contribution to our understanding of the host-parasite interaction as regards heme homeostasis ., The biological significance of this finding and the potential role of this gene family in parasitism are discussed in terms of the parasite biology and prospects for application in disease intervention ., This study adhered strictly to the recommendations in the Fundamental Guidelines for Proper Conduct of Animal Experiment and Related Activities in Academic Research Institutions under the jurisdiction of the Ministry of Education , Culture , Sports , Science and Technology , Japan ( Notice No: 71 ) ., All animal experiments were approved by Nagasaki University Board of Animal Research , according to Japanese guidelines for use of experimental animals ( Approval No: 0809050699 ) ., Six to eight weeks old Female BALB/c mice were purchased from SLC Inc ., Labs , Japan ., The CLAWN strain miniature pigs were from Japan Farm , Kagoshima , Japan ., The miniature pigs were infected percutaneously with 200 S . japonicum cercariae ., Multiple alignments were performed using NCBI BLAST and Multialin Interface 31 ., Post translational modifications were predicted using YingOYang 1 . 2 32 ., Molecular structure modeling was performed by fold recognition and ab-initio structure prediction methods using Protein Homology/Analogy Recognition Engine ( Phyre v2 . 0 ) 33 and Rosetta Full Chain Protein Structure Prediction Server 34 ., Ligand binding analysis to identify potential ligands and their binding sites in the folded protein was performed using 3DLigandSite server 35 ., The modeled structures were analyzed using Discovery Suite 3 . 5 Molecular Visualizer , while the modeled receptor-ligand interactions were analyzed on the PyMol Molecular Graphics System , Version 1 . 6 ( Schrodinger , LLC ) ., Total mRNA was purified from parasite egg , sporocyst , cercaria and schistosomula using Micro-to-Midi total RNA purification system ( Invitrogen , USA ) , and from adult worms using NucleoSpin RNA II kit ( Macherey-Nagel , Germany ) ., Reverse transcription and amplification of the double stranded cDNAs were performed using Ovation Pico WTA System v2 ( NuGEN , USA ) ., For each candidate gene and the reference gene ( S . japonicum β-Actin ) , PCR fragment was first cloned into pCR2 . 1 cloning vector and the resulting constructs used as templates for qPCR standards and for estimation of copy numbers ., Relative expression of candidate genes in different developmental stages of the parasite was quantified using SYBR Premix Ex Taq II Reagents ( Takara , Japan ) ., Real-time PCR and data analysis were performed on AB 7500 Real-Time PCR Systems v2 . 0 . 5 ., The complete coding sequences of the candidates were amplified and cloned into the TOPO TA cloning site of the expression vector pcDNA4/HisMax and expressed in BL21 E . coli cells , and FreeStyle 293 expression system ( Invitrogen , USA ) for binding assays ., We took advantage of His6 tag to purify the recombinant proteins using TALON Metal Affinity Resins ( Clontech , USA ) ., Purified proteins were concentrated and imidazole elution buffer exchanged using Amicon Ultra Centrifugal Filters ( Millipore , USA ) ., Size exclusion gel filtration was performed using Sephadex G-50 medium ( GE healthcare , USA ) ., For heme-binding assays , purified proteins from FreeStyle 293 cells were treated with enterokinase to remove tags and purified with EK-Away resin ( Invitrogen , USA ) ., Polyclonal mouse sera were produced against recombinant antigens by subcutaneous immunization of mice with 25 µg of purified recombinant proteins in 50 µl PBS , mixed with an equal volume of Gerbu Adjuvant 100 ( GERBU Biotechnik , Denmark ) , on days 0 , 21 and 42 ., Two weeks after the last inoculation , mice were exsanguinated to collect sera and spleens were aseptically obtained for monoclonal antibody preparation using the Clonacell-HY system ( Stemcell Technologies , USA ) , according to manufacturers instructions ., The monoclonal antibodies were biotinylated using the one-step antibody biotinylation kit ( Mitenylbiotech , USA ) ., Freshly perfused adult S . japonicum were washed three times in PBS ( pH 7 . 4 ) and fixed in 4% neutral paraformaldehyde at 4°C until use ., The samples were alcohol dehydrated , embedded in paraffin , cut into 5–7 µm thin sections and then mounted on microscope glass slides ., Paraffin sections were deparaffinized by incubating for 10 min in two changes of xylene and rehydrated by sequential 10 min incubations in 100% , 95% , 70% and 50% ethanol , before rinsing in two changes of double deionized water ., Schistosomulae were prepared by mechanical transformation and washed in Hanks solution ., After washing with distilled water , the juvenile worms were fixed in cold acetone for 2 hours ., Two drops of acetone fixed schistosomulae were added to poly-L-lysine coated glass slides and dried overnight ., Immunoperoxidase technique was then performed as in adult worm sections ., Immunoperoxidase staining and immunofluorescence assays were performed using minor modifications to the method detailed by 36 ., Briefly , the sections for immunoperoxidase staining were treated with 3% H2O2 in PBS for 30 min to destroy endogenous peroxidase ., All sections were blocked for non-specific binding with 5% skim milk in PBS for 1 h , and then incubated for 2 h at room temperature with biotinylated monoclonal antibody or immune sera as indicated in each case ., After washing three times in PBS pH 7 . 4 for 5 min each , the sections were incubated in FITC conjugated secondary antibody for immune sera IFA ., For biotinylated mAB IFA and immunoperoxidase assays , sections were incubated for 30 mins with streptavidin-FITC ( 1∶500 ) and streptavidin-HRP ( 1∶500 ) solution respectively ., The immunoperoxidase sections were washed in PBS and treated with diaminobenzidine tetrahydrochloride ( DAB ) chromogen , according to manufacturers instructions ( Dako , Japan ) ., After counterstaining immunoperoxidase sections with Mayer hematoxylin , all the sections were washed , dehydrated by passage through alcohol and xylene , mounted , and viewed under Keyence All-in-one Fluorescence Microscope ( Keyence , USA ) ., Pre-immune serum was used as negative control ., For glycoprotein detection assay , SDS-PAGE fractionated purified recombinant proteins were stained using the Pierce Glycoprotein Staining Kit ( Thermo Scientific , USA ) ., We utilized array type sugar chip ( SUDx-Biotec , Japan ) ; which is an array of 48 structurally defined sugar chains ( glycans ) immobilized on a thin gold chip to analyze the interactions of the SEA-domain proteins with glycans using SPR imaging 37 ., The surface plasmon is excited when light is focused on the opposite side of the chip ., The reflective light is measurable and is altered in response to binding of the proteins to the immobilized glycans ., This alteration of the surface plasmon ( expressed as resonance units , RU ) is directly proportional to change in bound mass of analytes ., Real time measuring of the SPR RU was used to monitor changes in the surface concentration or amount of bound analytes ( protein ) ., One of the benefits of this SPR system is that the weak interactions , which are easily washed out in the regular array technology and therefore not recognized , can also be monitored in real time ., We used this method to detect real-time biological interactions between several glycans and the characterized SEA-domain proteins ., For assessing the specificity and affinity of the protein-glycan interactions , we used chondroitin sulfate GAG chip to measure the association and dissociation kinetics in real time to determine KD of the binding ., Hemin-agarose binding assay was applied to study heme binding as detailed by 38 ., Briefly , 200 µl of hemin-agarose ( Sigma-Aldrich , USA ) was washed three times in 1 ml of 100 mM NaCl-25 mM Tris-HCl ( pH 7 . 4 ) with centrifugation done at 750×g for 5 min ., Hemin-agarose was incubated with protein ( 20 µg ) for 1 h at 37°C with gentle mixing ., After 4 washes to remove unbound proteins , the beads were incubated for 2 min with elution solution ( 2% ( wt/vol ) SDS and 1% ( vol/vol ) β-mercaptoethanol in 500 mM Tris HCl , pH 6 . 8 ) , boiled at 100°C for 5 min; centrifuged , and the supernatant analyzed by SDS-PAGE ., Binding assay based on the peroxidase activity of bound heme was performed as detailed by 38 , 39 ., Briefly , micro-titer plate coated with serial dilutions of the recombinant protein was incubated with hemin ( 20 µg/100 µl ) at 37°C for 1 h ., The unbound hemin was removed and the wells were washed three times with PBS ( pH 7 . 3 ) ., 50 µl of ready-to-use substrate tetramethylbenzidine/H2O2 ( TMB ) ( Bangalore-Genei , India ) was added and the reaction stopped after 15 min with addition of equal volume of 1N H2SO4 ., The OD450 was determined in an ELISA plate reader ( Bio-Rad , USA ) ., The amount of hemin bound to protein was calculated from a linear graph of the peroxidase activities of known concentrations of hemin ., Optical absorption spectrometric studies were performed on Hitachi U-3900H spectrophotometer according to method detailed by 40 ., Briefly , the binding of proteins to heme was titrated by adding increasing amount of the protein ( 0–28 µM ) to 10 µM of heme in 40% dimethyl sulfoxide ( DMSO ) buffered with 20 mM HEPES ( pH 7 . 4 ) ., Difference in absorption spectra over a range of 350 to 700 nm was recorded ., We used the increase in absorbance at Soret peak ( 412 nm ) to monitor the formation of the protein heme complex ., The heme binding curve was constructed by plotting the change in absorbance at the Soret peak ( ΔA412 ) versus the protein concentrations ., The heme-binding curve was fitted using one site specific binding with Hill slope model on GraphPad Prism , v5 . 00 ., Data analysis was performed on GraphPad Prism , v5 . 00 ., Mann-Whitney test was used to compare differences between two groups , while Kruskal-Wallis test was applied to compare differences among several groups ., All plotted data are means with error bars representing standard deviation ( SD ) ., Statistical significance was designated as p<0 . 05 ., PFAM: PF01390 , SCOP: 82671 , SCOP: 54861 , PDB: 2e7v , PDB: 2acm , PDB: 1ivz , GenBank: AY570748 , GenBank: AY570737 , GenBank: AY570742 ., We had identified a novel gene family with similar signal sequence and promoter regions among SST isolated cDNAs ( Figure S1A ) 26 , and showed that this gene family had originated from retrotransposon-mediated gene duplication mechanism 27 ., Although several transcripts from ∼27 duplicons were found to belong to this family , we could not readily identify the molecular functions of these genes since no sequence homolog was readily identifiable in any other organism 27 ., Consequently , we utilized comparative structural homology modeling to identify features and domains that could predict the putative molecular functions of the encoded proteins ., Firstly , protein topology indicated that while all the members of this family bear similar signal sequence and are thus trafficked to the surface; some also contain C-terminal transmembrane regions , akin to type-I transmembrane proteins ( Figure S1B ) ., The molecular folding patterns of the proteins were modeled simultaneously in Phyre 2 and Rosetta using fold recognition and ab-initio structure predictions ( Figure S1C ) ., These programs create sequence alignment profiles from PSI-Blasts followed by scanning of ‘fold library’ to identify remote structural homologs from experimentally determined structures in PDB and SCOP databases 33 , 34 ., The secondary structure components showed antiparallel arrangement of β-sheets , backed by α-helices ( Figure 1A ) , typical of ferredoxin-like folds ., Interestingly , models from both programs identified an extracellular domain of ∼120 amino acids common among this family , with striking similar folding pattern as SEA-domain ( sea urchin protein , enterokinase and agrin ) PFAM: PF01390; SCOP: 82671 ( Figure 1A and Table S1 ) ., SEA-domain is a domain with ferredoxin-like fold SCOP: 54861 , found in several proteins of diverse functions in different organisms 28–30 , 41 , 42 ., Notably , crystal structure of the SEA-domain of transmembrane protease serine II ( TMPRSS2 ) of Mus musculus PDB: 2e7v was the highest scoring template at over 95% confidence , according to which the shown structures were modeled ., For clarity , only the original SST identified candidates are shown as representative of the family ( Figure 1A ) ., The structural models for all members of the gene family are summarized in Table S1 ., Other high scoring homologs at over 95% precision were the SEA-domains of Mucin 1 PDB: 2acm and Mucin 16 PDB: 1ivz ., To validate the models , rigid body superposition with the highest scoring template PDB: 2e7v was performed ., The result showed Cα and main chain root mean square deviations ( RMSD ) of 0 . 680 Å and 0 . 838 Å respectively for SjCP3842 , a representative member of this gene family ( Figure 1B ) ., Similar low RMSD values were recorded for the other candidates ., Ramachandran plot ( φ/ψ ) of conformation angles for each residue showed over 98% of the residues in the favored region , with less than 2% in the outlier region ., These results indicate the reliability of the predicted models ( Figure 1B ) ., A reciprocal ‘BackPhyre’ using the modeled structures to scan over 25 genomes also mapped the domain to SEA-domains at over 95% confidence , albeit with limited protein sequence homology ., The low sequence similarity ( Figure 1C ) observed from alignments of this extracellular domain with two major SEA-domains ( MUC1 and TMPRSS2 ) could imply that this structural similarity is at least partly independent of amino acid sequence homology 29 ., As a matter of fact , SEA-domains are primarily defined by their characteristic folding pattern , extracellular localization on transmembrane proteins , their ability to assist or regulate binding to glycans , and their presence in proteins with O-linked glycans 28 , 29 , 41 ., As expected , multiple O-glycosylation sites were identified by posttranslational modification prediction ., We also confirmed that the expressed proteins contain O-linked glycans using glycoprotein detection assay ( Figure S2 ) ., Equally , two conserved cysteine residues are present in all the candidates ( Figure S1A ) , which could be structurally important by providing disulfide bridges in the folded protein ., Further evidence to classify the identified domain as SEA-module was the identification of the typical glycine-serine amino acid consensus ( frpG/Svvv ) 30 auto-cleavage site of SEA-domains ( Figure 1C ) ., Some SEA-domain proteins have been shown to undergo auto-cleavage , although the resulting subunits remain non-covalently associated in the native state 30 , 41 , 42 ., This cleavage site is usually located within the bend between β2 and β3 sheets 30 as we equally observed ( red arrow in Figure 1 A and C ) ., In addition , the SDS fractionated recombinant protein ( shown later ) contained extra bands of expected molecular weight as the potential cleavage products ., Taken together , these results provide multiple grounds to classify this extracellular domain as SEA-domain ., To provide lead to the possible molecular function of the gene products , we subjected the modeled structures to ligand binding site identification using 3DLigandSite 35 ., This program uses protein structure to search a structural library to identify homologous structures with bound ligands , which are then superimposed on the protein structure to predict potential ligand binding sites 35 ., Interestingly , a binding site was observed for Fe-protoporphyrin-IX ( heme ) at significantly high precision ( Figure S3 ) ., Binding sites for energy transfer coenzymes including ATP , and several metal ions ( Mg , Zn , Cu ) binding sites were also identified ., The heme-binding site was predicted based on 178 heme ligands present in 177 homologous structures with bound heme ( Figure S3 ) ., Analysis of the modeled heme-binding pocket of SjCP3842 showed that the vinyl end of the amphiphilic heme is inserted into a hydrophobic cavity created between α2 and α3 helices , and β2 and β3 sheets ( Figure 2 A and B ) ., Many of the interacting residues in the binding pocket are conserved among the members of this protein family ( labeled in red in Figure S4B ) , consistent with binding of a heme group ., The hydrophilic propionate end ( red sphere ) of heme is rather facing away from the hydrophobic pocket ( Figure 2 A and B ) , with one propionate group engaged in electrostatic interactions with a nitrogen atom in Arg-157 side chain ( Figure 2C ) ., The phenyl rings of three conserved phenylalanine residues ( Phe-80 , Phe-140 and Phe-156 ) and one other phenylalanine ( Phe-143 ) engage in pi-stacking interactions with the heme Pyrrole rings , which further stabilize heme-binding ( Figure S4B ) ., There were also polar contacts between heme and Thr-79 , Tyr-83 , His-147 and His-149 ( Figure S4B ) , and several hydrogen bond interactions within the binding site ., Consistent with binding to heme , we readily identified potential axial ligands for heme iron , indicating hexa-coordination state involving two possible pairs ., The imidazole group on His-149 side chain ( bond distance of 2 . 0 Å ) is the putative proximal ligand with either His-147 ( Figure 2C ) or the thioether group on Met-50 ( Figure S4C ) as the distal ligand of heme iron ., However , the exact pair of axial ligands or the possibility of simultaneously binding two molecules of heme needs to be experimentally clarified ., Similar binding site characteristics were observed in another characterized candidate ( SjCP1531 ) ., However , the iron is coordinated to Tyr-154 as its axial ligand ( Figure S5 ) ., We investigated whether this gene family is differentially expressed among developmental stages of S . japonicum by stage specific mRNA expression using real time PCR ., All other in-vitro based characterization was limited to three candidates: SjCP3842 GenBank: AY570748 , SjCP1084 GenBank: AY570737 and SjCP1531 GenBank: AY570742 ., Relative expression of each candidate gene was quantified and expressed as copy number per nanogram of cDNA ( Figure 3 and Table S2 ) ., There was differential expression of the three genes among developmental stages of the parasite , with SjCP3842 expressed at higher levels relative to the other two characterized candidates ( Figure 3 and Table S2 ) ., SjCP3842 was overtly expressed in the adult stage ( 5680±370 . 9 ) , although at a higher level in female adult worm ( 4846±302 . 1 ) as compared to the male worms ( 2000±453 . 9 ) ., The expression levels in the snail intermediate inhabiting sporocyst ( 2474±627 . 2 ) and infective cercaria ( 2871±98 . 4 ) stages were also relatively high as compared to somula ( 543 . 4±64 . 1 ) and egg stage ( 252±370 . 1 ) ., SjCP3842 was expressed at the minimal level in the egg stage ( Figure 3A ) ., Conversely , SjCP1084 was mainly expressed in the egg stage in relation to other stages ., However , the expression levels of SjCP1531 in all stages of the parasite were relatively low and mainly expressed at the egg and adult stages ( Figure 3C and Table S2 ) ., To confirm expression at protein level , we expressed recombinant proteins , generated and used specific immune sera to identify the native proteins in parasite crude extracts ., The complete coding regions of the genes were amplified from S . japonicum adult worm cDNA library and cloned into the expression vector , pcDNA4-HisMax ., For recombinant protein expression , the plasmid constructs were transformed into Freestyle 293 and BL21 E . coli cells ., The recombinant proteins used for biochemical assays were expressed in Freestyle 293 cells to ensure proper folding and post translational modification ., The proteins were found to exist as oligomers in the native state as seen in the multiple bands of additive ∼30 kDa subunits observed both on SDS-PAGE ( Figure 4A ) , western blots probed with anti-HisG antibody ( Figure 4 B and C ) , and by multiple peaks from size exclusion chromatography fractions ( Figure 4D ) , all showing the tetramer as the native state ., Similar oligomeric state was also predicted by structural modeling ( Figure S1D ) ., Oligomerization may have been mediated by the disulfide bridges on two conserved cysteine residues common among the members of this family ( Figure S1A ) ., Other extra bands are of same molecular weight as the expected SEA-domain auto-cleavage products ( Figure 1C ) ., To confirm native expression and to show potential antigenicity of the candidates during schistosomiasis , immunoblotting and ELISA techniques were applied ., Parasite egg ( SEA ) and adult worm ( SWA ) crude antigen preparations were blotted and probed with the polyclonal immune sera ( α-SjCP3842 , α-SjCP1531 and α-SjCP1084 ) ., Blotted protein fractions of sizes similar to both the subunits ( ∼30 kDa ) and tetramer ( ∼120 kDa ) reacted specifically with the immune sera ( Figure 4E ) ., Also , the recombinant proteins specifically reacted with sera from S . japonicum infected miniature pigs , with significantly high titers of IgG in ELISA ( Figure 4F ) ., These results indicate that this gene family is actually expressed in the parasite , appear functional and potentially antigenic during schistosomiasis ., In addition to their characteristic folding pattern , SEA-domains are known to assist or regulate binding to carbohydrate moieties ., We assessed interactions of the characterized SEA-domain proteins with glycans using recombinant proteins and array type sugar chips in a Surface Plasmon Resonance ( SPR ) system 37 ., The SPR signal ( expressed in resonance units , RU ) is proportional to the amount of protein analytes bound to the sugar chains immobilized on the sensor chip in a 48 glycans array ., The SPR imaging showed specific binding to sulfated GAGs with relatively high affinity ., There was disproportionately high specific binding to chondroitin sulfate , dermatan sulfate ( CS-B ) , heparin , dextran sulfate and other sulfated GAGs ( Figure 5 ) ., SjCP1084 and SjCP1531 have similar glycan binding pattern while SjCP3842 showed relatively less glycan binding capacity but also preferentially binds sulfated GAGs ( Figure 5 ) ., We further confirmed the specificity and affinity of protein-GAG interactions by using chondroitin sulfate GAG ( CS-GAG ) chip containing all possible sulfated disaccharides subunits of chondroitin sulfate , and different concentrations of the protein as analytes ., The glycan array format of the CS-GAG chip used and the SPR imaging of the glycan binding assays are shown in a supplementary file ( Figure S6 A and B ) ., The binding kinetics of the carbohydrate-protein interactions showed significant binding affinity to CS-GAGs , with dissociation constant ( KD ) within the range of receptor-ligand interactions ( Figure S6 C and D ) ., Figure S6C shows the detailed sensorgram and the binding curve of the interaction between SjCP1084 and chondroitin sulfate E ( KD\u200a=\u200a9 . 84×10−9 M ) , as representative of the binding kinetics data ., The other KD values for the interactions of SjCP1084 and SjCP1531 with different sulfated disaccharides of chondroitin sulfate are summarized in Figure S6D , showing values within nanomolar range ., These results indicate the specificity and affinity of the observed protein-glycan interactions ., To validate the structure based heme-binding model , we showed heme-binding properties of this family in-vitro , by three independent methods: hemin-agarose binding assay , heme-dependent peroxidase activity of protein-hemin complexes and optical UV absorption spectroscopy ., First , we showed using SjCP3842 that the purified recombinant protein has potential to bind heme on hemin-agarose beads ., The eluted fraction showed evidence of specific binding of the protein to heme ( Figure 6A ) ., Same experiment performed using unconjugated Sepharose 4B as negative control did not show any trace of the protein in the eluted fraction ., Heme binding assay was repeated using the three characterized candidates and similar specific binding was consistently observed after immunoblotting using immune sera ( Figure 6B ) ., To confirm this observation in the native state , hemin-agarose beads were incubated with parasite adult worm crude antigen ( SWA ) to isolate the total heme-binding protein fractions in the parasite ., The fractions were blotted and probed with monoclonal antibody against SjCP3842 ( Figure 6C ) ., The result clearly showed the presence of the protein in the parasite heme-binding protein fractions ., The multiple bands are expected molecular weights of the monomer , dimer and tetramer ., The fact that binding was ablated by the reducing effect of β-mercaptoethanol and denaturing effect of sodium dodecyl sulfate ( SDS ) used for elution suggests that the observed heme-binding property is at least partly non-covalently mediated by structure of the folded proteins ., To estimate the amount of heme bound by the protein , we assayed the heme-dependent peroxidase activity of the protein-hemin complex using SjCP3842 ., We first estimated the peroxidase activities of known concentrations of hemin , and used the resulting standard curve ( linear graph ) to estimate the amount of heme bound by the characterized heme-binding protein based on the peroxidase activity of bound heme ( Figure 6D ) ., The result showed that the amount of bound heme increased with increasing protein concentration , reaching saturation at about 2 µg of protein , when 1 µg of hemin was bound ( Figure 6D ) ., To further assess the binding affinity of the protein-heme interaction , optical absorption spectra of the protein-heme complex was monitored by differential titration of 10 µM of heme with increasing concentrations of the protein ( 0 to 28 µM ) ( Figure 6E ) ., The Soret absorption peak for heme alone was characteristically broad and was initially 388 nm prior to addition of the protein ( broken lines ) ., The Soret absorption maximum was red shifted to 412 nm on addition of protein and absorbance at this peak increased gradually depending on accumulation of protein-he | Introduction, Materials and Methods, Results, Discussion | We previously identified a novel gene family dispersed in the genome of Schistosoma japonicum by retrotransposon-mediated gene duplication mechanism ., Although many transcripts were identified , no homolog was readily identifiable from sequence information ., Here , we utilized structural homology modeling and biochemical methods to identify remote homologs , and characterized the gene products as SEA ( sea-urchin sperm protein , enterokinase and agrin ) -domain containing proteins ., A common extracellular domain in this family was structurally similar to SEA-domain ., SEA-domain is primarily a structural domain , known to assist or regulate binding to glycans ., Recombinant proteins from three members of this gene family specifically interacted with glycosaminoglycans with high affinity , with potential implication in ligand acquisition and immune evasion ., Similar approach was used to identify a heme-binding site on the SEA-domain ., The heme-binding mode showed heme molecule inserted into a hydrophobic pocket , with heme iron putatively coordinated to two histidine axial ligands ., Heme-binding properties were confirmed using biochemical assays and UV-visible absorption spectroscopy , which showed high affinity heme-binding ( KD\u200a=\u200a1 . 605×10−6 M ) and cognate spectroscopic attributes of hexa-coordinated heme iron ., The native proteins were oligomers , antigenic , and are localized on adult worm teguments and gastrodermis; major host-parasite interfaces and site for heme detoxification and acquisition ., The results suggest potential role , at least in the nucleation step of heme crystallization ( hemozoin formation ) , and as receptors for heme uptake ., Survival strategies exploited by parasites , including heme homeostasis mechanism in hemoparasites , are paramount for successful parasitism ., Thus , assessing prospects for application in disease intervention is warranted . | While isolating membrane-bound and secreted proteins as targets for Schistosoma japonicum vaccine , we identified a novel potentially functional gene family which had originated by a gene duplication mechanism ., Here , we integrated structural homology modeling and biochemical methods to show that this gene family encodes proteins with sea-urchin sperm protein , enterokinase and agrin ( SEA ) –domain , with heme-binding properties ., Typical of SEA-structural domains , the characterized proteins specifically interacted with glycosaminoglycans ( GAGs ) , with implication in ligand gathering and immune-evasion ., Consistent with modeled heme-binding pocket , we observed high affinity heme-binding and spectroscopic attributes of hexa-coordinated heme iron ., Localization of the native gene-products on adult worm tegument and gastrodermis , host interfaces for heme-sequestration and acquisition , suggests potential roles for this gene family in heme-detoxification and heme-iron uptake . | biomacromolecule-ligand interactions, medicine, biochemistry, infectious diseases, genome analysis tools, genome databases, functional genomics, proteoglycans, sequence databases, schistosomiasis, genome evolution, comparative genomics, biology, genomics, gene prediction, parasitic diseases, glycobiology | null |
journal.pgen.1004651 | 2,014 | Coexistence and Within-Host Evolution of Diversified Lineages of Hypermutable Pseudomonas aeruginosa in Long-term Cystic Fibrosis Infections | The opportunistic pathogen Pseudomonas aeruginosa is found in many environments and can cause acute or chronic infections in a range of hosts from protozoans to plants to humans 1 , 2 ., In particular , patients with cystic fibrosis ( CF ) are highly susceptible to chronic colonization by P . aeruginosa , which is frequently fatal because of a persistent inflammatory response leading to gradual decline of lung function 3 , 4 ., In most cases , following a period of recurrent colonizations , a single strain of P . aeruginosa becomes predominant and persists for the rest of the patients life 5 , 6 ., Genetic adaptation has been shown to play a major role in successful establishment of long-term chronic P . aeruginosa infections of CF patients , and natural selection acts on these bacteria in CF airways to accommodate the fixation of mutations that cause beneficial phenotypic changes 7 , 8 , 9 ., The selected phenotypes display traits that differ from those of environmental isolates but are common in populations found in CF patients , suggesting repeatable patterns of long-term adaptation to the CF lung 10 , 11 , 12 ., A trait frequently observed in chronic infections is an increased mutation rate leading to a mutator phenotype 13 , 14 ., P . aeruginosa from chronically infected CF airways was the first natural model to reveal a high proportion of mutators in contrast to reported proportions in acute infections 13 ., Hypermutability in CF P . aeruginosa is due primarily to inactivation of the mismatch repair system ( MRS ) through lost function of the antimutator mutS and mutL genes 15 , and 36–54% of CF patients have been shown to be infected by mutator isolates 13 , 16 , 17 , 18 , 19 ., Theoretical and experimental approaches have attempted to explain the selection of MRS-mutators as the result of co-selection ( hitchhiking ) with linked beneficial mutations 20 , 21 , 22 , 23 , 24 , 25 , and their overrepresentation as a consequence of high recombination rates 26 ., Mutators have been linked to the development of antibiotic resistance both in vitro and in vivo 13 , 27 , 28 , 29 , 30 , 31 , and have been reported to enhance genetic adaptation to CF airways through increased accumulation of new mutations 32 ., However , comparisons between mutators and normo-mutators did not reveal any association between hypermutability and a particular distribution of mutations among genes , even for antibiotic resistance-related genes 32 ., No study to date has linked hypermutability in CF adaptation to any specific adaptive mutation 17 , 32 , 33 , nor to any key adaptive trait in the transition to a chronic state of infection 33 ., Our previous studies demonstrated a role of MRS deficiency in the acquisition of CF-related phenotypes under in vitro conditions such as mucoid conversion 34 , lasR inactivation 35 , 36 , and enhanced adaptability in biofilms 37 – all hallmarks of P . aeruginosa chronic airway infection ., We also reported the ability of MRS deficiency to bias mutagenic pathways toward DNA simple sequence repeats ( SSRs ) , which gave specific mutational spectra under both in vitro 34 , 38 and in vivo conditions 18 , 39 ., In view of the widespread effect of hypermutability on the process of adaptation to the CF lung 32 , it is important to elucidate the evolution of MRS-mutator strains in the course of CF chronic lung infections ., Previous genome analyses of longitudinally collected P . aeruginosa from CF patients demonstrated intra-patient genomic diversity of clonal isolates , suggesting that within-host P . aeruginosa population dynamics are driven by clonal competition ( clonal interference ) and/or niche specialization ( adaptive radiation ) 18 , 40 , 41 , 42 ., To further investigate these processes , Chung et al . compared the genomes of pairs of randomly selected contemporary isolates sampled from three chronically infected adult CF patients , and found that the pairs were differentiated by 1 , 54 , and 344 SNPs , respectively ., In the latter case , both isolates were mutators 43 ., Although mutators are frequently found in CF infections , no whole-genome studies have focused on the within-host evolution of mutators ., Similarly , diversity of within-patient pathogen populations is relevant to planning of clinical intervention strategies , elucidation of transmission networks , and understanding of evolutionary processes , but no study to date has involved genome sequencing of a sufficiently large collection of P . aeruginosa isolates taken from the same patient at the same time point to facilitate an in-depth analysis of population diversity ., We combined two distinct strategies for a genome-wide analysis of P . aeruginosa MRS-mutators:, ( i ) a longitudinal analysis of two separate clonal lineages of mutators obtained from two CF patients;, ( ii ) a within-host population analysis of a large collection of isolates obtained from a single sputum sample from each of the patients , to provide a snapshot of mutator population structure in the CF lung at a single time point ., Whole-genome sequencing of 27 P . aeruginosa isolates allowed us to quantify the nature and extent of the genomic changes of MRS-mutator clones and provide a panorama of the high genomic diversity that shapes the structure of P . aeruginosa mutator populations during long-term adaptation to the CF airway environment ., To quantitatively describe the evolutionary processes of MRS-deficient strains during chronic airway infections , we performed longitudinal and cross-sectional analyses of clonal P . aeruginosa isolates collected from two CF patients , referred to here as CFA and CFD ( Figure 1 , and Materials and Methods ) ., The cross-sectional study included 90 isolates obtained from a single sputum sample from each patient ., These large collections were used to investigate the clonal genomic diversity within mutator populations in a single host at a single time point ., Two different non-epidemic P . aeruginosa strains were collected from geographically distant locations , Argentina ( CFA ) and Denmark ( CFD ) , in which different therapeutic protocols are applied ., We were thus able to analyze two independent mutator populations whose evolutionary histories were presumably subjected to common as well as patient-specific selective pressures ., The collection from CFA included:, ( i ) Two sequential isolates obtained in 2004 ( CFA_2004/01 ) and 2007 ( CFA_2007/01 ) ., These isolates were characterized as MRS-mutators because they harbored missense mutations in the mutS and mutL genes and showed increased mutation rate ( Table 1 and Text S1 ) ., ( ii ) A collection of 90 P . aeruginosa isolates obtained from a single sputum sample in 2010 ( CFA_2010 ) ., The collection from CFD included:, ( i ) One normo-mutable isolate obtained in 1991 ( CFD_1991/01 ) ., ( ii ) Two sequential MRS-deficient mutators from 1995 ( CFD_1995/01 ) and 2002 ( CFD_2002/01 ) ( Table 1 and Text S1 ) that harbored the same mutS missense mutation 33 ., ( iii ) A collection of 90 P . aeruginosa isolates obtained from a single sputum sample in 2011 ( CFD_2011 ) ., The CFA and CFD collections covered periods of 6 and 20 yrs in the patients lives , respectively ., Based on previous studies indicating a doubling time of 115 min for P . aeruginosa in sputum 44 , we estimated that ∼36 , 500 and ∼91 , 400 duplication events occurred between the first and last isolates collected from CFA and CFD , respectively ., Genotypic characterization of the two collections by various molecular methods ( Materials and Methods ) showed that each patient was chronically infected by a single , unrelated P . aeruginosa clone that persisted throughout the study period ., The hexadecimal codes 45 for the SNP patterns of the CFA and CFD isolates analyzed are 2C32 and 249A , respectively ., The proportion of MRS-deficient mutators present in each patient was determined based on the rifampicin mutation frequency of the 90 P . aeruginosa isolates of the CFA_2010 and CFD_2011 panels ., All the CFA_2010 isolates showed mutation frequencies ( 1 . 7×10−5–7 . 6×10−6 ) consistent with a strong mutator phenotype ., Similarly , in the CFD_2011 panel , strong mutator isolates ( 1 . 4×10−5–8 . 3×10−6 ) comprised ∼94% of the population ., The remaining 6% showed mutation frequencies close to those observed in the prototypic wild-type normo-mutable strain PAO1 ( 3×10−8–1×10−7 ) ., These findings were supported by the observed prevalence of mutators in 90 isolates obtained ≥6 months later from new sputum samples ( CFA_2011 and CFD_2012 ) ; 100% of CFA and 90% of CFD isolates displayed a strong mutator phenotype ., To our knowledge , these are the highest proportions of mutators reported to date in large intra-patient populations of P . aeruginosa isolated from CF patients ., A previous study , which analyzed sets of 40 isolates per sputum sample from 10 CF patients , reported proportions of 27% or less 8 ., These findings indicate that P . aeruginosa can persist in chronic airway infections without the mutator phenotype , but in certain CF populations , such as those described here , MRS-deficient isolates may be prevalent and dominate the entire infecting population ., To analyze the genomic evolution of CFA and CFD P . aeruginosa mutator lineages , we performed whole-genome sequencing of 13 CFA and 14 CFD isolates ., From the CFA collection , we selected and sequenced the initial isolate CFA_2004/01 , the intermediate CFA_2007/01 , and 11 mutator isolates chosen randomly from the CFA_2010 population ( CFA_2010/01 , CFA_2010/11 , CFA_2010/26 , CFA_2010/31 , CFA_2010/32 , CFA_2010/40 , CFA_2010/43 , CFA_2010/72 , CFA_2010/78 , CFA_2010/82 , CFA_2010/87 ) ., From the CFD collection , we selected and sequenced the initial normo-mutable isolate CFD_1991/01 , the intermediate mutators CFD_1995/01 and CFD_2002/01 , and 11 isolates from the CFD_2011 population consisting of five normo-mutators ( CFD_2011/04 , CFD_2011/11 , CFD_2011/45 , CFD_2011/57 , CFD_2011/95 ) and six randomly selected mutators ( CFD_2011/27 , CFD_2011/28 , CFD_2011/33 , CFD_2011/34 , CFD_2011/83 , CFD_2011/94 ) ., The reads of CFA_2004/01 and CFD_1991/01 were assembled de novo , yielding genomes of 6 , 294 , 248 bp and 6 , 313 , 855 bp , respectively ( Table S1 ) , which were used as references in subsequent analyses ., The sequences of the remaining CFA and CFD isolates were aligned against the corresponding references to assess the genetic changes accumulated in the two mutator lineages during the infection process ( Table S2 ) ., Both lineages accumulated a high number of mutations during their evolution in CF airways in comparison with previously reported normo-mutable CF isolates such as the DK2 and PA14 clones 18 , 40 ., The CFA collection had a total of 2 , 578 single-nucleotide polymorphisms ( SNPs ) and 544 1- to 10-bp insertion/deletion mutations ( microindels ) ., The CFD collection had a total of 5 , 710 SNPs and 1 , 078 microindels ( Table S3 ) ., We applied Bayesian statistical analysis to infer time-measured phylogenies 46 , resulting in estimated mutation rates of 106 SNPs/yr ( 4 . 2×10−9 SNPs/bp per generation ) for CFA and 89 SNPs/yr ( 3 . 2×10−9 SNPs/bp per generation ) for CFD ., These findings indicate a mutation rate ∼40-fold higher than that ( 2 . 6 SNPs/yr ) reported previously for normo-mutable isolates obtained from CF chronic infections 18 ., SNPs were used as phylogenetic markers to perform a maximum-parsimony reconstruction of the evolutionary relationship of CFA and CFD isolate groups , and to evaluate temporal changes in their population genetic structure ( Figure 2 ) ., Alleles of P . aeruginosa reference strain PAO1 were used to root the trees ., In both cases , essentially all SNPs ( >99 . 5% ) supported single phylogenetic trees ., Interestingly , high genetic diversity was observed in both CFA and CFD P . aeruginosa intra-patient populations ., CFA_2010 and CFD_2011 contemporary clones were grouped together into three and four distinguishable clusters , respectively: Clusters II , III , and IV for CFA ( Figure 2A ) and Clusters I , IV , V , and VI for CFD ( Figure 2B ) trees ., Clusters were composed of genetically similar isolates , whereas more extensive genetic dissimilarities were observed between clusters ., The branch lengths among clusters differed substantially , indicating uneven mutational loads in coexisting P . aeruginosa intra-patient populations ., All CFD normo-mutable isolates were grouped together in Cluster IV ( Figure 2B ) ., In spite of the normo-mutable phenotype , this cluster shared common branches with the intermediate mutator isolates CFD_1995/01 and CFD_2002/01 and with their contemporary mutator clones ( branches D , F , and H ) ( Figure 2B ) ., These findings suggest that the normo-mutators arose from a mutator population of branch J at some point ., Cluster IV showed the highest accumulation of mutations during the infection process , despite the low mutation frequencies of its members ( Figure 2B and Table S3 ) ., The estimated time points of the most recent common ancestor ( MRCA ) of the CFA and CFD populations were 2002 and 1988 , respectively ., Interestingly , each of these estimates coincided with the year at which the patient was diagnosed as chronically infected ( Figure 1 ) ., This finding indicates that the divergent sub-lineages coexisted for many years – the same as the colonization period of the patient ( Figure 2 ) ., The question arose whether the genetic diversity observed in the two CF intra-patient populations was associated with differing repertoires of mutated genes among contemporary isolates , or whether most mutations occurred in common genes among the clones ., To distinguish between these possibilities , we first determined the set of genes of each CFA and CFD isolate that were altered by nonsynonymous SNPs and microindels ., Minimum Spanning Trees ( MSTs ) were then constructed to illustrate the relationships among contemporary isolates and their corresponding ancestors based on the number of distinctive mutated genes ., The CFA_2010 intra-patient P . aeruginosa mutator population ( Figure 3A ) was distributed in three main clusters , whereas the CFD_2011 population ( Figure 3B ) showed four clusters with all normo-mutable isolates grouped together ., The structures of the two MSTs indicate high genetic diversification and a scenario in which even contemporary CFA and CFD mutator populations diversified through mostly different evolutionary pathways ., We classified the SNPs according to their distribution in coding and noncoding regions and their effect in translation , to assess the selective forces acting on the CFA and CFD P . aeruginosa lineages ., Most of the SNPs in CFA and CFD were found to occur within coding regions , and 58 . 0% and 60 . 2% ( respectively ) corresponded to missense mutations ( Figure S1 ) ., The rates of nonsynonymous to synonymous mutations ( dN/dS ) were 0 . 68 for CFA and 0 . 79 for CFD lineages ., Thus , most of the SNPs that became fixed in the P . aeruginosa mutator lineages were neutral mutations , with most of each genome showing a signature of purifying selection and/or genetic drift ( dN/dS<1; P\u200a=\u200a5 . 2×10−56 and P\u200a=\u200a1 . 5×10−47 , respectively ) ., However , it is conceivable that positively selected genes remain “hidden” among the much larger number of genes that have accumulated mutations by genetic drift ., Evolving populations of P . aeruginosa presumably accumulate adaptive mutations in response to the human host environment in which they propagate ., We would therefore expect to observe parallelism in the adaptive genetic routes of the different lineages ., To confirm such convergent evolution , we attempted to identify genes that underwent parallel mutation in the 10 sub-lineages ( four CFA sub-lineages and six CFD sub-lineages ) that coexisted over many years ( Figure 2 ) ., We analyzed our dataset by selecting those genes that were independently mutated in at least half of the parallel evolving sub-lineages ( see Materials and Methods ) ., Forty genes were found to be frequently mutated across the sub-lineages ( Table 2 ) , suggesting that the parallel mutation of these genes was due to positive selection for mutations ., Consistently , the signature for selection for SNPs accumulated in the 40 genes ( dN/dS\u200a=\u200a0 . 97 ) was significantly higher than the ratio obtained for SNPs affecting all other genes ( dN/dS\u200a=\u200a0 . 75; P\u200a=\u200a0 . 029 by Fishers exact test ) , suggesting that these mutations were positively selected during evolution ., Analysis of the 40 genes was further focused on those that were non-synonymously mutated in at least half of the 10 sub-lineages ( Figure 4 ) ., Several of these genes were associated with functions related to CF host adaptation ., In particular , ftsI , ampC , fusA1 , mexY , PA1874 , and PA0788 are involved in resistance to antibiotics commonly used in CF therapies , i . e . , betalactams , aminoglycoside , quinolones , chloramphenicol , trimethoprim , and imipenem 29 , 47 , 48 , 49 , 50 , 51 , 52 , 53 ., Even though the GacA/GacS system is required for activation of genes involved in chronic persistence , gacS mutants are prone to generate stable and stress-tolerant small colony variants ( SCVs ) when growing in biofilms , exposed to stress factors 54 , 55 , or in vivo 56 , suggesting that the absence of GacS may confer some additional advantage for persistence in the CF lung ., fusA1 encodes the elongation factor EF-G1A , which confers resistance to the antibiotic argyrin in P . aeruginosa 57 , 58 ., Chung et al . recently reported independent fusA1 mutations in two CF patients and suggested that these mutations are involved in regulation of virulence through a ppGpp-dependent stringent response 43 ., On the other hand , the gene pslA is involved in biofilm formation 59 , and cupC3 is associated with motility/attachment 60 ., Lack of motility is a trait frequently observed in isolates from chronically colonized patients , and may give P . aeruginosa a survival advantage in chronic CF infection by enabling it to resist phagocytosis and conserve energy 61 ., Alterations in several genes related to bacterial catabolism ( e . g . , aceE , gcvP1 , soxA , xdhB , PA0794 ) were also observed , suggesting that the inactivation of certain metabolic functions may be a common trait related to CF host adaptation ( see below ) ., The concurrent alteration of specific genes or functions related to adaptation to the CF airway environment provides strong evidence for parallel evolution not only across CFA and CFD lineages , but also across intra-patient coexisting sub-lineages ., Certain genes were convergently but exclusively mutated among CFD sub-lineages , e . g . , ampC ( beta-lactamase precursor ) , PA0788 ( penicillin binding protein ) , and PA3271 ( two-component sensor ) ., These findings suggest the occurrence of in-host parallel evolutionary processes resulting from specific selective pressures from differential antibiotic treatments ., Mutations in the global regulators mucA , algT , rpoN , and lasR are related primarily to adaptation to the CF airway environment 10 , 17 , 33 , 62 , 63 , 64 ., However , our analyses did not reveal such mutations because they arose in the ancestral isolates before diversification into sub-lineages ( Table S4 ) ., The entire population from CFA had a mutation in mucA , whereas the population from CFD had mutations in lasR and rpoN ( Table S4 ) ., These findings indicate that mutations in these regulator genes were specifically fixed in the respective bacterial populations during early in-host evolution ., Our previous in vitro studies showed that , in a MRS-deficient background , G∶C SSRs constitute hotspots capable of biasing mutagenesis toward a specific genetic pathway 34 , 38 ., Our recent studies of P . aeruginosa PACS2 and epidemic DK2 strains demonstrated the same phenomenon at a genome-wide level in CF in vivo chronic airway infection 18 , 39 ., The present study design allows examination of such a genome-wide effect of biased mutagenesis in large populations obtained at single time points , and observation of SSR instability in coexisting isolates ., Analysis of types of mutations occurring in both the CFA_2010 and CFD_2011 isolates revealed a mutational spectrum typical of MRS-deficient strains ., Transitions ( ∼80% ) and small indels ( 1–4 bp ) ( ∼16% ) were the most frequently observed mutations in both collections ( Figure 5C ) ., The most prevalent transition was G∶C→A∶T , accounting for 65% and 62% of total transitions in CFA and CFD lineages , respectively ., Of the 1–4 bp indels , >80% were located within a homopolymeric SSR , and ∼75% were located specifically in G∶C SSRs ( Figure 5A ) ., In contrast , indels in A∶T SSRs accounted in average for 5 . 2% and 6 . 7% of the 1–4 bp indels in CFA and CFD isolates , respectively ( Figure 5A ) ., Based on this strong skewing of MRS spectra toward small indels in G∶C SSRs , we selected homopolymeric G∶C SSRs of ≥6 bp , which were mutated in at least half of the coexisting isolates in both CFA and CFD lineages , and analyzed their mutational dynamics at the intra-population level ., Eleven of these highly mutated G∶C SSRs harbored 2–5 distinct indel mutations accounting for independent mutational events ( Figure 5B ) ., A single SSR was observed to be either unaltered or modified by different indel mutations even in coexisting isolates from the same cluster ., Analysis of these 11 G∶C SSRs in the normo-mutable CFD_1991/01 isolate showed no mutations ., Using genome data available online , we evaluated the occurrence of indel mutations in these G∶C SSRs in 12 normo-mutable P . aeruginosa strains ( PAO1 , PA14 , M18 , NCGM2 . S1 , B136-33 , RP73 , 39016 , PACSC2 , 2192 , C3719 , DK2 , LESB58; the latter five are normo-mutable isolates obtained from CF infections ) , whose genomes have been sequenced and are available in the Pseudomonas Genome Database ( www . pseudomonas . com ) 69 ., Our survey revealed that these 12 strains harbored no indel mutations in the analyzed G∶C SSRs , even though large G∶C SSRs are considered to be “hotspots” for mutagenesis ., One of the identified homopolymers is located in a gene ( PA4071/PADK2_03970 ) which has been previously suggested to be preferentially mutated in mutators and to represent a mutator-specific target of adaptive mutations 18 ., These findings suggest a scenario in which MRS-deficient populations generates a vast of genetic diversity due to G∶C SSR instability ., In this scenario , genes containing large G∶C SSRs constitute continual sources of genetic diversification primarily in mutator bacterial populations ., We evaluated the dynamics of phenotypic changes in the 27 P . aeruginosa CFA and CFD isolates by determining global catabolic activities ( the “catabolome” ) ., Biolog phenotype microarrays were used to monitor the catabolic profiles of each isolate with various C and N sources ( Table S7 ) ., The total catabolic functions in the isolates were greatly reduced ( average reduction 73 . 5% and 63 . 8% , respectively ) in comparison with those of the CFA_2004/01 and CFD_1991/01 ancestors ( Figure 6 ) ., This extensive loss of functions led to homogeneous populations in both the CFA and CFD lineages , with slight catabolome variation among clones ., Catabolic function reduction thus appears to be a phenotypic pattern shared by CFA and CFD mutator lineages ., Accordingly , genes related to catabolism were convergently mutated in both the CFA and CFD lineages ( Figure 4 ) ., This phenomenon may be partially responsible for the decreased catabolic phenotype ., This study provides a complete panorama of the genomic diversity that shapes the structure of P . aeruginosa mutator populations during long-term adaptation to the CF airway environment ., We combined a longitudinal study with an extensive cross-sectional approach , including multiple isolates obtained from single sputum samples , which allowed in-depth analysis of population diversity ( Figure 1 ) ., We utilized P . aeruginosa panel collections from two chronically infected CF patients , CFA ( Argentinian ) and CFD ( Danish ) , with time spans of 6 yrs and 20 yrs , respectively , from initial to later stages of chronic infection ., Our comprehensive study design included whole-genome sequencing and high-throughput phenotypic approaches , calculation of mutation frequencies , phylogenetic estimation of time points of sub-lineage diversification , and analysis of mutS and mutL genes to obtain a wide-ranging depiction of hypermutability in CF ., We expected ( and confirmed ) that each of the two patients was infected by a single non-epidemic P . aeruginosa clone that did not present , during the initial stages of infection , the pathoadaptive mutations displayed by epidemic clones 42 , 70 , 71 , 72 ., We were therefore confident that our analysis addressed specific and independent in-host evolutionary processes ., Mutator strains were highly prevalent in both patients , essentially dominating the populations ., The proportion of mutators was ≥90% in the single time point 90-isolate collections , indicating that mutators , once selected , dominated the CFA and CFD infecting populations ., The observed prevalence of within-patient mutators was much higher than the values reported in previous studies 8 , 13 ., These findings indicate that although P . aeruginosa may persist throughout the course of chronic infection without ever acquiring the mutator phenotype , mutator strains may become prevalent and even dominate the whole population under certain yet-unknown conditions ., This concept is supported by the observation that two patients of different ages from geographically distant locations , infected with different non-epidemic P . aeruginosa clones and subjected to different therapeutic protocols , underwent overlapping evolutionary trajectories that led to complete domination of mutators ., Recent reports have demonstrated high diversity at the phenotypic level among P . aeruginosa populations from CF lung infections 8 , 9 , 73 ., However , there have been no genome-wide studies of such diversity in bacterial populations from the same clinical samples ., The global picture of genetic structure of intra-patient mutator populations in the present study reveals significant genomic diversity driven by high accumulation of mutations ( Figures 2 and 3 ) , reflected by the typical MRS spectra ( Figure 5 ) ., The distribution and combination of thousands of mutations result in a unique genotype for every isolate , allowing long-term persistence in the CF airway environment ., The observed genomic variation into the CFA and CFD lineages indicates that the population structure in each case was not determined by homogeneous single dominating clones , but occurred through multiple evolutionary genetic pathways that adapted equally to the CF airway environment and allowed the coexistence of diverse subpopulations for many years ., We determined that this high genomic diversity , to an equal degree in the two patients , spread out from the establishment of chronic infection ., Interestingly , MRS genes were also characterized by the coexistence of multiple polymorphisms ., However , underlying the polymorphisms within the MRS genes , there is an ancestral mutation that was fixed in each CFA and CFD population and is apparently responsible for hypermutability ., These findings suggest the existence of common selective forces acting on MRS inactivation in the two patients ., The long-term evolution of the P . aeruginosa CFA and CFD lineages was signed mainly by purifying selection and/or genetic drift ., There are conflicting reports regarding whether genomic evolution of CF isolates shows signatures of positive selection 10 or ( in contrast ) genetic drift and/or purifying selection 32 ., Our results strongly support the latter concept; i . e . , that genomic signatures of purifying selection and/or genetic drift are not inherent consequences of mutators , but are characteristic of the genetic adaptation processes underlying P . aeruginosa persistence in chronic lung infections ., According to our observations , the large number of mutations were for the most part distributed randomly among the P . aeruginosa mutator genome ( Figure 3 ) ., However , we also identified a group of genes that were convergently mutated in multiple genomes by independent mutational events ( Table 2 and Figure 4 ) ., Most of these genes code for functions related to pathogenicity ( e . g . , antibiotic resistance , virulence , motility , attachment ) , suggesting that they were positively selected as beneficial mutations ., We note that five of the genes ( ampC , ftsI , fusA1 , PA3271 , PA2018 ) were also found to be mutated in isolates obtained from the epidemic DK2 clone 18 , providing evidence of parallel evolution for certain specific traits among different P . aeruginosa lineages ., As we have reported recently for the PACS2 39 and epidemic DK2 P . aeruginosa strains 18 , the impact of hypermutability on the evolution of the CFA and CFD lineages is reflected by the high tendency of G∶C SSR-containing sequences to be mutated ( Figure 5A ) ., This finding confirms that genes which maintain G∶C SSRs in their coding region and/or in neighboring regulatory sequences are highly unstable in an MRS-deficient background and may be mutator-specific targets of adaptive mutations ., This concept is extended here by the demonstration that large G∶C SSRs , as DNA sequences per se , are highly polymorphic in single time point populations , indicating that they are continual sources for diversification ( Figure 5B ) ., This SSR-driven diversity is not observed in genomes of other P . aeruginosa strains , even of normo-mutable CF clones ., Our present and previous results 18 , 34 , 38 , 39 , taken together , demonstrate a clear association between MRS-deficiency and G∶C SSR instability , which exerts a global effect along the entire genome ., The impact of hypermutability during evolution of P . aeruginosa in the CF airway environment is not simply a major , rapid acquisition of mutations in quantitative terms ., In contrast with SNPs , indels that are not multiples of three produce frameshifts in the coding sequence of genes and thereby affect gene function ., Indels in G∶C SSRs may play an important role in the evolutionary process and in relation to mutator competitiveness ., Along this line , nine of the 11 analyzed G∶C SSRs ( Figure 5B ) were located in coding regions of the genome ., Five of these genes are predicted to encode for hypothetical proteins with no assigned function ., On the other hand , some G∶C SSRs were positioned in genes functionally related to transcriptional regulators ( PA1490 ) , adaptation-protection ( PA1127 ) , membrane proteins , and transport of small molecules ( PA1626 , PA2203 ) ., A small percentage ( 6% ) of the CFD population has a reduced mutation frequency similar to that of normo-mutable strains ., This subpopulation , which is grouped in a single cluster ( Cluster IV in Figure 2 ) , had the highest accumulation of mutations observed in the whole CFD collection ( Table S3 ) ., This observation posed the question whether the mutational load of these clones is too heavy to continue supporting a mutator phenotype ., However , the normo-mutable subpopulation carried the same mutS loss-of-function mutation ( Table 1 ) as the mutator isolates ., Phylogenetic analysis indicated that isolates from Cluster IV had arisen from a mutator subpopulation at some undetermined point in branch J ( Figure 2 ) ., Our sequencing data suggested that the most feasible explanation is the emergence of secondary mutations , in genes not belonging to the MRS , that compensate for mutS hypermutability , since neither reversion of the original −CG1551 nor duplication of the mutS gene was observed in normo-mutable genomes ., Alth | Introduction, Results, Discussion, Materials and Methods | The advent of high-throughput sequencing techniques has made it possible to follow the genomic evolution of pathogenic bacteria by comparing longitudinally collected bacteria sampled from human hosts ., Such studies in the context of chronic airway infections by Pseudomonas aeruginosa in cystic fibrosis ( CF ) patients have indicated high bacterial population diversity ., Such diversity may be driven by hypermutability resulting from DNA mismatch repair system ( MRS ) deficiency , a common trait evolved by P . aeruginosa strains in CF infections ., No studies to date have utilized whole-genome sequencing to investigate within-host population diversity or long-term evolution of mutators in CF airways ., We sequenced the genomes of 13 and 14 isolates of P . aeruginosa mutator populations from an Argentinian and a Danish CF patient , respectively ., Our collection of isolates spanned 6 and 20 years of patient infection history , respectively ., We sequenced 11 isolates from a single sample from each patient to allow in-depth analysis of population diversity ., Each patient was infected by clonal populations of bacteria that were dominated by mutators ., The in vivo mutation rate of the populations was ∼100 SNPs/year–∼40-fold higher than rates in normo-mutable populations ., Comparison of the genomes of 11 isolates from the same sample showed extensive within-patient genomic diversification; the populations were composed of different sub-lineages that had coexisted for many years since the initial colonization of the patient ., Analysis of the mutations identified genes that underwent convergent evolution across lineages and sub-lineages , suggesting that the genes were targeted by mutation to optimize pathogenic fitness ., Parallel evolution was observed in reduction of overall catabolic capacity of the populations ., These findings are useful for understanding the evolution of pathogen populations and identifying new targets for control of chronic infections . | Patients with cystic fibrosis ( CF ) are often colonized by a single clone of the common , widespread bacterium Pseudomonas aeruginosa , resulting in chronic airway infections ., Long-term persistence of the bacteria involves the emergence and selection of multiple phenotypic variants ., Among these are “mutator” variants characterized by increased mutation rates resulting from the inactivation of DNA repair systems ., The genetic evolution of mutators during the course of chronic infection is poorly understood , and the effects of hypermutability on bacterial population structure have not been studied using genomic approaches ., We evaluated the genomic changes undergone by mutator populations of P . aeruginosa obtained from single sputum samples from two chronically infected CF patients , and found that mutators completely dominated the infecting population in both patients ., These populations displayed high genomic diversity based on vast accumulation of stochastic mutations ., Our results are in contrast to the concept of a homogeneous population consisting of a single dominant clone; rather , they support a model of populations structured by diverse subpopulations that coexist within the patient ., Certain genes involved in adaptation were highly and convergently mutated in both lineages , suggesting that these genes were beneficial and potentially responsible for the co-selection of mutator alleles . | bacterial genomics, organismal evolution, microbial mutation, microbial evolution, genetics, biology and life sciences, microbiology, genomics, evolutionary biology, bacterial evolution, microbial genomics | null |
journal.pbio.1001903 | 2,014 | Non-associative Potentiation of Perisomatic Inhibition Alters the Temporal Coding of Neocortical Layer 5 Pyramidal Neurons | In the cerebral cortex , fast GABAergic inhibition is tightly coupled to excitation both temporally and in strength ., This constant balance of opposing forces is necessary for the correct development of cortical sensory receptive fields 1 and allows for the generation and tuning of cortical network activity underlying cognitive functions and complex behaviors 2 ., Indeed , it has been proposed that alterations of this equilibrium result in devastating neurological and/or psychiatric diseases , such as epilepsy , schizophrenia , and autism 3 ., Studies have shown that dynamic cellular mechanism are capable of compensating changes in synaptic excitation in order to maintain a particular excitation-to-inhibition ( E/I ) ratio intact , for example , either by weakening of feed-forward inhibition 4 or persistently enhancing inhibitory neurons excitability 5 ., Nevertheless , perturbations in the E/I balance can play a key role in sensory learning and receptive field reorganization 6 , 7 , suggesting it may be necessary to unlock the restrictive gate on the E/I balance ., However , no such cellular mechanisms have been demonstrated ., Moreover , the E/I ratio is remarkably different across cortical layers , resulting in layer-specific suppression or augmentation of pyramidal neuron output in response to sustained input activation 8 ., Thus , E/I ratios can be state-dependent and regulated according to computational requirements of specific microcircuit pathways ., In principle , short- and long-term forms of synaptic plasticity of either inhibitory or excitatory neurotransmission could be responsible for dynamically altering the E/I ratio of specific cortical networks ., This is especially true for cortical GABAergic synapses as they originate from a rich diversity of interneuron types 9 , 10 , which may differentially modulate the excitatory information flow along the dendro-somatic axis of pyramidal neurons ., In this context , alteration of the E/I ratio might have important and specific consequences in input–output transformations of pyramidal neurons and their ability to integrate and relay different salient features of sensory information ., Although the E/I ratio is usually referred to as a “global” balance , it is not known whether specific inhibitory circuits can induce region-specific unlocking of this equilibrium ., Interestingly , we have previously found that layer 2/3 pyramidal neurons can self-tune their excitability and inhibitory synaptic strength solely in response to their own activity 11 ., Whether this mechanism can alter the E/I balance is not known ., Here we show that in contrast to layer 2/3 , single layer 5 pyramidal neurons activity alone can alter E/I balance by inducing long-term potentiation of perisomatic inhibitory GABAergic transmission ( LTPi ) while leaving the strength of glutamatergic inputs unchanged ., Moreover , this plasticity is specific for inhibition originating from parvalbumin ( PV ) -positive basket cells and not somatostatin ( SST ) -expressing interneurons , which target distal dendrites ., Physiological burst-firing patterns of pyramidal neurons are sufficient to induce retrograde signaling of nitric oxide ( NO ) , which increases GABA release from NO-sensitive PV presynaptic terminals ., This non-associative potentiation of perisomatic GABAergic synapses results in an efficient layer 5 alteration of the balance between excitation and inhibition , reducing firing probability and , importantly , markedly sharpening the time window of synaptic integration ., This activity-dependent auto-modulation of layer 5 neocortical pyramidal neurons is ideally suited to enhance sparseness and improve the precision of time-coded information processing in a region-specific manner ., We examined whether layer 5 pyramidal neurons can modulate the strength of GABAergic synapses by postsynaptic depolarization similarly to layer 2/3 pyramidal neurons 11 , and if also glutamatergic transmission could be altered by postynaptic depolarization protocols ., Inhibitory postsynaptic currents ( IPSCs ) onto layer 5 pyramidal neurons were evoked by extracellularly stimulating their perisomatic afferents , in the continuous presence of the ionotropic glutamate receptor antagonist DNQX ( 10 µm ) ., Surprisingly , in contrast to layer 2/3 pyramidal neurons 11 , which responded to repeated somatic depolarizing steps with LTDi , a similar protocol ( ten 5-s long steps to 0 mV , repeated every 30 s from a holding potential of −70 mV ) induced a robust increase in the amplitude of eIPSCs onto layer 5 pyramidal neurons ., LTPi persisted for >30 min ( eIPSCs baseline , 260 . 1±24 . 03 pA; eIPSCs 20 min after steps , 517 . 4±77 . 50 pA , n\u200a=\u200a16 , p\u200a=\u200a0 . 0045 , paired t test; Figure 1A , B; normalized changes of eIPSCs , see Materials and Methods; ΔeIPSCs amplitude\u200a=\u200a129 . 0±40 . 7% , Figure S1A ) , and interestingly , it occurred in the absence of any presynaptic stimulation during somatic depolarizing steps ( non-associative LTPi ) ., An increase in eIPSCs amplitude of at least 50% of eIPSCs baseline amplitude was present in 71 out of 101 ( control or vehicle ) tested layer 5 pyramidal neurons ( 71 . 7%; e . g . Figure 1C ) ., Importantly , LTPi-inducing protocols failed to induce long-term plasticity of glutamatergic excitatory synaptic responses , which were isolated in the continuous presence of the GABAAR antagonist gabazine ( baseline , 161 . 6±22 . 08 pA; after steps , 178±21 . 86 pA , n\u200a=\u200a10 , p\u200a=\u200a0 . 3355 , paired t test; Figure 1D and 1E ) ., This potentiation of inhibitory but not excitatory synapses likely results in an unbalanced E/I ratio ( see below ) ., Although layer 5 pyramidal neurons fire rather irregularly during awake asynchronous states , they commonly display high-frequency ( >100 Hz ) burst firing depending on the behavioral state of the animal 12 , 13 ., We therefore tested if short bursts of APs ( induced by somatic current injections ) could increase GABAergic synaptic strength , similarly to somatic depolarizations in voltage clamp ., We recorded pharmacologically isolated evoked inhibitory postsynaptic potentials ( eIPSPs ) in layer 5 pyramidal neurons , in current-clamp mode with physiological intracellular chloride ( see Materials and Methods ) ., Repeated bursts of APs ( 5–10 spikes , at 50 or 100 Hz; Figure 2D ) efficiently increased GABAergic transmission onto layer 5 pyramidal neurons 1 . 8±0 . 4 versus 2 . 92±0 . 63 mV , baseline versus after 20 min after AP bursts ( 10 AP at 50 Hz ) , respectively , n\u200a=\u200a9 , p\u200a=\u200a0 . 004 , Wilcoxon signed rank test; Figure 2A–D ., Interestingly , repeated 1-s-long AP bursts at 50 Hz failed to induce GABAergic plasticity ( 3 . 95±0 . 86 versus 3 . 53±0 . 66 mV , baseline versus after 20 min after AP bursts , respectively; n\u200a=\u200a7 , p\u200a=\u200a0 . 29 Wilcoxon signed rank test; Figure 2C and 2D ) ., These experiments indicate that LTPi can be induced in current clamp by short postsynaptic bursts of APs alone ., Taken together , these results show a non-associative form of LTP of inhibitory synapses onto layer 5 pyramidal neurons: inhibition is increased by postsynaptic activity without the requirement of concomitant presynaptic stimulation ., In cortical structures , including the neocortex , perisomatic and dendritic GABAergic inhibition is provided by distinct interneuron types 9 , 14 ., We therefore sought to identify if LTPi was a general property of GABAergic synapses or if it was present at a specific inhibitory circuit ., First , perisomatic and dendritic GABAergic synapses were evoked in the continuous presence of DNQX ( 10 µM ) in the same neuron by placing two stimulation electrodes near pyramidal neurons cell bodies ( proximal , perisomatic stimulation ) and at a distal ( ∼400 µm ) dendritic location , respectively ( Figure 3A ) ., The integrity of dendrites was confirmed by visual inspection under IR-DIC video microscopy ., In some cases , neurons were filled with the fluorescent dye Alexa 594 ( 20 µM ) or with neurobiotin for post hoc histological reconstructions ( unpublished data ) ., Perisomatic IPSPs could be reliably potentiated by repeated bursts of 5 APs at 100 Hz ( proximal baseline , 1 . 42±0 . 23 mV; proximal after AP bursts , 2 . 9±0 . 57 mV , n\u200a=\u200a8 , p\u200a=\u200a0 . 0145 , paired t test; Figure 3A and 3B ) ., Layer 5 pyramidal cells were depolarized using AP trains in current clamp and in physiological chloride , as described in Figure 2 ., Interestingly , distal IPSPs were unaffected by the same postsynaptic firing protocol ( distal baseline , 1 . 149±0 . 28 mV; distal after AP bursts , 0 . 88±0 . 17 , n\u200a=\u200a8 , p\u200a=\u200a0 . 0549 , paired t test; Figure 3A , B ) ., To confirm synaptic activation of dendritic and perisomatic inhibition , in some experiments we gently cut pyramidal neuron dendrites using a knife pipette , at the end of experiments ., This procedure resulted in the disappearance of distally evoked IPSPs , leaving perisomatic responses unaltered ( Figure 3A ) ., In the neocortex , perisomatic and dendritic inhibition are provided by different interneuron classes 9 , 15 , 16 ., We tested plasticity of inhibition originating from PV-positive , fast-spiking ( FS ) basket cells , and SST-positive interneurons ., The former target the perisomatic region of pyramidal neurons , whereas the latter target the distal portion of their apical dendrites 15 , 17 ., To identify GABAergic transmission originating from PV interneurons , we used paired recordings between PV cells and pyramidal neurons ( using PV-Cre::RCE mice , Figure 3C–E; see Materials and Methods 18 , 19 ) ., Conversely , to selectively activate SST-cell IPSCs , we expressed the light-activated channel channelrhodopsin 2 ( ChR2 ) using viral vectors in SST-Cre mice ( see Materials and Methods 20 ) ., Optogenetic activation of SST interneurons invariably produced a response that was abolished by gabazine and had a relatively slow rise time ( 2 . 9±0 . 1 ms , n\u200a=\u200a9; Figures S2D and S3C , D ) , as compared to optogenetically evoked IPSCs from PV interneurons ( 2 . 0±0 . 1 ms; n\u200a=\u200a6 , p<0 . 05 , Mann–Whitney test; Figure S3C–D ) and consistent with dendritic electrotonic filtering ., A residual , minimal inward light-induced current was present in gabazine when both SST and PV neurons were photostimulated ( Figures S2D and S3E ) ., This residual current was abolished by 0 . 5 µM TTX ( Figure S3E ) as previously reported 21 ., In paired recordings between PV cells and pyramidal neurons , repeated depolarizing steps of postsynaptic pyramidal neurons potentiated unitary IPSCs ( uIPSCs ) in 7 out of 10 pairs ( ΔuIPSC LTPi , 59 . 18±24 . 63% , n\u200a=\u200a10 , p\u200a=\u200a0 . 0371 , Wilcoxon signed rank test; Figure 3D–E ) ., Importantly , non-depolarized PV–pyramidal neuron pairs showed no significant change over time ( −13 . 18±19 . 15% , n\u200a=\u200a5 , p>0 . 05 , Wilcoxon signed rank test; Figure 3E ) ., In contrast , LTPi-inducing stimuli failed to trigger plasticity of SST-cell IPSCs , induced by brief ( 2 ms ) flashes of blue light ( λ\u200a=\u200a470 nm ) ( baseline , 230 . 1±46 . 7; after steps , 239 . 9±47 . 86 pA , n\u200a=\u200a9 , p\u200a=\u200a0 . 57 , paired t test; Figure 3F–H ) ., Altogether , these results indicate that LTPi is interneuron-selective , as it affects perisomatic GABAergic synapses from PV basket cells , leaving dendritic inhibition from SST interneurons intact ., Previous results indicated that postsynaptic depolarization of layer 5 visual cortical pyramidal neurons from hyperpolarized membrane potential ( −90 mV ) can induce long-term plasticity of GABAergic neurotransmission due to alterations of postsynaptic trafficking of GABAARs 22 ., On the other hand , a recent study reported a non-Hebbian ( i . e . , non-associative ) presynaptic form of GABAergic plasticity in the thalamus 23 ., Therefore we decided to investigate the locus of expression of the LTPi described here ., We found that LTPi ( 10 out of 16 cells , Figure 1C and Figure S1A ) was accompanied by a significant increase in the frequency of spontaneous ( s ) IPSCs ( baseline , 6 . 9±0 . 99 Hz; after steps , 8 . 9±1 . 04 Hz , n\u200a=\u200a10 , p\u200a=\u200a0 . 0135 , paired t test; Figure 4A , B , left panel ) with no changes in their amplitudes ( baseline , 37 . 66±5 . 5 pA; after steps , 35 . 74±5 . 9 pA , n\u200a=\u200a10 , p\u200a=\u200a0 . 49 , paired t test; Figure 4B , right panel ) ., If LTPi resulted from increased GABAAR function at postsynaptic sites , a change in quantal synaptic event amplitudes would be apparent ., In fact , amplitudes of miniature ( m ) IPSCs ( recorded in 0 . 5 µM TTX ) were unchanged by somatic LTPi-inducing depolarizing steps ( baseline , 15 . 63±1 . 8 pA; after steps , 15 . 90±1 . 8 pA , n\u200a=\u200a10 , p\u200a=\u200a0 . 75 , paired t test; Figure 4C ) ., Conversely , similarly to sIPSCs , mIPSC frequency also increased in response to LTPi-inducing depolarizing steps ( 13 . 29±3 . 8 versus 19 . 39±5 . 9 Hz , baseline versus 20 min after steps; n\u200a=\u200a10 , p<0 . 05 , paired t test; Figure 4D ) ., Importantly , mIPSCs had very fast rise times ( <1 ms; Figure S1D ) , indicating that inhibitory quantal events were mostly perisomatic , as suggested by much faster IPSC rise times from PV as compared to SST cells ( p<0 . 01; Figure S3C–D ) ., No change of rise-time distribution was observed after LTPi-inducing stimuli ( Figure S1D ) ., Accordingly , coefficient of variation ( CV ) analysis of uIPSCs obtained from connected PV-basket cells and pyramidal neurons revealed that five out of seven pairs exhibiting LTPi had a purely presynaptic locus of expression ( Figure 4E , F ) ., Moreover , the ratio of PV-cell–induced uIPSCs elicited at a short time interval ( 20 ms , paired-pulse ratio , or PPR ) significantly decreased after LTPi ( baseline , 1 . 1±0 . 06; after steps , 0 . 9±0 . 03 , n\u200a=\u200a7 , p\u200a=\u200a0 . 01 , Wilcoxon signed rank test; Figure 4G , H ) ., Also extracellularly evoked IPSCs increased their CV and decreased their PPR , following LTPi ( Figure S1B , C , D ) , consistent with a presynaptic locus of plasticity ., Importantly , these parameters remained unchanged in cells that did not express LTPi ( Figure S1B and C ) ., To examine whether postsynaptic GABAergic plasticity also contributes to LTPi 22 , we performed single photon photolysis of the caged GABA compound Rubi–GABA ( 20 µM ) before and after LTPi , using a 5 µm laser spot ( λ\u200a=\u200a488 nm ) positioned at the perisomatic region of layer 5 pyramidal neuron ., Photolysis-evoked IPSCs ( pIPSCs ) were elicited with 1 ms laser pulses , producing baseline current amplitudes ranging between 25 and 160 pA ., Following the same somatic depolarization used to induce LTPi , we did not detect alterations in the amplitude of pIPSCs ( 85 . 22±10 . 9 pA baseline and 97 . 92±14 . 73 pA after step depolarization , n\u200a=\u200a11; p>0 . 05 , paired t test; Figure 4I–L ) , ruling out a postsynaptic locus of LTPi ., Taken together , these results show that LTP of perisomatic inhibitory synapses is expressed primarily presynaptically ., What are the cellular mechanisms underlying LTPi ?, Elevation of postsynaptic calcium concentration ( Ca2+ ) is often involved in GABAergic synaptic plasticity 24 ., To prevent postsynaptic Ca2+ elevations in pyramidal neurons , we included the Ca2+chelator 1 , 2-bis- ( o-aminophenoxy ) -ethane-N , N , N′ , N′-tetraaceticacid , tetraacetoxymethyl ester ( BAPTA , 20 mM ) in the intracellular pipette solution ., In this condition , LTPi was prevented ( IPSCs , 289 . 7±26 . 69 versus 243 . 0±32 . 64 pA , before versus 20 min after depolarizing steps , respectively; n\u200a=\u200a9 , p\u200a=\u200a0 . 14 , paired t test; Figure 5A , B and Figure S4A ) ., Importantly , LTPi was present in control conditions , even when induced after up to 20 min of intracellular dialysis , following patch rupture ( IPSCs , 289 . 3±43 . 93 versus 608 . 6±115 . 5 pA , before versus 20 min after depolarizing steps , respectively; n\u200a=\u200a9 , p\u200a=\u200a0 . 0034 , paired t test; Figure 5A , B and Figure S4A ) ., To confirm that intracellular Ca2+ elevations is required for the strengthening of inhibitory synapses originating from PV cells , we expressed ChR2 in PV cells and elicited IPSCs originating from this cell type selectively ., We confirmed LTPi following photostimulation in control conditions ( p<0 . 05; n\u200a=\u200a6 Wilcoxon signed rank; Figure S3A–B ) , similarly to results illustrated in Figure 3C–E ., Importantly , intracellular perfusion of BAPTA completely abolished LTPi ( Figure S3A–B; p>0 . 05 , n\u200a=\u200a4 ) similarly to results shown in Figure 5A–B ., In addition , these experiments confirm that photostimulated IPSCs can undergo LTPi ., Voltage-gated Ca2+ channels ( VGCCs ) and ionotropic glutamate NMDA receptors are efficient sources of postsynaptic Ca2+ , classically involved in synaptic plasticity ., We found that the L-type Ca2+ channel blocker nifedipine ( 10 µM ) prevented LTPi ( IPSCs , 240 . 2±30 . 02 versus 280 . 5±29 . 92 pA , before versus 20 min after depolarizing steps , respectively; n\u200a=\u200a16 , p\u200a=\u200a0 . 168 , paired t test; Figure 5C–E ) , whereas blockade of NMDARs with D-APV ( 50 µM ) had no effect on this form of GABAergic plasticity ( IPSCs , 277 . 3±36 . 02 versus 652 . 9±114 . 3 pA , before versus after 20 min after depolarizing steps , respectively; n\u200a=\u200a11 , p\u200a=\u200a0 . 0019 , paired t test; Figure 5C–E ) ., Overall , these data show that postsynaptic Ca2+ signaling via L-type Ca2+ channels is important for LTPi induction ., Ca2+-dependent postsynaptic induction of persistent changes of presynaptic GABA release suggests the involvement of retrograde synaptic signaling ., Two major molecular messengers have been indicated as responsible for retrograde synaptic signaling and plasticity: endocannabinoids and NO 11 , 23 , 25 ., We found that CB1 blockade by AM-251 ( 2 µM ) was ineffective in preventing LTPi ( Figure S4B ) ., However , when NO production was prevented by preincubation and constant perfusion of cortical slices with the general NO synthase inhibitor Nφ-nitro-L-arginine methyl ester ( L-NAME , 100 µM ) , LTPi was blocked ( IPSCs , 337 . 3±43 . 38 versus 473 . 9±115 . 2 pA before versus 20 min after depolarizing steps , n\u200a=\u200a11 , p\u200a=\u200a0 . 22 , paired t test; Figure 6A and Figure S4B ) ., Importantly , LTPi could be reliably induced in interleaved control experiments , incubating slices with the L-NAME vehicle ( IPSCs , 257 . 0±43 . 38 versus 660 . 2±119 . 7 pA , before versus 20 min after depolarizing steps , n\u200a=\u200a15 , p\u200a=\u200a0 . 001 , paired t test; Figure 6A and Figure S4B ) ., Accordingly , LTPi was prevented by intracellular perfusion of L-NAME via the patch pipette ( p>0 . 05 , n\u200a=\u200a11; unpublished data ) ., Moreover , application of the NO donor S-nitroso-N-acetylpenicillamine ( SNAP , 200 µM ) , in the continuous presence of the phosphodiesterases inhibitor ( IBMX , 200 µM ) , induced an increase of eIPSCs ( 184 . 6±32 . 25 versus 515 . 3±151 . 8 pA , before versus after SNAP , respectively; n\u200a=\u200a9 , p\u200a=\u200a0 . 039 , Wilcoxon signed rank test; Figure 6B and 6C ) ., IBMX was used to prevent nonspecific cGMP degradation 26 ., Pharmacological inhibition of the canonical NO receptor guanylylcyclase ( GC ) with 1H-{1 , 2 , 4}oxadiazolo{4 , 3-a}quinoxalin-&-dione ( ODQ , 10 µM ) completely blocked the induction of LTPi ( control , 206 . 6±31 . 36 pA versus 414 . 6±69 . 79 pA , before versus after depolarizing steps , n\u200a=\u200a13 , p\u200a=\u200a0 . 0064 , paired t test; ODQ , 261 . 6±24 . 01 pA versus 274 . 4±38 . 92 , before versus after depolarizing steps , n\u200a=\u200a13 , p\u200a=\u200a0 . 724 , paired t test pA; Figure 6D and Figure S4C ) ., Interestingly , when GC activity was impaired by ODQ 5 min after induction of LTPi , its maintenance was preserved ( p<0 . 01 Wilcoxon signed rank test; Figure S5A–C ) , suggesting that constant GC activity is not required for the expression of this form of plasticity ., Protein kinase G ( PKG ) was shown to be involved in the expression of NO-dependent GABAergic plasticity 27 , 28 ., Accordingly , when we blocked PKG with the inhibitor KT5823 ( 500 nM ) , LTPi was prevented , in fact producing a significant reduction of eIPSCs after the steps ( control , 253 . 6±47 . 84 pA versus 442 . 8±66 . 60 pA , before versus after depolarizing steps , n\u200a=\u200a11 , p\u200a=\u200a0 . 007 , paired t test; KT5828 , 213 . 8±30 . 31 pA versus 150 . 8±29 . 89 pA , before versus after depolarizing steps , n\u200a=\u200a10 , p\u200a=\u200a0 . 17 , paired t test; Figure 6E and Figure S4D ) ., Importantly , all drugs that were used here to affect various steps of NO signaling did not affect basal GABAergic synaptic transmission ( p>0 . 05 in all cases , Figure S5D ) ., If NO is involved in LTPi , then it should diffuse to synapses impinging neighboring neurons , as it has been previously shown 29 ., We therefore performed simultaneous recordings of two layer 5 pyramidal neurons , separated by various distances ., Depolarization of one postsynaptic pyramidal neuron ( PN1-test , Figure 7A and 7C ) invariably induced a long-term increase of eIPSC amplitudes , as expected ( PN1 LTPi amplitude\u200a=\u200a170±40 . 74% , n\u200a=\u200a11 , p\u200a=\u200a0 . 001 , Wilcoxon signed-rank test ) ., Interestingly , a significant , persistent increase of GABAergic transmission was also observed on a second , unperturbed cell ( PN2 , Figure 7A–C ) if it was within 50 µm from the depolarized cell ( PN2 LTPi\u200a=\u200a70±19 . 28% , n\u200a=\u200a7 , p\u200a=\u200a0 . 0156 , Wilcoxon signed-rank test ) ., In contrast , when the second pyramidal neuron was farther than 50 µm from PN1 ( Figure 7B–C ) , GABAergic transmission was unaffected by PN1 depolarizations ( PN2 LTPi\u200a=\u200a10±7 . 7% , n\u200a=\u200a5 , p\u200a=\u200a0 . 375 , Wilcoxon signed-rank test ) ., Overall , these results confirm the involvement of NO , which , as a gaseous diffusible messenger , can affect synapses impinging neighboring neurons within the cortical circuit ., We have shown that postsynaptic activity does not potentiate dendritic inhibition ( Figure 3 ) ., To investigate whether the absence of LTPi at dendritic GABAergic synapses was due to a decrease in the dendritic Ca2+ produced by back-propagating APs ( bAPs ) at distal synapses , we performed two-photon Ca2+ imaging while delivering LTPi-inducing bursts of APs ., Using the low-affinity calcium indicator OGB-5N , we found that the peak intracellular Ca2+ transient produced by a train of 5 bAPs at 100 Hz decreased along the pyramidal neuron apical dendrite , but not to zero ., At ∼500 µm from the soma , corresponding to the location of distal stimulations , the peak change in Ca2+ was still 50% of that in the proximal dendrite ( ΔF/F 100 µm , 0 . 41±0 . 04 versus ΔF/F 500 µm , 0 . 22±0 . 06 , n\u200a=\u200a9 , p\u200a=\u200a0 . 027 , Wilcoxon matched pairs signed rank test; Figure 8A–C ) ., However , if the 50% smaller dendritic Ca2+ transient at distal dendrites is the major cause for the lack of LTPi , increasing dendritic Ca2+ might reveal LTPi at distal synapses ., We therefore depolarized pyramidal neurons in voltage clamp , in the presence of intracellular cesium to block K+ channels , a condition that permits robust depolarization of distal dendrites 30 ., We observed LTPi of perisomatic but not dendritic GABAergic synapses ( Figure S6A–B ) , suggesting that the lack of LTPi at distal inhibitory synapses is not due to reduced Ca2+ entry in distal dendrites , but due to a difference downstream ., We considered whether dendrite-targeting interneurons forming distal dendritic GABAergic synapses might lack sensitivity to NO ., Therefore , we applied the NO donor SNAP ( 200 µM , in the continuous presence of the phosphodiesterases inhibitor IBMX , 200 µM ) while stimulating dendritic IPSCs that were isolated pharmacologically ., We found that , in contrast to perisomatic IPSCs ( Figure 6B , C ) , dendritic GABAergic responses were insensitive to NO ( IPSC amplitudes , 100±39 . 37 and 73 . 05±17 . 05 pA; before and 20 min after SNAP application; n\u200a=\u200a7 , p\u200a=\u200a0 . 58 , Wilcoxon signed rank test; Figure 8D ) ., To test whether NO-mediated signaling changes GABAergic strength via alteration of presynaptic excitability or alterations in the presynaptic AP waveform of PV cells 5 , 31 , 32 , we tested whether somatic excitability and presynaptic Ca2+ dynamics are altered in response to the NO donor SNAP ., LTPi induction did not alter resting membrane potential , membrane resistance , firing dynamics , nor somatic AP waveform ( p>0 . 05 in all cases; Figure S7A–G ) ., Yet somatic and axonal APs can result from substantially different ion channels ., If the terminal AP waveform is changed by NO , this should be reflected by an altered magnitude of Ca2+ entry into the presynaptic bouton ., However , two-photon imaging of single AP-evoked presynaptic Ca2+ transients in PV-cell boutons , did not reveal NO-dependent alterations in their amplitude ( p<0 . 05; Figure S7H–K ) ., These experiments , in addition to LTPi-mediated increase of mIPSC frequency ( recorded in TTX ) , suggest that the expression of LTPi is downstream of Ca2+ entry , rather than a mechanism mediated by changes in PV-cell excitability ., Altogether , these data indicate that LTPi depends on retrograde NO signaling , which increases GABA release onto depolarized and nearby pyramidal neurons through a GC-dependent PKG activation ., Moreover , the lack of dendritic LTPi is due to lack of NO sensitivity of dendrite-targeting interneurons and not failure to intracellular Ca2+ propagation in distal dendrites ., LTPi-inducing protocols failed to induce long-term plasticity of glutamatergic excitatory synapses ( Figure 1D , E ) , suggesting that LTPi-induced alterations of E/I ratio might modulate the computational properties of pyramidal neurons ., Therefore , in current-clamp mode , with physiological intracellular chloride and leaving excitation intact , we evoked EPSP-IPSP sequences ( composite PSPs , Figure 9A , top panel ) ., LTPi-inducing burst firing produced a significant change in the composite PSP waveform ., Overall , the peak of the depolarizing ( EPSP ) component was unchanged ( baseline , 2 . 1±0 . 19 mV; after AP bursts , 1 . 9±0 . 22 mV , n\u200a=\u200a11 , p\u200a=\u200a0 . 0615 , paired t test; Figure S8A–B ) , but the PSP area significantly decreased as a consequence of potentiation of the hyperpolarizing ( IPSP ) component ( baseline , 51 . 36±8 . 1 mV/ms; after AP bursts , 19 . 30±6 . 1 mV/ms , n\u200a=\u200a11 , p\u200a=\u200a0 . 0017 , paired t test; Figure 9A , Figure S8A–B ) ., Interestingly , however , in some cases , LTPi led to the complete disappearance of the excitatory portion of the composite synaptic response ( Figure S8A , example 2 ) ., Importantly , LTPi strongly reduced the E/I ratio , measured as the EPSP area divided by the total composite PSP area ( baseline , 0 . 74±0 . 08; after AP bursts , 0 . 29±0 . 09 , n\u200a=\u200a11 , p\u200a=\u200a0 . 0020 , Wilcoxon matched pairs signed rank test; Figure 9A , bottom panel ) ., To measure synaptic integration we then injected postsynaptic pyramidal neurons with artificial excitatory postsynaptic currents ( aEPSCs ) , producing artificial ( a ) EPSPs ( Figure 9B , inset ) ., Using this approach , we could measure synaptic integration of temporally controlled , fixed-amplitude synaptic events 33 ., Indeed , aEPSCs were injected at different intervals from the recorded composite evoked PSPs ( Figure 9B ) ., When aEPSPs and composite PSPs occurred simultaneously ( time zero ) , they summated similarly before and after induction of LTPi ( normalized synaptic summation , 0 . 79±0 . 085 versus 0 . 67±0 . 139 , baseline versus after AP bursts; n\u200a=\u200a12 , q\u200a=\u200a0 . 1195 , F ( 11 , 10 ) =\u200a8 . 9 , p>0 . 05 , one-way ANOVA followed by Bonferronis multiple comparison test; Figure 9B ) ., Interestingly , however , a significant narrowing of the integration window was observed , after LTPi induction , at 5–10 ms time intervals ( normalized summation at 5 ms , 0 . 7±0 . 16 versus 0 . 14±0 . 26 , baseline versus after AP bursts; normalized summation at 10 ms , 0 . 55±0 . 17 versus 0 . 001±0 . 2219 , baseline versus after AP bursts; n\u200a=\u200a12 , q\u200a=\u200a3 . 087 , F ( 11 , 10 ) =\u200a8 . 9 , p<0 . 05 and p<0 . 01 for 5 and 10 ms , respectively , one-way ANOVA followed by Bonferronis multiple comparison test; Figure 9B ) ., We reasoned that , because distal GABAergic synapses do not express LTPi ( Figure 3 ) , activation of distal inputs can be reliably used to measure synaptic integration , before and after potentiation of perisomatic inhibition ., We evoked dendritic and perisomatic synaptic responses in the same pyramidal neuron , by stimulating distal and proximal afferents , respectively ( Figure 9D ) ., Separate activation of these two pathways was confirmed by the lack of short-term plasticity , when they were activated in voltage clamp at brief time intervals ( Figure S8C–E ) ., Also in these experiments , LTPi altered proximal PSP waveform ( p\u200a=\u200a0 . 3187 for PSP peak and p<0 . 05 for PSP areas , before and after LTPi induction; Kruskal–Wallis test; n\u200a=\u200a12; Figure 9D ) ., No significant changes were observed at distal PSP before and after LTPi induction ( p>0 . 05; n\u200a=\u200a12; Figure 9D–E ) ., Moreover , the E/I ratio decreased at proximal synapses after LTP induction , but it was unaltered at distal synapses ( p<0 . 05 baseline versus after bursts for proximal stimulation , and p>0 . 05 for distal stimulation; Kruskal–Wallis test , n\u200a=\u200a12; Figure 9E , right panel ) ., Even in this case , LTPi did not alter summation at time zero ( F ( 9 , 84 ) =\u200a5 . 116 , p>0 . 05 one-way ANOVA followed by Bonferronis multiple comparison test; Figure 9F–G ) , but PSP summation was significantly reduced at 8–12 ms intervals following potentiation of GABAergic synapses ( normalized summation , 0 . 65±0 . 16 versus 0 . 14±0 . 26 , baseline versus AP bursts , respectively , q\u200a=\u200a3 . 997 , F ( 9 , 84 ) =\u200a5 . 116 , p<0 . 01 , one-way ANOVA followed by Bonferronis multiple comparison test , n\u200a=\u200a12; Figure 9F–G ) ., Overall , these experiments indicate that layer 5 pyramidal neurons can alter their ability of summating temporally dispersed synaptic events in response to self-induced potentiation of GABAergic proximal synapses ., How does this alteration of synaptic integration window translate into spike output of layer 5 pyramidal neurons ?, Perisomatic E/I ratio is strongly reduced after LTPi induction , thereby likely contributing to a modification in the spike probability of pyramidal neurons ., To test this hypothesis , we stimulated perisomatic synaptic afferents to layer 5 pyramidal neurons in short trains ( 5 pulses at 25 Hz ) ., Stimulation intensity was adjusted in order to evoke sporadic firing as a result of EPSP summation ( Figure 10A ) ., Spike probability was calculated as the number of APs divided by the number of trials at each individual stimulus ., We found that the spike probability dramatically decreased after intracellularly evoked , LTPi-inducing AP bursts ( 5 APs at 25 Hz , repeated 10 times every 1 . 5 s; spike probability , 0 . 45±0 . 05 versus 0 . 23±0 . 06 , control versus LTPi , respectively; n\u200a=\u200a12 , p\u200a=\u200a0 . 0043 , paired t test; Figure 10B–C ) ., The presence of LTPi was confirmed as a change of composite PSP waveform ( as in Figure 9A , C and Figure 10A ) ., Interestingly , decrease of discharge probability was absent in a subset of cells that did not express LTPi ( spike probability , 0 . 42±0 . 03 versus 0 . 46±0 . 02 , baseline versus after bursts; n\u200a=\u200a4 , p\u200a=\u200a0 . 25 Wicoxon signed rank test ) ., Importantly , EPSP trains were evo | Introduction, Results, Discussion, Materials and Methods | In the neocortex , the coexistence of temporally locked excitation and inhibition governs complex network activity underlying cognitive functions , and is believed to be altered in several brain diseases ., Here we show that this equilibrium can be unlocked by increased activity of layer 5 pyramidal neurons of the mouse neocortex ., Somatic depolarization or short bursts of action potentials of layer 5 pyramidal neurons induced a selective long-term potentiation of GABAergic synapses ( LTPi ) without affecting glutamatergic inputs ., Remarkably , LTPi was selective for perisomatic inhibition from parvalbumin basket cells , leaving dendritic inhibition intact ., It relied on retrograde signaling of nitric oxide , which persistently altered presynaptic GABA release and diffused to inhibitory synapses impinging on adjacent pyramidal neurons ., LTPi reduced the time window of synaptic summation and increased the temporal precision of spike generation ., Thus , increases in single cortical pyramidal neuron activity can induce an interneuron-selective GABAergic plasticity effectively altering the computation of temporally coded information . | The proper activity of cortical neurons ( the brain cells responsible for memory and consciousness ) relies on the precise integration of excitatory and inhibitory inputs ., The excitation and inhibition ( E/I ) ratio has to remain constant both in time and strength to prevent neurological and psychiatric diseases ., Fast inhibitory synaptic inputs to cortical pyramidal neurons originate from a rich diversity of GABAergic interneurons that operate a strict division of labor by differentially targeting precise regions of the pyramidal neurons ., Here , we show that large pyramidal neurons of neocortical layer 5 can unlock the E/I ratio in response to their own activity ., Excitatory activity of pyramidal neurons , in the form of membrane depolarization or trains of action potentials , induces a Ca2+-dependent mobilization of nitric oxide , which diffuses to inhibitory synapses and triggers a persistent enhancement of GABA release ., Notably , this potentiation of inhibition is specific for synapses originating from parvalbumin ( PV ) -expressing interneurons that target mainly the perisomatic region of pyramidal neurons ., Long-term potentiation of perisomatic inhibition , in turn , changes the ability of pyramidal neurons to integrate excitatory inputs as well as the temporal properties of their own action potential output ., Selective plasticity of perisomatic inhibition can thus play a crucial role in cortical activity , such as sensory processing and integration . | calcium imaging, biochemistry, signal transduction, molecular neuroscience, specimen preparation and treatment, neurochemistry, mechanical treatment of specimens, specimen disruption, cell biology, neural networks, biology and life sciences, neurotransmitters, electroporation, cell signaling, neuroscience, neuroimaging, research and analysis methods | Long-term potentiation of inhibitory GABAergic transmission controls synaptic integration and action potential generation of specific neocortical neurons. |
journal.ppat.0030111 | 2,007 | Differential Regulation of Caspase-1 Activation, Pyroptosis, and Autophagy via Ipaf and ASC in Shigella-Infected Macrophages | An effective immune response against microbial pathogens relies on the ability of the host to sense the presence of the infectious agent as well as the ability to destroy the invading pathogen ., The presence of infection is detected through pathogen recognition molecules that sense unique microbial components called pathogen-associated molecular patterns ( PAMPs ) 1 , 2 ., The recognition of bacterial PAMPs is mediated by several host molecules , including Toll-like receptors ( TLRs ) that are present on the cell surface and endosomal compartments , as well as nucleotide-binding oligomerization domain ( NOD ) -like receptors ( NLRs ) that sense the presence of PAMPs in the cytosol 1 , 2 ., The NLR protein family contains more than 20 members , including Nod1 , Nod2 , cryopyrin ( also called as Nalp3 ) , Nalp1 , and Ipaf ., NLR proteins contain C-terminal leucine-rich repeats that are linked to microbial recognition , a centrally located NOD domain that mediates oligomerization , and an N-terminal effector domain that includes caspase-activation and recruitment domain or pyrin domain 3–5 ., Cryopyrin and Ipaf have been implicated in caspase-1 activation and interleukin ( IL ) -1β processing induced by TLR agonists , gout-associated uric acid crystals , and specific bacterial infection 6–9 ., Ipaf has been shown to mediate caspase-1 activation , IL-1β processing , and caspase-1-dependent cytotoxicity induced by intracellular Salmonella or Legionella 10–14 ., Caspase-1 activation and IL-1β processing induced through Nalp3 or Ipaf also required the adaptor protein ASC ( apoptosis-associated speck-like protein containing a C-terminal caspase recruitment domain ) , which is thought to be important for the formation of the inflammasome , a multiprotein complex that mediates caspase-1 activation 6 , 10 , 15 ., NLR proteins such as cryopyrin and Ipaf play a crucial role in processing mature IL-1β ( also IL-18 ) , which are important inflammatory cytokines in host defense against infection and pathogenesis of inflammatory disorders 16–18 ., Shigella are highly adapted human pathogens that cause bacillary dysentery ( also referred to as shigellosis ) ., The prominent pathogenic feature of Shigella is their ability to invade a variety of host cells , including epithelial cells , macrophages , and dendritic cells , which leads to severe inflammatory responses in intestinal tissue ., Internalized Shigella multiply in the cytoplasm of epithelial cells and induce actin polymerization at one pole of the bacterium , allowing intracellular bacteria to move within the cytoplasm and to spread into adjacent epithelial cells 19 , 20 ., Shigella also invade resident macrophages in the intestinal tissue , and the bacteria escape from the phagosome into the cytosol ., Infected macrophages undergo caspase-1-mediated cell death , termed pyroptosis , which is a newly identified pathway of programmed cell death associated with an inflammatory response that is accompanied by plasma membrane permeability and nuclear condensation 21–23 ., In the macrophage cytosol , Shigella induce pyroptosis through activation of caspase-1 that is dependent on IpaB , a protein secreted via the type III secretion system ( TTSS ) or by lipopolysaccharide ( LPS ) moiety released from the bacteria , leading to the processing and secretion of IL-1β 23 , 24 ., As a result , pro-inflammatory chemokines and cytokines produced by the macrophages and epithelial cells infected with Shigella elicit strong inflammation in the intestinal tissue ., Flagellin , a major subunit of the flagellum , is a prerequisite for pyroptosis and caspase-1 activation of infected macrophages with Salmonella and Legionella 11 , 12 , 14 , 25 , 26 ., In the case of Salmonella , Ipaf has been shown to be essential for caspase-1 activation and pyroptosis by sensing flagellin in the cytosol 11 , 12 ., In this study , we examined the mechanisms that regulate caspase-1 activation or cell death induced by Shigella in macrophages ., Surprisingly , we find that Ipaf mediates caspase-1 activation and cell death independently of flagellin in Shigella-infected macrophages ., In addition , autophagy was induced by Shigella infection , and this process was negatively regulated by Ipaf and caspase-1 , but not ASC ., Flagellin is required for pyroptosis and caspase-1 activation of macrophages infected with Salmonella and Legionella 11 , 12 , 14 , 25 , 26 ., On the other hand , Shigella flexneri strains are not motile and are non-flagellated bacteria ., To examine whether S . flexneri express flagellin , genome sequence information from the strains 2457T 27 and 301 28 was compared in silico with that of S . enterica serovar typhimurium LT2 29 ., Forty-two genes are associated with flagella formation in Salmonella , including fliC and fljB , two genes that encode flagellin proteins 30 ., Several genes that are known to be essential for flagellar assembly in Salmonella were absent in Shigella ., Eight genes ( flgC , F , K , L , flhD , fliF , J , and P ) were absent in the 2457T strain , and seven genes ( flgC , D , F , K , L , flhD , and fliF ) were deleted in the 301 Shigella strain ., In particular , we found that flhD , the master regulatory gene that controls the transcription of flagellar genes 31 , was not present in S . flexneri ., These results suggest that both flagellar assembly and flagellin expression is deficient in Shigella ., Consistent with this notion , Salmonella , but not Shigella , expressed FliC by western blot analysis with anti-FliC antibody ( Figure 1A ) ., In addition , the expression of the Shigella fliC gene was not detected by RT-PCR analysis ( Figure 1B ) ., However , the open reading frame of the Shigella flagellin gene ( fliC ) was intact in that expression of flagellin was induced under the inducible promoter after fliC plasmid complementation ( Figure 1A ) ., We next examined the involvement of FliC in Shigella-induced pyroptosis ., Mouse bone marrow–derived macrophages ( BMMs ) were infected with wild-type Shigella , ΔfliC mutant , or ΔfliC mutant strain complemented with fliC ( ΔfliC/fliC ) , and cell death was examined by lactose dehydrogenase ( LDH ) release in infected cells ., All the strains examined had similar kinetics of LDH release ( Figure 1C ) ., Thus , FliC expression is not essential for pyroptosis induction in Shigella-infected BMMs ., Moreover , infection of BMMs with wild-type Shigella and ΔfliC mutant , but not TTSS-deficient mutant ( S325 , mxiA::Tn5 ) , induced caspase-1 activation and caspase-1-mediated IL-1β processing ( Figure 1D and 1E ) ., These results indicate that an intact TTSS , but not flagellin , is required for caspase-1 activation and IL-1β processing in Shigella-infected BMMs ., Ipaf and ASC are required for caspase-1 activation in Salmonella-infected macrophages 10–12 ., To gain insight into the molecular mechanism responsible for caspase-1 activation induced by Shigella infection , we analyzed caspase-1 activation in infected BMMs isolated from wild-type , caspase-1-deficient , Ipaf-deficient , or ASC-deficient mice ( Figure 2 ) ., After infection , the processed p10 fragment from procaspase-1 was detected in wild-type BMMs , but this proteolytic cleavage was impaired in Ipaf-deficient and ASC-deficient BMMs ( Figure 2C and 2E ) ., Similarly , IL-1β processing was impaired in Ipaf-deficient and ASC-deficient BMMs , but not in wild-type BMMs ( Figure 2D and 2F ) ., At 2 or 3 h post infection , low levels of mature IL-1β were detected after infection of Ipaf-deficient BMMs , and at an earlier time in ASC-deficient cells , suggesting that in addition to caspase-1 , other proteases can contribute to proIL-1β cleavage ., This is consistent with detection of low levels of mature IL-1β in caspase-1-deficient BMMs at later time points ( Figure 2B ) ., The presence of residual proIL-1β cleavage at earlier time points in ASC-deficient macrophages may reflect increased cell death in response to Shigella when compared to Ipaf- and caspase-1-deficient macrophages ( see below ) ., It is also possible that other NLRs may contribute to caspase-1 activation in Ipaf- or ASC-deficient BMMs during Shigella infection ., These events are reminiscent of the Salmonella system in which at high multiplicity of infection ( MOI ) , there is residual IL-1β secretion that is induced independently of cytosolic flagellin and , presumably , of Ipaf 12 ., The uptake of bacteria was similar in wild-type , caspase-1-deficient , Ipaf-deficient , and ASC-deficient BMMs ( unpublished data ) , suggesting that deficient internalization of Shigella was not responsible for the phenotype ., These results indicate that Ipaf and ASC are important for Shigella-inducing caspase-1 activation and subsequent IL-1β processing ., Furthermore , unlike in Salmonella , the activation of the Ipaf-ASC inflammasome is independent of flagellin ., BMMs derived from Ipaf-deficient mice , but not ASC-deficient mice , are resistant to caspase-1-dependent Salmonella-induced pyroptosis 10–12 ., To investigate the role of Ipaf and ASC in Shigella-inducing pyroptosis , we analyzed LDH release from infected wild-type , Ipaf-deficient , and ASC-deficient BMMs ., We found that LDH release , a marker of pyroptosis , was abrogated in Ipaf-deficient BMMs within 2 h of Shigella infection when compared to wild-type BMMs , but the release was induced after 3 h of infection and by 5 h was comparable to that of wild-type BMMs ( Figure 3B ) ., The kinetics of LDH release induced by Shigella in Ipaf-deficient BMMs was similar to that in caspase-1-deficient BMMs ( Figure 3A ) 23 ., However , the kinetics of LDH release induced by Shigella in ASC-deficient BMMs was indistinguishable from that observed in wild-type BMMs ( Figure 3C ) , suggesting that ASC is not important in cell death induced by Shigella , despite ASC being required for caspase-1 activation ( Figure 2E ) ., Notably , Ipaf- and ASC-deficient BMMs did not undergo rapid LDH release when infected with the Shigella TTSS mutant ( Figure 3E and 3F ) , indicating that an intact TTSS is required for Ipaf-dependent induction of pyroptosis ., Thus , Ipaf , but not ASC , is required for the rapid pyroptotic response induced by Shigella infection in BMMs ., Furthermore , caspase-1 activation may be dispensable for pyroptotic cell death induced by Shigella infection in the absence of ASC ., Autophagy is induced by diverse death stimuli , including that associated with caspase-independent death , but the regulation of autophagy triggered by bacterial infection is poorly understood 32 ., In pathogen-infected cells , autophagy appears to function as a host defense mechanism that can be subverted by certain virulent bacteria to enhance their intracellular replication 33–37 ., To study the role of Ipaf and ASC in autophagy , we first examined whether autophagy is induced in Shigella-infected BMMs ., To assess autophagy , wild-type and caspase-1-deficient BMMs were transfected with GFP-LC3 ( Atg8 , a marker protein of autophagy ) using a retroviral vector , and the GFP-LC3 labeling pattern was visualized by fluorescence microscopy ., Autophagy induced by amino acid starvation was not affected by caspase-1 , because a similar number of GFP-LC3 aggregates that are typically associated with the formation of autophagosomal vesicles were observed in wild-type and caspase-1- , Ipaf- , and ASC-deficient BMMs ( Figure 4A and 4B ) ., As another approach , endogenous LC3-I to LC3-II conversion , which is an indicator of autophagosome maturation , was examined by western blotting after rapamycin treatment to induce autophagy 38–40 ., Although LC3-II was more abundant than LC3-I in steady state in all BMMs , LC3 conversions ( an increase of the amounts of LC3-II ) were actually observed in wild-type and caspase-1- , Ipaf- , and ASC-deficient BMMs ( Figure 4C ) , suggesting that rapamycin-induced autophagy is not affected by these genetic deficiencies ., The cellular localization of GFP-LC3 was examined 30 min after infection with Shigella , since at this early time the membrane integrity of the majority of wild-type BMMs was retained ( unpublished data ) ., As shown in Figure 5A and 5B , nearly 20% of intracellular Shigella was associated with accumulated GFP-LC3 in wild-type BMMs , whereas the percentage increased to about 90% in caspase-1-deficient BMMs ., No accumulation of GFP alone was observed in infected cells ., These results indicate that the absence of caspase-1 promotes autophagosome maturation induced by Shigella infection ., Interestingly , a large number of GFP-LC3-containing vesicles , which were not associated with bacteria , were observed in caspase-1-deficient BMMs infected with Shigella ( Figure 5A ) , suggesting that endogenous autophagy was also activated during infection ., The endogenous LC-I to LC3-II conversion was also detected by Shigella infection in caspase-1-deficient BMMs ( Figure 6 ) ., In epithelial cells , internalized Shigella can escape from autophagy by secreting the IcsB effector , which interferes with the interaction of host Atg5 with bacterial surface protein VirG 41 ., VirG is not only a bacterial factor essential for actin polymerization , but also a molecular target of host autophagy ., Indeed , ΔvirG , a Shigella mutant lacking VirG , did not induce GFP-LC3 accumulation in epithelial cells 41 ., Notably , we found that unlike that observed in epithelial cells , ΔvirG induced a similar level of GFP-LC3 aggregates as wild-type Shigella in BMMs ( Figures 5A , 5B , and 6 ) ., These results indicate that factors other than VirG are involved in autophagy formation in caspase-1-deficient BMMs ., In both wild-type and caspase-1-deficient BMMs , GFP-LC3 accumulation around the phagocytosed Shigella TTSS mutant ( S325 ) and endogenous LC3 conversion were not induced ( Figures 5A , 5B , and 6 ) , suggesting that bacterial escape from the phagosome is required for autophagosome maturation in Shigella-infected BMMs ., We next examined the role of Ipaf and ASC in autophagosome maturation in Shigella-infected BMMs ., Similar to that observed in caspase-1-deficient BMMs , GFP-LC3 accumulation and endogenous LC3-I to LC3-II conversion were enhanced after Shigella infection in Ipaf-deficient BMMs when compared to wild-type cells ( Figures 6 , 7A , and 7B ) ., Because caspase-1 activation induced by Shigella is deficient in caspase-1- and Ipaf-deficient BMMs , these results suggested that caspase-1 activation inhibits the induction of Shigella-induced autophagy ., However , when ASC-deficient BMMs were infected with Shigella , the levels of autophagy associated with intracellular bacteria were similar to those observed in wild-type BMMs ( Figures 6 , 7A , and 6B ) , indicating that autophagosome maturation is not enhanced by ASC deficiency , even though caspase-1 activation is abrogated upon Shigella infection ( Figure 2E ) ., GFP-LC3-associated autophagic vesicles triggered by amino acid starvation and endogenous LC3-I to LC3-II conversion by rapamycin treatment were still induced in ASC-deficient BMMs , suggesting that the autophagic machinery is intact in the absence of ASC ., Together with the results presented in Figure 3 , these results indicate that Ipaf and ASC differentially regulate the induction of autophagy and suggest that autophagy in caspase-1- and Ipaf-deficient BMMs is associated with resistance to pyroptotic cell death ., To begin to examine the functional role of autophagy in Shigella-induced cell death of infected macrophages , we incubated caspase-1- or Ipaf-deficient BMMs with 3-methyladenine ( MA ) , a well known inhibitor of autophagy , after Shigella infection ., Because 3-MA inhibits uptake of bacteria by macrophages 42 , the compound was added after phagocytosis of bacteria by BMMs ( 10 min after infection ) ., As shown in Figure 8A , the addition of 3-MA did not affect pyroptosis induced by Shigella infection in wild-type BMMs ., Also , viability and multiplication of intracellular Shigella were not significantly affected by addition of 3-MA under microscopic observation ( unpublished data ) ., The treatment with 3-MA enhanced the LDH release from caspase-1- and Ipaf-deficient BMMs infected with Shigella ( Figure 8B and 8C ) , suggesting that the inhibition of autophagy promotes Shigella-induced cell death in caspase-1- and Ipaf-deficient BMMs ., In contrast , the addition of 3-MA did not affect LDH release from macrophages infected with the TTSS mutant ( Figure 8D–8F ) , indicating that the cytosolic invasion is required for 3-MA to enhance membrane permeability associated with pyroptosis ., These results suggest that autophagy induced by Shigella infection protect infected macrophages from pyroptosis ., Intracellular pathogenic bacteria trigger immune responses distinct from extracellular bacteria , which are mainly recognized by TLRs ., Recent reports have indicated that cytosolic recognition of flagellin by Salmonella and Legionella mediates caspase-1 activation and IL-1β maturation 11 , 12 , 14 , 25 , 26 ., The host protein Ipaf is required for activation of caspase-1 and IL-1β processing as well as for the inducement of rapid cell death through the sensing of intracellular flagellin during Salmonella and Legionella infection 11 , 12 , 14 ., In this study , we demonstrate that Ipaf and its adaptor protein ASC are required for caspase-1 activation and IL-1β processing in Shigella-infected macrophages , but these processes , unlike in Salmonella , are independent of flagellin ., The results suggest that unknown bacterial factor ( s ) are released from intracytosolic Shigella or secreted via the TTSS and are sensed directly or indirectly by Ipaf to promote caspase-1 activation ., Shigella-induced activation of caspase-1 was previously attributed to IpaB 24 ., Since IpaB is an integral component of the TTSS transmembrane pore complex that is inserted into the host cell membrane , the ipaB mutant is unable to translocate many effector proteins via TTSS ., Thus , it is difficult or impossible to definitively attribute caspase-1 activation to IpaB alone by the use of the Shigella ipaB mutant ., In Salmonella , the IpaB homolog SipB has been suggested to directly interact with caspase-1 and mediate its activation 43 ., However , caspase-1 activation induced by Salmonella depends on the sensing of intracellular flagellin by Ipaf , but not SipB , in that flagellin mutants do not induce caspase-1 activation even though their SipB function is intact 11 , 12 ., Because the TTSS in both Shigella and Salmonella forms a pore in the membrane of infected macrophages , the TTSS apparatus may induce a potassium ion efflux or another activity across the membrane , a signal that has been suggested as activating Nalp3 through the activation of the purigenic P2X7 receptor 6 ., However , Nalp3 plays no function in caspase-1 activation induced by Salmonella infection 6 or Shigella infection ( unpublished data ) ., It was suggested that the recognition of intracellular flagellin by Ipaf is indirect 11 , but the molecular mechanism by which flagellin is sensed by Ipaf is unclear ., Thus , it is possible that both flagellin and the Ipaf-activating factor of Shigella interact with a common signaling machinery , and this host factor ( s ) is sensed by Ipaf to mediate caspase-1 activation ., Further studies are needed to fully understand the molecular mechanism by which intracellular Shigella induces caspase-1 activation through the Ipaf-ASC-caspase-1 inflammasome ., The induction of caspase-1-independent pyroptotic cell death by Shigella infection was induced in Ipaf-deficient BMMs as well as in caspase-1-deficient BMMs 23 ., We initially assumed that this phenotype was due to the lack of caspase-1 activity ., However , ASC-deficient BMMs were not resistant to pyroptosis induced by Shigella despite the absence of caspase-1 activation ., These results indicate that the function of Ipaf and ASC differ in a subtle manner and suggest that pyroptosis can proceed in the absence of caspase-1 activation ., One possibility is that in caspase-1-deficient or Ipaf-deficient macrophages , anti-pyroptotic signals might be induced , leading to transient protection of macrophages from pyroptosis caused by bacterial infection ., In this model , ASC might promote such a survival signal in the absence of caspase-1 or Ipaf ., In certain experimental systems , ASC is known to mediate NF-κB activation 4 , 44 , and thus NF-κB or another activity induced via ASC independently of caspase-1 might provide survival signals to counter the induction of pyroptosis in Shigella-infected macrophages ., We found that the induction of autophagy was facilitated by Shigella infection in the absence of caspase-1- or Ipaf-deficient BMMs but not in ASC-deficient BMMs ., Because autophagy induced by amino acid starvation or by rapamycin treatment was normally induced in the absence of caspase-1 or Ipaf , the results indicate that the autophagic machinery is intact in the mutant cells and that caspase-1 activation inhibits autophagy formation in wild-type BMMs infected with Shigella ., We hypothesized that , in wild-type macrophages , activated caspase-1 may degrade some factors that are essential for the induction of autophagy pathway , and that inhibition of autophagy and consequent rapid pyroptosis may serve to promote efficient induction of host inflammatory responses ., Previous studies suggested that Naip5 , another NLR family member , regulates autophagy in mouse macrophages infected with Legionella pneumophila 40 ., The mechanism by which Naip5 and Ipaf/ASC/caspase-1 regulate autophagy in response to bacterial infection remains poorly understood ., However , it is likely that these NLR proteins act through different mechanisms , as recent studies suggest that Ipaf , but not Naip5 , controls caspase-1 activation 45 ., The connection between caspase activation and autophagy are complex in that both events shared regulatory and mechanistic components 35 ., Our results indicate that 3-MA , an inhibitor of autophagy , enhances cell death , thus raising the possibility that autophagy induction protects macrophages from cell death caused by Shigella infection ., The mechanism by which Shigella or other intracellular bacteria trigger autophagy remains poorly understood ., We have found no role for Shigella VirG in the induction of autophagy , in contrast to that reported in epithelial cells infected with Shigella 41 ., It is likely that components released from intracellular bacteria into the host cytosol activate autophagy in that the Shigella TTSS mutant did not activate autophagy ., Our results raise the possibility that caspase-1 activation and necrotic cell death , the two important activities for induction of pyroptosis after Shigella infection , represent independent phenomena ., Our results also suggest that , at least in part , the delayed cell death observed in caspase-1-deficient BMMs may be a consequence of induction of autophagy ., Further studies are needed to understand the molecular link between caspase-1 activation , pyroptosis , and autophagy , as well as their role in regulating host innate immune responses against intracellular bacteria ., The wild-type S . flexneri 2a YSH6000 strain has been described previously 46 ., Shigella mutants , S325 ( mxiA::Tn5 ) 47 , and a cell-to-cell spreading deficient virG null mutant ( ΔvirG ) 41 , were used in this study ., The fliC mutant ( ΔfliC ) was constructed by allele replacement strategies according to the procedures described previously 48 ., The wild-type S . enterica serovar Typhimurium SR-11 χ3181 and the isogenic fliA::Tn10 were provided by H . Matsui ( Kitasato Institute for Life Science , Tokyo , Japan ) 49 ., The Shigella and Salmonella strains were grown routinely in brain-heart infusion broth ( Becton Dickinson , http://www . bd . com/ ) or Luria-Bertani broth , respectively ., A fliC gene of Shigella was cloned downstream of the ptac promotor of expression vector pTB101-Tp 50 ., FliC expression was driven by adding of 10 μM IPTG in the bacterial culture for 1 h ., 3-MA was purchased from Sigma ( http://www . sigmaaldrich . com/ ) ., Anti-FliC antibody was provided from H . Matsui ( Kitasato Institute for Life Science , Tokyo , Japan ) ., The following antibodies were obtained commercially: rabbit anti-mouse caspase-1 ( Santa Cruz Biotechnology , http://www . scbt . com/ ) , goat anti-mouse IL-1β ( R&D Systems , http://www . rndsystems . com/ ) , and rabbit anti-LC3 antibody ( MBL International , http://www . mblintl . com/ ) ., The bacteria grown to the exponential phase were stabilized by addition of RNAprotect bacteria reagent ( Qiagen , http://www . qiagen . com/ ) ., Total RNA from the cells was prepared by using RNeasy mini kit and RNAse-free DNase ( Qiagen ) according to the protocols of the manufacturer , and converted to cDNA with ReverTra Ace ( Toyobo Life Science , http://www . toyobo . co . jp/ ) as a template for PCR reactions ., The primers for amplification of cDNA fragments were as follows: fliC , forward primer , 5′-CGTATTAACAGCGCGAAGGA-3′ , reverse primer , 5′-AGACAGAACCTGCTGCGGTA-3′; phoA , forward primer , 5′-ATGTCACGGCCGAGACTTATAG-3′ , reverse primer , 5′-GTGAATATCGACGCCCAGCG-3′; ipaB , forward primer , 5′-GCAGCAGTCGTTCTCGTAGC-3′ , reverse primer , 5′-TCAAGCAGTAGTTTGTTGCAAAATTG-3′ ., BMMs were prepared from the femurs and tibias of caspase-1-deficient mice 23 , Ipaf-deficient mice 11 , and ASC-deficient mice 51 by culture for 5 d in 10% FCS-RPMI 1640 supplemented with 30% L-cell supernatant ., All mice are C57BL/6 background or backcrossed with C57BL/6 mice ., Mice were housed in a pathogen-free facility ., Animal studies used protocols approved by the University of Michigan Committee on Use and Care of Animals ( Ann Arbor , Michigan , United States ) , and the Animal Care and Use Committee of the Institute of Medical Science , University of Tokyo ( Tokyo , Japan ) ., BMMs were seeded at 5 × 105 cells in 24-well plates containing 10% FCS-RPMI 1640 ., The cells were infected with Shigella at an MOI of ∼10 per cell ., The plates were centrifuged at 600g for 10 min to synchronize the stage of infection , and gentamicin ( 100 μg/ml ) and kanamycin ( 60 μg/ml ) were added 30 min later ., At the times indicated after infection , the LDH activity of the culture supernatants of infected cells was measured by using a CytoTox 96 assay kit ( Promega , http://www . promega . com/ ) according to the manufacturers protocol ., For immunofluorescence study , the infected cells were fixed and immunostained as described previously 23 , and they were analyzed with a confocal laser-scanning microscope ( LSM510; Carl Zeiss , http://www . zeiss . com/ ) ., BMMs seeded at 1 × 106 cells in 6-well plates were infected with Shigella at an MOI of ∼10 per cell ., Cells were lysed and combined with supernatants precipitated with 10% trichloroacetic acid ., The samples were loaded onto 15% SDS-PAGE , and the cleaved form of caspase-1 and IL-1β were detected with anti-caspase-1 or anti- IL-1β antibody , respectively ., Plat-E cells were transfected with pMX-puro-GFP or pMX-puro-GFP- rat LC3 using FuGENE 6 ( Roche , http://www . roche . com/ ) 41 , 52 ., Two-day cultures of BM cells were transfected with resulting retrovirus and cultured for an additional 3 d ., In our experiments , GFP-transfected cells were 40%–50% and GFP-LC3-transfected cells were 30%–40% after recombinant virus infection , respectively ., GFP and GFP-LC3 expression in BMMs were confirmed by the observation using a confocal laser-scanning microscope with the same threshold level ., For induction of endogenous autophagy in amino acid–starved conditions , the BMMs were incubated with Earles Balanced Salt Solution buffer ( Sigma ) for 2 h ., To score autophagosome formation , a macrophages was defined as positive if it contained >10 donut-like shaped GFP-LC3-labeled structures ., For induction of autophagy by rapamycin treatment , macrophages without retrovirus infection were incubated with rapamycin ( 25 μg/ml; LC Laboratories , http://www . lclabs . com/ ) ., At the indicated time , total cell lysates were prepared and analyzed by western blotting for detecting LC3-I to LC3-II conversion ., Statistical analyses were performed by the Mann–Whitney U test ., Differences were considered significant at p < 0 . 05 . | Introduction, Results, Discussion, Materials and Methods | Shigella infection , the cause of bacillary dysentery , induces caspase-1 activation and cell death in macrophages , but the precise mechanisms of this activation remain poorly understood ., We demonstrate here that caspase-1 activation and IL-1β processing induced by Shigella are mediated through Ipaf , a cytosolic pattern-recognition receptor of the nucleotide-binding oligomerization domain ( NOD ) -like receptor ( NLR ) family , and the adaptor protein apoptosis-associated speck-like protein containing a C-terminal caspase recruitment domain ( ASC ) ., We also show that Ipaf was critical for pyroptosis , a specialized form of caspase-1-dependent cell death induced in macrophages by bacterial infection , whereas ASC was dispensable ., Unlike that observed in Salmonella and Legionella , caspase-1 activation induced by Shigella infection was independent of flagellin ., Notably , infection of macrophages with Shigella induced autophagy , which was dramatically increased by the absence of caspase-1 or Ipaf , but not ASC ., Autophagy induced by Shigella required an intact bacterial type III secretion system but not VirG protein , a bacterial factor required for autophagy in epithelial-infected cells ., Treatment of macrophages with 3-methyladenine , an inhibitor of autophagy , enhanced pyroptosis induced by Shigella infection , suggesting that autophagy protects infected macrophages from pyroptosis ., Thus , Ipaf plays a critical role in caspase-1 activation induced by Shigella independently of flagellin ., Furthermore , the absence of Ipaf or caspase-1 , but not ASC , regulates pyroptosis and the induction of autophagy in Shigella-infected macrophages , providing a novel function for NLR proteins in bacterial–host interactions . | Shigella are bacterial pathogens that are the cause of bacillary dysentery known as shigellosis ., A crucial aspect of the propensity of Shigella to cause diseases lies in its ability to invade the cytoplasm of epithelial cells as well as macrophages ., The bacterial invasion of macrophages induces pyroptosis , the proinflammatory cell death associated with caspase-1 activation ., Activated caspase-1 then cleaves and activates prointerleukin ( proIL ) -1β and proIL-18 , which are proinflammatory cytokines involved in host inflammatory responses ., However , the precise mechanisms of caspase-1 activation induced by Shigella infection remain poorly understood ., Ipaf , a cytosolic pattern-recognition receptor of the nucleotide-binding oligomerization domain ( NOD ) -like receptor ( NLR ) family , is a crucial host factor that activates caspase-1 through the sensing of flagellin produced by some bacteria , such as Salmonella or Legionella ., We discovered that Ipaf and the adaptor protein ASC are required for caspase-1 activation induced by non-flagellated Shigella infection ., Thus , Ipaf and ASC mediate caspase-1 activation by sensing an unknown bacterial factor , but not flagellin ., Autophagy , a cellular system for eliminating intracellular pathogens , was dramatically enhanced in Shigella-infected macrophages by the absence of caspase-1 or Ipaf , but not ASC ., The inhibition of autophagy promoted Shigella-induced cell death , suggesting that autophagy protects infected macrophages from pyroptosis ., This study provides evidence that in Shigella-infected macrophages , autophagy is inhibited by Ipaf and caspase-1 , but positively regulated by ASC , providing a novel function for NLR proteins in bacterial–host interactions . | infectious diseases, cell biology, mammals, immunology, mus (mouse), eubacteria | null |
journal.pcbi.1002319 | 2,011 | Recovering Protein-Protein and Domain-Domain Interactions from Aggregation of IP-MS Proteomics of Coregulator Complexes | CoRegs are members of multi-protein complexes transiently assembled for regulation of gene expression 1 ., Assembly of these complexes is affected by ligands that bind to nuclear receptors ( NRs ) , such as steroids , retinoids , and glucocorticoids 2–5 ., CoRegs complexes exist in many combinations that are determined by post-translational modifications ( PTMs ) and presence of accessory proteins 6 , 7 ., To date , over 300 CoRegs have been characterized in mammalian cells 8 and it has been shown that CoRegs complexes control a multitude of cellular processes , including metabolism , cell growth , homeostasis and stress responses 6 , 9 , 10 ., Many CoRegs complexes are considered master regulators of cell differentiation during embryonic and post-developmental stages 10 , 11 , and evidence suggests that malfunction of these proteins can lead to the pathogenesis of endocrine-related cancers 3 , 12 and diabetes 13 ., Importantly , it is believed that development of better chemical modulators of CoRegs will lead to a ‘new generation’ of drugs with higher efficacy and selectivity 14 , 15 ., To accelerate research in the area of CoRegs signaling , the Nuclear Receptor Signaling Atlas ( NURSA ) 16 have been applying systematic proteomic and genomic profiling related to CoRegs 17 , 18 ., Recently , the NURSA consortium released a massive high-throughput ( HT ) IP/MS study reporting results from 3 , 290 related sets of proteomics pull-down experiments 19 ., The results from these experiments are protein identifications with semi-quantitative spectral count measurements , which can be used to approximate protein enrichment in individual IPs ., Multiple IP experiments that sample different protein complex subunits can be integrated to gain a global picture of protein complex composition 20–22 ., Several prior studies applied to human cells have proposed strategies to reconstruct protein complexes by combining results from HT-IP/MS 23–28 ., Some of the results from such studies have been processed by algorithms that probabilistically predict binary protein-protein interactions ( PPIs ) ., In some cases , such predictions were validated using known PPIs from the literature , where in few cases predicted interactions were further validated experimentally ., For example , Washburn and colleagues implemented the multidimensional protein identification technology ( MudPIT ) method to pull down complexes using 27 bait proteins from the Mediator complex to suggest 557 probabilistic interactions between the baits and their pulled preys 23 ., They used the Jaccard distance to integrate protein co-occurrence in the different experiments , and compared their ‘high-confidence’ interactions with those listed in a literature-based database , the human protein reference database ( HPRD ) 29 ., Experimentally , the study validated few predicted interactions using co-IP and western blots ., In a follow up study , different clustering approaches to extract sub-complexes from related affinity purification ( AP ) -MS experiments using three distance measures: Manhattan , Euclidian , and Correlation Coefficient for clustering are described 30 ., The aforementioned work , and other similar prior studies , ranked predicted associations and provided probabilities for interactions between baits and preys , building on the explicit nature of bait-prey relationship in epitope-based purifications ., However , due to secondary cross-reacting proteins , bait-prey relationships are rarely explicit in IPs carried out with primary antibodies ., Hence , here we developed and compared different ways , coded into mathematical functions , to score prey-prey interactions from a large , recently published , HT-IP/MS dataset ., The equations predict direct protein-protein interactions between prey proteins without considering the specific baits ., We also used the same equations to predict domain-domain interactions underlying the protein-protein interactions ., We evaluated the performance of these equations using known protein-protein and domain-domain interactions from the literature and validated one protein-protein interaction experimentally , and one domain-domain interaction using computational docking ., By combining the data from the 3 , 290 IP-MS experiments collected by NURSA we predicted binary interactions between prey proteins and their domains ., We offer a global view of CoRegs complexes in human cells , and provide the predicted networks for exploration on the web through a web-based application with downloadable tables freely available at http://maayanlab . net/HT-IP-MS-2-PPI-DDI/ ., A detailed description of the IP-MS procedure can be found in references 19 , 26 and the list of experiments in Dataset S1 ., The data we analyzed is provided as supporting material tables for reference 19 ., These supporting tables contain GeneIDs for identified protein products , as well as the spectral count ( SPC ) measurements , and ‘abundance’ values , defined as SPCs/MW , where MW is the molecular weight for the largest isoform of the gene product ., The latter normalization approximately accounts for the number of peptides expected from a protein ., Abundance is logically similar to the normalized spectral abundance factor ( NSAF ) scores previously proposed 30 , except the values are not scaled per experiment ., To score prey-prey interactions from the HT-IP/MS data table , containing the ranks of proteins from the 3 , 290 IP-MS experiments , we evaluated existing and developed new equations implemented as algorithms in MATLAB and Java ., Sørensen similarity coefficient ( Sor ) provides a symmetric similarity coefficient for comparing two finite sets ., The coefficient ranges between 0 and 1 , where 0 denotes no similarity , and 1 denotes identical sets ., The Sørensen coefficient is calculated as the ratio of the cardinality of shared members between two sets and the sum of the cardinalities of the same sets ., ( 1 ) The Sørenson coefficient was applied to determine the likelihood that proteins A and B directly interact ., MA and MB are the sets of all experiments that reported either protein A , B or both as present in the lists of pulled prey proteins ., MA , B are lists where both A and B are present ., Pearsons Correlation coefficient ( Pr ) characterizes the linear dependency of two variables ., Here we used the Pearsons Correlation coefficient to quantify the correlation the SPC scores of two proteins across all IP/MS experiments ., ( 2 ) ρA , B is the Pearsons Correlation coefficient between proteins A and B where Q denotes the reported ‘abundance’ which is SPC/MW ( MW , molecular weight ) ., and are the column vectors of Q at indices and ., is the covariance and and are the standard deviations of and ., Equation 3 ( E3 ) was developed through an intuitive manual symbolic search for functions that perform well , based on benchmarking , using known protein-protein interactions ., E3 calculates a ratio between the sum of the SPC scores in experiment j ( ) and the difference between the ranks of protein pairs based on their SPC scores in the same experiment ., The average E3 scores across all experiments is the final score that is used to quantify the likelihood that two prey proteins interact ., The rationale behind the E3 equation is to reward pairs of proteins that have similar SPC scores and similar ranks across all experiments , rewarding pairs of proteins with high SPC scores that appear in the same complexes ., ( 3 ) The AB correlation was also developed through an intuitive manual symbolic search for functions that perform well based on benchmarking using known protein-protein interactions ., The AB correlation computes the mean of the product of SPC scores normalized by dividing by the sum of mean SPC scores across all experiments ., ( 4 ) The AB method also rewards pairs of proteins that have higher SPC scores in the same subset of experiments ., To evaluate the predicted prey-prey protein interactions using the four equations , we used an updated version of the human literature-based protein-protein interactome we developed for the program Genes2Networks 31 ., The PPIs are from 12 databases: HPRD 29 , MINT 32 , DIP 33 , MIPS 34 , PDZBase 35 , PPID 36 , BIND 37 , Reactome 38 , BioGRID 39 , SNAVI 40 , Stelzl et al . 41 , and Vidal and co-workers 42 ., These databases contain direct physical interactions for mouse , rat , and human proteins containing 11 , 438 proteins connected through 84 , 047 interactions extracted manually from publications ., We converted all IDs to human IDs using homologene ( http://www . ncbi . nlm . nih . gov/homologene ) ., To identify domains for proteins , we used the Pfam domain database release 24 . 0 ., The file ‘Pfam-A . full . gz’ was downloaded from: ftp://ftp . sanger . ac . uk/pub/databases/Pfam/releases/Pfam24 . 0/on November 1st 2010 ., Domain-domain interactions ( DDI ) were obtained from the Domine database 43 ., The Domine database contains 26 , 219 domain-domain interactions ., Among these domain-domain interactions , 6 , 634 were inferred from the protein data bank ( PDB ) and 21 , 620 were computationally predicted by one or more of 13 prediction methods ., In order to score domain-domain interactions , we developed a prediction vector containing a combined score for all predicted PPIs that contain domain-pairs at each side of a scored PPI ., We assigned the score of the predicted PPI to the DDI score ., Antibodies for STRN , also called Striatin , are polyclonal rabbit , and were purchased from Millipore Corp ., Antibodies for CTTNBP2NL were purchased from GeneTex ., MCF-7 cells were lysed in immunopreciptation buffer containing Hepes ( 50 mM , pH 7 . 4 ) , NaCl ( 150 mM ) , EDTA ( 1 mM ) , Tween-20 ( 0 . 1% ) , glycerol ( 10% ) and protease inhibitors ., The lysates were pre-cleared in the presence of rabbit IgG and protein A beads ., The input sample was collected after pre-clearing ., Samples were rotated overnight with IgG or Striatin antibody and subsequently incubated for two hours with Protein-A beads ., The washed protein-containing beads were denatured and analyzed by Western blot ., The MolSoft ICM software was used to perform the domain-domain docking simulation ., ICM uses a two-step method: pseudo-Brownian rigid-body docking followed by biased probability Monte Carlo minimization of the ligand side-chains , to sample conformational space in order to identify the global energy minimum for a given interaction 44 ., For this specific simulation , the protein PPP2R1A ( PDB ID: 1B3U ) , the receptor , was kept rigid , while conformations of the ligand STK25 ( PDB ID: 2XIK ) were sampled around the receptor and corresponding docking scores were retrieved ., Domains were then examined for interactions based on these scores ., We analyzed the experimental data from 3 , 290 IP-MS experiments targeting 1 , 083 antigens ( bait proteins ) using 1 , 796 different antibodies ., These experiments detected 11 , 485 non-redundant proteins ( Dataset S1 ) ., Some of the baits were pulled-down with several different antibodies ., Some of the experiments with the same baits and antibodies were repeated several times but conducted under different conditions , i . e . , stimulated/un-stimulated cells , or different cell types ., Complexes are mostly isolated from nuclear fractions but some experiments use cytosolic fractions ., Summary of the experimental conditions , cell types , antibodies and baits used , counts of normalized peptides identified in each experiment per protein , and size of the lists of proteins identified in each experiment can be directly obtained from the primary publication provided as reference 19 ., IP-MS proteomics profiling have several known experimental challenges that need to be considered when applying functional global analyses on such data ., First , it is well established that the proteins identified in such experiments are enriched for highly abundant and “sticky” proteins ., This results in numerous proteins appearing frequently in almost all pull-downs regardless of the cell type , cellular fraction or experimental conditions ., To address this we used a list of “non-specific” proteins to filter protein identifications that appear frequently in many pull-downs ( Dataset S1 ) ., For all further analyses we removed these proteins from the results ., Such a “non-specific” protein list can be useful as a guideline for filtering other IP-MS proteomics data applied to human cells ., However , it should be noted that the concept of filtering IP-MS proteomics data based on a “non-specific” list is only meant as a guide ., The sticky non-relevant proteins may play an important biological role that would be missed by removing them ., In general , proteins that appear in the list are enriched in heat shock , ribosomal , and heterogeneous nuclear ribonucleoproteins ( hnRNPs ) ., Also , the majority of proteins on the non-specific list were selected based on the purifications from nuclear extracts , so some abundant cytosolic proteins may be over represented in the protein-protein and domain-domain interaction predictions since these may not have been removed ., In order to integrate and visualize the results from the 3 , 290 IP-MS experiments , we first used the Jaccard Distance ( JD ) to construct a CoRegs complex similarity graph were nodes represent protein lists from each experiment and links represent overlap between experiments ( Fig . S1 ) ., Nodes and links are preserved in the network if the similarity is greater than the Jaccard distance of 0 . 7 ., This retained 491 experiments and 2233 links between them , which are a small portion of all possible experiments and their similarities ( Fig . S2A ) ., On average , pull-down experiments reported the identification of ∼30–200 proteins but the distribution has a heavy tail with few experiments identifying over 1000 proteins ( Fig . S2B ) ., Our aim in this study is to assign confidence scores to binary prey-prey protein-protein and domain-domain interactions by integrating information from the 3 , 290 IP-MS experiments ., The rationale for this approach is that the experiments , reporting lists of ∼30–200 proteins for each pull-down , taken together , provide enough information to reconstruct high-fidelity , small-sized complexes and potentially enough to recover direct physical interactions between pairs of proteins and domains ., We reasoned that if we use all the information across all experiments to score each pair of proteins for potential direct interaction , we will be able to identify novel associations in addition to recovering known interactions better than by chance ., In contrast with most prior methods that focused on scoring bait-prey interactions , our equations predict interactions between prey proteins that commonly reappear together in different pull-downs ., Although the data collected for this study was aimed at the recovery of interactions between the intended antigens ( baits ) and other proteins , the majority of primary antibodies cross-react with multiple secondary antigens and those antigens interact with other proteins ., This complicates bait-prey scoring of HT-IP/MS data ., Yet , logically , if two proteins reappear together at the top of lists in many different pull-downs , we can guess that they may physically interact regardless of which baits were used to pull them down , making it possible to predict likely binary interactions by utilizing the spectral counts , not just co-occurrence ., To encode such logic into mathematical functions we devised four scoring schemes , each attempting to address the problem in a slightly different way ., To evaluate the performance of the four scoring schemes we used known PPIs we consolidated from online databases 31 ., The overall schema for this approach is depicted in Fig . 1 ., To compare the performance of the different scoring methods we visualized the results as either receiver operator curve ( ROC ) ( Fig . S3 ) , random walks ( Fig . S4 ) , or a sliding window ( Fig . S5 ) ., Visualization of overlap between a ranked list and a gene set using a random walk was borrowed from the popular Gene-Set Enrichment Analysis method 45 ., The three equations AB , E3 , and Pr can be combined with the Sørenson coefficient to slightly improve the predictions by the AB and E3 equations , and significantly improve the predictions made with the Pr equation ., AB and E3 perform best when combined with the Sørenson coefficient because these equations take into account the quantitative levels of the peptides , rewarding interactions that appear on top of the same pull-downs and penalizing potential interactions where the two proteins are not present in the same pull-down , or when one protein appears at the top and the other at the bottom ., The different methods recover different sets of interactions and in some cases complement each other , suggesting perhaps that a combined weighted score may provide better results than using a single equation ( Fig . S6 , Dataset S2 ) ., Next , we used ball-and-stick diagrams to visualize the results across all experiments ., We first visualized all overlapping interactions listed in the top 10% of predicted protein-protein interactions by each method ( AB , E3 and Pr combined with Sor ) ., This resulted in a network made of 2 , 509 proteins ( nodes ) and 28 , 886 interactions ( edges ) ( Fig . 2 ) ., Using Cytoscapes organic visualization algorithm , the hubs of this network self-organize into an interesting hierarchical structure that may reflect their complex formation relationship ., This network provides a global view of the CoRegs interactome , allowing zoom-in to view the identity of high confidence predicted protein-protein interactions and the complexes that these interactions form ., To accomplish this zoom-in view , we increased the threshold to only include interactions from the top 1% of predicted interactions by all three scoring methods and include only three-node cliques ., Three-node cliques are triangles in the network topology where three proteins are connected to each other with a maximum of three links ., The resultant network contains 543 proteins and 1 , 893 interactions organized into 63 tightly connected protein complexes containing 3 to 25 proteins ( Fig . 3 ) ., Many of the interactions and complexes that emerged are already known from low-throughput protein-protein interactions studies ., However , some of the complexes within this network and many of the predicted protein interactions are novel ., As a proof of concept , we focused on one predicted complex where most of the members of the complex were exclusively prey proteins in all experiments , and most interactions in the complex were not previously known ( Fig . 4A ) ., The complex contains ten densely connected proteins with the protein STRN in the center , predicted to interact with all other nine members ., STRN , STRN3 and STRN4 are scaffolding proteins with a calmodulin binding domain ., Interestingly CTTNBP2NL has been previously reported with STRN and STRN3 in another IP/MS study 46 ., To experimentally validate one of the interactions within this complex we used IP and western blotting to demonstrate a direct interaction between STRN and CTTNBP2NL which is another member of the predicted complex ( Fig . 4B ) ., We chose this interaction based on antibody availability ., Our experiment clearly shows that the two proteins interact ., Such a demonstration of physical interaction experimentally does not prove that our prediction method works well , but it demonstrates how predicted interactions can be further validated experimentally ., To prove that the predictions are of high quality , many such experiments need to be performed with appropriate controls to show statistically that the combined equations can predict , with high fidelity , physical interactions ., Before analyzing all of the 3 , 290 IP-MS experiments published by Malovannaya et al 19 , we had access to a subset of the data before it was published ., Therefore , we developed our analysis methods on a subset of 114 IP-MS experiments that are a fraction of the entire set of the 3 , 290 IP-MS experiments ., In order to integrate and visualize the results from these 114 IP-MS experiments , similarly to the network shown in Fig . S1 , we created the Jaccard Distance ( JD ) CoRegs complex similarity graph ( Fig . S7 ) ., Most of these initial 114 experiments used Estrogen Receptor α ( ESR1 ) and nuclear receptor co-activator 3 ( NCOA3 ) , also called SRC3 , as baits in different cellular conditions ., Both proteins play an important role in breast cancer , where SRC3 serves as the main co-activator of estradiol-dependent ESR1 47 , 48 ., The experiments that used ESR1 and NCOA3 as baits resulted in similar protein lists ( clusters in the subnetwork in Fig . S7 ) compared with the other pull-downs ., Using the same prediction combined scores with the three equations , with lower thresholds , we identified five distinct high confidence complexes we named: SMARC , CSTF , RCOR , MBD , and SIN3A ( Fig . S8 ) ., These five complexes have been previously reported in the Corum database 49 and some have been functionally characterized ( Fig . S9 ) ., Specifically , the SMARC complex highly overlaps with complex IDs 238 , 714 , 803 , and 806 in Corum , a database of reported protein complexes 49 ., The CSTF complex is listed as complex number 1147 in Corum , RCOR is listed as 626 , and MBD and SIN3A have associated IDs with highly overlapping entries for complexes in Corum ., The SMARC and CSTF complexes were recovered mostly from ESR1 pull-down experiments , while the other three complexes are formed by combinations of many other types of baits ., Notably , the SMARC and CSTF complexes are nearly mutually exclusive to two different antibodies targeting ESR1 , and are recovered in the control experiment from HeLa cells that do not express ESR1 ., Thus , one antibody is likely cross-reacting with a member of the SMARC complex , whereas the other antibody cross-reacts with a member of the CSTF complex ( Fig . S10 ) ., This result highlights the importance of protein complex reconstruction from HT-IP/MS based on prey-prey co-occurrence alone , independently of the intended baits ., Since PPIs are often the result of interactions between the structural domains of the interacting proteins , and since we know most of those domains for all pulled prey proteins based on their amino-acid sequences , we can use the scores for PPIs to also score and rank domain-domain interactions ( DDIs ) ., The scoring of domain interactions is slightly more complex since most proteins have several different domains and the domains can appear more than once within the same protein ., To resolve this we used the score for PPIs containing domains between all possible domain pairs from each side of the PPI and normalized the score across all the domains ( see methods ) ., The aggregated score for all DDIs was accumulated across and within all 3 , 290 IP-MS experiments ., The idea of predicting DDIs from PPIs is not new 50–52 ., DDIs can also be predicted using structural biology methods or by evolutionary conservation of sequences across organisms 53 ., To evaluate which PPI scoring method works best to predict DDIs , we compared the predicted scores for DDIs with reported DDIs from the Domine database ., The Domine database contains both structurally observed and computationally predicted DDIs 43 ., ROC curves and random-walk plots were used to evaluate DDI predictions , similarly to how we evaluated the PPI prediction methods ( Fig . S11 and S12 , Dataset S3 ) ., The plots show that we can reliably recover known and predicted DDIs ., In addition to the four equations used to score PPIs , we introduced another scoring scheme , λ , for scoring DDIs ., λ is an index that counts the number of times two predicted interacting prey proteins have a domain on each side of the PPI ., Such an index improves DDI predictions ., In addition to the type of analysis we did for PPIs , we also attempted to further combine different prediction methods to optimize DDI predictions ., Finally we visualize our predicted DDIs with known DDIs as a network diagram to visually explore interactions among all domains ( Fig . S13 ) and within the STRN centered complex identified by the PPIs predictions ( Fig . 5A ) ., To further validate one of the predicted DDIs we pursued a computational structural biology approach ., We attempted to dock the PKinase domain of STK25 to the HEAT domain of PPP2R1A ., We chose these two proteins because they had a crystal structure in PDB ., Although the DDI is already listed in Domine , the prediction of this DDI interaction is based on sequence and homology ., Hence there is no direct evidence of such interaction between these two proteins and their domains ., Using the Molsoft ICM software we obtained a docking score of −46 . 75 kcal/mol ., This score is considered high and as such confirms the interaction ., By examining the confirmation of this interaction it appears that the Pkinase domain of the STK25 protein binds to the HEAT domain of PPP21RA ., The energy gap of approximately 2 kcal/mol ( ICM score units ) between the best obtained and next consecutive docking score clearly suggests strong recognition of the HEAT domain by the Pkinase domain ( Fig . 5B–D ) ., In this study we combined results from 3 , 290 experiments that identified nuclear protein complexes in human cells using IP-MS ., We implemented and evaluated four different equations assessing their ability to predict direct physical PPIs from the aggregated proteomics data using known PPIs from the literature ., The highest scoring predictions were visualized as networks with many densely connected clusters that are likely made of real protein complexes ., The prediction scores for potential interactions could be considered as surrogates to real affinity constants ., However , since we do not know the exact quantities of proteins , it is not possible to compute exact dissociation constants ., Such binding constants can be useful for dynamical simulations where we could stochastically trace the transient dynamics of CoRegs complex formation in-silico ., Scoring PPIs by only using the prey measurements may become more robust as more IP-MS experiments are published ., However , careful attention should be given to weighting the repetitiveness of experiments so interactions from similar pull-downs , if repeated , are not mistakenly given higher scores ., Regardless of possible limitations , the ability to recover direct PPIs based on such a massive dataset is an important step toward utilizing HT/IP-MS datasets for reconstructing networks and generating hypotheses ., In addition , we show that the equations can be extended to predict interactions between structural domains ., We also demonstrated two ways to further validate predicted PPIs and DDIs , using experimental and computational approaches ., In summary , our analyses explored new methodologies for scoring PPIs and DDIs using data from related IP-MS experiments , providing many hypotheses about mammalian CoRegs complexes formation , and allowing users to explore novel complexes , PPIs and DDIs online at http://maayanlab . net/HT-IP-MS-2-PPI-DDI/ ., This resource can help us advance the catalogue of transcriptional regulation by CoRegs in normal and diseased mammalian cells . | Introduction, Methods, Results, Discussion | Coregulator proteins ( CoRegs ) are part of multi-protein complexes that transiently assemble with transcription factors and chromatin modifiers to regulate gene expression ., In this study we analyzed data from 3 , 290 immuno-precipitations ( IP ) followed by mass spectrometry ( MS ) applied to human cell lines aimed at identifying CoRegs complexes ., Using the semi-quantitative spectral counts , we scored binary protein-protein and domain-domain associations with several equations ., Unlike previous applications , our methods scored prey-prey protein-protein interactions regardless of the baits used ., We also predicted domain-domain interactions underlying predicted protein-protein interactions ., The quality of predicted protein-protein and domain-domain interactions was evaluated using known binary interactions from the literature , whereas one protein-protein interaction , between STRN and CTTNBP2NL , was validated experimentally; and one domain-domain interaction , between the HEAT domain of PPP2R1A and the Pkinase domain of STK25 , was validated using molecular docking simulations ., The scoring schemes presented here recovered known , and predicted many new , complexes , protein-protein , and domain-domain interactions ., The networks that resulted from the predictions are provided as a web-based interactive application at http://maayanlab . net/HT-IP-MS-2-PPI-DDI/ . | In response to various extracellular stimuli , protein complexes are transiently assembled within the nucleus of cells to regulate gene transcription in a context dependent manner ., Here we analyzed data from 3 , 290 proteomics experiments that used as bait different member proteins from regulatory complexes with different antibodies ., Such proteomics experiments attempt to characterize complex membership for other proteins that associate with bait proteins ., However , the experiments are noisy and aggregation of the data from many pull-down experiments is computationally challenging ., To this end we developed and evaluated several equations that score pair-wise interactions based on co-occurrence in different but related pull-down experiments ., We compared and evaluated the scoring methods and combined them to recover known , and discover new , complexes and protein-protein interactions ., We also applied the same equations to predict domain-domain interactions that might underlie the protein interactions and complex formation ., As a proof of concept , we experimentally validated one predicted protein-protein interaction and one predicted domain-domain interaction using different methods ., Such rich information about binary interactions between proteins and domains should advance our knowledge of transcriptional regulation by CoRegs in normal and diseased human cells . | systems biology, spectrometric identification of proteins, biochemistry, mathematics, protein interactions, protein abundance, statistics, regulatory networks, biology, computational biology, signaling networks, proteomics, biostatistics | null |
journal.pntd.0007536 | 2,019 | Elucidating diversity in the class composition of the minicircle hypervariable region of Trypanosoma cruzi: New perspectives on typing and kDNA inheritance | The protozoan parasite Trypanosoma cruzi ( Kinetoplastea: Trypanosomatidae ) is the causative agent of Chagas disease ., This parasite infects millions of people throughout its distribution in Latin America ., Chagas disease can display a broad pathological spectrum , including potentially fatal cardiological and gastrointestinal dysfunctions 1 ., T . cruzi is a monophyletic taxon showing a remarkable genetic heterogeneity , with at least six phylogenetic lineages formally recognised as Discrete Typing Units ( DTUs ) , TcI–TcVI 2 , 3; and a seventh lineage , named TcBat 4–6 ., The genetic diversity of T . cruzi was firstly revealed by Multilocus Enzyme Electrophoresis 7 , 8 and posteriorly by very diverse techniques including Multilocus Sequence Typing ( MLST ) 9–12 , microsatellite typing ( MLMT ) 13–18 , target-specific PCR 19–21 , PCR-RFLP 22 , 23 , PCR-DNA blotting with hybridization assays 24–26 , and recently by amplicon deep sequencing 27 , 28 ., The different approaches have their own advantages and disadvantages and bring out the genetic diversity of T . cruzi at different levels ., Approaches that allow direct typing from biological samples ( blood , tissues , etc . ) , avoiding parasite culture , are more suitable for clinical and epidemiological studies ., However , nowadays there is no methodologic approach with a good sensibility , specificity and reproducibility for direct typing on biological samples ., Because there is usually a low number of parasites in infected tissues or blood samples , genetic markers with high number of copies are required to achieve good sensitivity of detection 29 ., In this regard , T . cruzi , as all the kinetoplastids , has a unique and large mitochondrion which contains a complex network of DNA , the kinetoplastic DNA ( kDNA ) ., The kDNA represents approximately 20–25% of the total cellular DNA in T . cruzi and consists of two kind of circular DNA molecules: maxicircles and minicircles ., Maxicircles contain mitochondrial genes characteristic of other eukaryotes 30 ., Minicircles are present in tens of thousands of copies 31 ., Each of them is organized into four highly conserved regions located 90° apart each other , and an equal number of hypervariable regions ( mHVRs ) interspersed between the conserved regions 32 ., The highly conserved regions of minicircles have been widely used as targets for molecular detection of T . cruzi DNA ., The used primers show a good sensitivity and specificity 29 and amplify a region of about 330 bp that totally include the mHVRs present between conserved regions ., This amplified region has been used in hybridization assays ( mHVR probes ) and DTU-specific hybridization was observed only between isolates belonging to the same genetic group 25 , 26 , 33–35 ., This specificity observed in hybridization assays suggests the presence of DTU specific sequences and even genotype-specific sequences ( i . e . sequences showing specificity at intra-DTU level ) ., However , technical limitations that existed until a few years ago for sequencing these highly variable kDNA regions , prevented the identification of the sequences in which the specificity relies ., Some attempts were made by cloning and sequencing some mHVRs 36 , 37 but the limited number of studied sequences were not enough to obtain a complete picture of the genetic diversity of these sequences ., Thus , the observed hybridization patterns between mHVRs continue being a black box system and the sequence diversity of T . cruzi mHVRs virtually unknown ., Beyond the potential utility for strain typing , studying mHVR diversity is also interesting because these sequences are involved in functions that are only known in kinetoplastids and in no other eukaryotic organism ., mHVRs code for short RNAs called guide RNAs ( gRNAs ) ., gRNAs are involved on edition of several mitochondrially-coded mRNAs ., This edition varies from addition of some Us to building almost the full open reading frame of the mRNA 38 , 39 ., In this sense , gRNAs can be inferred from sequences of the mitochondrial mRNAs and diversity on edition among strains can be addressed 40 ., In addition , studying mHVR diversity can shed light on how such sequences evolve and how they are inherited ., Here , we propose an amplicon deep sequencing approach that allows an accurate knowledge of the sequence diversity of the hypervariable region of kDNA minicircles of T . cruzi and opens the possibility of functional and evolutionary studies ., This approach can be also used as a typing method for hundreds of samples at time ., DNA from nine cloned T . cruzi strains belonging to the six main DTUs was examined in this study ( Table 1 ) ., All the strains were typified by using an optimized Multilocus Sequence Typing scheme based on four gene fragments ( HMCOAR , GPI , TcMPX and RHO1 ) according to Diosque et al . 7 , in order to confirm DTU for each strain ., In order to amplify the minicircles hypervariable region , kDNA specific primers 121 ( 5’-ACACTCTTTCCCTACACGACGCTCTTCCGATCTAAATAATGTACGGG ( T/G ) GAGATGCATGA-3’ ) and 122 ( 5’-GTGACTGGAGTTCAGACGTGTGCTCTTCCGATCTGGTTCGATTGGGGTTGGTGTAATATA-3’ ) were modified by adding an oligo adapter to be used in an Illumina platform ., The mHVR libraries were generated by a one-step PCR performed in 5 μl reaction volumes containing 5 ng of DNA , 250 nM of each primer , 2 μM of barcode primers , 5 U of Fast Start High Fidelity Enzyme Blend ( Roche ) , 0 . 50 μl of 10X buffer ( supplied with the Fast Start High Fidelity Enzyme Blend ) , 25 nM of MgCl2 ( Roche ) , 0 . 25 μl of DMSO ( Roche ) , 10 mM of PCR grade nucleotide mix ( Roche ) ., The PCR reaction was carried out on a Veriti Thermal Cycler ( Life Technologies ) and ran as follow: an initial denaturation step ( 10 min at 95°C ) , 10 cycles ( 95°C for 15 seconds , 60°C 30 seconds , 72°C 1 min ) , 2 cycles ( 95°C for 15 seconds , 80°C 30 seconds , 60°C 30 seconds , 72°C 1 min ) , 8 cycles ( 95°C for 15 seconds , 60°C 30 seconds , 72°C 1 min ) , 2 cycles ( 95°C for 15 seconds , 80°C 30 seconds , 60°C 30 seconds , 72°C 1 min ) , 8 cycles ( 95°C for 15 seconds , 60°C 30 seconds , 72°C 1 min ) and 5 cycles ( 95°C for 15 seconds , 80°C 30 seconds , 60°C 30 seconds , 72°C 1 min ) ., Amplicons were then purified using the magnetic beads Agencourt AMPure XP-PCR Purification ( Beckman Genomics , USA ) ., The concentration of the purified amplicons was controlled using Qubit Fluorometer 2 . 0 ( Invitrogen , USA ) ., All libraries were validated using the Fragment Analyzer system ( Advanced Analytical Technologies , USA ) ., The average size of the mHVR amplicons was ~480bp ., All samples were then pooled and prepared according to the manufacturers recommendations ( Illumina Protocols: Sequencing Library Preparation ) and sequenced on an Illumina MiSeq using a 500 cycle v2 kit ( Illumina , San Diego , USA ) to produce amplicons of approximately ~480 bp in length ( 250 bp paired-end reads ) ., The raw data set has been deposited in the NCBI SRA database ( BioProject ID: PRJNA514922 ) ., A total of 22 , 092 , 382 paired reads were obtained by amplicon sequencing of the mHVR from nine strains belonging to six DTUs ., A total of 14 , 766 , 753 sequences were retained ( an average of ≈1 . 4 million of sequences per strain ) after trimming low quality ends , merging paired reads ( forward and reverse ) , elimination of chimeric reads and filtering by base quality ( S1 Table ) ., Surviving sequences were clustered according to different identity thresholds ( 85% , 90% and 95% ) ( Table 2 , S2 and S3 Tables ) ., The number of mHVR clusters for each strain was very similar using different thresholds ( with differences less than 10% in all comparisons between 85% and 95% thresholds ) ., However , clustering at 85% threshold returned few more mHVR clusters than clustering at 90% and 95% identity ( See Table 2 , S2 and S3 Tables ) ., In addition , most clusters were highly divergent among them ( S1 Fig ) ., At any threshold , the number of mHVR clusters was variable among strains and DTUs ( Table 2 ) , ranging from 71 ( Mncl2 –TcV ) to 373 ( X109/2 –TcIII ) clusters ., Additionally , strong intra-DTU variations in the number of clusters were observed in strains of TcI and TcII ( Table 2 ) ., Finally , rarefactions of each dataset discarded that these differences among strains are the effect of different sequencing depths ( Table 2 , S2 and S3 Tables ) ., Strains belonging to TcIV , TcV and TcVI showed some dominant clusters containing a high proportion of reads ( i . e . the cluster size ) ( Fig 1 ) ., The sum of the six most abundant clusters in TcIV , TcV and TcVI represent in all cases more than 50% of the clustered sequences ( 80 . 9% and 69 . 1% in the TcV strains LL014R1 and MNcl2 , respectively; 58 . 7% in TcIV strain CANIIIcl1; and 52 . 5% in the TcVI strain LL015P68R0cl4 ) ., Even more , in LL014R1 and MNcl2 ( TcV strains ) the most abundant cluster represented the 29 . 7% and 17 . 8% of the total mHVR , respectively ., Instead , none of the clusters present in TcI , TcII and TcIII strains represented more than 5 . 2% ., This higher diversity in TcI , TcII and TcIII is also revealed by a higher Simpson diversity index than other DTUs ( Table 2 ) ., Moreover , intra-DTU differences in mHVR cluster diversity were observed in TcII ., Particularly , Tu18cl93 had relatively less cluster diversity than Esmeraldo ( Table 2 and Fig 1 ) ., As expected , shared mHVR clusters were mostly observed in strains belonging to the same DTU ., However , the percentage of shared clusters was highly variable depending on DTU ., TcV strains ( LL014R1 and MNcl2 ) showed the higher proportion of shared clusters ( 97 . 3%; 72/74 ) ., However , we observed strong differences in the cluster sizes ( Fig 2C ) although a positive correlation was detected ( correlation coefficient , r = 0 . 75 ) and some shared clusters were highly abundant in both strains ( Fig 2C ) ., TcI strains ( PalDa20cl3 and TEV55cl1 ) shared 17 . 5% ( 83/475 ) , and TcII strains ( Tu18cl93 and Esmeraldo ) shared 7 . 1% ( 33/466 ) ., Conversely , when we look for shared mHVR clusters between strains belonging to different DTUs , we detected none or few shared clusters ( Fig 2D–2I and S2 Fig ) ., The Bray-Curtis dissimilarity between strains was calculated using mHVR clusters conformed at the different identity thresholds ( 85% , 90% and 95% ) ., Such dissimilarities were used to analyze principal coordinates ( PCoA ) and to build UPGMA trees ( Fig 3 and S3 Fig ) ., Strains from the same DTU clustered together ( Fig, 3 ) despite the high dissimilarities between strains belonging to the same DTU ( Fig 3C ) ., These high dissimilarities between strains belonging to the same DTU determine that the three first axis in the PCoA explain just 49 . 1% of the variance ., TcV strains clustered distant from other DTUs ., TcIII and TcIV strains clustered near to each other ., Interestingly , TcVI strain was placed between TcII and TcIII in the PCoA ., Moreover , TcVI was clustered with TcII in the UPGMA tree ( Fig 3C ) ., Such results are not in agreement with the hypothesis of uniparental inheritance of the minicircles in the hybrid TcVI , which comes from hybridization between TcII and TcIII ., Consequently , we analyzed shared clusters between TcII , TcIII and the hybrids DTUs ( TcV and TcVI ) in order to analyze the hypotheses of uniparental or biparental inheritance of minicircles ., We used a 90% identity threshold in order to be more confident about the identity by descendance of the clusters ., We observed that TcV and TcVI share 11/530 and 19/559 mHVR clusters with TcII , respectively ., Likewise , TcV and TcVI shared 12/429 and 9/469 mHVR clusters with TcIII , respectively ( Fig 4 ) ., Instead , TcII and TcIII share only 2 mHVR clusters between them out of a total number of clusters of 842 combining TcII and TcIII ., These results suggest that minicircle inheritance is biparental ., In addition , TcV and TcVI shared more mHVR clusters with their parental DTUs than between them ( Fig, 4 ) which is concordant with the hypothesis of independent origins of TcV and TcVI ., In order to test if parallel amplicon sequencing would be useful for simultaneous typing of hundreds of strains , we first evaluated rarefaction curves ., In general , the minimum number of reads required to detect at least 95% of the observed clusters was 20 , 000 filtered reads ., The only exception was MNcl2 , which required 40 , 000 filtered reads ., Increasing the number of reads per sample beyond 20 , 000 slightly increased the number of detected mHVR clusters ( Fig 5A ) ., In addition , we evaluated the minimum number of reads required to observe the right DTU assignment described in Fig 2 ., As few as 10 , 000 reads were enough to accurate clustering of the strains ( Fig 5B and 5C ) at 100% of the rarefactions ., Amplicon sequencing of the mHVR could be useful to identify intra-DTU clusters , particularly in TcV or TcVI where strains may have the same composition of mHVR clusters but with high differences in abundance of each one ., In order to develop future methods to assign strains to intra-DTU clusters is pre-requisite that amplicon sequencing can be reproducible to determine mHVR cluster abundance ., Consequently , we assessed reproducibility by sequencing and comparing two independent PCR reactions of the mHVR in LL015P68R0cl4 strain ( TcVI ) ., High correlation in cluster abundances in different PCRs of the same sample was observed ( r = 0 . 999 for the three different identity thresholds ) ( Fig 5D ) ., Here , we made a deep amplicon sequencing of the hypervariable region of kDNA minicircles in the six main lineages ( DTUs ) of T . cruzi ., To the best of our knowledge , this is the first time that these kDNA regions were sequenced at millions of reads of depth ., Our results shed light on different and very interesting aspects of these intriguing DNA sequences ., We accurately show the level of sequence diversity of mHVR within strains , between strains belonging to the same DTU , and between strains belonging to different DTUs ., Although it was already known that mHVR were highly diverse 36 , the magnitude of this diversity at the intra- and inter-DTU level has not been demonstrated with the high precision provided by an NGS approach , as we made here ., We propose a method for typing/elucidating intra-specific diversity of T . cruzi based on the deep sequencing of the hypervariable region of kDNA minicircles ., The idea is based on the outdated but highly sensitive method of mHVR probes 25 , 26 , 35 , 46–48 ., Such probes are useful to detect T . cruzi diversity in biological samples ., However , this methodology has the disadvantages of being technically cumbersome , relying on visual interpretation of bands and requiring representative strains of the diversity of T . cruzi in every assay ( used as probes ) ., The deep amplicon sequencing approach proposed here is reproducible and based on objective sequence data which can be stored in online databases ., Also , the method is multiplexable for hundreds of samples at time and it would be directly applied to biological samples as the mHVR probes ., The method may be potentially useful to address epidemiological questions about associations between intra-specific diversity and variability in clinical manifestations of the chronic disease or the different rates of congenital transmission in different endemic areas ., Such questions have been unsuccessfully addressed using molecular markers with low resolution and/or low sensitivity on biological samples ., We determined that around 20 , 000 filtered reads are enough to reveal most mHVR diversity in a strain and theoretically allowing for running hundreds of samples in a single run of a MiSeq with costs similar or lower than MLST ., However , a wider set of strains belonging to the six main lineages must be studied ., In addition , new bioinformatic methods of analysis will be required for a direct application of the method to biological samples ., In order to develop such typing method , we preliminarily analyzed and compared the diversity of mHVR sequences in reference strains of six DTUs and at millions of reads of sequencing depth ., We observed that strains of the same DTU share more mHVR clusters than strains of different DTUs ., However , unprecedented high differences in mHVR cluster composition was observed for strains of the same DTU with less than 20% of shared mHVR clusters in TcI and TcII ., Instead , almost all mHVR clusters were shared between different TcV strains ., In addition , the patterns of DTU specificity observed by using mHVR probes may be explained in TcV and TcVI by the presence of some shared and abundant clusters ., Instead , considering the higher diversity and low abundance of clusters in TcI , TcII and TcIII , the global pattern of sequences is probably the responsible of specificity in the hybridization assays involving these DTUs ., Interestingly , our data revealed that diversity of mHVR sequences was variable even within a DTU ., This was particularly evident in TcII , where the number of mHVR clusters in Esmeraldo strain doubled that of Tu18cl93 ., Such differences may be caused by long times in culture as it has been observed for other trypanosomatids 40 , 49 ., However , both strains were isolated in the eighties and although it is possible that they had different times in culture , such times would be not very different ( i . e . not in the order of decades ) ., According to this , we suppose that the observed difference in mHVR diversity between the two TcII strains is not due to long time in culture ., In support of the hypothesis of no influence of the time in culture , we observed no differences in mHVR diversity between the two TcV strains examined , despite they have very different times of isolation and maintenance mode in the laboratory ., One of them was isolated in the 1980s and subjected to long periods of maintenance in culture ( Mncl2 ) ; and the other TcV strain ( LL014R1 ) was isolated in 2008 and maintained in triatomine-mouse passages ., Our results also shed some light on the evolutionary mechanism determining the large genetic distances in mHVR sequences among strains and DTUs ., The focus should be first placed on TcV strains which are identical according to MLST and which shared most mHVR clusters ., Despite this , they strongly varied in relative frequencies of mHVR clusters ., Such variations cannot be attributed to simple stochasticity of the PCR amplification because we observed good correlation between different PCR reactions from the same sample ( Fig 5D ) ., Consequently , it is probable that minicircle diversity is mainly driven by genetic drift ., We propose that when two strains diverge , the frequencies of mHVR cluster varies stochastically , some clusters increasing their relative frequency and other decreasing it ., The next step can be seen in strains of TcI which are more genetically distant than the TcV ones ., Such TcI strains show clusters with high abundance in one strain and with very low ( or null ) abundance in the other one ( look at most clusters located on the axes in Fig 2 ) ., Therefore , some clusters will be lost if such lost is not deleterious ( i . e . replaced by a different mHVR class that codes a gRNA editing the same mRNA fragment ) ., Thus , strains would diverge by variations in frequency of the mHVR classes faster than by changes in their sequences ., These variations in the frequency of mHVR classes probably are not under selective pressure ., mHVR frequency variations are apparently allowed because the effective edition of the mRNA is not dependent on the abundance of a minicircle 50 , 51 ., Variations in the frequency of mHVR classes have been also inferred for T . brucei and Leishmania 52 and by a theoretical study assuming random or partially random segregation of minicircles 53 ., With the purpose of developing in the future DTU specific PCRs , we analyzed if different DTUs share common mHVR clusters ., Telleria et al . 36 did not detected shared sequences between DTUs probably because the low sequencing depth ., With a different approach , Velazquez et al . 37 detected that most abundant mHVR classes in CL-Brener ( TcVI ) were also present in other DTUs but in a considerably lower frequency ., We detected shared mHVRs between different DTUs but we did not detect any sequence shared by the six DTUs ., Interestingly , we observed shared clusters between TcVI and TcIII ( 2 . 1% ) ., This is expected considering that TcIII is a parental DTU of the hybrid TcVI and maxicircle sequences of TcIII are closely related to the TcVI ones 54–58 ., However , the TcVI strain also shared 2 . 5% of mHVR clusters with Esmeraldo strain ( belonging to TcII , the other parental DTU of TcVI ) ., Something similar is observed for the also hybrid DTU TcV ( Fig 3 ) ., Instead , only 2 mHVR clusters were shared between TcII and TcIII strains ( 0 . 2% ) ., This clearly suggests that although maxicircles have apparently uniparental inheritance in TcV and TcVI , minicircles were probably inherited from both parentals and some of them persisted for 60 , 000 years since hybridization 59 ., Biparental inheritance of minicircles and maxicircles has been proposed for Trypanosoma brucei hybrids 60–62 ., In this parasite , it has been observed that maxicircle and minicircle inheritance is biparental in hybrids ., However , maxicircles ( 20–50 copies ) are homogenized by genetic drift resulting in the loss of whole maxicircles of one parental in few generations ., However , minicircles have much more copies and they resist the fixation effect of genetic drift for more time ., Consequently , maxicircle inheritance is biparental and just seems to be uniparental due to genetic drift ., As consequence of the biparental inheritance of minicircles , it has been proposed that such inheritance may help to preserve mHVR diversity in T . brucei preventing the effect of the drift , and even that T . brucei requires genetic exchange to prevent the deleterious effect of loss of essential minicircle classes 53 ., Nevertheless , genetic exchange has remained elusive to be detected in T . cruzi ., Experimental hybrids obtained by Gaunt and coworkers showed that maxicircles are from one parental but minicircles were not analyzed 63 and kDNA inheritance was still not addressed in more recent experimental hybrids 64 ., In addition , the frequency of genetic exchange may be variable among different DTUs ., TcV and TcVI ( which display a clearly clonal genetic structure at population level ) 9 , 10 , 12 , 57 have very low mHVR diversity ., Instead , TcI , TcII and TcIII , for which genetic exchange has been proposed in the nature 11 , 13 , 15 , 65 , have higher mHVR diversity ., Moreover , our data may help elucidate the origin of hybrid DTUs ., It has been proposed that TcV and TcVI are the result of a single hybridization event between TcII and TcIII and both DTUs diverged posteriorly 66 , 67 ., However , the alternative hypothesis ( two independent hybridization ) gain weight in the last years ., Particularly , Multilocus Microsatellite Typing ( MLMT ) and Multilocus Sequence Typing ( MLST ) analyses favored the two independent hybridizations hypothesis 57 , 59 ., Considering biparental inheritance , and assuming a single hybridization event , the two hybrid DTUs ( TcV and TcVI ) should share more mHVR classes between them than with the parentals ., However , our analyses show the contrary with very few classes shared between TcV and TcVI ( Fig 4 ) ., This result supports independent hybridizations for the origin of TcV and TcVI ., Alternatively , because both DTUs would have lost many mHVR clusters , the high divergence among them may have been caused by simple stochasticity , although is less likely ., Interestingly , if minicircle are biparentally inherited it is expected that they will behave like the nuclear genes ., So , it is expected that nuclear phylogenies will be similar to the mHVR phylogeny and both discordant to maxicircle phylogeny in cases of hybridization or introgression ., However , some hypotheses about events that occurred very distant in time ( e . g . mitochondrial introgression in the origin of TcIII 57–58 ) might not be addressed by mHVR-based phylogenies because the almost null number of shared mHVR clusters between some DTUs ., Concluding , massive amplicon sequencing of the mHVR is reproducible and suitable for typing hundreds of T . cruzi strains at time because few thousands of reads are required per sample ., However , some drawbacks still need solution ., The main problem in biological samples are mixed infections of different genotypes or DTUs which are very frequent 48 ., However , such problem can be overpassed by developing new bioinformatic methods comparing mHVR composition of a sample against a reference mHVR database which should collect information about the diversity in the DTUs of T . cruzi ., In addition , the develop of an online database where mHVR representative sequences are stored is needed ., We are currently working on such items ., In addition , some rare events of mitochondrial introgression observed in natural populations of T . cruzi lead to discordant typing between nuclear and maxicircle markers 16 , 68 , 69 ., However , it is unknown the effect of mitochondrial introgression on minicircles ., In this sense , a Multilocus deep Sequence Typing ( MLdST ) may be good alternative and a second step ., The deep sequencing of amplicons of the mHVR plus satDNA ( a 195 bp sequence with 105 sequences per genome ) 70 may help elucidate such rare events and may increase sensitivity for typing on biological samples . | Introduction, Materials and methods, Accession numbers, Results, Discussion | Trypanosoma cruzi , the protozoan causative of Chagas disease , is classified into six main Discrete Typing Units ( DTUs ) : TcI-TcVI ., This parasite has around 105 copies of the minicircle hypervariable region ( mHVR ) in their kinetoplastic DNA ( kDNA ) ., The genetic diversity of the mHVR is virtually unknown ., However , cross-hybridization assays using mHVRs showed hybridization only between isolates belonging to the same genetic group ., Nowadays there is no methodologic approach with a good sensibility , specificity and reproducibility for direct typing on biological samples ., Due to its high copy number and apparently high diversity , mHVR becomes a good target for typing ., Around 22 million reads , obtained by amplicon sequencing of the mHVR , were analyzed for nine strains belonging to six T . cruzi DTUs ., The number and diversity of mHVR clusters was variable among DTUs and even within a DTU ., However , strains of the same DTU shared more mHVR clusters than strains of different DTUs and clustered together ., In addition , hybrid DTUs ( TcV and TcVI ) shared similar percentages ( 1 . 9–3 . 4% ) of mHVR clusters with their parentals ( TcII and TcIII ) ., Conversely , just 0 . 2% of clusters were shared between TcII and TcIII suggesting biparental inheritance of the kDNA in hybrids ., Sequencing at low depth ( 20 , 000–40 , 000 reads ) also revealed 95% of the mHVR clusters for each of the analyzed strains ., Finally , the method revealed good correlation in cluster identity and abundance between different replications of the experiment ( r = 0 . 999 ) ., Our work sheds light on the sequence diversity of mHVRs at intra and inter-DTU level ., The mHVR amplicon sequencing workflow described here is a reproducible technique , that allows multiplexed analysis of hundreds of strains and results promissory for direct typing on biological samples in a future ., In addition , such approach may help to gain knowledge on the mechanisms of the minicircle evolution and phylogenetic relationships among strains . | Chagas disease is an important public health problem in Latin America showing a wide diversity of clinical manifestations and epidemiological patterns ., It is caused by the parasite Trypanosoma cruzi ., This parasite is genetically diverse and classified into six main lineages ., However , the relationship between intra-specific genetic diversity and clinical or epidemiological features is not clear , mainly because low sensitivity for direct typing on biological samples ., For this reason , genetic markers with high copy number are required to achieve sensitivity ., Here , we deep sequenced and analyzed a DNA region present in the large mitochondria of the parasite ( named as mHVR , 105 copies per parasite ) from strains belonging to the six main lineages in order to analyze mHVR diversity and to evaluate its usefulness for typing ., Despite the high sequence diversity , strains of the same lineage shared more sequences than strains of different lineages ., Curiously , hybrid lineages shared mHVR sequences with both parents suggesting that mHVR ( and DNA minicircles from the mitochondria ) are inherited from both parentals ., The mHVR amplicon sequencing workflow proposed here is reproducible and , potentially , it would be useful for typing hundreds of biological samples at time ., It also provides a valuable approach to perform evolutionary and functional studies . | taxonomy, medicine and health sciences, parasitic diseases, parasitic protozoans, phylogenetics, data management, protozoans, mitochondria, molecular biology techniques, bioenergetics, cellular structures and organelles, genetic epidemiology, research and analysis methods, sequence analysis, computer and information sciences, kinetoplasts, artificial gene amplification and extension, bioinformatics, epidemiology, biological databases, evolutionary systematics, molecular biology, trypanosoma cruzi, biochemistry, trypanosoma, eukaryota, sequence databases, cell biology, polymerase chain reaction, database and informatics methods, biology and life sciences, energy-producing organelles, evolutionary biology, organisms | null |
journal.pgen.1008193 | 2,019 | A Flp-SUMO hybrid recombinase reveals multi-layered copy number control of a selfish DNA element through post-translational modification | The yeast 2-micron plasmid , nearly ubiquitous among Saccharomyces yeast strains , is a highly optimized extrachromosomal selfish DNA element 1–4 ., The plasmid resides in the nucleus , offers no apparent fitness advantage to its host , and does not impose any significant disadvantage at its normal copy number of 40–60 molecules per haploid chromosome set ., The compact plasmid genome ( ~6 . 3 kbp ) is organized into two functional modules , one devoted to stable propagation ( the plasmid partitioning system ) and the other to copy number maintenance ( the plasmid amplification system ) ., The partitioning system 5–7 , comprised of the plasmid-coded Rep1 and Rep2 proteins together with a cis-acting locus STB , promotes nearly equal segregation of plasmid molecules duplicated by the host replication machinery into mother and daughter cells ., Current evidence is consistent with a ‘hitchhiking model’ in which the plasmid utilizes chromosomes as a vehicle for segregation by physically associating with them 8–11 ., In this respect , the plasmid resembles the episomes of mammalian papilloma and gammaherpes viruses that also resort to chromosome-tethering for stable maintenance during prolonged periods of latent infection 12–20 ., It is possible that selfish genomes inhabiting evolutionarily distant hosts have independently converged on the common strategy of chromosome-coupled segregation as a means for self-preservation ., The plasmid amplification system , consisting of the plasmid-coded Flp site-specific recombinase and its target FRT sites arranged in head-to-head orientation within the plasmid genome , counteracts any reduction in copy number resulting from rare missegregation events 21 , 22 ., Amplification is thought to be triggered by a Flp-mediated recombination event coordinated with bi-directional plasmid replication—DNA inversion within a plasmid monomer or resolution within a plasmid dimer—that reconfigures the mode of replication ( Fig 1A and 1B ) 21 , 23 ., A second recombination event can restore normal fork movement , and terminate amplification ., The amplified plasmid concatemer may be resolved into monomers by Flp or by the host’s homologous recombination machinery ., Positive and negative transcriptional regulation of FLP by plasmid-coded proteins—the putative Rep1-Rep2 repressor and its antagonist Raf1—ensures a prompt amplification response when needed without causing a runaway increase in plasmid copy number 24–27 ., Thus , self-imposed moderation of selfishness is an integral element in the survival strategy of the yeast plasmid 2 , 28 , 29 ., Interestingly , Raf1 appears to play a dual role in plasmid physiology , contributing to both plasmid stability and copy number control ., In addition to blocking the assembly of the Rep1-Rep2 repressor complex , Raf1 is involved in promoting the organization of the Rep1-Rep2-STB partitioning complex 26 , 30 ., The post-translational protein modification machinery of the host also contributes to the regulation of 2-micron plasmid stability and copy number 31–33 ., Impaired sumoylation of Rep1 and Rep2 interferes with their STB-association , and adversely affects plasmid segregation 32 ., Deficient SUMO conjugation to Flp raises its steady-state levels , leading to hyper-amplification of the plasmid ., The resulting increase in plasmid load causes cell cycle delays and reduced replicative life-span 31 , 34 , 35 ., The 2-micron plasmid exemplifies the collaborative roles of self-regulation and host-mediated regulation in the coexistence of a selfish DNA element and its host genome with minimal mutual conflicts between them ., High plasmid copy number and attendant cell death phenotypes are produced by a variety of mutations in protein components associated with SUMO conjugation and deconjugation steps , and with ubiquitin-dependent degradation of sumoylated proteins ., These mutations map to E3 ligases ( siz1Δ , siz2Δ ) , the SUMO maturase/deconjugase ( ulp1 or nib1 ) , a SUMO-targeted ubiquitin ligase ( slx5Δ , slx8Δ ) and certain NPC ( nuclear pore complex ) proteins required for normal cellular localization of Ulp1 31 , 33 , 34 , 36–38 ., These mutants exhibit a marked differential killing effect on yeast strains harboring the 2-micron plasmid Cir+ versus those lacking the plasmid Cir0 ., The misregulated amplification of plasmid DNA likely stems from enhanced single strand nicks at the plasmid FRT sites due to elevated Flp levels ( Fig 1C ) 31 , 33 ., Conversion of the nick into a double strand break by encountering an advancing replication fork can trigger strand invasion by the broken end into an intact circular plasmid , to be followed by break-induced replication ( BIR ) ( Fig 1C ) ., In principle , BIR in the circular template may persist through multiple rounds , producing large plasmid concatemers ., BIR-mediated aberrant amplification is supported by the significant reduction in a high molecular weight DNA form of the plasmid in the absence of Pol32 or of Rad proteins required for known BIR pathways 33 ., The reaction is formally analogous to the alternative ( telomerase-independent ) pathway for lengthening of telomeres via telomere mini-circles as templates , which occurs in yeast , many transformed cell lines and in certain human cancers 39 , 40 ., We wished to address whether , in addition to lowering Flp levels , the SUMO modification of Flp may also modulate its DNA recognition and/or catalytic properties ., To circumvent the technical challenges posed by the low level of the in vivo modification , we utilized a Flp-SUMO fusion protein in which SUMO residues 1–96 are joined in frame to the carboxyl-terminus of Flp ., By demonstrating the nearly identical behavior Flp-SUMO and physiologically sumoylated Flp in a variety of in vivo experimental contexts , we validated the utility of the fusion protein in directly probing the effects of SUMO modification on the physicochemical interactions of Flp with the FRT site ., In vitro assays using purified Flp and Flp-SUMO revealed that Flp-SUMO binds FRT less efficiently than Flp and with weaker cooperativity , and is preferentially excluded from FRT in the presence of Flp ., Consequently , the fusion protein is less active in FRT x FRT recombination than Flp , and the lower activity is reflected in both the strand cleavage and strand joining steps of recombination ., The in vitro results are corroborated by an in vivo assay for DNA damage induced by Flp and Flp-SUMO at FRT sites , consistent with the lower occupancy of these sites ( or accelerated exit from them ) by the fusion protein ., The present results , in conjunction with previously published reports 24 , 25 , 27 , 31 , 33 , suggest a tripartite mechanism for the copy number control of the 2-micron plasmid involving gene expression , protein turnover and protein activity ., The first is imposed by the plasmid itself , while the other two are instituted by the host ., Collectively , they provide a paradigm for the bilateral interactions through which selfish DNA elements and their host organisms strike a fine balance between the fitness advantage gained by such an element from high copy number and the fitness cost incurred by the host—and thus indirectly by the element—from the extra genetic load ., The list of yeast strains and plasmids utilized in this study is given in S1 and S2 Tables , respectively ., The specific figures and table depicting the experiments in which they were employed are also indicated ., The presence or absence of the native 2-micron plasmid in a given strain is denoted as Cir+ or Cir0 , respectively ., This designation does not include ARS-based or 2-micron-derived plasmid constructs ., The genotype of a strain containing such engineered plasmids , but not the native plasmid , is still referred to as Cir0 with the resident plasmid spelled out ., The yeast plasmids ( S1 Fig ) used for genetic assays were constructed by a strategy analogous to that described previously 41 ., The rationale is to generate two requisite linear DNA fragments in vitro by PCR amplification , and allow them to self-assemble the desired circular plasmid in vivo by homologous recombination/repair ., Recombination is facilitated by overlapping sequences that these fragments carry at their ends ., A suitable marker ( ADE2 ) contained in one of the fragments permits the selection of plasmid-containing cells ., We used a constant DNA fragment corresponding to the A-form of the plasmid 42 that included , in sequential order , the 2-micron plasmid RAF1 gene , ADE2 inserted into the plasmid HpaI site , STB , ORI , a copy of the inverted repeat , and the REP2 gene ., The other variable fragment included the REP1 gene , the second copy of the inverted repeat , and the FLP gene with the incorporated modifications ., Sequences adjoining REP1 and RAF1 provided homology at one end ., Homology at the other end came from sequences adjacent to FLP and REP2 ., An equimolar mixture of the constant fragment with one of the variable fragments was used to transform an ade2 Cir0 yeast strain to adenine prototrophy ., DNA samples isolated from a subset of the transformants were analyzed by PCR to identify those that contained the correct plasmid ., Critical regions of the plasmid , including the modified FLP locus and the recombination junction regions , were further verified by DNA sequencing ., Once a parental strain harboring the correct plasmid was established , subsequent transfer of the plasmid to other recipient strains was performed by transformation using isolated total DNA ., Integration of exogenous DNA cassettes into a specific chromosome locale was accomplished by one of three methods based on homology-dependent double strand break repair: ( 1 ) using a linearized integrative plasmid cut within the region of homology , ( 2 ) using PCR-amplified DNA fragments with flanking homology , or ( 3 ) using the CRISPR ( Cas9-sgRNA ) technology ., The first two methods required selection of a marker included in the incoming/editing DNA; no selection was required for the third method ., All constructs were authenticated by DNA sequencing ., Native Flp , Flp-HA-His8 and Flp-SUMO-HA-His8 , as well as mutant derivatives of the tagged proteins , were overexpressed in E . coli cells using the pBAD system ( Invitrogen ) ., Purification of untagged Flp was carried out using previously described procedures 43–45 ., Purification of the tagged proteins included an additional first step of nickel chromatography , followed by dialysis to remove the imidazole present in the elution buffer ., The final preparations were ≥ 85% pure , as judged by SDS-PAGE and densitometric scanning of the Coomassie Blue stained bands ., Protein concentrations in the final preparations were determined using the Bradford assay ., Overnight cultures were prepared from purified single transformant colonies containing individual 2-micron circle-derived plasmids by growing them selectively ( in medium lacking adenine ) at 30°C ., These cultures were diluted in YEPD medium ( n = 0; 104 cells/ml ) and grown for 10 generations ( n = 10 ) at 30°C ., Aliquots were plated out from n = 0 and n = 10 cultures on YEPD medium ., Nibbled and mini-colonies were counted after incubating the plates for 5 days at 26°C ., A founder cell that had lost the plasmid , and the plasmid-borne ADE2 marker , gave rise to a fully red ( non-sectored ) smooth colony ., Such colonies were excluded from the total population in calculating the fraction of nibbled colonies ., Plasmid loss rate per generation ‘I’ ( for instability ) was estimated from plates incubated at 30°C based on fully red and total colony counts ., I = ( 1/10 ) x ln ( f0/f10 ) 46 , where f0 and f10 are the fractions of plasmid containing cells ( yielding colonies other than the fully red ones ) at n = 0 and n = 10 , respectively ., The sample size for the individual estimates of plasmid loss rate and the fraction of nibbled or mini-colonies was a minimum of 800 colonies ., The assays were performed as at least three repetitions ., The PCR protocols followed those described by Chen et al . 31 , and utilized the same plasmid and reference chromosomal amplicons as well as the primer pairs described by them ., The amplification reactions were carried out with ABI PRISM 7900HT SDS using the SYBR Green Master Mix ( Applied Biosystems ) ., The number of cycles required to reach the CT number ( preset threshold ) for each DNA sample was calculated from six separate experiments ., The relative change in the copy number of a plasmid between two strains , normalized to the chromosomal reference sequence , was calculated by the 2-ΔΔCT analysis 47 ., The expression cassette for Flp-SUMO controlled by the GAL1 promoter was inserted at the TRP1 locus ( thus disrupting it ) on chromosome IV in Cir0 wild type , siz1Δ siz2Δ , slx5Δ and slx8Δ strains ., Aliquots of raffinose-grown overnight cultures were inoculated into raffinose medium , grown to mid-log phase at 30°C , and induction was performed by transferring them to 2% galactose with continued incubation at 30°C ., The control ( uninduced ) cells were transferred from raffinose to glucose medium and incubated at 30°C ., At 2hr , cells were spun down , washed , and suspended in TE buffer before adding cycloheximide ( 100 μg/ml ) to arrest protein synthesis ., Cells removed at intervals over a 60 min time course were treated with lysis buffer ( 50 mM HEPES , pH 7 . 0; 75 mM KCl . , 1 mM MgCl2 , 1 mM EGTA , 0 . 5% Triton X-100 , I mM DTT , 1 mM PMSF and one protease inhibitor tablet from Roche/50 ml ) ., Cell extracts prepared by bead beating ( 5 min; 4°C ) were fractionated by 12% SDS-PAGE , and analyzed by quantitative western blotting ., Flp-SUMO bands were detected by anti-HA antibody ( BioLegend ) at 1:1000 dilution , and normalized against actin bands visualized using anti-β-actin antibody ( Gene Tex ) at 1:1000 dilution ., Proteasome function was inhibited with MG-132 according to published procedures 48 ., The following modifications were made to the standard protocols for measuring protein turnover ., Overnight raffinose cultures were grown in synthetic medium without ammonium sulfate , and supplemented with 0 . 1% proline as well as other appropriate amino acids ., In addition , the re-inoculation medium for obtaining mid-log phase cells for galactose induction included 0 . 003% SDS ., The induction period was 2 hr , with 75 μm MG-132 ( Biomol , Plymouth Meeting , PA ) being added at 90 min ., Control cells received an equivalent volume of DMSO , the solvent for MG-132 ., The rest of the procedure—cycloheximide treatment , preparation and fractionation of cell extracts , and western blotting—was performed as described under the previous section on protein stability assays ., The experimental strains were derived from Cir0 wild type and siz1Δ siz2Δ strains expressing the GAL1 promoter driven Flp-SUMO ( see the section above on protein stability assays ) or from an analogous set of strains in which Flp-SUMO was replaced by Flp ., The plasmid pADE2-Flp ( S1 Fig ) was introduced into these strains , and maintained by adenine selection ., A CEN-TRP1-plasmid expressing Rad52-YFP from the native RAD52 promoter was also maintained in them by selection ., The conditions for Flp or Flp SUMO induction were the same as those described for protein stability estimates in the absence of MG-132 ., Cells induced for 2 hr in galactose and the corresponding uninduced control cells ( 2 hr in glucose ) were collected , washed and fixed in formaldehyde for scoring fluorescent foci ., Each set of assays was repeated three times ., The binding assays were performed in 30 μl individual mixtures incubated on ice for 20 min using the buffer conditions described by Prasad et al . 49 ., The substrate DNA fragment ( 0 . 05 pmol per binding reaction ) was 262 bp long , and contained one FRT site ., Aliquots were fractionated by electrophoresis in 5% polyacrylamide gels ( 29:1 crosslinking ) at 4°C in 1x TBE duffer ., The bound complexes and the unbound substrate were visualized by autoradiography or phosphor imaging ., The conditions for in vitro recombination were similar to those described previously 50 , 51 ., Each 30 μl reaction mixture contained 0 . 2 pmol plasmid substrate ( with two FRT sites oriented head-to-tail ) and 1 pmol of purified Flp or Flp-SUMO ., At the end of the 30°C incubation period ( from 0 . 5 to 30 min ) , the reactions were stopped by treatment with 0 . 2% SDS ( final concentration ) followed by proteinase K treatment ( 50 μg/reaction sample ) ., DNA purified by chloroform-phenol extraction and ethanol precipitation was digested with NdeI and EcoRV ., The digestion products were separated by 1% agarose gel electrophoresis , and DNA bands were visualized by ethidium bromide staining ., Strand cleavage and strand joining reactions were carried out in the recombination buffer with 0 . 05 pmol of the respective 32P-labeled half-site substrates per reaction and Flp or Flp-SUMO ranging from 0 . 2 pmol to 2 pmol ., At the end of 30 min incubation at 30°C , reactions were stopped by adding 0 . 2% SDS , and processed without proteinase K treatment ., The cleavage and joining reactions were analyzed by electrophoresis in 12% SDS-polyacrylamide ( 29:1 crosslinking ) and 12% polyacrylamide-urea ( 19:1 crosslinking ) gels , followed by phosphorimaging or autoradiography ., First , tubes were set up in pairs on ice with one tube within a pair containing the 32P-labeled half-site plus the R191A mutant , and the other containing the same labeled half-site plus the Y343F mutant ., The amounts of half-site and protein in each tube were 0 . 05 pmol and 0 . 5 pmol , respectively , in 15 μl of 1 . 5x recombination buffer ., Following 10 min on ice to allow full occupancy of the half-site by protein , 7 . 5 μl each of the binding mixture were withdrawn from each set of paired tubes , and transferred simultaneously to fresh tubes ( maintained at 30°C ) containing 2 . 5 pmol of an unlabeled DNA fragment with one FRT site in 15 μl 0 . 5x recombination buffer ., The contents were gently mixed in each tube and incubated for 30 min ., Except for the difference in the substrate half-sites , the strand cleavage and joining reactions were similar in other respects ., The reactions were analyzed by SDS-polyacrylamide gel ( 12% ) electrophoresis ( for cleavage ) and by polyacrylamide-urea gel ( 12% ) electrophoresis ( for joining ) ., Radioactively labeled DNA bands , captured on a phosphor storage screen ( Bio-Rad ) , were scanned using a Typhoon Trio Phosphorimager ( GE-Healthcare ) ., Unlabeled DNA bands were visualized in agarose gels by ethidium bromide staining ., Protein bands were detected in western blot analyses using PierceTM ECL protocol ( ThermoFisher Scientific ) ., Image analysis and quantitation of band intensities were performed using the software Quantity One ( Bio-Rad; version 4 . 5 . 1 ) ., For recombination , strand cleavage and strand joining assays , the extent of reaction was estimated as the ratio of the intensity of product band ( s ) to the sum of the intensities of substrate and product bands ., For DNA binding , the ratios of the bound C-I and C-II complexes to the sum of C-I , C-II and unbound DNA were determined ., Protein bands were quantitated against actin as the internal control ., Multiple exposures were used to compensate for large intensity differences between individual bands ., Appropriate correction factors were applied , based on the linear ranges of intensity variation ., Impairment in the regulation of Flp-mediated amplification leads to high 2-micron plasmid copy number 31 , 35 , 52 , which induces characteristic nibbling at colony edges ., This phenotype , which is more conspicuous at 20°C than at 30°C , is due to differences in plasmid copy number in individual cell lineages , resulting in variable growth inhibition and cell mortality among them ., Loss of plasmid restores normal growth and smooth edges ., Over time , plasmid-free Cir0 cells tend to rise in the population ., Thus , colony morphology and plasmid loss rates are reliable reporters of the mean plasmid load carried by cells , and indirectly of the Flp level/activity in them ., The steady-state level of sumoylated Flp in a wild type strain is ~10% of total Flp 31 ., The predominant site of SUMO conjugation , mediated by Siz1 and Siz2 , is Lys-375 31 located < 50 amino acids upstream of the carboxyl-terminus ( Ile-423 ) ., Replacement of Lys-375 by arginine partially recapitulates the effects of siz1Δ siz2Δ in a wild type background , yielding ~4-fold increase in Flp , ~2-fold higher plasmid copy number , and consistently more abnormal colonies on plates incubated at 20°C 31 ., Given the relative proximity of Lys-375 and Ile-423 , we suspected that Flp containing the SUMO moiety ( amino acids 1–96 ) as a carboxyl-terminal extension is likely to functionally mimic Flp ( K375-SUMO ) ., If so , Flp-SUMO may justifiably be utilized as a surrogate for Flp ( K375-SUMO ) in addressing potential differences between native and modified Flp in their relative stability in vivo as well as their DNA recognition and catalytic properties in vivo and/or in vitro ., It is nearly impossible to study exclusively the naturally sumoylated Flp in a cell , as it would be diluted out by the excess unmodified version ., In addition , the extent of the modification ( ~10% ) makes it technically quite challenging to obtain sufficient quantities of Flp ( K375-SUMO ) for in vitro analyses ., In order to test whether Flp-SUMO can redress the effects of siz1Δ siz2Δ on colony morphology , we transformed Cir0 strains with 2-micron plasmid derivatives engineered to express Flp or Flp-SUMO under the control of the native FLP-promoter ( S1 Fig ) ., Except for manipulations of the FLP locus and an insertion of the ADE2 marker , the reporter plasmids retained the overall organization of the native 2-micron plasmid genome ., Note that the expressed Flp , Flp-SUMO and their variants contained the HA-His8 epitope tag at their carboxyl-termini ., For simplicity , these proteins as well as the plasmids expressing them are referred to without mentioning the tag ., The large fraction of nibbled colonies ( > 75% at 26°C ) in the siz1Δ siz2Δ strain containing pADE2-Flp was substantially reduced ( ~1% ) when the strain harbored pADE2-Flp-SUMO ( Fig 2A and 2B; S3 Table ) ., The enlarged images of colonies shown above S3 Table highlight the difference between smooth and nibbled edges ., The frequency of mini-colonies in the population , signifying highly retarded cell growth or extensive cell death , also showed a corresponding reduction ( from ~18% to ~3% ) ( S3 Table ) ., The nibbled- and mini-colony phenotypes were consistent with a ~5-fold increase in the mean copy number of pADE2-Flp in siz1Δ siz2Δ compared to the wild type ( S2 Fig ) ., For comparison , the increase in the native 2-micron plasmid copy number in the mutant strain was ~10-fold ( S2 Fig ) ., As Flp-SUMO is recombination-competent ( S3 Fig ) , the lack of nibbling in siz1Δ siz2Δ harboring pADE2-Flp-SUMO was not due to Flp-SUMO being inactive in generating FRT-nicks , which are intermediates in the recombination reaction ( and potential initiators of BIR ) ., The catalytic variant Flp ( H305L ) expressed by pADE2-Flp ( H305L ) ( S1 Fig ) is strongly defective in strand joining in vitro and in vivo in yeast 53–55 , and is expected to cause an accumulation of the FRT-nicked intermediate ., The variant is inactive in recombination ., The presence of pADE2-Flp ( H305L ) , contrary to expectation , produced few nibbled ( 1–2% ) or mini-colonies ( ~3% ) either in the wild type or in the siz1Δ siz2Δ strain at 26°C ( Fig 2C; S3 Table ) ., Presumably , sumoylation of Flp ( H305L ) at Lys-375 in the wild type strain was sufficient to suppress excessive FRT-nicking ., The lack of nibbling even in the siz1Δ siz2Δ strain was likely due to high pADE2-Flp ( H305L ) missegregation , signified by the relative abundance of red ( ade2 ) and red-sectored colonies ( Fig 2C ) ( see also plasmid loss rates in Fig 2D ) ., As a result , mother cultures ( n = 0 in Fig 2 ) would be enriched in cells with low copy numbers of pADE2-Flp-SUMO as well as plasmid-free cells capable of resuming growth when provided with adenine ., Such cells would be further enriched during non-selective growth ( from n = 0 to n = 10 ) due to their fitness advantage ., In fact , overexpression of Flp ( H305L ) in Cir+ cells is a convenient method for rapidly curing them of the endogenous 2-micron plasmid 56 ., In sum , the aberrant cell growth typical of under-sumoylation of Flp is alleviated by the expression of Flp-SUMO from the native FLP-promoter ., The lack of anticipated nibbling in the siz1Δ siz2Δ strain from Flp ( H305L ) expression is the result of the high rate of plasmid loss induced by this Flp variant ., As equal segregation of 2-micron plasmid molecules occurs in physical association with chromosomes 8–11 , the high molecular weight hyper-amplified plasmid concatemers in a siz1Δ siz2Δ strain are likely to interfere with this process ., Furthermore , the deficiency in sumoylation of Rep1 and Rep2 partitioning proteins might have an additional effect on segregation 32 ., Plasmid-free cells , having higher fitness , will tend to outgrow their plasmid-containing counterparts ., ‘Apparent’ plasmid stability during non-selective growth provides a reasonable test of the potential salutary effect of Flp-SUMO under conditions that proscribe normal sumoylation of Flp ., The pADE2-Flp plasmid showed a significantly higher loss rate in the siz1Δ siz2Δ strain compared to the wild type ( Fig 2D; left pair of histograms ) ., By contrast , there was a modest improvement in the stability of pADE2-Flp-SUMO in the mutant compared to the wild type ( Fig 2D; middle pair of histograms ) ., The lower basal stability of pADE2-Flp-SUMO than pADE2-Flp in the wild type strain ( Fig 2D; green histograms of the left and middle pairs ) might result from the larger size of pADE2-Flp-SUMO , the particular modification of the FLP locus , or from potential additional sumoylation at Lys-375 ( which would be ameliorated by siz1Δ siz2Δ ) ., Consistent with the expected increase in steady state DNA damage at FRT and the attendant plasmid hyper-amplification , pADE2-Flp ( H305L ) was more unstable than pADE2-Flp in the wild type strain ( Fig 2D; green histograms of the left and right pairs ) ., The instability of pADE2-Flp ( H305L ) was worsened by siz1Δ siz2Δ ( Fig 2D; right pair of histograms ) , suggesting that Flp ( H305L ) , analogous to Flp , is also downregulated via sumoylation ., The comparable loss rates of pADE2-Flp ( H305L ) and pADE2-Flp in the wild type and the siz1Δ siz2Δ strains , respectively , ( Fig 2D; orange histogram of the left pair and green histogram of the right pair ) might appear to suggest that normal sumoylation of Flp ( H305L ) and strongly reduced sumoylation of Flp are more or less equivalent with respect to the amount of strand nicks that the two proteins produce at FRT ., However , unlike Flp , Flp ( H305L ) cannot resolve an amplified plasmid concatemer by recombination , nor can it counter plasmid missegregation by recombination-mediated amplification ( Fig 1A and 1B ) ., These factors may aggravate the instability of pADE2-Flp ( H305L ) in the siz1Δ siz2Δ strain ., The plasmid stability results demonstrate that the increased formation and expansion of plasmid-free cells triggered by FRT-nicks in a siz1Δ siz2Δ host is strongly suppressed when this strain expresses Flp-SUMO instead of Flp or Flp ( H305L ) ., The BIR pathway promotes the repair of one-ended double strand breaks—such as those resulting from the arrival of a replication fork at a Flp-nicked FRT site 31 , 33 , 39 , 57 , 58 ., Consistent with the mechanism diagrammed in Fig 1C , the formation of amplified high molecular weight plasmid DNA is dependent on strand cleavage by Flp , and requires Pol32 as well as Rad proteins involved in BIR 33 ., The initial D-loop intermediate of BIR formed by strand invasion of homologous DNA may be processed into a replication fork ., Alternatively , it may mature into a Holliday junction by convergence with a replication fork from the opposite direction ( Fig 3A; left ) ., The coalescence of two D-loops expanding in opposite directions would generate a double Holliday junction ( Fig 3A; right ) ., The organization of the two FRT sites and the location of the bi-directional replication origin within the 2-micron plasmid provide opportunities for a Flp-mediated BIR D-loop to meet a replication fork , or a second such D-loop , approaching it head-on ( Fig 3A ) ., As the Flp-induced single strand nicks may occur at one or both of the plasmid FRT sites , and on either DNA strand within a site , head-to-tail configuration of two D-loops or a D-loop and a replication fork is also possible ., Crystal structures and in vitro experiments rule out double strand breaks at FRT by Flp 50 , 59–61 , minimizing the probability of two-ended double strand break repair in FRT DNA damage ., The resolution of specialized DNA structures such as Holliday junctions and D-loops in yeast require Yen1 and/or Mus81-Mms4 activities 53 ., Induction of Flp ( H305L ) in a strain containing an FRT site inserted between two strong replication origins in a chromosome results in poor viability in the yen1Δ mus81Δ background 53 ., We utilized isogenic Cir0 and Cir+ strains harboring an identical chromosomal insertion of FRT to test whether the presence of additional copies of plasmid-borne FRT sites would aggravate the Flp ( H305L ) —yen1Δ or the Flp ( H305L ) -mus81Δ effect , and whether cell survival can be improved by replacing Flp ( H305L ) by Flp ( H305L ) -SUMO ., In the Cir0 strain , mus81Δ caused a decrease in colony forming units upon Flp ( H305L ) induction , with yen1Δ and yen1Δ mus81Δ displaying a stronger effect ( Fig 3B ) ., The loss of viable colonies from each single mutation as well as the double mutation was magnified in the Cir+ strain ( Fig 3C ) ., Expression of Flp ( H305L ) -SUMO in place of Flp ( H305L ) restored cell survival in the single mutants to nearly the same level as in the wild type ( Fig 3D ) ., The palliative response to Flp ( H305L ) -SUMO , though not as strong , was evident in the double mutant as well ( Fig 3D ) ., In the absence of Mus81 or Yen1 or both , the branched intermediates of Flp ( H305L ) -induced BIR ( three or four-way DNA junctions ) appear to accumulate , to the detriment of the cell ., The apparent reduction of these intermediates in the presence of Flp ( H305L ) -SUMO is consistent with an abatement in the formation of unsealed strand nicks at FRT that precedes the BIR events ., Furthermore , these results corroborate the previous inference that the DNA damage induced by Flp ( H305L ) at plasmid FRT sites may be masked by accelerated plasmid loss from cells ., The Slx5-Slx8 STUbL ( sumo targeted ubiquitin ligase ) regulates a wide range of cellular functions in yeast that include gene expression , quality control of nuclear proteins , DNA damage repair , and chromosome stability 62–66 ., The coupling to ubiquitin-proteasome systems via Slx5-Slx8-mediated recognition of conjugated SUMO , or native surface features that mimic SUMO , may bring about not only the degradation of particular target proteins but also the functional re-localization of multi-subunit protein machines ., Examples include proteolysis of the Mot1 transcription factor 67 or the Matα2 repressor 68 , and the relocation of double strand DNA breaks in G1 cells to repair centers stationed at the nuclear periphery 69 ., Ubiquitin-dependent proteolysis of sumoylated Flp appears to be one important mechanism for the post-translational regulation of Flp ., The levels of Flp , including its SUMO-conjugated form , are increased in slx5Δ and slx8Δ strains 33 ., We therefore tested whether the beneficial effects of Flp-SUMO observed in the siz1Δ siz2Δ strain would be reversed in an slx5Δ or slx8Δ mutant ., The slx5Δ an | Introduction, Materials and methods, Results, Discussion | Mechanisms for highly efficient chromosome-associated equal segregation , and for maintenance of steady state copy number , are at the heart of the evolutionary success of the 2-micron plasmid as a stable multi-copy extra-chromosomal selfish DNA element present in the yeast nucleus ., The Flp site-specific recombination system housed by the plasmid , which is central to plasmid copy number maintenance , is regulated at multiple levels ., Transcription of the FLP gene is fine-tuned by the repressor function of the plasmid-coded partitioning proteins Rep1 and Rep2 and their antagonist Raf1 , which is also plasmid-coded ., In addition , the Flp protein is regulated by the host’s post-translational modification machinery ., Utilizing a Flp-SUMO fusion protein , which functionally mimics naturally sumoylated Flp , we demonstrate that the modification signals ubiquitination of Flp , followed by its proteasome-mediated degradation ., Furthermore , reduced binding affinity and cooperativity of the modified Flp decrease its association with the plasmid FRT ( Flp recombination target ) sites , and/or increase its dissociation from them ., The resulting attenuation of strand cleavage and recombination events safeguards against runaway increase in plasmid copy number , which is deleterious to the host—and indirectly—to the plasmid ., These results have broader relevance to potential mechanisms by which selfish genomes minimize fitness conflicts with host genomes by holding in check the extra genetic load they pose . | Plasmids of budding yeasts , exemplified by the 2-micron plasmid of Saccharomyces cerevisiae , and mammalian papilloma and gammaherpes viruses typify eukaryotic extra-chromosomal selfish DNA elements ., The plasmid and the viral episomes , despite the long evolutionary divergence of their hosts , share striking similarities in lifestyles ., These include the ability to segregate to daughter cells by hitchhiking on chromosomes and to switch from cell cycle regulated replication to iterative replication for copy number maintenance ., While selfish elements , including those integrated into chromosomes , rely on their hosts’ genetic potential for long-term survival , their genetic load is carefully regulated to minimize fitness conflicts with the hosts ., Our study focuses on the Flp site-specific recombinase , which is central to the copy number control of the 2-micron plasmid and whose steady state levels are regulated through transcriptional control by plasmid coded proteins and through post-translational modification by the host’s sumoylation machinery ., We demonstrate that sumoylation , in addition , attenuates the catalytic activity of Flp by diminishing its DNA binding affinity and inter-monomer cooperativity , providing another layer of protection against runaway increase in plasmid copy number ., Population control by self-imposed and host-mediated mechanisms is likely a general strategy among selfish elements to ensure nearly conflict-free coexistence with host genomes . | recombination reactions, monomers, plasmids, dna-binding proteins, plasmid construction, sumoylation, dna replication, genetic elements, forms of dna, dna construction, dna, molecular biology techniques, research and analysis methods, polymer chemistry, proteins, chemistry, recombinant proteins, molecular biology, biochemistry, post-translational modification, nucleic acids, genetics, biology and life sciences, chemical reactions, physical sciences, genomics, mobile genetic elements | null |
journal.pcbi.1005839 | 2,017 | Metabomatching: Using genetic association to identify metabolites in proton NMR spectroscopy | Since the seminal metabolome-wide genome-wide association study ( mGWAS ) by Gieger et al . in 2008 1 , mGWASes performed on blood and urine spectral metabolome phenotypes have uncovered an increasing part of the heritable variability of the human metabolome through the discovery of hundreds of genetically influenced metabolome phenotypes 2–4 ., Most mGWASes use estimated metabolite concentrations as phenotypes 1 , 5–10 ., In such targeted mGWASes , metabolite concentrations are obtained by quantification 11 of spectral metabolome data produced by mass spectrometry ( MS ) or nuclear magnetic resonance ( NMR ) spectroscopy ., While targeted approaches pave the way for reproducible metabolomics , only a fraction of the measured metabolome data is quantified into metabolite concentrations due to the arduous nature of metabolite identification 12–16 ., In Rueedi et al . 17 , we used an untargeted approach 18–20: we binned then normalized the NMR data , and tested the resulting bin intensities , which we called metabolome features , for association with genotypes ., We then sought metabolite identification only for significantly associated metabolome features ., To do so , we employed an inherent characteristic of the untargeted approach: genetic spiking ., If the genetic component of a metabolite concentration is detected in the untargeted mGWAS , then the relevant genotype will associate with metabolome features that correspond to the peaks of the NMR spectrum of the metabolite ., Much as metabolite spiking does by flooding a sample with a metabolite of interest , genetic spiking isolates , by genetic association , the spectrum of the genetically influenced metabolite ., However , whereas the aim of metabolite spiking is to determine an unknown spectrum for a known metabolite , we developed metabomatching to use genetic spiking to identify an unknown metabolite from a known spectrum ., We previously showed that metabolite identification using the metabomatching procedure works in principle 17 , 20 , and applied it to identify the metabolite involved in a novel SNP-feature association ., Here , we further develop metabomatching , present its core concepts and data , perform numerical simulations , and evaluate its performance on two sets of mGWAS data ., We also present the metabomatching software , describe its implementation and settings , and highlight the best practices and pitfalls of its application ., The default spectral database used by metabomatching is acquired from the Human Metabolome DataBase 24 ( HMDB ) , which lists experimental proton NMR spectra for 835 metabolites ., In HMDB , the spectrum of a metabolite is described in two ways: as a list of peaks , and as a list of multiplets ( see Fig 1B ) ., A peak is defined by a spectral position , expressed as a chemical shift in parts per million ( ppm ) , and a relative NMR intensity , that is the peak height expressed relative to the highest peak in the spectrum ., A multiplet is defined by a chemical shift range , and a proton count ., Peaks group into clusters , and for each such cluster in the peak description , there is , generally , a corresponding multiplet in the multiplet description whose range encloses the cluster ., Furthermore , the area under the curve delimiting the peaks of a cluster can be related to the proton count of the corresponding multiplet 25 ., The two descriptions are usually , but not always , coherent ., Alternatively , metabomatching can use a database acquired from the Biological Magnetic Resonance dataBank 26 ( BMRB ) , which lists experimental proton NMR spectra for 670 metabolites ., In BMRB , the spectrum of a metabolite is described only as a list of peaks ., Each metabolite , however , may have several peak description spectra , obtained in different experiments ., Both HMDB and BMRB collect information on any metabolites found in the human body ., As a result , many of the spectra contained in the full spectral databases may be irrelevant for a specific mGWAS , typically because the corresponding metabolites may not be contained in the studied biofluid ., Metabomatching can therefore also use specific subsets of the full spectral databases ., For urine , the spectral database is derived from the urine metabolome database ( UMDB ) 27 and contains proton NMR spectra for 180 metabolites , 124 if based on BMRB ., For serum , the spectral database is derived from the work of Gowda et al . 12 and contains proton NMR spectra for 67 metabolites if based on HMDB , 49 if based on BMRB ., For the comparison of pseudospectra to reference spectra , we introduce a feature match set Fδ ( m ) for every metabolite m in the reference database ., Fδ ( m ) is defined to contain all features f within a neighborhood of δ ppm of any spectrum peak listed in the peak description of m ( see Fig 1C ) ., For the pseudospectrum of a given SNP r and the spectrum of every metabolite m , we compute the match sum, ∑ f ∈ F δ ( m ) β r f 2 s r f 2 , ( 1 ), with βrf the effect size and srf the standard error of the association between SNP r and feature f ., Even though the features are usually not independent , we consider the match sum to be χ2-distributed with |Fδ ( m ) | degrees of freedom , so as to define the score for the tested metabolite as the negative logarithm of the corresponding p-value ., As a result , while we use the scores to rank metabolites for a given SNP , the scores do not inform on the statistical significance of a spectrum-pseudospectrum match , nor do we compare scores obtained for the pseudospectra of different SNPs ., Because multiplet descriptions of the reference NMR spectra in HMDB can significantly differ from peak descriptions , they can be considered as composing a separate spectral database ., To use this set , or its corresponding biofluid specific subsets , metabomatching can be run in multiplet mode , instead of the standard peak mode described above ., The match set Fδ ( m ) used to compute the match sum ( 1 ) for m is then defined to contain all features f falling in , or within δ of , any multiplet range of metabolite m ( see Fig 1C ) ., Because multiplet ranges tend to pad their corresponding peak cluster , the neighborhood parameter δ takes a smaller value in multiplet mode than in peak mode ., The resulting match sets are then comparable to their peak mode counterparts , in general ., However , differences between the two descriptions , in cluster position , size , or even presence , occur for about 10% of metabolites in HMDB ., These differences can significantly affect metabomatching results ., Metabolome features that are common to the spectrum of a metabolite present in the study samples correlate , and metabomatching can be set to take this correlation into account ., The correlation is strongest among neighboring features , which may be common to multiple metabolites of spectra containing similar peak clusters , but also appears in features corresponding to peaks in different spectrum clusters ., Heuristically however , only the correlation between neighboring features is detrimental to metabomatching , and decorrelation is therefore applied only to feature neighborhoods ., Given the user-provided feature-feature correlation matrix C ^ , match sum ( 1 ) is then modified to, ∑ f , g ∈ F δ ( m ) β r f s r f C δ ; f g - 1 β r g s r g , ( 2 ), where C δ ; f g ≐ ( 1 - λ ) C ^ f g J δ ; f g + λ I f g provides decorrelation , with λ ∈ 0 , 1 the shrinkage parameter 28 , which serves to regularize ., Restriction to feature neighborhoods results from the block diagonal matrix Jδ , with Jδ;fg = 1 if f and g are members of the same neighborhood , that is if they are connected by a sequence of features in Fδ ( m ) each at most 2δ ppm apart , and I the identity matrix ., Metabomatching includes two variants for cases where a SNP affects a pair of metabolites: 2-compound metabomatching if the effects are of equal directions , and ±-metabomatching if the effects are of opposite directions ., For 2-compound metabomatching , we compute the match sum for pairs of metabolites by running the sum over pair match sets , defined as Fδ ( m1 , m2 ) ≐ Fδ ( m1 ) ∪ Fδ ( m2 ) ., Metabolite pairs are accordingly scored and ranked ., In ±-metabomatching , standard ( 1-compound ) metabomatching is run separately for each effect direction , setting to 0 the effect size for associations in the other direction that exceed a user-provided p-value threshold ., 2-compound and ±-metabomatching can be combined into ±-2-compound metabomatching for SNPs affecting at most one pair of metabolites in each direction ., By squaring β/s in match sum ( 1 ) or ( 2 ) , χ2-scoring increases signal to noise ratio , both by amplifying the contribution of strongly associated features to metabomatching scores , and by ignoring effect directions ., This increase applies indiscriminately , however , and may actually favor competing metabolites more than the metabolite to identify ., Therefore , for pseudospectra where this increase is not necessary , such as those produced in mGWASes of high statistical power , for example , stronger matches may be obtained with Z-scoring ., Here , scores are computed according to the match sum, ∑ f ∈ F δ ( m ) β r f s r f ( 3 ), which we consider to be normally distributed , under the null hypothesis , with zero mean and variance |Fδ ( m ) | , even though the sampled features are not independent ., To apply decorrelation in Z-scoring metabomatching , match sum ( 3 ) is not modified , but the variance is computed as |∑fg Cδ;fg| , with Cδ the block diagonal matrix as previously defined ., As in χ2-scoring , multiplet-mode and 2-compound variants applied by using the corresponding match sets in match sum ( 3 ) ., Because Z-scoring is explicitly sensitive to effect directions , ±-metabomatching is not required for SNPs affecting two metabolites with opposite effect directions ., However , the separate presentation of results of ±-metabomatching may be useful in cases where the effect sizes are such as to cause metabolites matched with one effect direction to systematically outrank metabolites matched with the other direction ., To summarize , metabomatching is run for a given pseudospectrum: against a set of match sets , defined by the selected spectral reference database , the mode , and neighborhood parameter δ; where appropriate , as 1-compound , 2-compound , ±- , or ±-2-compound variant; and depending on performance , with or without decorrelation , and with χ2- or Z-scoring ., Metabomatching outputs the score for each metabolite in the spectral database , and produces a figure showing the pseudospectrum and the spectra of the highest ranked candidate metabolites ( Fig 2 ) ., We bin the chemical shift range 0 , 10 uniformly , in 0 . 01 ppm increments , and round reference spectra to the binning ., We express the spectrum of each metabolite as a vector hm , with h j m the height of the peak in bin j , set to 0 if the spectrum does not include bin j , and define the size of the spectrum as s m ≐ ∑ j h j m ., To model the genetic association between a SNP and metabolite m , we randomly assign a genotype gi ∈ {0 , 1 , 2} to each individual ( i ∈ 1 , 400 ) , according to a minor allele frequency of 0 . 2 , and build the feature metabolome M0 of elements, M i j 0 ≐ β h j m g i + N ( 0 , 1 ) ., ( 4 ), Because the number of individuals , the minor allele frequency and the amplitude of noise are fixed , the strength of the association is controlled fully by the choice of effect size β ., We then associate the metabolome M0 with the genotype g , and apply metabomatching to the resulting pseudospectrum ., For each metabolite , we repeat this procedure 1 000 times , and compute r 90 m , the 90th percentile over the 1 000 ranks of m ., We consider metabomatching successful for m if r 90 m = 1 ., From the results of this simple model , shown in Fig 3A for UMDB , we can make two important observations ., First , that if the effect size is large enough , metabomatching can identify any metabolite ., Second , that the performance of metabomatching , characterized here by r 90 m , is strongly correlated with the spectrum size ., We then add genetic noise to the model , in the form of Na randomly drawn features that also associate , with a randomly drawn direction , with genotype g ., We insert these genetic noise features in the model by adding the terms aj ∈ {−1 , 0 , 1} , such that ∑j |aj| = Na , when building the feature metabolome Mα of elements, M i j α ≐ β h j m g i + α β g i a j + N ( 0 , 1 ) ,, where α < 1 ., As the amount Na , or amplitude α , of genetic noise increases , metabolite m faces wider , respectively stronger , competition from other metabolites in the spectral database ., When β is small , random noise still determines metabomatching performance , and r 90 m is similar to that for metabolome M0 shown in Fig 3A ., When β is large , however , genetic noise dominates ., As shown in Fig 3B ( and S4 Fig for other settings and for both UMDB and HMDB ) , metabomatching can then no longer identify all metabolites consistently , because other metabolites in the database outscore m by matching genetic noise features ., Some of these other metabolites may obtain their score from genetic noise features only , but true competition for m is provided by metabolites that match both genetic noise features and features of m ., Because these competing metabolites have spectra similar to the spectrum of m , they tend to be viable metabomatching candidates ., For metabolites with a single peak f , we can count the number of metabolites of match set that contain f to determine the size of this competing group ., In Fig 3C , we show this number , η ( f ) for UMDB , and in Fig 3D we see that r 90 m for metabolites of size 1 correlates strongly with η ., For larger spectra , where we take η for the lead feature ( the one of height 1 ) , the correlation holds , but η is less representative of the size of the competing group ., We first tested metabomatching on pseudospectra obtained in the urine NMR mGWAS 17 in the CoLaus study 29 ., NMR data were aligned , normalized , and uniformly binned in 0 . 005 ppm increments ., The resulting untargeted metabolome contained 1 , 276 features for 835 individuals ., As references , we used SNP-metabolite associations that were previously reported in targeted mGWASes on urine NMR 8 , 18 , 20 with a p-value below 10−5 and involving a metabolite for which an NMR spectrum is listed in UMDB ., If a CoLaus SNP located within 500kb of the reference SNP associated with p < 10−6 with at least one feature contained in the match set of the reference metabolite ( with δ = 0 . 03 in peak mode , 0 . 01 in multiplet mode ) , we considered the CoLaus SNP pseudospectrum testable , and assumed the reference metabolite to be the metabolite underlying the SNP-feature association ., This resulted in nine testable pseudospectra , each with a single reference metabolite ., Metabomatching with default settings ( peak mode , χ2-scoring , and without decorrelation ) , and using the urine specific UMDB reference database , was successful for eight of the nine testable pseudospectra , ranking the reference metabolite first three times and in the top ten five times ( column P C X of Table 1 , detailed results in S1 Fig ) ., For the SOSTDC1 SNP , the pseudospectrum ( S1C Fig ) shows strong inflation across almost the entire chemical shift range , making metabomatching fail systematically ., Metabomatching in multiplet mode performed better overall ( column M C X ) , ranking the reference metabolites first six times and second twice , though the performance was qualitatively different only for the HPD SNP pseudospectrum , for which the testable association involved a different reference metabolite ( S1F and S1G Fig ) ., Decorrelation had little effect on rankings , in either mode , provided a shrinkage parameter λ greater than 0 . 1 was used ( results for λ = 0 . 5 in Table 1 columns P D X and M D X , other values of λ in S1 Table ) ., Z-scoring metabomatching properly ranked the reference metabolites for the pseudospectra characterized by the strongest associations , that is those for SNPs in AGXT2 , PYROXD2 , and SLC7A9 ., Pseudospectra with weaker associations fared worse , with Z-scoring metabomatching ranks significantly lower than their χ2-scoring counterparts , except for the UPS9 pseudospectrum ., For the SLC6A20 and SLC6A13 pseudospectra , the reference metabolite is outranked by a number of metabolites of spectra that obtain their score by matching a group of strongly correlated features ., Applying decorrelation reduces this correlation-based score , thereby significantly improving the rank of the reference metabolite in both peak- and multiplet-mode ( see S3 Fig ) ., Using the full HMDB spectral database ( S2 Table ) , metabomatching ranked the reference metabolites for PYROXD2 , PNMT , HPD markedly lower , due to stronger competition among the larger pool of candidate metabolites ., Using the UMRB or BMRB spectral databases ( S2 Table ) , metabomatching ranks the reference metabolite for PNMT lower , for PYROXD2 higher , but is otherwise comparable to UMDB or HMDB , respectively ., We then tested metabomatching on pseudospectra obtained in the urine mGWAS 20 in the SHIP study 30 ., NMR data were normalized , binned in 0 . 0005 ppm increments , then processed with FOCUS 31 ., The resulting untargeted metabolome contained 166 features for 3 , 861 individuals ., In addition , NMR data were manually annotated using Chenomx NMR Suite 7 . 0 ., The resulting targeted metabolome contained the concentrations of 59 metabolites for the same 3 , 861 individuals ., Having both metabolome features and metabolite concentrations in the same sample allowed for the direct comparison of SNP-metabolite association results via metabomatching with targeted metabolite quantification followed by association ., We considered the pseudospectrum of a SNP associating with p < 10−6 with both a metabolite and at least one feature contained in the metabolite spectrum testable ., This resulted in nineteen testable SNP-metabolite associations involving fourteen SNPs ., Because testing is in the same samples , and because of the higher sample size of the study , metabomatching results for SHIP pseudospectra are more nuanced than they were for CoLaus pseudospectra ., For the nine SNPs that associate with a single metabolite , metabomatching in default settings ranked the reference metabolite first five times , and in the top ten four times ( see Table 2 column P C X , detailed results in S2 Fig ) ., For the CPS1 and HPD SNPs , which associate with two metabolites each , metabomatching ranked one metabolite first , the second in the top ten , and 2-compound metabomatching ranked the reference metabolite pair first ., The pseudospectra for the three remaining SNPs are more complex ., While the NAT2 SNP only associates with formate , its pseudospectrum ( S2N Fig ) indicates the presence of additional associations , in both effect directions ., We therefore applied ±-2-compound metabomatching ( S2O Fig ) , which ranks a metabolite pair that includes formate first , in the β > 0 direction ., With associations with three reference metabolites , the PNMT SNP pseudosepctrum ( S2U Fig ) is too complex for metabomatching , or 2-compound metabomatching , to provide any of the reference metabolites as plausible candidates ., The SLC6A19 SNP pseudospectrum ( S2J Fig ) is similar to the PNMT SNP pseudospectrum , but with weaker associations ., Because the secondary associations are closer to the noise background , metabomatching still provides top ten ranks for the two reference metabolites ., 2-compound metabomatching , however , does not properly rank the reference pair ., Metabomatching in multiplet mode shows similar results for most SNPs ( column M C X ) ., However , for the CPS1 , XYLB , HPD SNPs , the multiplet ranges describing the spectra of the respective reference metabolites are wide ( between 0 . 16 and 0 . 28 ppm ) even though each range encloses only a single peak ., The resulting multiplet-mode neighborhoods have a higher number of degrees of freedom than their peak-mode counterparts , yet produce similar sum values ., This lowers the scores of the reference metabolites , which are then outranked by competing metabolites , particularly in 2-compound metabomatching ( S2D , S2E , S2G , S2S and S2T Fig ) ., Z-scoring metabomatching underperforms χ2-scoring overall ( columns P C Z and M C Z ) , yet Z-scoring ranks obtained for the reference metabolites are close to their corresponding χ2-scoring ranks for all but two pseudospectra ., For the CPS1 and DMGDH pseudospectra , the association of the lead feature is too weak to compensate for the associations of opposite effect direction of other features captured by the match sets of the corresponding reference metabolites ( see S2B and S2L Fig ) ., The resulting penalties incurred under Z-scoring produce low reference metabolite ranks ., FOCUS combines neighboring features into a single representative feature , obtained either by peak picking or by integration of the NMR curve in the neighborhood ., As a result , the effect on metabomatching ranks of correlation in feature neighborhoods is weaker because neighborhoods contain fewer features after FOCUS processing ., Correspondingly , ranks with decorrelation are essentially equal to ranks without decorrelation ( columns P D X , P D Z , M D X , and M D Z ) ., Using the full HMDB spectral database ( S3 Table ) , metabomatching ranked the reference metabolites for SLC6A20 , SLC7A9 markedly lower ., Using the UMRB or BMRB spectral databases ( S3 Table ) , metabomatching ranks the reference metabolites for SLC6A19 , SLC6A13 , and PNMT higher , but is otherwise comparable to UMDB or HMDB , respectively ., While not yet as widespread as the targeted approach , the untargeted approach to metabolome-wide genome-wide association studies has already shown compelling results ., Because it analyses all measured metabolome features , the untargeted approach more fully exploits experimental data and may discover genetically determined metabolites that were missed , because they eluded identification , by a targeted approach ., By focusing the identification effort on the comparatively few metabolites found to be genetically determined , the untargeted approach also presents the pragmatic advantage of shortening the path from spectral metabolome data to mGWAS results ., Metabomatching further reduces this identification effort , by combining genetic spiking information with spectral reference data to assign candidate metabolites to genetically associated metabolome features ., In addition , because identification through genetic spiking is not an in-sample procedure , metabomatching becomes of particular interest when applied in an mGWAS that combines untargeted and targeted approaches ., In such a combined mGWAS , metabomatching can both provide an independent line of evidence for in-sample identifications of metabolites , and inform on the identity of metabolites that were missed by the targeted approach because they eluded in-sample identification ., Naturally , while focus was placed here , and in previous applications of metabomatching , on pseudospectra resulting from genetic association with NMR features , metabomatching is not limited to genome-wide association studies ., Any trait that influences , or is influenced by , metabolome features produces an association pseudospectrum to which metabomatching can assign candidates ., Notably , metabolome-wide association studies , analyzing the effects of the metabolome on organismal traits , would similarly benefit from both the untargeted approach and metabomatching ., The performance of metabomatching is inherently linked to the strength of genetic spiking and the quality of spectral databases ., With increasing mGWAS sample sizes , and the continuing efforts to establish spectral databases that are more complete and better annotated , both conditions are expected to improve ., Metabomatching is therefore not only likely to become a valuable tool for exploring the links to metabolites of listed spectrum , but may also provide impetus to complete databases of spectral information for human metabolites , reducing instances where no good match can be found ., Metabomatching is written for Matlab and compatible with octave ., Documentation and code can be obtained from the metabomatching website http://www . unil . ch/cbg/index . php ?, title=metabomatching or GitHub ., Metabomatching is also available as a docker container , and within the metabolomics e-infrastructure PhenoMeNal http://phenomenal-h2020 . eu . | Introduction, Materials and methods, Results, Discussion | A metabolome-wide genome-wide association study ( mGWAS ) aims to discover the effects of genetic variants on metabolome phenotypes ., Most mGWASes use as phenotypes concentrations of limited sets of metabolites that can be identified and quantified from spectral information ., In contrast , in an untargeted mGWAS both identification and quantification are forgone and , instead , all measured metabolome features are tested for association with genetic variants ., While the untargeted approach does not discard data that may have eluded identification , the interpretation of associated features remains a challenge ., To address this issue , we developed metabomatching to identify the metabolites underlying significant associations observed in untargeted mGWASes on proton NMR metabolome data ., Metabomatching capitalizes on genetic spiking , the concept that because metabolome features associated with a genetic variant tend to correspond to the peaks of the NMR spectrum of the underlying metabolite , genetic association can allow for identification ., Applied to the untargeted mGWASes in the SHIP and CoLaus cohorts and using 180 reference NMR spectra of the urine metabolome database , metabomatching successfully identified the underlying metabolite in 14 of 19 , and 8 of 9 associations , respectively ., The accuracy and efficiency of our method make it a strong contender for facilitating or complementing metabolomics analyses in large cohorts , where the availability of genetic , or other data , enables our approach , but targeted quantification is limited . | Metabolome-wide genome-wide association studies aim to discover how genetic variation affects metabolome traits ., Such studies typically follow an acquire-identify-associate procedure: metabolome data are acquired experimentally , metabolites are identified in the experimental data and their concentrations quantified , and the metabolite concentrations are tested for association with genetic variants ., The untargeted approach follows instead an acquire-associate-identify procedure: the experimental data are binned into metabolome features , and the features tested directly for genetic association ., When the metabolome is measured by proton NMR spectroscopy , genetically associated features tend to correspond to peaks in the NMR spectrum of the underlying metabolites ., This inherent property of the untargeted approach acts as a genetic spiking which informs on the identities of involved metabolites ., Metabomatching is a method that uses genetic spiking information to identify the metabolite candidates , listed in a spectral database , most likely to underlie observed feature associations ., Here , we present the method and its software , and evaluate its performance . | protons, genome-wide association studies, medicine and health sciences, body fluids, nmr spectroscopy, urine, metabolomics, metabolites, genome analysis, molecular genetics, research and analysis methods, genomics, nucleons, molecular biology, physics, biochemistry, proton nmr spectroscopy, anatomy, nuclear physics, physiology, genetics, biology and life sciences, physical sciences, computational biology, metabolism, spectrum analysis techniques, human genetics | null |
journal.pcbi.1003380 | 2,013 | Time Scales in Epigenetic Dynamics and Phenotypic Heterogeneity of Embryonic Stem Cells | Embryonic stem cells ( ESCs ) are pluripotent having the ability to differentiate into a variety of lineages , while in suitable culture conditions they proliferate indefinitely by maintaining pluripotency ., These self-renewing ESCs are distinguished by the marker proteins including Sox2 , Oct4 and Nanog ( SON ) 1–4 ., SON are transcription factors ( TFs ) which directly or indirectly promote the expression of themselves by constituting an overall positive feedback network 5–10 , among which Nanog is an essential factor working as a gatekeeper for pluripotency 11 , 12 ., Here , a remarkable feature is the large cell-to-cell variation of the level of Nanog in the self-renewing isogenic population of ESCs 13–15 ., Since a distinct downregulation of Nanog is associated with the differentiation of ESCs into mesendoderm or neural ectoderm lineages 16 , the heterogeneous Nanog expression can be intimately related to the process of fate decision of individual cells 14 , 17 ., The molecular mechanism and biological implication of this phenotypic fluctuation of ESCs , however , have not yet been clarified ., In this paper we address this problem by constructing a model of the regulatory network of core genes in mouse ESCs ., One can figure out , at a glance , several scenarios which may explain the phenotypic heterogeneity ., A simple scenario relies on the possible enhancement of fluctuation of the signal received by a cell: Since the reception of factors such as leukemia inhibitory factor ( Lif ) by a cell is stochastic , it necessarily bears fluctuation , which might be enhanced through the signal cascade to stochastically activate Nanog 18 ., The other possible mechanism is based on the presumed self-activation of Nanog 6 , 19 , which may lead to the fluctuating pulsative expression of Nanog 14 , 20 ., With these mechanisms , however , another key factor , Oct4 , should also exhibit the large fluctuation since Oct4 is activated by the reception of Lif and the Oct4 expression is maintained through mutually activating interactions among SON ., Contrary to this expectation , the observed expression of Oct4 is rather homogeneous 14 , 17 ., A possible resolution of this inconsistency is to assume that some unknown factors which can bind to the Oct4 locus suppress fluctuation of the Oct4 expression 20 ., There has been , however , no direct experimental observation yet for the existence of such regulatory factors , and therefore , in this paper we look for the other mechanism without relying on this assumption ., For modeling the gene regulatory dynamics , not only the topological wiring diagram among genes but also the rates of reactions in the regulatory network should be quantified ., These estimated rates , however , have very different values depending on the type of reactions , and hence it is strongly desired to develop the theoretical framework to treat effects of coexistence of the distributed timescales 21 , 22 ., In simple bacterial cells , for example , the DNA-protein binding/unbinding is often much faster than the protein-copy number change , so that the fast DNA-state change can be regarded as equilibrated and the dynamical interference between the fluctuation of gene switching and the fluctuation of protein-copy number can be neglected ., By borrowing the wording from condensed-matter physics , this separation of fast and slow processes should be referred to as the “adiabatic” separation ., Theoretical studies have shown that when the adiabatic limit is not the case , the kinetic flow of the coupled stochastic dynamics of gene switching and protein-copy number change is described as “eddy current” 23 , which gives rise to a variety of unexpected dynamical effects in gene regulation 23–28 ., Indeed , it has been suggested that the transition of Bacillus subtilis into the competence period should be due to the non-adiabatic gene switching in the excitatory self-activating gene network 29 ., In eukaryotic cells , processes of gene switching are much more complex including the assembly of transcriptional apparatus ( TA ) 30–33 , the transition from the poised state of TA to the elongation state 34 , chemical modifications of nucleosomes 35–39 , and the structural reorganization of chromosomes 40–42 ., Such epigenetic change of the gene state can be much slower than the bacterial DNA state change , and their timescales are often comparable with or longer than the timescale of the protein-copy number change , so that the non-adiabatic effects should play significant roles in eukaryotic cells ., Many marker genes of ESCs have been identified 43 , among which SON regulate many other genes , and hence , the SON network has been regarded as the central network to maintain pluripotency 8 , 9 , 43 ., Models of the core SON network of ESCs have been developed 14 , 16 , 20 , 43 , 44 , but all of these models have been based on the assumption that the gene state is determined by the fast equilibrated binding/unbinding of TF to/from the gene locus: The assumption of the adiabatic limit has been adopted in all the previous models and the slow non-adiabatic switching dynamics has not been explicitly taken into account ., In this paper , we discuss ESCs by focusing on the non-adiabatic effects , the effects of slow epigenetic processes , and we propose a hypothesis that the non-adiabatic switching in the core gene-network explains the large fluctuation of Nanog expression ., By using the landscape picture , we discuss the roles of this non-adiabatic switching in the cell-fate decision of ESCs ., We first discuss ESCs in media containing Lif and other agents ., Lif activates c-Myc 60 , which activates SON by keeping the histone code of lineage-specific genes repressive 34 ., We simulate this culture by adopting the null rate for turning the histone-code active as for , and ( See Fig . 2 and Methods for the definition of parameters ) ., First simulated is the case that the formation/dissolution of TA is adiabatic with with ., As a typical value to satisfy this inequality , we use for all ., Distributions of the expression level of SON simulated with this parameterization are shown in Figs ., 3A and 3C ., We can see that the simulated distribution of the expression level of Nanog shows a single peak and the simulated distributions of Sox2 and Oct4 are double peaked at their finite values of expression level with some additional populations at the zero expression ., These features are different from those observed in experiments: Compared are the distributions of cell population in a culture plotted as functions of the expression level of SON ., The observed distribution of Oct4 is single peaked ( Fig . 3D ) 14 , the distribution of Sox2 is similar to that of Oct4 14 , and the observed distribution of Nanog shows two peaks ( Fig . 3D ) 11 , 14 ., The observed two-peak distribution of Nanog indicates that the fluctuation of Nanog is dominated by transitions between two states; the high-Nanog ( HN ) and low-Nanog ( LN ) states 11 , 33 ., The simulated Nanog distribution with the adiabatic TA formation/dissolution apparently disagrees with this observed two-peaked Nanog distribution ., The assumption of the adiabatic TA formation/dissolution with used in the above simulation is questionable when we consider the following features of Nanog expression: First , the TA of Nanog consists of the fairly large ( kb ) region of DNA 61 , which should make the formation/dissolution of TA a rather complex process ., Second , the allelic regulation of Nanog 51 indicates that the chromosome organization on the nuclear scale regulates the Nanog expression ., This observation is also consistent with the recent finding that the loci of genes of the pluripotent factors are spatially in proximity to the Nanog locus in an ESC-specific manner 62 , indicating that the nuclear scale organization of chromosomes is involved in the activation of Nanog in ESCs ., For such complex and spatially extended processes for TA at the Nanog locus , it should be reasonable to assume that the timescale of TA formation/dissolution is as long as the cell cycle period ., To find the plausible values for the rate of TA formation ( ) and the rate of TA dissolution ( ) at the locus , we performed a massive parameter search by generating more than 1 , 000 scattered points on the two-dimensional plane of and with ., The score for each of generated parameter sets was calculated by averaging 10 , 000 trajectories simulated with the corresponding parameter set , where the score is the number of the experimentally observed features that the simulated data reproduced ., The features used to count the score include ( 1 ) bimodality of the distribution of expression level of Nanog , ( 2 ) the ratio of the copy-number at the HN state to that at the LN state , ( 3 ) the ratio of the peak height at the HN state to that at the LN state , ( 4 ) the single-peak distribution of expression level of Oct4 , and so on ., The score calculated in this way is plotted in Fig . 4 for 1 , 125 parameter sets ., See Massive parameter search subsection in Methods section for more details on the definition of the score ., Search of the other parameter set is shown in Fig . S1 ., Results of Fig . 4 indicate that the normalized rate of TA formation should be around and the normalized rate of TA dissolution should be for to reproduce the experimentally observed heterogeneous expression levels in ESCs ., Here , the result of was needed to reproduce the observed feature that the HN peak height is larger than the LN peak height ., Since the biologically reasonable lower bound of is the frequency of cell cycle , we use the lowest allowed value of in the subsequent analyses by keeping for , and Oct4 ., The precise values of other parameters are explained in Parameters subsection in Methods section ., The simulated distributions of SON with this non-adiabatic switching of Nanog are shown in Figs ., 3B and 3C ., The simulated width of peaks is narrower than the observed one because in simulation the extrinsic noise due to the cell-cycle oscillation and the fluctuating reception of Lif are neglected for simplicity ., The overall features of the distributions , however , agree well with the experimental data 14: Nanog shows a clear two-peak distribution and the Oct4 distribution has an asymmetric single major peak ., Shown in Figs ., 5A and 5B is the temporal change of distributions of Nanog calculated by starting from the ensemble of cells either in the HN or LN state at Day, 0 . Within several days , the single-peaked distribution of cells in either of the HN or LN state recovers the two-peak features , which reproduces the experimentally observed temporal relaxation 11 , 14 ., This relaxation indicates that ESCs show dynamical transitions between HN and LN states with timescale of a few days ., The agreement between the observed and simulated timescales of transitions between HN and LN states indicates the validity of the small for the slow switching at the Nanog locus , and hence the data in Fig . 5 should rule out the other hypothetical models which can yield a bimodal Nanog distribution but with the large ., A possible origin of the slow non-adiabatic switching of Nanog is the large scale chromatin reorganization in the formation/dissolution of TA of Nanog ., This assumption of slow switching explains the observed two-peak distribution and the dynamical transition of the expression level of Nanog , and is also consistent with the single-peak distribution of Oct4 ., Thus , the assumption of the slow non-adiabatic switching of Nanog explains the observed phenotypic heterogeneity of ESCs ., Given the consistent model for the heterogeneity of ESCs , it is interesting to analyze how cells initiate differentiation ., To simulate cells that can differentiate , the rate to turn the histone-code active , , is increased to have a finite value for and ., for is also turned finite but kept small because in embryo , the distinct expression of Cdx2 is the event prior to the formation of inner cell mass from which ESCs are prepared , so that it is plausible to assume that the methylated DNA or the collective action of regulating factors inhibits the histone code of Cdx2 from being active in ESCs ( See subsection Parameters in Methods ) ., Examples of trajectory simulated with this parametrization are shown in Figs ., 6A and 6B ., The trajectory in Fig . 6A wanders around several transient states but neither Cdx2 nor Gata6 dominates during this wandering: Cells are jumping among the states by maintaining the features of ESC ., In Fig . 6B , on the other hand , the trajectory escapes from the ESC states to reach the Gata6 dominant state which is a gateway to the primitive endoderm lineage ., In both Figs ., 6A and 6B , the trajectories are not the continuous drifts but consist of sojourns and jumps ., This feature allows us to represent each trajectory as a sequence of transitions among “cell states”: Using the feature that the copy number of each factor , , shows a multiple-peak distribution ( Fig . S2 ) , we divide each distribution into a few parts , each of which is named in an abbreviated way as HN ( high-level Nanog ) , MN ( middle-level Nanog ) , LN ( low-level Nanog ) , S ( high-level Sox2 ) , LS ( low-level Sox2 ) , etc ., The thresholds used to divide the distributions are summarized in Table, 1 . Then , cell states are defined by thus discretized distributions and also by a set of the histone states ., The trajectory is regarded as a sequence of transitions among those cell states ., With this coarse-grained representation of trajectories , the mean waiting time for transition from the th to the th cell states can be estimated as , and the mean transition rate is defined by ( See subsection Transition diagram in Methods for the detailed explanation on ) ., In the case that the trajectory stays for a long duration at each cell state to erase its dynamical memory , this coarse-grained dynamics can be regarded as Markovian , or in other words , the transition probability from the th to th states is not affected by which state the trajectory visited before reaching the th state ., It is suggested from Figs ., 6A and 6B that the trajectories stay at each cell state long enough to show many oscillations during the stay , but the more quantitative test should be necessary to judge whether the coarse-grained dynamics is indeed Markovian or not ., We leave this test as a future problem and proceed further in this paper to show how the transition diagram and the landscape view capture the important features of transitions among cell states ., Drawn in Fig . 6C is the diagram of transitions among thus defined cell states , where the value of is shown on the link from the th to th cell states ., In Fig . 6C the cell states in which all of Sox2 , Oct4 , and Nanog ( SON ) are expressed are regarded as the pluripotent states ( or the ESC states ) though the level of Nanog fluctuates largely among these states and sometimes Cdx2 or Gata6 coexists with SON ., These ESC states are connected by loops of transitions and hence the cells wander among ESC states to wait for a chance to escape from the ESC states ., Trajectories that have escaped from the ESC states go through the network of transitions among the intermediate states in which one or two of SON are lacking ., From these intermediate states , cells reach the state in which Gata6 dominates ., In some cell states , Cdx2 appears as fluctuation but the small value of prevents Cdx2 from dominating the state ., It should be interesting to examine the validity of these predictions with the experimental observations: By quantifying the expression level of important factors , we will be able to define cell states from the experimental data ., Then , we can check whether the differentiation is the process of jumping among these states ., Though there is a global trend of kinetic flows from the ESC states to the differentiated states , the predicted pathways are not single but comprise the network of flows ., It should be important to compare the predicted distribution of pathways as in Fig . 6C with the distribution of pathways experimentally observed by following the fate of individual cells in the culture ., To analyze dynamics of differentiation , the epigenetic landscape that underlies transitions among cell states should provide a useful perspective 63 , 64 ., Here , the landscape is derived from the transition diagram by using the analogy with the free energy surface in equilibrium dynamics ., In equilibrium dynamics , by using the transition-state theory formula , the rate of transition from to th states should be proportional to where , and and are the dimension-less free-energy like quantity at the th state and at the transition state between the and th states , respectively ., We use this analogy to equilibrium dynamics by fitting the calculated rates to this transition-state theory formula to obtain the free-energy like quantity and ., When the transition diagram has a tree-like structure without a loop , we can determine of each state one by one by fitting the simulated rates to this formula ., We use this analogy with equilibrium dynamics as far as possible to draw the landscape of non-equilibrium transitions ., This method of fitting , however , apparently breaks down when the transition network contains one or more loops: When the transition network contains a loop , for example , we may attempt to determine of states in the loop one by one by starting from the th state in the loop with the landscape value , but at the end of traverse along the loop , we return to the initial th state with a different value of from the original ., In this way , the fitting to the transition-state theory formula is inconsistent along loops ., This inconsistency can be resolved when we explicitly consider the non-equilibrium feature of dynamics by introducing the curl flux of transition kinetics 65–68 ., Thus , the kinetic process along each loop can be expressed by the combination of the landscape and a kinetic flow curling along the loop ., Transitions , therefore , are described by the combined representation of landscape and non-equilibrium curl flux ., An example of a looped diagram having curl fluxes is shown in Fig . 7 ., From of this diagram , the free-energy like quantity at the th cell state and at the barrier between the and th cell states are calculated for , and , and curl fluxes and are obtained simultaneously ., See subsection Epigenetic landscape in Methods for the explanation on how to calculate , , and from of Fig . 7 ., In Fig . 8 the landscapes and curl fluxes calculated from the simulated in the differentiation processes are shown on the two-dimensional plane with the coordinates of and ., Here , is the label of the discretized expression level of the th factor , which is defined to have the larger value for the higher expression level ., and , therefore , represent the degree of closeness to the trophectoderm and primitive endoderm lineages , respectively ., The precise values of are chosen for obtaining good visibility of Fig . 8 , and are explained in Table, 1 . In Fig . 8 , the calculated and are plotted by assigning and for and , and and are interpolated by a smooth surface in the two-dimensional space of and ., The landscape corresponding to the diagram of Fig . 6C is shown in Fig . 8A ., We see that the ESC states distribute on a flat basin in the region of small and : ESCs wander around this basin driven by both the fluctuations satisfying the balance between the forward and reverse transitions and the kinetic flow of curl flux that breaks the balance ., ESCs start differentiation as they move along the valley stretching toward the Gata6 dominant state ., Transitions among intermediate states along this valley are also accompanied by the weak non-equilibrium curl flux ., In Figs ., 8B and 8C , the artificial depletion of Oct4 is simulated with the decreased rate of synthesis of Oct4 ., Since Oct4 and Cdx2 work in an antagonistic way , the depletion of Oct4 results in the stronger expression of Cdx2 , which leads ESCs to the trophectoderm linage: With the decrease of the rate of Oct4 synthesis to 25% ( Fig . 8B ) and 10% ( Fig . 8C ) of the value in Fig . 8A , the landscape changes its shape by extending the valley toward the Cdx2 dominant state ., In Fig . 8B two valleys to primitive endoderm and trophectoderm coexist with the curl flux on the basin of ESC states remaining , and in Fig . 8C the valley to trophectoderm dominates ., These results are consistent with the experimentally observed induction of the trophectoderm lineage through the reduction of Oct4 31 ., Shown in Figs ., 8D–8F are landscapes calculated with the assumption of the fast Nanog switching: ., With this fast Nanog switching , the flat basin of the ESC states disappears , the curl flux in ESC states becomes localized , and ESCs quickly differentiate toward primitive endoderm ( Fig . 8D ) ., The curl flux on the ESC basin , therefore , originates from the slow Nanog switching ., In other words , the eddy current associated with the non-adiabatic switching 23 manifests itself in the curl flux on the epigenetic landscape ., Difference between the slow and fast Nanog switching becomes more evident upon the reduction of Oct4 ( Figs . 8E and 8F ) ., With the fast Nanog switching , two valleys do not represent the distinct cell fate but they are directly connected to each other by the frequent transdifferentiation ( Fig . 8F ) ., This obscured differentiation arises from the averaged intermediate amount of Nanog synthesis under the fast Nanog switching ., With the intermediate level of Nanog , the alleles of the lineage-specific genes tend to take the intermediate histone code as and or and ., This intermediate level of activation of both Cdx2 and Gata6 increases the frequency of the transdifferentiation ., With the slow Nanog switching , on the other hand , the histones of Gata6 and Cdx2 become either active with or repressive with , and such a clear-cut histone switching decreases the probability of the mixed expression of Cdx2 and Gata6 ., In this way the simulated results suggest that the distinct cell fate decision is based on the slow Nanog switching , so that the phenotypic heterogeneity of ESCs is necessary for the stable differentiation ., The present quantification of epigenetic landscapes showed that the model naturally reproduces the observed differentiation to primitive endoderm 10 ., The model also reproduces the induced differentiation to trophectoderm observed when the Oct4 expression is artificially suppressed 3 ., It should be interesting to further examine possibility of the predicted transdifferentiation due to the fast Nanog switching ., We developed a model of epigenetic dynamics and proposed a hypothesis that the timescale of formation/dissolution of TA decisively affects the self-renewal and differentiation of mouse ESCs ., These effects can be checked experimentally by artificially varying the timescale of formation/dissolution of TA ., The slower rate of formation/dissolution of TA for Oct4 , for example , should give rise to the multi-peak distribution of Oct4 , which should also affect the epigenetic landscape and non-equilibrium curl fluxes on the landscape ., Further important is the application of the present ideas to engineering differentiation ., Overexpression or repression of specific genes should alter the epigenetic landscape and curl fluxes , so that the calculation and observation of landscape and curl fluxes should provide a guideline for designing the process of cell differentiation ., An intriguing question is the effect of variation of the number of working alleles in a cell ., In the present simulation , following the report for the single non-silenced Nanog allele in each ESC 51 , only the single Nanog locus was considered in the simulation , which explained the bimodal Nanog distribution when the Nanog switching was slow ., Assuming that both two alleles are working independently owing to the invalidated allelic regulation , we have three peaks in the Nanog distribution corresponding to the ‘high-high’ , ‘high-low’ and ‘low-low’ levels of expression for two alleles of Nanog with the slow Nanog switching as shown in Fig . S3 ., This predicted three-peak distribution could be experimentally tested in ESCs , though the more careful investigation is needed on the possible correlation between the allelic regulation and the regulation of the timescale of gene switching ., The core part of the network relations among genes in the present paper was built from the experimental observations , but there are experimental suggestions still not taken into account in the present model ., For example , a recent report suggested the auto-repression of Nanog 45 ., This suggested interaction can affect the transition dynamics between the HN and LN states , which should be examined by simulation ., The validity of the assumptions used in the present modeling of epigenetic dynamics should be checked by examining how the results are modified when the model is further extended ., In the present model , three processes having the different timescales were considered; TF binding/unbinding , TA formation/dissolution , and the histone code modification ., Each of these processes consists of multiple sub-processes , and therefore if the model is extended with the finer resolution , the involved timescales should have more variety 69 ., The TA formation/dissolution , for example , may involve assembly of mediators and RNA polymerase , phosphorylation of these factors , chromatin looping , and the large scale change in the chromosome positioning in nucleus ., In the present model , we treat them in a coarse-grained manner by representing the TA state with which takes a value between 0 and 1 ., By treating these multiple processes explicitly , we may be able to construct a more quantitative model that can be compared with experiments in more details , and the validity of the level of coarse-graining in the present model could be checked through such comparison ., We should stress , however , that the main conclusions on the importance of design of timescales of regulations and the usefulness of combined representation of landscape and non-equilibrium curl fluxes do not depend on the molecular details ., Indeed , the simplified mathematical or statistical physical models to capture the essential features of landscapes and curl fluxes should be useful ., The dynamical systems models , for example , emphasize the importance of oscillations in the gene network 70 , which conforms with the view presented here on the importance of rotating curl fluxes ., Another important direction to improve the present model is to take into account the core genes that guide ESCs to primitive ectoderm which further differentiates into the primary germ layers ., To develop the reliable models , the effects of cell-cell communication and cell cycles should be also taken into account ., Especially , the cell-cell communication should play important roles to stabilize the cell type of colony of interacting cells 70 , 71 ., The model developed in the present paper was based on the assumption that the partial effect of cell-cell communication is implicitly taken into account by the mutual inhibition between Cdx2 and Gata6 ( See Methods section ) ., In order to analyze the differentiation process more quantitatively , the model needs to be extended to explicitly treat the effects of cell-cell communication ., Those more elaborate models , together with the simplified statistical mechanical models , should reveal the rich phenomena in ESCs and differentiation processes ., The model consists of interactions among six genes ., Those interactions are inferred from the experimental data , which are complemented with various levels of assumptions as explained below ., In the following , the assumptions used are categorized into Level A , Level B , Level C , and Others ., The aim of the present study is not to claim the validity of those assumptions , but to clarify the mechanisms of epigenetic dynamics by using a set of biologically consistent assumptions ., The interactions considered in Fig . 1 were inferred from the discussions below , which are numbered in the same way as interactions designated in Fig . 1: Level A . Microarray or other genetic experimental techniques revealed the correlation or anti-correlation between expression levels of two genes , and the chromatin immunoprecipitation or other biochemical data showed the binding of one factor to the locus of the other gene ., These data support the assumption that the transcription factor ( TF ) synthesized from one gene directly regulates the other gene ., The Level A assumptions give the backbone of the present model of the regulatory network ., 1 . Each of Oct4 , Sox2 and Nanog loci has the Oct/Sox enhancer region 7 , 72 , on which Oct4 and Sox2 bind together to form the Oct4-Sox2 complex to activate Oct4 , Sox2 , or Nanog 4 , 7 , 72 ., There are two possible ways of binding though they are not mutually exclusive; The Oct4-Sox2 complex is formed before they bind to DNA , or Oct4 and Sox2 bind to the adjacent sites of DNA to form the complex after binding ., These two ways of binding are different in their cooperativity in the binding process ., However , since the cooperativity of binding is masked by the cooperative formation/dissolution of transcription apparatus ( TA ) in the present model , these two ways of binding do not give significant difference in the switching behavior ., We use , for simplicity , the latter assumption of forming complex after binding to DNA , but represent the effects of complex formation by assuming that the binding of either one of Oct4 or Sox2 is not enough but the binding of both two factors are needed for forming TA ( We assume that the formation of the Oct4-Nanog complex is another route to form TA ) ., 2 . Gcnf binds to the Oct4 and Nanog loci to repress them 48 ., 3 . The Oct4-Cdx2 complex represses both Oct4 and Cdx2 47 , 73 ., 4 . Nanog binds directly to the Gata6 locus to repress it 13 ., 5 . Because the binding of Oct4 to the Nanog locus is necessary for forming the higher order structure of chromatin at the Nanog locus 61 and the binding site of Oct4 is adjacent to the binding site of Nanog at the Nanog locus 6 , we expect that the Oct4-Nanog complex formed on the chromatin is necessary for building the TA of Nanog ., 6 . Nanog promotes the expression of Oct4 5 and both Nanog and Oct4 directly bind to the Oct4 locus 6 ., Because the binding site of Oct4 is in proximity of the binding site of Nanog at the Oct4 locus 6 , we assume the promotion of the formation of TA of Oct4 through the binding of Oct4-Nanog complex on the Oct4 locus ., 7 . Nanog is suggested to promote expression of Sox2 8 , 74 and both Nanog and Oct4 directly bind to the Sox2 locus 6 ., Because the binding site of Oct4 is in proximity of the binding site of Nanog at the Sox2 locus 6 , we assume that the formation of TA of Sox2 is promoted by the binding of Oct4-Nanog complex on the Sox2 locus ., Level B . Genetic experimental data showed the correlation or anti-correlation between expression levels of two genes , but the direct evidence for the physical interactions between two genes are not yet obtained ., In this case , the interactions can be indirect through the other unidentified factors ., Even in that case , we may assume the hypothetical direct interaction between two genes in the model ., Such assumption is reasonable in the coarse-grained model , in which the multiple detailed molecular processes are summarized into one process ., 8 . Excess expression of Oct4 reduces the expression level of Nanog 75 ., We assume in the model that the Nanog locus has multiple binding sites of Oct4 and the occupation of the part of those sites is necessary for the formation of TA , but the occupation of all sites increases th | Introduction, Results, Discussion, Methods | A remarkable feature of the self-renewing population of embryonic stem cells ( ESCs ) is their phenotypic heterogeneity: Nanog and other marker proteins of ESCs show large cell-to-cell variation in their expression level , which should significantly influence the differentiation process of individual cells ., The molecular mechanism and biological implication of this heterogeneity , however , still remain elusive ., We address this problem by constructing a model of the core gene-network of mouse ESCs ., The model takes account of processes of binding/unbinding of transcription factors , formation/dissolution of transcription apparatus , and modification of histone code at each locus of genes in the network ., These processes are hierarchically interrelated to each other forming the dynamical feedback loops ., By simulating stochastic dynamics of this model , we show that the phenotypic heterogeneity of ESCs can be explained when the chromatin at the Nanog locus undergoes the large scale reorganization in formation/dissolution of transcription apparatus , which should have the timescale similar to the cell cycle period ., With this slow transcriptional switching of Nanog , the simulated ESCs fluctuate among multiple transient states , which can trigger the differentiation into the lineage-specific cell states ., From the simulated transitions among cell states , the epigenetic landscape underlying transitions is calculated ., The slow Nanog switching gives rise to the wide basin of ESC states in the landscape ., The bimodal Nanog distribution arising from the kinetic flow running through this ESC basin prevents transdifferentiation and promotes the definite decision of the cell fate ., These results show that the distribution of timescales of the regulatory processes is decisively important to characterize the fluctuation of cells and their differentiation process ., The analyses through the epigenetic landscape and the kinetic flow on the landscape should provide a guideline to engineer cell differentiation . | Embryonic stem cells ( ESCs ) can proliferate indefinitely by keeping pluripotency , i . e . , the ability to differentiate into any cell-lineage ., ESCs , therefore , have been the focus of intense biological and medical interests ., A remarkable feature of ESCs is their phenotypic heterogeneity: ESCs show large cell-to-cell fluctuation in the expression level of Nanog , which is a key factor to maintain pluripotency ., Since Nanog regulates many genes in ESCs , this fluctuation should seriously affect individual cells when they start differentiation ., In this paper we analyze this phenotypic fluctuation by simulating the stochastic dynamics of gene network in ESCs ., The model takes account of the mutually interrelated processes of gene regulation such as binding/unbinding of transcription factors , formation/dissolution of transcription apparatus , and histone-code modification ., We show the distribution of timescales of these processes is decisively important to characterize the dynamical behavior of the gene network , and that the slow formation/dissolution of transcription apparatus at the Nanog locus explains the observed large fluctuation of ESCs ., The epigenetic landscapes are calculated based on the stochastic simulation , and the role of the phenotypic fluctuation in the differentiation process is analyzed through the landscape picture . | physics, biochemical simulations, gene networks, statistical mechanics, gene regulation, gene expression, biophysics theory, molecular genetics, genetics, biology, computational biology, biophysics | null |
journal.pntd.0004050 | 2,015 | Influence of Genetic Ancestry on INDEL Markers of NFKβ1, CASP8, PAR1, IL4 and CYP19A1 Genes in Leprosy Patients | Leprosy is an insidious infectious disease caused by the obligate intracellular bacteria Mycobacterium leprae that affects the skin and peripheral nerves , causing a chronic granulomatous infection 1 ., Leprosy patients may be classified in two major groups , based on clinical manifestations using a simple system introduced by the WHO ( World Health Organization ) in 1982 ., Paucibacillary ( PB ) is the primary characteristic of Tuberculoid ( TT ) leprosy and is characterized by a few lesions and scarce bacilli , and Multibacillary ( MB ) is the primary characteristic of anergic Lepromatous ( LL ) leprosy ., From an epidemiological perspective , the situation in Brazil is critical because , along with India and Indonesia , it has the highest rate of new cases detected worldwide 2 , 3 , 4 ., In addition to the system introduced by WHO in 1982 , the use of histological and immunological criteria as described by Ridley-Jopling further improves definition of Borderline cases ., According to this classification , TT ( tuberculoid-tuberculoid ) patients , who have the PB type , exhibit a strong cellular immune response ( CIR ) mediated by Th1 , and a negative skin smear test ., In contrast , LL ( lepromatous-lepromatous ) patients have a weak or absent CIR and a highly positive skin smear associated to an humoral immune response ., In the middle of this spectrum are a large number of borderline patients , which together with LL comprise the MB pole , with symptoms varying from weak to strong CIR and negative to positive skin smears 5 , 6 ., The regulation of the host immune response and manifestation of disease clinical between types PB ( better ) and MB ( severe ) involves cytokine and others mediators produced by various subtypes of T cells ., In PB , an inflammatory immune response is mediated by Th1 cells that express pro-inflammatory interleukins that stimulate macrophages and phagocytosis mechanisms to inhibit bacillary growth and kill mycobacteria 2 , 7–9 ., On the other hand , MB patients have an intense Th2 immune response with production of anti-inflammatory cytokines in addition to the specific anti-PGL-1 ( phenolic glycolipid 1 ) antibody ., This mechanism does not block bacillary growth and contributes to the host’s inability to resist the development of severe disease 2 , 8 , 9–11 ., Recent studies have investigated genetic markers , usually innate immune response genes , as possible susceptibility factors for leprosy because the SNPs in these genes can modulated the host immune response and consequently lower host resistance to bacillus growth 6 , 12 , 13 ., However , few studies have investigated INDEL polymorphisms ( insertion-deletion ) in immune response genes in leprosy ., Moreover , such polymorphisms present interesting features as genetic markers because, i ) INDELs are spread throughout the human genome ,, ii ) INDELs derive from a single event ( they do not present homoplasy ) ,, iii ) small INDELs can be analyzed using short amplicons , which improves amplification of degraded DNA and facilitates multiplexing reaction ,, iv ) INDELs can create abrupt changes in the normal function of the gene and, v ) INDELs can be easily genotyped using a simple dye-labeling electrophoretic approach 14 ., The current study select eight INDEL in seven genes ( CYP19A1 , NFKβ1 , IL1α , CASP8 , UGT1A1 , PAR1 , CYP2E1 , and IL4 ) , which have relation with the immune response modulation in leprosy patients , beside literature that demonstrate these molecular markers like functional polymorphisms that alter transcriptional activity of the gene , and consequently the immunological phenotype against the bacilli ., Additionally these INDELs can be able to contribution to construction a possible panel of susceptibility markers ., However , from the genetic point of view , Brazil is recognized as having one of the most heterogeneous populations in the world , with important genetic information being contributed by three main continental groups , Europeans , Africans and Amerindian , resulting in a genetically very diverse modern Brazilian population 15 ., Therefore , analysis of genetic markers in complex diseases may result in spurious results due to population substructure 16 , and it is important to perform the genomic ancestry control , especially in populations with a high degree of interethnic admixture 14 ., The objective of this study was to investigate eight INDEL polymorphisms in seven genes involved in modulation of the host immune response , including CYP19A1 rs11575899 , NFKβ1 rs28362491 , IL1α rs3783553 , CASP8 rs3834129 , UGT1A1 rs8175347 , PAR1 rs11267092 , and CYP2E1 INDEL 96pb , besides one VNTR ( variable number tandem repeat ) of 70 bp on intron 3 of IL4 rs79071878 in a group consisting of 141 leprosy patients and 180 healthy individuals , to identify possible susceptibility markers of leprosy and evaluate the influence of genetic ancestry on disease risk ., The project was approved by the Pará Federal University ethics committee ( N° 197/07 ) ., We investigated 141 leprosy patients who attended the Dr Marcello Candia Reference Unit in Sanitary Dermatology of the State of Pará ( UREMC ) , in Marituba , Pará , Brazil between January 2008 and December 2009 ., All patients were informed about the study before they signed informed consent forms ., Since 2002 , UREMC registered between 308 and 472 leprosy patients ( mean: 408 cases per year ) ., Of the 765 leprosy cases registered in 2008 and 2009 alone , 141 ( 18 . 43% ) were randomly selected for this study ., These patients were divided according to Ridley-Jopling classification 5 into Paucibacillary ( TT: PB 31 ) and Multibacillary ( BT , BB , BL and LL: MB 110 ) groups ., A total of 180 healthy individuals who were unrelated , without leprosy or other chronic diseases and from the same geographic area as each other were chosen for the control group ., Leprosy patient’s descriptions were made previously 6 ., These subjects were asked to participate in the study after being informed about the study objectives and signing informed consent forms ., DNA extraction was performed as previously described by phenol-chloroform method 6 , 17 ., The DNA concentration was determined by spectrophotometry ( Themo Scientific NanoDrop 1000 , NanoDrop Technologies , Wilmington , US ) ., Individual interethnic admixture was estimated using a panel of 48 ancestry informative markers ( AIMs ) as previously described 6 , 14 ., The allelic frequencies between healthy individuals and leprosy patients and between PB and MB patients were estimated by gene counting ., Deviation from the Hardy-Weinberg equilibrium was assessed using chi-squared tests , using the Arlequin v3 . 5 software 18 , and p-value of HWE was corrected by Bonferroni methods ., Differences between leprosy patients and healthy individuals and between PB and MB patients with respect to age , gender and genetic ancestry were estimated using Student’s t-Test , Fisher’s exact test and Mann-Whitney tests , respectively ., The association of markers between groups was analyzed by logistic regression tests , all the test were corrected by FDR ( False Discovery Rate ) method , and all tests were performed using the statistical package under R calculation ., A two-tailed p-value < 0 . 05 was considered statistically significant ., The individual contributions of European , African and Amerindian genetic ancestry were estimated using the STRUCTURE 2 . 3 . 3 program assuming three parental populations ( European , African and Amerindian ) , a burn-in period of 200 , 000 , and 200 , 000 Markov Chain Monte Carlo repetitions after burn-in 16 ., The differences in allelic frequencies between leprosy cases and the healthy individuals for markers analyzed following an adjustment for population stratification was performed using the STRAT software program with 10 , 000 simulations 16 ., The data of clinical and demographic distribution of leprosy patients and healthy individuals is shown in Table, 1 . The mean age was higher in healthy individuals ( 55 . 7±12 versus 43 . 3±21 , p<0 . 001 ) , and male patients were more frequent among leprosy patients ( 97 68 . 8% versus 65 36 . 1% , p<0 . 001 ) ., Analysis of ethnicity showed that the mean frequency of Africans was higher among leprosy patients ( 0 . 284 versus 0 . 236 , p<0 . 001 ) and Europeans were more frequent in healthy individuals ( 0 . 461 versus 0 . 427 , p = 0 . 004 ) ., The frequencies of INDELs for the eight ( 8 ) genes analyzed in leprosy patients and healthy individuals are show in Table, 2 . For the polymorphism in IL4 ( VNTR of 70 bp ) , only two alleles were identified in the sample ., One allele had two repeats of 70 bp ( allele A1 ) and the other had three repeats of 70 bp ( allele A2 ) , suggesting theses alleles are biallelic markers ., All the polymorphisms analyzed were according to the Hardy Weinberg equilibrium , therefore the association analysis were performed with regression logistic test and differences in allelic frequencies were corrected by frequencies of ancestry markers informative ., When the INDELs were analyzed by logistic regression , the genes NFKβ1 and PAR1 showed statistically significant differences associated with the presence of the DEL allele ( p = 0 . 016 and p = 0 . 022 , respectively ) and both were associated like protection factors to not developing the disease ( ORIC95% = 0 . 500 . 27–0 . 88 and ORIC95% = 0 . 350 . 14–0 . 86 , respectively ) , for these genes was found a dominance effect DEL allele , that increase your protection capacity in general population ., The CASP8 showed significant differences associated with the presence of the DEL/DEL homozygous genotype and was associated with a risk factor for leprosy development ( p = 0 . 017; ORIC95% = 2 . 331 . 16–4 . 69 ) ( Table 3 ) ., The analysis of allele frequency differences was then corrected for the influence of genetic ancestry on population structure , and the results showed that the DEL allele of PAR1 gene and the allele A1 of IL4 is more frequent in healthy individuals ( p = 0 . 018 and p = 0 . 019 , respectively ) ( Table 3 ) , these results shown the importance of statistical correction in admixture population , in order to exhibit differences covert by structure population ., Table 4 summarizes the clinical and demographic characteristics of leprosy patients grouped according to clinical manifestation in PB ( Paucibacillary ) and MB ( Multibacillary ) groups , and the only significant difference was observed for age ( p = 0 . 003 ) , with a higher mean age in MB patients ( 45 . 7±22 versus 34 . 9±15 ) ., When the INDELs were analyzed by logistic regression , NFKβ1 showed significant differences like risk factor associated with the presence of the allele DEL in MB patients ( p = 0 . 024; ORIC95% = 2 . 641 . 13–6 . 19 ) , of contradictory way the dominance effect of DEL allele seem protect against the development of leprosy , but when the disease is established your effect seem inefficient to combat to bacilli ., PAR1 showed significant differences associated with the presence of homozygous DEL/DEL genotype in PB patients ( p = 0 . 031; ORIC95% = 0 . 410 . 17–0 . 96 ) ( Table 5 ) ., The analysis of allele frequency differences were corrected for population structure and showed that the DEL allele of CASP8 is more frequent in PB patients ( p = 0 . 003 ) , while the DEL allele of CYP19A1 is more frequent in MB patients ( p = 0 . 007 ) ( Table 5 ) ., Fig 1 shows the OR ( odds ratio ) values obtained from leprosy patients and healthy individuals within groups having distinct level of ancestry composition ., The figure shows that greater frequency of European ethnic between the groups ( leprosy patients and healthy individuals ) , higher is the risk for developing leprosy , while the smaller the frequency of the African ethnic , lower is the risk for developing leprosy ., No statistically significant values were obtained for the analysis of the Amerindian group ., These results are better understood on frequencies distribution , according with range of ancestry contribution ( S1 Table ) ., For African ancestry 99 . 4% of health individual is closed between 0% and 50% of African contribution ( range that have p<0 . 05 on Fig 1 ) , moreover the contribution range of 10% to 30% is closed 81 . 7% of health individual , in this range the Fig 1 have more decline of OR value , that showed the higher protection effect of African ancestry ., To European ancestry , 61 . 7% of leprosy patients is closed between 40% to 80% of European contribution , while 87 . 2% of health individuals is closed between 0% to 50% of European contribution ., Additionally , for the contribution range between 60% to 80% we observed 17% of all patients , while no healthy individual was observed this range , these data show that leprosy patients have higher European contribution compared with healthy individuals ., Take together the Fig 1 and S1 Table shown that to leprosy patients of an admixture population , like Brazil , African ethnic generates protection against the development of disease , and the opposite is also truth for European ethnic ., NF-κB belongs to family of protein transcription factors that modulate many inflammatory processes ., In the resting state , IκBα ( inhibitor of NF-kβ activity ) sequesters NF-κB in the cytoplasm and prevents its activity , but in response to specific stimuli , IkBα is ubiquitinated and degraded allowing NF-kB to migrate to the nucleus and stimulate the transcription of proinflammatory genes 19 , 20 ., The allele DEL ( rs28362491 ) has been shown to be associated with a decrease of transcriptional activity of variety genes of immune response 21 and with auto immune disease such as Systemic Sclerosis 22 and lupus erythematosus 23 ., The role of NF-kβ in leprosy is not clear , and studies linked to expression of NF-kβ have suggested that lower expression is common in leprosy patients 24 , 25 ., Our results suggest that the DEL carries genotype induces protection against leprosy ( Table 3 ) , although a comparison of PB and MB patients also suggests that DEL behaves like a risk factor for the development of the severe clinical form of MB ( Table 5 ) ., Because the transcription of NF-kβ is mediated by specific stimuli , such as the presence of M . leprae 24 , it is conceivable that the presence of DEL confers risk to MB leprosy ., PAR1 is a receptor of the PAR family of proteins that belong to a unique group of G protein—coupled receptors ., In particular , PAR1 protein is present in a variety of cells like platelets , endothelia , epithelial , neurons , fibroblasts , smooth muscle , leukocytes and tumor lines 26 ., This receptor has been shown to be involved in many natural physiological processes , that involve inflammation like the systems cardiovascular , respiratory and central nervous and in embryogenesis , cancer and inflammation 27 ., PAR1 suppresses T helper type 1 ( Th1 ) and T helper type 17 ( Th17 ) cells and the secretion of IL-12 and IL-23 , thereby resulting in the inhibition of pro-inflammatory responses 28 ., The allele of insertion ( INS ) of INDEL studied ( rs11267092 ) has been shown to increase gene transcription 29 and therefore , it is a risk allele for leprosy ., Our results suggest that the presence of DEL induces protection against leprosy ( Table 3 ) , and the DEL/DEL genotype confers protection against the development of clinical forms of MB ( Table 5 ) , thus this genotype of PAR1 gene can suppresses cellular infiltration and increase both Th1 and Th17 responses to infection ., Moreover , analyses of macrophages revealed that secretion of IL-12 and IL-13 , two cytokine that play role key on cellular immunity Th1 and Th17 , can be suppressed by PAR1 activation ., Furthermore , PAR1 can suppress interferon regulatory factor 5 ( IRF5 ) , that play role key like transcription factor for IL-12 and IL-23 , which modulates the sub sets of cellular immunity ., Thereby the suppression of IRF5 and IL-12/23 secretion by PAR1 gene , can provides a novel mechanism by which the host suppresses the Th1 and Th17response to infection , and dysregulation of this process can likely an important factor in the susceptibility of some individuals to leprosy 28 ., Macrophages with a high load of M . leprae have been shown to undergo apoptosis , and this mechanism is under the control of cytokines 30 ., In leprosy patients , the immune system is overburdened with bacilli , and most likely the continuous activation of T cells by circulating M . leprae antigens leads to apoptosis and to a reduction of peripheral lymphocytes and other immune effector cells in these patients with the regulation of apoptosis involved in the stimulation and activation of caspase-8 31 ., The allele DEL ( rs3834129 ) cause a decrease in CASP8 transcription and a reduction in apoptosis 32 , thereby improving the bacillary load ., Our results suggested that the DEL/DEL genotype ( Table 3 ) and the high frequency of DEL allele ( Table 5 ) can raise the bacillary load and thus confers a risk to leprosy development ., Interleukin-4 ( IL-4 ) is a key cytokine secreted by Th2 lymphocytes , eosinophils and mast cells that induces the activation and differentiation of B cells and the development of the Th2 subset of lymphocytes , which is ineffective in combating leprosy 33 ., Our analysis of the VNTR on intron 3 of the IL4 gene ( rs79071878 ) revealed two common alleles with two ( A1 ) and three ( A2 ) tandem repeats ., Of these , A2 allele is known to be a high producer of IL-4 34 ., Our results indicate that allele A2 is more frequent in leprosy patients compared to healthy individuals , consistent with the fact that higher levels of IL4 would be ineffective in controlling the growth of bacilli ( Table 3 ) ., The conversion of androgens to estrogens , catalyzed by aromatase encoded by the CYP19A1 gene , is the primary pathway of estrogen production in humans 35 ., The levels of these hormones are important in leprosy patients and it has been demonstrated that androgen levels are significantly lower in leprosy patients compared to healthy control subjects 36 ., Moreover , there is an inverse correlation between plasma androgen levels and secretion of inflammatory cytokines , suggesting that high plasma androgen levels can be less effective in inhibiting bacillus growth 37 ., The DEL allele ( rs11575899 ) has previously been reported to have a negative effect on aromatase activity 38 , and our results show that the DEL allele is more frequent in MB patients ( Table 5 ) ., We hypothesize that the DEL allele can decrease aromatase activity and increase androgen levels , resulting in an overall reduction in effective combat of bacillary growth and development of the severe clinical form MB ., It is unclear whether leprosy originated in Asia or Africa ., However , leprosy is believed to have been introduced into Europe from India , and the incidence was high in Europe during the Middle Ages until approximately 1870 when the number of cases dramatically reduced because of socioeconomic development 39 , 40 , 41 ., It is believed that leprosy was introduced in Brazil primarily by the Spanish and Portuguese 41 ., Estimates indicate that before the arrival of colonizers , approximately 2 . 5 million natives lived in Brazil , and during the European immigration in the first three centuries , approximately 500 , 000 individuals came from Portugal and approximately 3 . 5 million Africans were brought into Brazil through slave trade 14 ., Therefore , there is evidence of a so-called directed admixture process involving predominantly European , Native American and African people 42–45 ., Our data indicates that the contribution of different ethnic groups to the composition of the current Brazilian population can generate different rates of risk for leprosy development according to the level of inter-ethnic composition of the individuals involved ., Our analysis suggests that an increase in European contribution increases the risk of leprosy development , while an increase in African contribution decreases the risk for leprosy development and the Amerindian contribution does not result in any statistically significant differences ( Fig 1 ) ., The introduction of leprosy in Brazil primarily can it be accredited to the slave trade , but no only for this reason ., Slaves were firstly there from Africa , and in succeeding years the number these slaves were increased , but was not common between they the clinical manifestation of leprosy , because these slaves were from region of the Africa where leprosy was comparatively rare ., Moreover isnt doubt that the Portuguese and , to a less degree , Dutch , French and Spaniards were responsible by introduction of leprosy in Brazil , on period of country colonization ., Additionally , data showed that as early , as 1419 , the disease was common in Portuguese and epidemiologically in this time the leprosy was very prevalent in Europe , and particularly in Portugal 41 ., Therefore , our data of risk of leprosy according the different ethnic groups compositions is consistent with the higher numbers of settlers Portuguese that came to Brazilian that probably increases the frequencies of alleles of susceptibility on Brazilian population 14 , 42–45 ., In other hand , the African contribution may have increase the frequencies of allele that confer protection against to leprosy ., Comparative analyses of the four M . leprae genomes ( India , Thailand , Brazil and US ) have revealed little clonal differences ., Thus , the patterns of global human migration routes , during the past 100 , 000 years , corroborate and suggest that leprosy probably originated in Africa 46 ., African-descendants in admixture populations can be less susceptible to the leprosy bacilli , probably because of genetic polymorphisms accumulated during these times , in gene that can modulate the immune response on infection combat ., Furthermore African humans are the more genetically diverse population in the world consequently , by selection bias , genetic polymorphisms accumulated that confer protection against disease , can be present in this population and your descendants ., Of point view epidemiological , the situation of African and Americas region is critical , and is associated the socioeconomic challenges related to the disease , but genetics components also are important to disease knowledge 4 ., Thus understanding of like genetic ancestry , in admixture population , can to influence genetic susceptibility is essential to avoid spurious results ., In conclusion , our study shows that the NFKβ1 rs28362491 , CASP8 rs3834129 , PAR1 rs11267092 and IL4 rs79071878 genes are possible markers for the susceptibility to development of leprosy and the severe clinical form MB ., Moreover , after correcting for population structure within an admixture population , the results show that different levels of ethnic group composition can generate different OR rates for leprosy susceptibility . | Introduction, Materials and Methods, Results, Discussion | Leprosy is an insidious infectious disease caused by the obligate intracellular bacteria Mycobacterium leprae , and host genetic factors can modulate the immune response and generate distinct categories of leprosy susceptibility that are also influenced by genetic ancestry ., We investigated the possible effects of CYP19A1 rs11575899 , NFKβ1 rs28362491 , IL1α rs3783553 , CASP8 rs3834129 , UGT1A1 rs8175347 , PAR1 rs11267092 , CYP2E1 INDEL 96pb and IL4 rs79071878 genes in a group of 141 leprosy patients and 180 healthy individuals ., The INDELs were typed by PCR Multiplex in ABI PRISM 3130 and analyzed with GeneMapper ID v3 . 2 ., The NFKβ1 , CASP8 , PAR1 and IL4 INDELs were associated with leprosy susceptibility , while NFKβ1 , CASP8 , PAR1 and CYP19A1 were associated with the MB ( Multibacilary ) clinical form of leprosy ., NFKβ1 rs28362491 , CASP8 rs3834129 , PAR1 rs11267092 and IL4 rs79071878 genes are potential markers for susceptibility to leprosy development , while the INDELs in NFKβ1 , CASP8 , PAR1 and CYP19A1 ( rs11575899 ) are potential markers for the severe clinical form MB ., Moreover , all of these markers are influenced by genetic ancestry , and European contribution increases the risk to leprosy development , in other hand an increase in African contribution generates protection against leprosy . | Leprosy is an infectious disease caused by Mycobacterium leprae , which can carry to skin lesions and affect peripheral nerves , which cause physical and motor injuries on the patients ., Moreover , leprosy , may be classified in two major groups , based on clinical manifestations in Paucibacillary ( PB ) or Multibacillary ( MB ) , and these phenotype may be influenced by host immune response; that can be controlled by genetics factors that can be useful like future panel of biomarkers to leprosy , and it’s related with the different genetic background of population studied ., Therefore , we conducted a study to evaluate seven INDEL polymorphisms in seven genes involved in modulation of the host immune response , and consequently can modulated o phenotype showed through the disease , to identify possible susceptibility markers of leprosy ., However this analysis can be spurious on presence of population structure , common in admixture population like the Brazilian , thus we evaluate like the influence of genetic ancestry can modulated the disease risk . | null | null |
journal.pgen.1001216 | 2,010 | A Functional Genomics Approach Identifies Candidate Effectors from the Aphid Species Myzus persicae (Green Peach Aphid) | Like most plant parasites , aphids require intimate associations with their host plants to gain access to nutrients ., Aphids predominantly feed from the plant phloem sieve elements , and use their stylets to navigate between the cells of different layers of leaf tissue during which plant defenses may be triggered ., Indeed , aphid feeding induces responses such as clogging of phloem sieve elements and callose formation , which are suppressed by the aphid in successful interactions with plant hosts 1 ., In addition , some aphid species can alter host plant phenotypes , by for example inducing the formation of galls or causing leaf curling 2 indicating that there is an active interplay between host and aphid at the molecular level ., During probing and feeding , aphids secrete two types of saliva: gelling saliva , which is thought to protect stylets during penetration , and watery saliva , which is secreted into various plant host cell types and the phloem 3 ., The secretion of aphid saliva directly into the host-stylet interface 4 , suggests that molecules present in the saliva may perturb plant cellular processes while aphids progress through different feeding stages ., Interestingly , the knock-down of the C002 salivary gene in Acyrthosiphon pisum ( pea aphid ) negatively impacts survival rates of this aphid on plant hosts 5 , 6 ., Furthermore , proteomics studies based on artificial aphid diets showed the presence of secreted proteins , including C002 , in aphid saliva indicating that these proteins are delivered inside the host plant during feeding 7 , 8 ., However , whether and how these aphid salivary proteins function in the plant host remains elusive ., Suppression of host defenses and altering host plant phenotypes is common in plant-pathogen interactions and involves secretion of molecules ( effectors ) that modulate host cell processes 9 , 10 ., Therefore it is likely that aphids , similar to plant pathogens , deliver effectors inside their hosts to manipulate host cell process enabling successful infestation of plants 9 ., Effector-mediated suppression of plant defenses , such as Pathogen-Associated Molecular Pattern ( PAMP ) -triggered immunity ( PTI ) , generally involves the targeting of a plant virulence target , or operative target 11 ., However , plant pathogen effectors that are deployed to suppress host defenses are recognized by plant disease resistance ( R ) proteins in particular host genotypes , resulting in effector-triggered immunity ( ETI ) 12 ., Interestingly , the R proteins that recognize plant pathogens and those that confer resistance to aphids , such as Mi-1 . 2 and Vat , share a similar structure , and contain a nucleotide binding site ( NBS ) domain and leucine rich repeat ( LRR ) regions 13–15 ., The Mi-1 . 2 resistance gene confers resistance in tomato to certain clones of Macrosiphum euphorbiae ( potato aphid ) , two whitefly biotypes , a psyllid , and three nematode species 16–19 , indicating that there is significant overlap in plant pathogen and aphid recognition in plants ., In addition , aphid resistance conferred by several resistance genes was shown to be race-specific 16 , 20 ., This suggests that depending on their genotype , certain aphid clones may be able to avoid and/or suppress plant defenses and fits with the gene-for-gene model in plant-pathogen interactions 21 ., Therefore , it is likely that not only plant pathogens , but also aphids , secrete effectors that in addition to targeting host cell processes may trigger ETI depending on the host genotype ., Plant pathogen effectors generally share the common feature of modulating host cell processes 22 ., Various assays have been developed to identify the functions of effectors from bacterial and eukaryotic filamentous plant pathogens 22–24 ., One important and common function of plant pathogen effectors is the suppression of PTI ., This activity is especially common among type III secretion system ( T3SS ) effectors ., For example , the large majority of Pseudomonas syringae DC3000 effectors can suppress PTI responses , including the oxidative burst 25 ., However , effectors from eukaryotic filamentous plant pathogens can also suppress PTI , as demonstrated for the AVR3a effector from Phytophthora infestans , which suppresses cell death induced by the PAMP-like elicitor INF1 26 , 27 ., Another activity of plant pathogen effectors is the induction of phenotypes in plants ., For example , several effectors , including CRN2 and INF1 , from the oomycete plant pathogen P . infestans induce cell death upon overexpression in planta 28 , 29 , whereas other effectors , like AvrB from P . syringae DC3000 induce chlorosis 30 ., Also , overexpression of effector proteins from plant pathogenic nematodes in host plants can affect plant phenotypes , as shown for the Heterodera glycines CLE protein Hg-SYV46 that alters host cell differentiation 31 ., As effectors exhibit functions important for pathogenicity , their deletion can have detrimental effects on pathogen virulence ., However , due to redundancy , the knock-down or deletion of single effectors does not always impact virulence ., On the other hand , overexpression of plant pathogen effectors can enhance pathogen virulence , as shown for active AvrPtoB , which enhances virulence to P . syringae DC3000 in Arabidopsis 32 , and for the H . schachtii effector 10A06 that , in addition to altering host plant morphology , increases nematode susceptibility in Arabidopsis 33 ., We exploited publicly available aphid salivary gland sequences to develop a functional genomics approach for the identification of candidate aphid effector proteins from the aphid species Myzus persicae ( green peach aphid ) based on common features of plant pathogen effectors ., Data mining of salivary gland expressed sequences tags ( ESTs ) identified 46 M . persicae predicted secreted proteins ., Functional analyses showed that one of these proteins , Mp10 , induced chlorosis and weak cell death in Nicotiana benthamiana , and suppressed the oxidative burst induced by the bacterial PAMP flg22 ., In addition , we developed a medium-throughput assay , based on transient overexpression in N . benthamiana , that allows screening for effects of aphid candidate effectors on aphid performance ., Using this screen , we identified two candidate effectors , Mp10 and Mp42 , that reduced aphid performance and one effector candidate , MpC002 , that enhanced aphid performance ., In summary , we found aphid secreted salivary proteins that share features with plant pathogen effectors and therefore may function as aphid effectors by perturbing host cellular processes ., We developed a functional genomics approach to identify candidate effectors from M . persicae using 3233 publicly available aphid salivary gland ESTs 34 ., We hypothesized that aphid effectors are most likely secreted proteins that are delivered into the saliva through the classical eukaryotic endoplasmic reticulum ( ER ) -Golgi pathway of the salivary glands ., A feature of proteins secreted through this pathway is the presence of an N-terminal signal peptide ., Therefore , we used the SignalP v3 . 0 program 35 to predict the presence of signal peptides in the amino acid sequences encoded by the open reading frames ( ORFs ) found in salivary gland ESTs ., Out of 5919 amino acid sequences corresponding to predicted ORFs , we identified 134 nonredundant sequences with signal peptide ( Figure 1A ) ., Out of these 134 proteins , 19 were predicted to contain a transmembrane domain in addition to the signal peptide , and are therefore likely to remain in the salivary gland membrane upon secretion ., Hence , 115 predicted secreted proteins remained ., In order to investigate the M . persicae candidate effector protein in functional assays , we started with the cloning of 46 candidates that corresponded to full-length sequences within the set of 115 candidates ., Effectors are subject to diversifying selection because of the co-evolutionary arms race between host and pathogen proteins 36 , 37 ., Therefore , we used the presence of amino acid polymorphisms among alignments of deduced protein sequences of M . persicae and A . pisum ESTs as an additional criterion ., Three candidates did not fulfill this criterion and were removed from our candidate set bringing the total to 43 candidates ., We applied a similar data mining approach as described above to 4517 publicly available salivary gland ESTs from A . pisum , thereby identifying 24 candidates ( Table S1 ) ., In the A . pisum salivary gland ESTs we predicted only 1751 ORFs , explaining the relatively low number of A . pisum candidates ., A total of three candidates were found in both M . persicae and A . pisum datasets ( combinations Mp1/Ap1 , Mp5/Ap7 and Mp16/Ap4 ) ., The remaining 21 non-overlapping A . pisum candidates were subjected to BLAST searches ( E value<10−15 ) against all available M . persicae ESTs to identify putative M . persicae homologs ., This led to the identification of three M . persicae sequences ( Mp3 , Mp54 and MpC002 ) that were added to the M . persicae candidate effector dataset bringing the total to 46 ( Figure 1A , Table S2 ) ., Interestingly , for two candidates , Mp39 and Mp49 , no similar sequences were present in the publicly available aphid sequence datasets , including the A . pisum genome sequence ( Table S2 ) ., Also , no homologs of these proteins were identified by BLAST searches against GenBank nucleotide and protein databases ( E value<10−5 ) ., This suggests these proteins may be specific to M . persicae ., A total of 11 candidates were shared between the independent salivary gland EST datasets from M . persicae and A . pisum but were not present in gut ESTs from M . persicae ( Table S2 ) providing support that the corresponding proteins may share a similar function in both these aphid species ., For four candidates matches were found in gut ESTs from M . persicae , suggesting these proteins may be derived from salivary gland contaminants in dissected gut tissues and not function uniquely in the salivary gland or saliva ., Indeed , gene expression analysis of Mp51 in various aphid tissues dissected from aphids fed on N . benthamiana confirmed that this gene is specifically expressed in the aphid gut ( Figure S1 ) ., In contrast , candidate effector genes Mp1 , Mp2 , Mp10 , Mp30 , Mp42 , Mp47 , Mp50 and MpCOO2 , were expressed in aphid heads and salivary glands but not in aphid guts ( Figure S1 ) , suggesting that their corresponding proteins are indeed produced in the salivary glands ., Furthermore , Mp1 and MpCOO2 were previously identified in saliva of M . persicae using a proteomics-based approach 7 confirming that these two proteins are secreted into aphid saliva ., To investigate the functions of the 46 effector candidates , we amplified the corresponding sequences encoding the mature proteins , without the signal peptide encoding sequences , from M . persicae cDNA for cloning ( Figure 1B ) ., To preserve the authentic sequence in the 3′ end of the ORF , we designed reverse primers in the 3′ untranslated regions ( UTRs ) based on EST sequences when possible ., Amplicons were cloned in a 35S cassette and corresponding constructs were transformed directly into Agrobacterium tumefaciens followed by sequencing ( Figure 1B ) ., Two out of the 46 candidates , Mp7 and Mp38 could not be amplified from M . persicae cDNA ., Of the remaining 44 candidates , four ( Mp6 , Mp17 , Mp33 and Mp35 ) were represented by two polymorphic forms , with polymorphisms within the mature protein portion ., Except for one of the polymorphic Mp6 sequences , all sequences were identical to those in the M . persicae EST databases ., To rule out that the polymorphism in Mp6 was due to PCR errors , we repeated the Mp6 PCR and sequencing several times on individual aphids with similar results ., Both forms of the four polymorphic candidates were cloned resulting in a total of 48 cloned M . persicae effector candidates ., Functional assays were performed based on transient over-expression in N . benthamiana to assess whether the M . persicae candidate effectors, 1 ) induce a phenotype in planta ,, 2 ) suppress PAMP-triggered immunity and, 3 ) affect the ability of M . persicae aphids to reproduce ( fecundity ) ( Figure 1B ) ., We assessed fecundity of M . persicae lineage RRes ( genotype O ) , which can utilize N . benthamiana as a host ., Several plant pathogen effectors induce a phenotype upon overexpression in planta , which may reflect their virulence activity 22 ., Hence , we performed transient overexpression of the effector candidates in N . benthamiana by agroinfiltration to screen for the induction of phenotypes ., Out of the 48 , one candidate effector , Mp10 , induced chlorosis starting from 2 days post inoculation ( dpi ) ( Figure 2A ) ., In addition , we observed local cell death in a low number of infiltration sites ( Figure S2A , S2B , S2C , S2D ) ., The phenotype was not affected by co-expression with the silencing suppressor p19 ( Figure S2E ) ., To independently confirm the phenotype , we expressed Mp10 in N . benthamiana using a Potato virus X ( PVX ) -based vector ( PVX-Mp10 ) ., Systemic PVX-based overexpression of Mp10 induced systemic chlorosis in N . benthamiana starting at 10 dpi ( Figure 2B ) ., This also suggests that the Mp10 response is not dependent on the presence of Agrobacterium ., To determine whether the response to Mp10 was specific to N . benthamiana , we infected N . tabacum , Solanum lycopersicum ( tomato ) and N . benthamiana plants with PVX-Mp10 in parallel ., Starting at around 10 dpi , systemic chlorosis was observed in N . benthamiana expressing PVX-Mp10 , but not in control PVX-infected plants ( Figure 2B ) ., Whereas mosaic symptoms were observed in S . lycopersicum , indicative of PVX infection , no Mp10-induced chlorosis was observed ( Figure 2C; Figure S3A , S3B ) ., Mp10 expression was confirmed by semi-quantitative RT-PCR in systemically PVX-Mp10 infected leaves of S . lycopersicum suggesting that the lack of symptoms is not due to a loss of the Mp10 sequence from PVX-Mp10 ( Figure 2E ) ., In contrast , N . tabacum plants infected with PVX-Mp10 did not show mosaic symptoms indicative of virus infection , while N . tabacum inoculated with PVX alone did ( Figure 2D; Figure S2B ) ., No Mp10 expression could be detected in leaves of N . tabacum plants inoculated with PVX-Mp10 , whereas expression of the viral coat protein was detected , indicating that PVX itself did systemically spread in N . tabacum ( Figure 2E ) ., In contrast , PVX-Mp42 did spread systemically in N . benthamiana , N . tabacum and S . lycopersicum , indicating that this aphid protein can be systemically expressed in these plant species using PVX ( Figure S4 ) ., It is possible that PVX-Mp10 may evoke an avirulence response in N . tabacum causing the selection of PVX without the Mp10 insert ., Loss of foreign gene fragments from the PVX genome has been reported previously and is most likely due to selection pressures forcing virus recombination 38 ., The lack of mosaic symptoms in PVX-Mp10-inoculated N . tabacum plants is possibly due to the initially low abundance of recombined PVX-virus as compared to the vector control ., The SGT1 protein , an ubiquitin-ligase associated protein , is required for plant cell death responses , including those involved in plant resistance 39 ., To investigate whether SGT1 is required for the Mp10 chlorosis response , we generated SGT1-silenced N . benthamiana plants using Tobacco rattle virus ( TRV ) -based virus-induced gene silencing ( VIGS ) ., Silenced plants ( treated with TRV-SGT1 ) and control plants ( treated with TRV ) ( Figure 2H ) were infiltrated with Agrobacterium strains expressing Mp10 or the positive control INF1 , an elicitin from P . infestans that induces cell death in control plants , but not in SGT1-silenced plants 40 ., Both the Mp10-induced chlorosis and the INF1-induced cell death were pronouncedly reduced in the SGT1-silenced plants , but not in the TRV-treated control plants ( Figure 2F and 2G ) , indicating SGT1 is required for these chlorosis and cell death responses ., Suppression of PTI induced by PAMPs like flg22 and chitin is a common feature of plant pathogen effectors ., To determine whether aphid candidate effectors can suppress PTI , we assessed whether any of our 48 candidates suppressed the oxidative burst response induced by the bacterial PAMP flg22 ., We decided to screen for suppression of the oxidative burst induced by flg22 only , as this PAMP gives a strong and consistent oxidative burst response in N . benthamiana , which is convenient for use in large screens ., N . benthamiana leaf discs overexpressing the effector candidate genes under control of the 35S promoter were challenged with the flg22 elicitor and the production of reactive oxygen species ( ROS ) was measured using a luminol-based assay 41 ., The bacterial effector AvrPtoB , a suppressor of the flg22-mediated oxidative burst response 42 , was included as a positive control ., We found that Mp10 suppresses the flg22-induced oxidative burst in leaf discs harvested 2 days post agroinfiltration ( three replicated experiments ) ( Figure 3A ) , whereas other candidate effectors did not ( data not shown ) ., Although the level of suppression by Mp10 was significant compared to that of the empty vector control , it was not as effective as AvrPtoB ., We tested whether Mp10 also suppressed the oxidative burst induced by a fungal PAMP , chitin , and found that while Mp10 suppressed the flg22 response , no suppression of the chitin-induced oxidative burst was observed ( Figure 3B ) ., Thus , Mp10 specifically suppresses the oxidative burst induced by the PAMP flg22 ., We developed a medium-throughput 24-well plate assay to assess M . persicae fecundity on N . benthamiana leaves transiently overexpressing the 48 candidate effectors ( Figure 4A ) ., Leaf discs were harvested from infiltrated leaves one day after agroinfiltration and placed upside down on water agar in 24-well plates ., Four first-instar nymphs were placed on each leaf disc and the plate was incubated up-side-down under a light source ., Aphids were moved every 6 days to plates with freshly infiltrated leaf discs , as expression levels of green fluorescent protein ( GFP ) in leaf discs were constant during 6 days and then tapered off ( Figure S5 ) ., The aphids placed initially on the leaf discs generally started producing nymphs after about 10–11 days ., Nymph production ( fecundity ) was assessed on day 12 , 14 and 17 by counting and removing newly produced nymphs on each leaf disc ., The total nymph production per adult was calculated and compared among the treatments and GFP and vector controls ., In our initial screens , in which candidate effector constructs were infiltrated on different leaves and not always side-by-side with the vector control , we identified 14 candidates that either enhanced or reduced aphid fecundity by one time the standard error compared to the empty vector ( EV ) control ( Figure S6 ) ., To confirm the effect on aphid fecundity of these 14 candidates , we conducted additional assays in which the candidates were infiltrated side-by-side with the vector control ( EV ) on the same leaves ., Two candidates , Mp10 and Mp42 , reduced aphid fecundity in three repeated confirmation assays compared to the vector control ( Figure 4B ) ., In addition , one candidate , MpC002 , enhanced aphid fecundity in three repeated confirmation assays compared to the vector control ( Figure 4B ) ., Transient overexpression of Mp10 did not induce chlorosis in leaf discs ( Figure S7 ) or leaves that were detached from the plant 24hrs after infiltration ( data not shown ) ., Thus , leaves need to be attached to the plant for chlorosis to occur and the chlorosis itself was therefore not likely responsible for the observed reduction in aphid performance ., In summary , we have developed a novel assay to screen for effects of in planta expressed aphid salivary proteins on aphid performance and thereby identified three candidates that potentially function as effectors by eliciting plant defenses or promoting aphid infestation of host plants ., To determine whether the candidates that alter aphid fecundity , ( i . e . Mp10 , Mp42 , and MpC002 ) share similarity to proteins of known or predicted function , we performed BLAST searches against the GenBank non-redundant ( nr ) protein database ( E value<10−5 ) ., One of the three candidates , Mp10 showed homology to an insect protein of predicted function , the olfactory segment D2-like protein ( OS-D2-like protein ) ., The OS-D2-like protein is a member of a family of chemosensory proteins in aphids that contain the conserved cysteine pattern CX6CX18CX2C 43 ., Mp10 also shows similarity to chemosensory proteins ( CSPs ) from other insects ( E value<10−5 ) , including the CSP5 protein from the mosquito Anopheles gambiae ( Figure 5A ) ., The four cysteines in Mp10 are conserved among different members of the CSP family 44 , 45 ( Figure 5A ) ., Among the aphid sequences similar to Mp10 , polymorphisms are predominantly present after the predicted signal peptide sequence , in the mature protein region ., For Mp42 and MpC002 , similar sequences were identified in the genome sequence of the aphid species A . pisum only , but these proteins have no similarities to proteins with known functions ., Alignment of Mp42 to a putative A . pisum homolog shows strong sequence divergence especially in the mature protein regions ( Figure 5B ) ., Finally , alignment of MpC002 to A . pisum C002 shows sequence divergence consisting of both amino acid polymorphisms and a 45 amino acid gap in A . pisum C002 after the predicted signal peptide sequence ( Figure 5C ) ., The presence of polymorphisms mainly in the mature protein regions may reflect that the functional domains of these proteins have diversified due to distinct selective pressures ., Aphids , like other plant parasites , deliver repertoires of proteins inside their hosts that function as effectors to modulate host cell processes ., These insects most likely secrete effectors into their saliva while progressing through the different plant cell layers during probing and feeding ., The identification and characterization of these proteins will reveal new insights into the molecular basis of plant-insect interactions ., Here , we have described a functional genomics pipeline to identify M . persicae effector candidates as well as various assays to determine whether the candidates share features with plant pathogen effectors ., Using this approach , we identified three candidate effectors , Mp10 and Mp42 , MpC002 that modulate host cell processes and affect aphid performance ., The induction of chlorosis and local cell death by Mp10 can reflect a genuine effector activity of this aphid salivary protein ., Ectopic expression of bacterial TTSS as well as filamentous plant pathogen effectors can affect host immunity and induce a variety of phenotypes in plants , ranging from chlorosis to necrosis 22 , 28 ., Both the P . syringae type III effectors AvrB 30 and HOPQ-1 46 induce chlorosis and for AvrB this activity is plant genotype specific 47 ., No Mp10 induced chlorosis was observed in tomato despite expression levels of PVX-Mp10 that were comparable to N . benthamiana ., This suggests that the Mp10 response was specific for N . benthamiana ., Interestingly , PVX-Mp10 was unable to infect N . tabacum , suggesting this protein may induce an unknown defense mechanism that is effective against PVX-Mp10 ., There are several possibilities that may explain the Mp10 phenotype in a biologically relevant context ., The first possibility is that the artificially high expression of Mp10 could lead to the induction of the chlorosis/local cell death phenotype and therefore this response could be an artifact of the Agrobacterium-mediated overexpression assay ., However , in this case we would expect that the induction of chlorosis and local cell death by Mp10 would be more widespread in various plant species , and would also be observed in N . benthamiana leaf discs or detached leaves ., Another possibility is that the high expression of Mp10 could lead to excessive targeting of the operative target as well as other host proteins leading to an exaggeration of the true virulence activity 22 ., Finally , the induction of chlorosis and local cell death could reflect avirulence activity of Mp10 ., Feeding of M . persicae is known to induce chlorosis and premature leaf senescence in plants , and this response is related to PAD4-mediated defense responses 48 ., Therefore , Mp10 may exhibit an avirulence activity specifically in Nicotiana spp resulting in chlorosis and local cell death ., The induction of chlorosis in N . benthamiana by P . syringae effector AvrB is thought to be due to weak activation of TAO1 , an NBS-LRR protein , and requires the plant-signaling component Rar1 49 ., We found that chlorosis induction by Mp10 requires the co-chaperone SGT1 , which is required for activation of NBS-LRR proteins and plant resistance responses 39 ., Therefore , Mp10 may activate an NBS-LRR resistance protein resulting in ETI ( further discussed below ) ., We also found that Mp10 suppressed the ROS response induced by flg22 , suggesting that suppression of PTI may be a feature shared by plant pathogens and insects ., Possibly , the flg22-induced signaling pathway may not be specific to bacteria as other ( non-bacterial ) PAMPs can induce this pathway ., Also , plants may have a PTI pathway ( s ) that is induced by an unknown insect PAMP ( s ) and partially overlaps with the signaling pathway induced by flg22 ., To date the role of perception of PAMP-like molecules in plant-insect interactions remains elusive ., However , chitin is a major structural component of the insect cuticle ., Degradation of chitin by plant chitinases generates fragments that induce PTI 50 ., Whether the chitin in the insect cuticle is degraded to induce plant defenses during interaction with host plants remains to be investigated ., It has been hypothesized that sheath saliva protects the insect stylets , which mainly consist of chitin , from triggering plant defenses 51–53 , potentially including PTI ., Recent studies showed that insect saliva of both chewing insects 54 and aphids 55 contains elicitors that induce defense responses in host plants ., The nature of these elicitors and their role in triggering PTI are unknown ., Despite the lack of an understanding of the role in perception of PAMP-like molecules in plant-insect interactions , our data suggest that an aphid salivary protein , Mp10 , can interfere with a specific PAMP response in a M . persicae host plant ., It is therefore possible that Mp10 is a genuine suppressor of PTI ., Alternatively , the overexpression of Mp10 may perturb a signaling component in the PTI pathway that is required for recognition of flg22 ., As Mp10 induces weak chlorosis starting from 2 dpi , it is possible that this response itself is responsible for loss of the oxidative burst response to PAMPs ., However , the Mp10 chlorosis response does not interfere with the oxidative burst triggered by chitin ., This suggests that the induction of chlorosis itself may not be sufficient to block the oxidative burst induced by flg22 , but that Mp10 specifically interferes with the flg22-triggered signaling cascade ., Despite the suppression of the flg22-mediated oxidative burst by Mp10 , its overexpression in N . benthamiana reduced aphid fecundity ., A plausible explanation for this contradictory observation is that Mp10 may activate an NBS-LRR resistance protein resulting in ETI , thereby reducing aphid performance ., Thus , the recognition of Mp10 , potentially through ETI , in Nicotiana spp may mask the true virulence activity of this protein ., If true , this recognition may be suppressed by other effectors during plant-aphid interactions so that Mp10 can exhibit its virulence function ., The leaf disc assay allowed us to generate vast amounts of functional data and directly implicated three effector candidates in plant-aphid interactions ., The differences in aphid fecundity observed in our screens were quite variable , requiring replication of experiments ., Despite the variation , Mp10 , Mp42 , and MpC002 showed consistent effects on aphid fecundity throughout the individual replicates ( data not show ) ., The fecundity was affected by Mp10 , Mp42 , and MpC002 by around 1–1 . 5 nymph produced per adult over a nymph production period of about 6 days ., Although these differences may seem small , they are expected to have a large impact on the population size of aphids ., Furthermore , M . persicae does not perform as well on N . benthamiana as it does on other hosts , such as Arabidopsis thaliana ., Despite the low reproduction level on N . benthamiana , the fecundity differences found in our screens are similar to those observed over a 2-day period on A . thaliana in a study by Pegadaraju et al . 56 which shows that overexpression of PAD4 reduced aphid fecundity by about 1 . 5 nymphs per adult ., The number of candidate effectors with an effect on aphid fecundity identified in this study may have been limited by our approach ., For example , when the amount of an effector secreted by the aphid is sufficient to modulate host cell processes to promote feeding , in planta overexpression may not necessarily further enhance this effect ., Also , there could be differences in plant responses to aphids in leaf discs versus whole plants as certain plant responses to aphids may require an intact plant transport system ., Despite these limitations , the development of a novel leaf disc-based assay allowed us to identify three effector candidates from the aphid species M . persicae ., Out of the three candidates that affect aphid fecundity in the leaf-disc assays , only Mp10 shows homology to a protein of predicted function , namely OS-D2 , a member of a family of predicted chemosensory proteins ., Insect chemosensory proteins ( CSP ) are thought to be involved in olfaction and gustation ., Indeed , several CSPs have been specifically found in chemosensory organs and are predicted to function in chemoperception 43 , 57 , 58 ., However , for some members of this large protein family functions have been identified in insect development 59 and leg regeneration 60 , suggesting that CSPs may have divergent functions ., This is further supported by gene expression studies , which show that some CSPs are specifically expressed in antenna 61 or mouthparts 62 , whereas others are expressed throughout the insect 63 ., CSPs are thought to bind small molecules , such as fatty acids , and for some members of this protein family there is evidence that they bind to pheromones 64 , 65 ., In the aphid species Megoura viciae a Mp10 homolog was found to be expressed in aphid heads without antenna , indicating that it is not an antenna specific CSP 43 ., Interestingly , in mosquitos , members of a family of odorant binding-related proteins , also with predicted functions in olfaction and gustation , are secreted into host cells to manipulate host physiology by for example scavenging host amines 66 ., Counteracting host amines has evolved in various blood-feeding insects independently through adaptation of members of the lipocalin or odorant-binding protein families 66 ., It is possible that also in plant feeding insects , proteins predicted to be involved in chemosensing are actually involved in early plant host recognition and plant host cell manipulation ., For Mp42 and MpC002 no homology was found to proteins of known or predicted function ., This is not surprising as most plant pathogen effectors described to date do not show similarity to proteins of known function based on amino acid alignments ., The reduction in aphid performance upon overexpression of Mp42 could reflect that Mp42 induces defense responses against aphids in the plant ., In c | Introduction, Results, Discussion, Methods | Aphids are amongst the most devastating sap-feeding insects of plants ., Like most plant parasites , aphids require intimate associations with their host plants to gain access to nutrients ., Aphid feeding induces responses such as clogging of phloem sieve elements and callose formation , which are suppressed by unknown molecules , probably proteins , in aphid saliva ., Therefore , it is likely that aphids , like plant pathogens , deliver proteins ( effectors ) inside their hosts to modulate host cell processes , suppress plant defenses , and promote infestation ., We exploited publicly available aphid salivary gland expressed sequence tags ( ESTs ) to apply a functional genomics approach for identification of candidate effectors from Myzus persicae ( green peach aphid ) , based on common features of plant pathogen effectors ., A total of 48 effector candidates were identified , cloned , and subjected to transient overexpression in Nicotiana benthamiana to assay for elicitation of a phenotype , suppression of the Pathogen-Associated Molecular Pattern ( PAMP ) –mediated oxidative burst , and effects on aphid reproductive performance ., We identified one candidate effector , Mp10 , which specifically induced chlorosis and local cell death in N . benthamiana and conferred avirulence to recombinant Potato virus X ( PVX ) expressing Mp10 , PVX-Mp10 , in N . tabacum , indicating that this protein may trigger plant defenses ., The ubiquitin-ligase associated protein SGT1 was required for the Mp10-mediated chlorosis response in N . benthamiana ., Mp10 also suppressed the oxidative burst induced by flg22 , but not by chitin ., Aphid fecundity assays revealed that in planta overexpression of Mp10 and Mp42 reduced aphid fecundity , whereas another effector candidate , MpC002 , enhanced aphid fecundity ., Thus , these results suggest that , although Mp10 suppresses flg22-triggered immunity , it triggers a defense response , resulting in an overall decrease in aphid performance in the fecundity assays ., Overall , we identified aphid salivary proteins that share features with plant pathogen effectors and therefore may function as aphid effectors by perturbing host cellular processes . | Aphids are insects that can induce feeding damage , achieve high population densities , and most importantly , transmit economically important plant diseases worldwide ., To develop durable approaches to control aphids , it is critical to understand how aphids interact with plants at the molecular level ., Aphid feeding induces plant defenses , which can be suppressed by aphid saliva ., Thus , aphids can alter plant cellular processes to promote infestation of plants ., Suppression of plant defenses is common in plant pathogens and involves secretion of effector proteins that modulate host cell processes ., Evidence suggests that aphids , like plant pathogens , deliver effectors inside their host cells to promote infestation ., However , the identity of these effectors and their functions remain elusive ., Here , we report a novel approach based on a combination of bioinformatics and functional assays to identify candidate effectors from the aphid species Myzus persicae ., Using this approach , we identified three candidate effectors that affect plant defense responses and/or aphid reproductive performance ., Further characterization of these candidates promises to reveal new insights into the plant cellular processes targeted by aphids . | plant biology/plant-biotic interactions, genetics and genomics/bioinformatics, genetics and genomics/functional genomics | null |
journal.pntd.0005276 | 2,017 | Comparative Ability of Oropsylla montana and Xenopsylla cheopis Fleas to Transmit Yersinia pestis by Two Different Mechanisms | Once fleas had been implicated as important vectors of the plague bacillus , Yersinia pestis , attention quickly turned to the mechanism and dynamics of flea-borne transmission ., The first experiments , conducted between 1904–1907 by members of the Indian Plague Research Commission and others , characterized what is now termed early-phase transmission 1–4 ., When fleas that had just fed on a rodent with terminal plague bacteremia were collected and used to challenge naïve rodents , they remained infective for about a week , with the transmission rate peaking about three days after the infectious blood meal and then waning ., Transmission very rarely occurred from challenge with a single flea , but challenges with 10 to 25 fleas resulted in a transmission rate of 20 to 70% ., Because simultaneous challenge with multiple fleas was required , early-phase transmission has also been referred to as mass transmission ., Although long assumed to be a form of mechanical transmission , recent studies suggest that it is not that simple 5; however , the mechanism of this early-phase transmission has yet to be determined ., In 1914 a second mode of transmission was discovered 6 , 7 ., This later phase of transmissibility occurs after Y . pestis forms a dense biofilm in the proventriculus , a valve in the flea foregut 8 ., As the biofilm grows and consolidates it interferes with the valvular function of the proventriculus ., Because the infected valve is unable to close completely , blood flowing into the midgut can flow back out again , carrying bacteria along with it , and be regurgitated into the bite site ., The proventriculus can eventually become completely blocked in some fleas , preventing any blood from reaching the midgut ., Inflowing blood is stopped by the bacterial mass that fills the proventriculus , and transmission occurs when blood mixed with bacteria from the surface of the biofilm is refluxed back into the bite site ., As they begin to starve , such completely blocked fleas make continuous , persistent attempts to feed , thereby increasing the probability of transmission ., Each feeding attempt by a single blocked X . cheopis flea has a 25 to 50% transmission success rate 9–13 ., The ecology of plague is complex , involving many different rodent-flea transmission cycles ., The relative competence of several flea vectors has been examined to varying degrees ( reviewed in 5 ) ., In general , these studies indicate that different flea species vary in their potential to become infected and subsequently transmit Y . pestis ., The comparative early-phase transmission efficiency of different flea vector species has been systematically evaluated in several recent studies 14–17 ., However , comparisons of the ability to transmit after the early phase are problematic because a variety of experimental methods have been used ., This has led to variable and sometimes contradictory results ., For example , some studies indicate that the North American ground squirrel flea Oropsylla montana transmits very poorly by the proventricular biofilm-dependent mechanism 9 , 10 , 18 , 19 , whereas other studies show it to transmit as efficiently as Xenopsylla cheopis , the flea often cited as the most efficient transmitter by that mechanism 12 , 20 ., A second unresolved issue is the relative ecologic importance of the two transmission mechanisms ., Transmission efficiency studies to date have concentrated on either the early-phase or the proventricular biofilm mechanism , making direct comparisons of their relative efficiency problematic , again due to the variety of experimental conditions used ., To begin to address these shortcomings , we developed a standardized , carefully controlled experimental model system to infect cohorts of fleas and to characterize infection , proventricular blockage , and transmission dynamics during a four-week period following a single infectious blood meal ., Comparative results for O . montana and X . cheopis establish that both develop proventricular blockage and transmit efficiently via the biofilm-dependent mechanism ., For both species , transmission efficiency in the early phase was lower than at later times after infection ., Y . pestis KIM6+ was inoculated from frozen stocks maintained at the Rocky Mountain Laboratories ( RML ) into 5 ml of brain-heart infusion ( BHI ) broth supplemented with 10 μg/ml hemin and incubated at 28°C ., After overnight incubation , 1 ml of this culture was used to inoculate 100 ml of BHI that was incubated at 37°C for 18 to 19 hours without aeration ., The KIM6+ strain lacks the Yersinia virulence plasmid and is attenuated in mammalian virulence but infects fleas normally 21 ., O . montana were from a laboratory colony originally established at the CDC , Fort Collins 22and maintained at RML since 2011 ., X . cheopis colonies were derived from fleas collected in Los Angeles , CA or Baltimore , MD and have been maintained at RML for ~ 10 to 25 years , respectively ., Fleas were pulled randomly from colonies and starved for four days ., Groups of about 300 fleas were allowed to feed through a mouse skin affixed to an artificial feeding device 21 containing 5 ml of heparinized mouse blood ( collected from RML Swiss Webster mice on site ) containing ~1 x 109/ml Y . pestis KIM6+ , or with KIM6+ that had been transformed with pAcGFP1 ( Clontech; Mountain View , CA ) , a plasmid containing a constitutively expressed green fluorescent protein ( GFP ) gene ., After the 1-hour feeding period , fleas were collected and immobilized by placing the tube containing them on ice ., Fleas were then arrayed on a chill table ( BioQuip; Rancho Dominguez , CA ) under a dissecting microscope , and only those that had taken an infectious blood meal , evidenced by the presence of fresh red blood in the midgut , were kept and used for experiments ., A sample of 20 female fleas was placed at -80°C for later determination of the initial infectious dose ., In conjunction with experiments to assess proventricular blockage , a second group of fleas from the same cohort used for infection was allowed to feed the same day on sterile mouse blood to serve as uninfected controls ., After feeding , ~100 infected and control fleas ( equal numbers of males and females ) were put into separate flea capsules 23 that were kept at 21°C and 75% relative humidity ., These fleas were allowed to feed on neonatal mice for 1 hour on days 2 , 6 , 9 , 13 , 16 , 20 , 23 , and 27 after infection ., An additional sample of 20 infected female fleas was maintained identically but collected and placed at -80°C on day 7 after infection ., A shallow layer of sterile sawdust was added to the bottom of capsules containing O . montana ., Following each maintenance feed , infected fleas were collected and examined microscopically for evidence of proventricular blockage , diagnosed by the presence of fresh red blood in the esophagus only , with none in the midgut ., Blocked fleas were segregated into a separate capsule after diagnosis for separate maintenance ., Mortality was also recorded on each maintenance feed day ., A sample of 20 surviving female fleas was placed at -80°C when the experiment was terminated on day 28 ., Three independent blockage experiments were conducted for both O . montana and X . cheopis ., After being diagnosed as blocked , O . montana fleas infected with the GFP+ strain were dissected and examined by fluorescence microscopy to verify proventricular blockage ., Dilutions of the infectious blood meal were plated on blood agar to determine the Y . pestis CFU/ml ., Samples of fleas that had been collected at different times after infection and stored at -80°C were thawed , surface sterilized , individually triturated , and then dilutions were plated in BHI soft agar overlays 24 ., All plates were incubated at 28° C for 48 hours prior to colony counts ., For transmission experiments , fleas that took an infectious blood meal were housed as described above ., On days 3 , 10 , 17 , 24 , and 31 , ~200 fleas ( approximately equal numbers of males and females ) were allowed to feed on sterile defibrinated rat blood ( BioreclamationIVT , New York ) in the artificial feeding system ., After 90 min , fleas were collected and examined microscopically as described above to determine how many had fed , and of those , how many were completely or partially blocked ., A sample of 20 unblocked female fleas that had fed were placed at -80°C ., Day 3 was the first feeding opportunity after the infectious blood meal and represents early-phase transmission ., In addition to the transmission test artificial feedings , fleas were also fed on neonatal mice on days 6 , 13 , 20 , and 27 ., Immediately after each transmission test feeding , blood was removed and the interior of the feeder was washed ten times with 3 ml PBS ., The entire volume of blood was distributively plated on blood agar plates ., Pooled washes were centrifuged 9 , 800 ×g for 30 min , most of the supernatant removed and the remainder was mixed thoroughly to resuspend any bacteria and spread onto blood agar plates ., During the period of peak transmission by O . montana ( 10 to 24 days after infection ) dilutions of the blood and wash samples , instead of the entire sample , were plated ., The external ( haired ) surface of the mouse skin membrane was disinfected with ethanol , cut into small pieces that were added to 1 ml PBS in a lysing matrix H tube and subjected to 2 min treatment in a FastPrep homogenizer ( MP Biomedicals , Santa Ana , CA ) to dislodge any transmitted bacteria associated with the interior surface of the mouse skin ., Skin sample supernatants were pooled and centrifuged at 20 , 000 ×g for 15 min ., Most of the supernatant was discarded and the remainder was vigorously mixed to resuspend any bacteria and then plated on blood agar ., Three independent transmission experiments were conducted with both O . montana and X . cheopis ., Fleas were placed in PBS on a glass microscope slide and dissected with a set of fine forceps ., The flea exoskeleton was removed and an 18 x 18 mm glass cover slip was gently placed over the top of the digestive tract ., Images of flea digestive tracts were obtained using a Nikon Eclipse E800 microscope with an Olympus DP72 camera and cellSens imaging software ., To calculate an esophagus:proventriculus ( E/PV ) width ratio , the width of the base of the esophagus was measured above the proventricular spines , where the musculature ends , and the proventriculus was measured at its widest position , near the midpoint of the valve ., All analyses were performed using GraphPad Prism 6 ( GraphPad Software Inc . , La Jolla , Ca . ) ., Statistical tests used and relevant P values are indicated in the figure legends ., All experiments involving animals were approved by the Rocky Mountain Laboratories , National Institute of Allergy and Infectious Diseases , National Institutes of Health Animal Care and Use Committee ( protocols 13–036 and 14–006 ) and were conducted in accordance with all National Institutes of Health guidelines ., Groups of fleas were infected by allowing them to feed on highly bacteremic blood in an artificial feeding device , and thereafter provided twice-weekly maintenance feeds on uninfected mice for four weeks ., Immediately after each maintenance feed , fleas were examined individually for the development of proventricular blockage , indicated by the presence of fresh red blood in the esophagus but not in the midgut ., Contrary to some previous reports , we found that O . montana readily became blocked , but at a lower incidence than X . cheopis ( mean blockage rates 22% and 36% , respectively; Fig 1A and Fig 2 ) ., However , O . montana tended to become blocked sooner ( mean = 9 . 6 days after infection ) than X . cheopis ( mean = 14 . 7 days ) and to survive longer after becoming blocked ( Fig 3 ) ., Blocked O . montana fleas survived up to 2 weeks after being diagnosed ( range = 1 to 16 days; mean = 7 days; median = 4 days ) , whereas blocked X . cheopis survived a maximum of 4 days ( range = 1 to 4 days; mean = 2 days; median = 1 day ) ., The mortality rate of uninfected control fleas at 28 days was 6 to 7% for both species ( Fig 1B ) ., Since blocked fleas die from starvation , the excess mortality of infected fleas is a surrogate indicator of blockage ., The majority of the mortality of infected X . cheopis ( 60% ) and O . montana ( 28% ) was in fact due to the death of blocked fleas , and reflects the difference in blockage rate between the two species ., The average number of Y . pestis that the fleas acquired in the infectious blood meal was 1 . 3 to 6 . 4 x 105 for all experiments ( Fig 1D ) ., By 7 days after infection , a significantly greater percentage of O . montana than X . cheopis had eliminated the infection ( Fig 1C ) ., However , the average bacterial load in the infected fleas was equivalent for both species at all time points ( Fig 1D ) ., In other experiments , groups of infected fleas were allowed to feed from a sterile blood reservoir at different times after infection ., After feeding , fleas were collected and examined for evidence of feeding and proventricular blockage , and the number of Y . pestis that the fleas had transmitted was determined ., The transmission test at 3 days after infection was the first feeding opportunity after the infectious blood meal , when transmission by the early-phase mechanism is maximum 1 , 2 , 5 ., Between each weekly transmission test thereafter , fleas were provided a separate maintenance feed on an uninfected mouse ., By the second transmission test on day 10 the fleas had had two uninfected blood meals since infection , and were beyond the early-phase transmission period ., Thus , transmission on days 10 to 31 after infection was via the proventricular biofilm-dependent mechanism ., Few Y . pestis were transmitted by the early-phase mechanism , and no transmission was detected in some experiments , even though over 100 fleas fed ( Fig 4 , Tables 1 and 2 ) ., Additional experiments were done to further evaluate early-phase transmission only ( 3 days after an infectious blood meal ) ., The number of Y . pestis transmitted was below the detection level in 7 of 16 trials and ranged from 1 to 164 CFU in the nine others ( Fig 4C ) ., Fleas of both species transmitted many more bacteria after the early phase , peaking 10 to 24 days after infection at ~104 CFU for X . cheopis and >105 CFU for O . montana ., Maxima of Y . pestis CFU transmitted ( 10 to 17 days after infection for O . montana and 17 to 24 days for X . cheopis ) correlated with the mean time for proventricular blockage to develop ( Fig 3A and 3B ) ., At most time points after infection , the cumulative number of Y . pestis CFU transmitted by O . montana cohorts was 10-fold or more higher than the number transmitted by X . cheopis ( Fig 4 , Tables 1 and 2 ) ., While examining the digestive tracts of dissected fleas , we noticed that the base of the esophagus , where it joins the proventriculus , appeared to be wider in O . montana than in X . cheopis ., This was especially evident in blocked O . montana , in which the base of the esophagus was often grossly distended ( Figs 2 and 5 ) ., To evaluate this quantitatively , we measured the esophageal and proventricular widths in uninfected and blocked specimens of the two species and calculated a ratio ., The esophagus:proventriculus ( E/PV ) width ratio was significantly greater for O . montana than X . cheopis ( Fig 5E ) ., In blocked fleas of both species , the Y . pestis proventricular biofilm extended into the esophagus ., This expanded the relative width of the esophagus in blocked vs . unblocked O . montana to a greater extent than in blocked vs . unblocked X . cheopis ., In some blocked O . montana , the esophagus was nearly as wide as the proventriculus ( mean E/PV ratio = 0 . 71; range = 0 . 60 to 0 . 84 ) ., In 1911 , McCoy demonstrated that O . montana was able to transmit Y . pestis during the early phase , and more recent work has indicated that O . montana and X . cheopis are comparable in early-phase transmission efficiency 16 , 25–30 ., X . cheopis has also consistently been shown to become blocked and transmit after the early phase ., However , conflicting results have been reported for the rate at which O . montana becomes blocked after infection and its corresponding ability to transmit by the proventricular biofilm-dependent mechanism ., Wheeler and Douglas found that O . montana was an even better vector than X . cheopis during the four-week period after the early phase 12 , 20 , but other studies of O . montana reported little or no blockage and transmission during this time frame 9 , 10 , 18 , 19 ., Based on the negative data , the recent literature now states that O . montana rarely becomes blocked 26 , 31 ., A first objective of this study was to resolve the contradictory conclusions of the earlier studies by using standardized methods to directly compare the post-infection blockage rates of O . montana and X . cheopis ., We found that obvious , complete blockage of O . montana is not a rare but a regular occurrence ( Fig 2 ) ., Differences among O . montana strains used for infection experiments has previously been suggested as a potential cause for conflicting results 10 ., However , our O . montana colony originated from one used for a study that reported a 0% blockage rate 19 ., A more likely reason for discordant results stems from the very high mortality rate associated with the studies that report negative results for O . montana blockage ., Two studies that reported a 0 to 3% blockage rate also recorded a 23 to 59% die-off during the first week after the infectious blood meal , 58 to 79% at two weeks , and few fleas surviving after four weeks 10 , 19 ., In initial experiments , we also noted high mortality , even of uninfected control O . montana , and low and inconsistent blockage rates ., Adult fleas maintained in proper conditions should live for a period of months , and we had previously observed that results with unhealthy X . cheopis cohorts ( indicated by a 4-week mortality rate of >25% for uninfected control fleas ) were unreliable and grossly underestimated the normal blockage rate observed for healthy X . cheopis ( mortality rate of uninfected controls <10% ) ., We found that unless O . montana were kept on a layer of sawdust , they became entangled via the long , curved pretarsal claws at the end of their legs , which led to stress-related mortality ., The addition of a sawdust substrate to capsules containing O . montana ( which is not necessary for X . cheopis ) eliminated this problem and reduced uninfected control flea mortality at four weeks from 60 to 75% to <10% ., Infected O . montana mortality was 20% at two weeks and 28% at four weeks after infection , correlating with the timing and incidence of blockage-induced mortality ( Fig 1 ) ., A high mortality rate during the first week after infection indicates that the fleas were unhealthy or stressed , complicating the interpretation of the results ., For this reason , experiments designed to compare long-term infection and transmission dynamics should include uninfected control fleas to verify normal viability ., Different flea species may require different maintenance conditions to maintain health ., The consistent high blockage and transmission rates reported for X . cheopis compared to other species may in part be due to the fact that X . cheopis is more easily adaptable to the laboratory ., A second objective was to compare the vector efficiency of O . montana and X . cheopis by the proventricular biofilm-dependent mechanism ., The vector efficiency of a given flea species is a product of several factors , including infection potential ( the percentage of individuals that become infected after feeding on blood containing Y . pestis ) , vector potential ( the % of infected fleas that develop a transmissible infection; i . e . , the % of infective fleas ) , and the transmission potential ( the average number of transmissions effected by a group of infective fleas ) 12 , 32 ., We measured the infection potential directly and found that it was lower for O . montana ( 45 to 75% ) than for X . cheopis ( 87 to 100%; Fig 1C ) ., This is consistent with previous reports that O . montana clears itself of infection at a higher rate than X . cheopis 10 , 19 , 20 ., Although complete blockage is not required for transmissibility , the blockage rate is often used as a surrogate marker for vector potential by the proventricular biofilm-dependent mechanism ., Like the infection rate , the blockage rate was also lower for O . montana ( 17 to 25% ) than for X . cheopis ( 30 to 40% ) ; however , blockage rates of stably infected fleas ( rather than the total that took an infectious blood meal ) were roughly equivalent for the two species ., In this study , we monitored transmission by a population of infected fleas , which allowed an overall comparison of transmission dynamics and kinetics ., To directly calculate the vector efficiency , it will also be necessary to monitor the vector potential and transmission potential of individual fleas after infection ., We are currently adapting our model system for this purpose ., Despite its lower infection and blockage rates , O . montana transmitted greater numbers of Y . pestis than did X . cheopis in mass transmission experiments , and transmission peaked earlier ., Several factors could account for these results ., O . montana developed proventricular blockage sooner than X . cheopis , and blocked O . montana survived significantly longer than blocked X . cheopis ., The blocking-surviving potential , defined as the mean day of death after becoming blocked divided by the mean day of becoming blocked after an infectious blood meal , has been described as an important component of flea vector efficiency 11 , 18 ., Based on the results shown in Fig 3 , the calculated blocking-survival potential of O . montana ( 0 . 73 ) is 5-fold higher than that of X . cheopis ( 0 . 14 ) ., It is also important to note that , as mentioned previously , complete blockage is not required for transmission—partially blocked fleas can also transmit efficiently 7 ., Partially blocked fleas , with fresh blood in the esophagus but some also in the midgut , were observed frequently during transmission trials ( Tables 1 and 2 ) ., Complete blockage appeared to develop more gradually in O . montana or was more ephemeral , as sometimes a small amount of blood appeared to seep through into the midgut when a flea previously diagnosed as completely blocked fed again ., Differences in foregut anatomy may also enhance transmission efficiency of O . montana ., The esophageal-proventricular junction is much broader in blocked O . montana than in blocked X . cheopis ( Fig 5 ) ., The increased esophageal distension observed in fully blocked O . montana would be expected to expose a greater surface area of the infectious biofilm to contact with incoming blood during a feeding attempt , potentially enhancing regurgitative transmission ., This may in part account for the greater number of CFUs recovered from O . montana mass transmission experiments ., We have hypothesized that foregut anatomical differences may also account for the larger numbers of Y . pestis transmitted by X . cheopis than by the cat flea Ctenocephalides felis 24 ., A third objective of this study was to evaluate , in the same cohorts of infected fleas , the relative importance of the two transmission mechanisms ., The transmission efficiency of both O . montana and X . cheopis , defined here as the number of Y . pestis transmitted per infective flea , was very low by the early-phase mechanism compared to later proventricular biofilm-dependent transmission ., In half of the day-3 early-phase transmission trials , no Y . pestis CFUs were recovered from the blood reservoir fed upon by >100 fleas ., Based on trials in which we added known numbers of Y . pestis to the blood in the feeding system , the expected recovery rate is ~95% ., Thus , we estimate that 0 to 5 CFU were transmitted in the negative early-phase transmission experiments ., Our results are in line with previous work indicating that early-phase transmission is inefficient ., In 1907 the Indian Plague Commission reported that only 1 of 67 fleas that had fed on a septicemic plague rat and then individually placed on naïve rats transmitted plague in the early phase 1 ., Recent studies estimate a ~0 to 10% probability of a single O . montana or X . cheopis flea transmitting by the early-phase mechanism 16 , 26–30 ., These estimates were based on both disease incidence and seroconversion following challenge by groups of ten fleas , indicating that the number of Y . pestis transmitted was sometimes at or below the LD50 , estimated at 1 to 10 CFU for the highly susceptible laboratory mice used ., The bite of a single blocked X . cheopis results in transmission 25 to 50% of the time 9–12 ., The number of CFU transmitted by the bite of a blocked X . cheopis is highly variable , ranging from <10 to several thousand 10 , 13 ., No data are available on the transmission efficiency of partially blocked fleas or the number of CFUs they transmit ., Our system allowed us to monitor transmission by cohorts of infected fleas ( “mass” transmission ) ., This allowed an overall comparison of transmission dynamics and kinetics and vector potential at the population level , but not the transmission rate ( the percentage of fleas that transmitted ) or transmission potential ( defined above ) ., Consistent with our results , however , a recent study estimated that the percent transmission efficiency of an individual O . montana flea is lower in the early phase compared to later time points after infection 28 ., Early-phase transmission has been proposed to largely account for epizootic spread by flea vectors that purportedly do not readily become blocked 33 , 34 ., However , it is recognized that comparative data regarding transmission by the proventricular biofilm ( “blockage” ) mechanism are limited and problematic , and warrant reexamination 5 , 20 , 35 ., A variety of experimental conditions and designs have been used with respect to infectious blood meal source , infectious dose , and flea maintenance conditions , all of which are known to influence infection and transmission dynamics ., In some cases , small numbers of fleas were used that were likely poorly adapted to laboratory conditions ., As in the case of O . montana , results have sometimes been inconsistent ., Furthermore , early-phase transmission has only been demonstrated from fleas that fed on blood with a very high bacteremia to highly susceptible laboratory rodents ( ID50 <10 CFU ) 36 ., The California ground squirrel , the major host of O . montana , reportedly has an ID50 of >250 CFU 37 ., Based on the previous reports that O . montana rarely blocks or transmits beyond the early phase , a recent study that modeled O . montana-ground squirrel plague made the assumption that only early-phase transmission was important , and that transmission beyond that was negligible 38 ., Our results indicating that very few bacteria are transmitted early ( less than the reported ID50 of ground squirrels ) , but that subsequent transmission is robust , suggest that the converse assumption is probably more realistic ., However , factors specific to different ecological settings and host and vector populations may also affect transmission dynamics ., In this study , we present a standardized , stringently controlled model system to more reliably compare vector efficiency and to monitor transmission dynamics of a population of infected fleas ., Key elements of this experimental system and their rationale include:, 1 ) Fleas are infected on a specific blood source containing comparable concentrations of Y . pestis ., Both the type of blood used and the bacteremia level significantly affect the infection potential and incidence of blockage 13 , 19 , 39 ., Use of an artificial feeding device allows the infectious blood meal to be matched in different experiments ., 2 ) A subset of the same flea cohort used for infection is fed on sterile blood of the same type for use as uninfected controls ., Infected and uninfected fleas are kept in the same environment and provided the same type and frequency of maintenance feeds ., The mortality of uninfected control fleas after 4 weeks should be low ., If not , the fleas were physiologically stressed and results based on them are unreliable ., Mortality of infected fleas should also be recorded , as it is a surrogate indicator of blockage-induced starvation ., 3 ) Fleas are individually examined immediately after the infectious blood meal and only those that took a full blood meal are included in the study ., The mean initial infectious dose acquired by the fleas is determined from a sample of these fleas , collected immediately after the infectious blood meal ., Some flea species may be reluctant to feed on other than their natural blood source , resulting in a lower infectious dose and subsequent lower infectivity rate ., 4 ) Fleas are examined microscopically immediately after each maintenance feed for evidence of partial or complete proventricular blockage using good quality optics and light source ., Fluorescence microscopy of the digestive tract dissected from fleas infected with GFP-expressing Y . pestis is very useful to determine proventricular blockage status ., This is particularly helpful for fleas that have a darkly pigmented exoskeleton that is less transparent to direct microscopic visualization of the esophagus , proventriculus , and midgut 24 ., 5 ) Infection rate and bacterial load are monitored by plate counts of flea samples collected at different times after the infectious blood meal ., 6 ) Transmission dynamics are monitored for a population that received the same infectious blood meal during a timeframe that encompasses both modes of transmission ., Our results resolve a long-standing controversy about the susceptibility of O . montana to become blocked and to transmit Y . pestis by the proventricular biofilm-dependent mechanism ., We previously used this experimental system to show that C . felis , normally a poor vector by either mechanism , readily becomes blocked and transmits if its usual daily feeding behavior is altered 24 ., The transmission dynamics of other flea vector species can be systematically reevaluated by using this system , with the important prerequisite that appropriate laboratory maintenance conditions can be established for them . | Introduction, Methods, Results, Discussion | Transmission of Yersinia pestis by flea bite can occur by two mechanisms ., After taking a blood meal from a bacteremic mammal , fleas have the potential to transmit the very next time they feed ., This early-phase transmission resembles mechanical transmission in some respects , but the mechanism is unknown ., Thereafter , transmission occurs after Yersinia pestis forms a biofilm in the proventricular valve in the flea foregut ., The biofilm can impede and sometimes completely block the ingestion of blood , resulting in regurgitative transmission of bacteria into the bite site ., In this study , we compared the relative efficiency of the two modes of transmission for Xenopsylla cheopis , a flea known to become completely blocked at a high rate , and Oropsylla montana , a flea that has been considered to rarely develop proventricular blockage ., Fleas that took an infectious blood meal containing Y . pestis were maintained and monitored for four weeks for infection and proventricular blockage ., The number of Y . pestis transmitted by groups of fleas by the two modes of transmission was also determined ., O . montana readily developed complete proventricular blockage , and large numbers of Y . pestis were transmitted by that mechanism both by it and by X . cheopis , a flea known to block at a high rate ., In contrast , few bacteria were transmitted in the early phase by either species ., A model system incorporating standardized experimental conditions and viability controls was developed to more reliably compare the infection , proventricular blockage and transmission dynamics of different flea vectors , and was used to resolve a long-standing uncertainty concerning the vector competence of O . montana ., Both X . cheopis and O . montana are fully capable of transmitting Y . pestis by the proventricular biofilm-dependent mechanism . | The ecology of plague is complex and its epidemiology is enigmatic ., Many different flea species are able to transmit Yersinia pestis , the plague bacillus , and they can transmit in two different ways ., Early-phase transmission can occur during the first week after a flea has fed on a diseased animal ., Thereafter , transmission occurs only as bacterial growth in the flea foregut interferes with and eventually blocks blood feeding ., Comparisons of the relative ability of different flea vectors to transmit have been problematic , and contradictory results have been reported for the ability of the ground squirrel flea Oropsylla montana to transmit beyond the early phase ., Our results show that O . montana readily develops foregut blockage , and transmission by that mechanism was as good as or better than observed for Xenopsylla cheopis , a flea known to block at a high rate ., In contrast , very few bacteria were transmitted in the early phase by either of these fleas compared to later times after infection , suggesting that early-phase transmission is pertinent only to highly susceptible animals ., Improved characterization of the transmission patterns of different flea vectors will aid in modeling plague incidence in its various natural settings . | death rates, united states, invertebrates, medicine and health sciences, plagues, body fluids, pathology and laboratory medicine, demography, pathogens, geographical locations, microbiology, vector-borne diseases, animals, north america, bacterial diseases, yersinia, fleas, bacteria, bacterial pathogens, digestive system, infectious diseases, yersinia pestis, medical microbiology, microbial pathogens, montana, insects, hematology, arthropoda, people and places, gastrointestinal tract, blood, anatomy, physiology, biology and life sciences, organisms, esophagus | null |
journal.pcbi.1004515 | 2,015 | At the Edge of Chaos: How Cerebellar Granular Layer Network Dynamics Can Provide the Basis for Temporal Filters | Many models of the cerebellum assume that the granular layer recodes its mossy-fiber inputs into a more diverse set of granule-cell outputs 1–4 ., It is further assumed that the recoded signals , which travel via granule-cell ascending axons and parallel fibers to Purkinje cells and molecular layer interneurons , are appropriately weighted using plastic synapses and then combined to produce the particular Purkinje cell outputs that are required for any given learning task ., Recoding in these models thus enables a given set of mossy-fiber inputs to generate one of a very wide variety of Purkinje cell outputs , giving the model demonstrable computational power ( e . g . 5 ) ., Although this framework is seen as plausible in broad outline ( e . g . 6 , 7 ) , the details of its workings are far from established 8 ., Relatively simple top-down models have shown that theoretically well-understood recoding schemes such as tapped delay lines , spectral timing , Gaussians , sinusoids , and exponentials can be effective , but do not establish how they could be implemented biologically ( references in 8–10 ) ., In contrast , more complex bottom-up models of recurrent inhibitory networks representing the connectivity between granule and Golgi cells are closer to biological plausibility , but have been used for very specific tasks such as eye-blink conditioning so that their general computational adequacy is unknown 11–20 ., In part this is because eyeblink conditioning requires a response only at the time the unconditioned stimulus arrives ., Eyelid ( or nictitating membrane ) position is not specified either for the period between the conditioned and unconditioned stimulus , or for the period ( possibly some hundreds of milliseconds ) after the unconditioned stimulus has been delivered ., In contrast , for a task such as the vestibulo-ocular reflex eye-position is very precisely specified for as long as the head is moving , and afterwards for as long as gaze has to be held constant ., Thus , cerebellar output—and hence granular-layer output—is more tightly constrained in motor-control tasks resembling the vestibulo-ocular reflex than in eyeblink conditioning 3 ., Here we combine elements of top-down and bottom-up approaches , by investigating whether the outputs of neural networks that incorporate the recurrent inhibition observed in the granular layer can be linearly combined to generate continuous filter functions which are computationally useful for example in vestibulo-ocular reflex adaptation 9 ., The split between a complex representation layer ( granular layer ) and a linear reconstruction layer ( perhaps corresponding to the plastic synapses between granule cells and Purkinje cells or molecular-layer interneurons ) is similar to the structure employed in reservoir computing 21 , and it is convenient to use terminology and methods from that field in analyzing these networks ( see Methods ) ., We begin by analyzing the case of a one-layer network with recurrent inhibition 15 ., This is simpler than the real granular layer in which feedback is provided via a second layer of Golgi cell interneurons , but is worth analyzing separately because it allows us to test the hypothesis , suggested by the reservoir computing metaphor , that the crucial parameter in determining the time extension of responses is the mean amount of feedback in the network , and how closely this parameter is tuned to the edge-of-chaos 22 ., This degree of tuning can be measured by the Lyapunov exponent ., Generally speaking , if there is very little recurrent feedback in a network , then responses will be highly stable and die away very quickly over time , while for large amounts of feedback the responses can be chaotic or even unstable ., The Lyapunov exponent ( see Methods ) is a quantitative measure of stability because it captures the rate of growth or decay of small perturbations ., In linear systems negative values imply stability , while positive values imply instability ., In non-linear systems , small , negative values of Lyapunov exponents can be especially interesting , since they can signal the ‘edge-of-chaos’ , where there are long-lasting and possibly complex responses to transient inputs ., We show that this is the interesting region for our reconstruction problem ., One novel feature of this contribution is its use of generic colored noise inputs , rather than the stereotyped pulse or step inputs that are usually considered ., These colored-noise inputs are essential for motor control applications such as the VOR , where they are needed to demonstrate that the filter can process generic vestibular signals ., A second novel feature is the use of statistical techniques that allow us to evaluate the ability of the network to approximate the range of linear filters required for these applications ., While previous work on reservoir networks focused on generic inhibitory and excitatory networks 22–30 this is the first work to systematically examine stability and reservoir performance in networks dominated by recurrent inhibition like the granular layer while also taking into account the effects of cerebellar network properties on filter approximations ., To achieve this we extend the model to two populations in order to represent inhibition via Golgi cells ., We also test the effect of other non-generic features of the cerebellum such as the newly discovered functional feature of Golgi cell inhibition by mGluR2 receptor activated GIRK channels 31 , 32 and Golgi cell afferent excitation often neglected in cerebellar simulations ., Furthermore we also evaluate the effect of output-feedback to the granular layer through the nucleocortical pathway ., The one-population model used in this study ( Fig 1A ) was based on that of Yamazaki and Tanaka 15 ., It consisted of Nz = 1000 granule cells , each receiving excitatory afferent inputs Ii ( t ) derived from the external signal x ( t ) , and recurrent inhibitory inputs from other cells ., The model neurons were firing-rate ( i . e . non-spiking ) , and the output zi ( t ) of the i-th neuron at time t was given by, zi ( t ) =Ii ( t ) −∑\u200bjNzAijwij∑\u200bs=1texp ( −t−sτw ) zj ( s−1 ) +nNi ( t ) +, ( 1 ), ( here the bracket notation +is used to set negative values to zero , preventing the firing-rate of a neuron from becoming negative ) ., This equation describes ( see Fig 1A ) memory-less rate-neurons connected by single-exponential synaptic process with time constant τw so that neuron i sums past inputs zj ( s − 1 ) , 1≤ s ≤ t from other neurons , exponentially weighted by distance s − t into the past ., Neuron j has synaptic weight Aij wij on neuron i where Aij was set to 1 with probability a and 0 otherwise , hence the parameter a controls the sparsity of the connectivity ., The connectivity strengths wij were drawn from a normal distribution with mean w and standard deviation vww , normalized by population size Nz , and constrained to be positive , so that wij\xa0=\xa02Nz ( w±vww ) + ., Each neuron received an excitatory input Ii ( t ) with additive noise nNi ( t ) ( here Ni ( t ) is a discrete white noise process with std ( N ) = 1/2 so that the added noise is smaller in magnitude than the noise amplitude n 95% of the time ) ., In the simulations , unless otherwise specified , we used the following default values for the parameters above ., The population size was Nz = 1000 ., The probability of connectivity was a = 0 . 4 ( close to the value 0 . 5 in Yamazaki and Tanaka 15 ) , and synaptic variability was set to zero ( vw = 0 ) ., The default input noise level was n = 0 . This model had a single time constant which Yamazaki and Tanaka 15 took to be equal to the membrane time constant of Golgi cells in their simulations of granular layer dynamics ., However it is not clear that this is the relevant time constant for a firing-rate model since the dynamics of the sub-threshold domain cannot be easily carried over into the supra-threshold ( spiking ) domain and are often counter intuitive ., While a still prevailing misconception is that long membrane time-constants are equal to a slow spike response , the exact opposite is the case: integrate-and-fire with an infinite time constant ( perfect integrators ) have the fastest response time to a current step and can respond almost instantaneously 33 ., Since temporal dynamics of neurons in a network are primarily determined by the time course of the synaptic currents 33–36 we have ignored membrane time constants in this and following models and instead related τw to the synaptic time constant of recurrent inhibition in the network ., We further want to note that the values for the synaptic time constants were not directly adjusted to replicate results for individual electrophysiological studies but rather kept at general values to study the effect on network output of interaction between different magnitudes of time constants ., This issue is considered further in the Discussion ., To allow for more realistic modeling of the dynamics of the granular layer we extended the one-population network of granule cells above to include inhibition via a population of interneurons corresponding to Golgi cells ( see Fig 1B ) ., In this model the firing-rates zi ( t ) of granule cells and qi ( t ) of Golgi cells were given by, zi ( t ) =Ii ( t ) −∑\u200bjNqAijwij∑\u200bs=1texp ( −t−sτw ) qj ( s−1 ) +L ( t ) +, ( 2 ), qi ( t ) =g⋅Ii ( t ) +∑jNzBij ( uij∑s=1texp ( −t−sτu ) zj ( s−1 ) −mij∑s=1texp ( −t−sτm ) zj ( s−1 ) ) +L ( t ) +, ( 3 ), The default sizes for the two populations were Nz = 1000 and Nq = 100 ., As before , the excitatory afferent input into a granule cell i was given by Ii ( t ) , however the two-population model also had direct afferent excitation gIi ( t ) of Golgi cells ., The factor g setting the level of excitation was set to 0 in the initial simulations , resulting in no afferent excitation for Golgi cells ., The output-feedback L ( t ) was 0 until later simulations ( see below ) ., The connectivity between the two populations was given by the random binary connection matrices W and U , however in this model the connectivity was not defined by a probability but by the convergence ratios cw = 4 between Golgi and granule cells and cu = 100 vice versa ., Thus exactly 4 randomly selected Golgi cells inhibited each granule cell and 100 randomly chosen granule cells were connected to each Golgi cell ., The weight of GABAergic inhibition between Golgi and granule cells was drawn from a normal distribution and normalized with wij\xa0=\xa02cw ( w±vww ) + ( default vw = 0 ) and the time constant of inhibition was given by τw ., Besides the glutamatergic excitatory connections between granule and Golgi cells with weight uij\xa0=\xa02cu ( u±vuu ) + ( default vu = 0 ) and time constant τu the model was extended to emulate the inhibitory effect of mGluR2 activated GIRK channels 31 with mij\xa0=\xa02cb ( m±vmm ) + ( default m = 0 , vm = 0 ) and time constant τm = 50ms ., Note that mGluR2 inhibition was not used until later simulations with m = 0 . 003 ., Additional simulations were conducted with only half of the Golgi cells receiving mGluR2 inhibition i . e . Pr ( m = 0 ) = 0 . 5 ., In all simulations u was set to 0 . 1 and normalized by the excitatory time constant resulting in u = 0 . 1/τu ., All network simulations were written in C and were integrated into Python by transforming them into dynamically linked extensions with the package distutils ., The stepsize in all simulations was dt = 1ms ., All results were analyzed using Python ., All models , methods and simulation results are available from the github repository https://github . com/croessert/ClosedLoopRoessertEtAl ., A snapshot of the model code can also be found on ModelDB: https://senselab . med . yale . edu/modeldb/ShowModel . asp ?, model=168950 ., Computational resources for the simulations were partially provided by the ICEBERG cluster ( University of Sheffield; access granted by the INSIGNEO Institute for in silico Medicine ) ., The modulated input to each cell was given by the excitatory input Ii ( t ) = I0i + f ∙ 0 . 1 ∙ I0i ∙ x ( t ) + ., Unless noted otherwise the input I0i was chosen from a normal distribution with mean 1 and default standard deviation vI = 0 . 1 ., To test increased input variability , standard deviation was increased to vI = 2 in a later experiment ., The factor f , randomly picked as either 1 or -1 defined whether the input was inverted or not ., This type of input coding , here termed “push-pull” coding can be routinely found for example in the vestibulo-cerebellum where half of the cells are ipsilateral preferring ( f = 1 , type I ) or contralateral preferring ( f = −1 , type II ) 37 ., In order to test the ability of the network to construct a linear filter with a given impulse response it is not sufficient to use impulse inputs alone , since this does not test linearity ( for example the response to two successive impulse inputs may not be the sum of the individual responses ) ., For this reason we also used random process inputs that mimic behavioral inputs ., The input signal x ( t ) consisted of 3 parts ( see Fig 1C ) ., The first part was a training sequence of a 5 second band-passed white noise signal ( low-passed with a maximum frequency of 20 Hz ) 38 chosen to mimic head velocity in the behaviorally relevant frequency range of 0–20 Hz 39 ., Additionally a 5 second silent signal ( x ( t ) = 0 ) was added to the training sequence to train a stable response ., Training with a segment of null data finds weights which not only give the appropriate impulse response but also produce zero output for zero input data , so that they reject spontaneous modulatory activity in the network ., Consecutively the previous signals were repeated with a different realization of the noise signal to test the quality of the filter construction ., The third part was an impulse test signal where x ( t ) = 0 apart from a brief pulse of 50 ms where x ( t ) = 1 . The colored noise signal was normalized to std ( x ) = 1/2 which ensured that the amplitude 0 . 1 ∙ I0i included the input 95% of the time ., To assess the ability of the network to implement linear filters that depend on the past history of the inputs , the output signals zi ( t ) of all granule cells during the training sequence were used to construct exponential ( leaky integrator ) filters y ( t ) = F * x ( t ) of increasing time constants as linear sums y ( t ) = ∑βizi ( t ) of granule cell outputs ., This can be regarded as the output of an artificial Purkinje cell that acts as a linear neuron ., In matrix terms ( writing time series in columns ) this expression can be written y −Zβ where the undetermined coefficients β are usually fitted by the method of least squares to minimize root sum square fitting error, ∥y−Zβ∥2=∑t ( y ( t ) −Σ\u200b\u200bβizi ( t ) ) 2\u200b, ( 4 ), However over-fitting of the data , due to the large output population , can make this method misleading and give excessively high estimates of reconstruction accuracy ., To avoid this problem we used the method of LASSO regression taken from the reservoir computing literature ., This is a robust fitting procedure that includes a regularization term to keep the reconstruction weights small 40 , 41 ., Here , the estimates are defined by β^\xa0=\xa0argminβy-Zβ2+αβ1 which is the least-squares minimization above with the additional constraint that the L1-norm ||β||1 = ∑βi of the parameter vector is also kept small ., In practice we find that up to about 90% of weights are effectively zero using this method ., In contrast to ridge regression that employs a L2 -norm penalty and is commonly used to prevent over-fitting in reservoir computing 27 LASSO regression produces very sparse weight distributions ., This corresponds well to the actual learning properties of the Purkinje cell , approximated as a linear neuron , in which optimality properties of the learning rule with respect to input noise force the majority of synapses to silence 42–45 ., We fitted three responses yj ( t ) = x ( t ) * Fj ( t ) with j = 1 , 2 , 3 and with Fj ( t ) = exp ( −t/τj ) being one of three exponential filters τ1 = 10ms , τ2 = 100ms or τ3 = 500ms ( see Fig 1D ) ., The regularization coefficient was set to α = 1e−4 which gave best maximum mean goodness-of-fit results for the one-population model with τw = 50ms ( not shown ) ., LASSO regression was implemented using the function sklearn . linear_model . Lasso ( ) from the python package scikit-learn 46 ., In general the estimated weights βi take both positive and negative values , which is not compatible with the interpretation of equation ( 4 ) above as parallel fiber synthesis by Purkinje cells ., The use of negative weights is usually justified by assuming a relay through inhibitory molecular interneurons 42 , 44 ., To test whether learning at parallel fibers alone is sufficient for the construction of filters from reservoir signals we additionally employed LASSO regression with only positive coefficients ( positive-LASSO ) as a comparison ., As a measure of the quality of filter construction , the weights estimated from the training sequence were used to construct the filtered responses in the test sequence and the goodness-of-fit between expected output and constructed output was computed for each filter using the squared Pearson correlation coefficient ( R2 ) 47 ( see Fig 1D ) ., For the final goodness-of-fit measure the mean of 10 networks with identical properties but with different random connections was computed ., A convenient way to analyze the stability or chaoticity of a dynamic system is the Lyapunov exponent λ ., It is a measure for the exponential deviation of a system resulting from a small disturbance 25 and a value larger than 0 indicates a chaotic system ., The Lyapunov exponent was measured empirically , similar to Legenstein and Maass 22 by calculating the average Euclidian distance dt\xa0=\xa0∑i\xa0=\xa01Nzzit-zi ( t ) 2 between all granule cell rates zi ( t ) from a simulation where x ( t ) = 0 and the rates zi ( t ) from a second simulation where the input was disturbed by a small amount at one time step , i . e . x ( 0 ) = 10−14 ., This state separation simulation was repeated for 10 randomly connected networks but otherwise identical parameters and λ was estimated from the mean average Euclidian distance d-t with λ\xa0=\xa0log2mean ( d-t\xa0=\xa02 . 01s:2 . 11s ) /mean ( d-t\xa0=\xa00 . 01s:0 . 11s ) /2s ., To estimate the transition between stability and chaos we were mainly interested in the sign of the Lyapunov exponent ., Although taking the mean of a 100 ms period and using a relatively large Δt of 2s 24 decreases the accuracy of the Lyapunov estimation , it was used here to prevent errors in the estimation of the sign ., The edge-of-chaos was defined as the point where λ crosses 0 for the first time when traversing in the direction of strong inhibition w to weak and therefore from high λ to low ., To model putative output-feedback to the reservoir via the nucleocortical pathway the signal L ( t ) = f ∙ oi ∙ −∑βizi ( t ) was injected into 20% of all granule and Golgi cells in the last simulations ., The factor f was randomly picked as either 1 or -1 to model 50% excitation and inhibition and the weight was drawn from a normal distribution with oi = 1e−4±1e−5+ ., In these simulations only the case for output-feedback of the slowest filter signal is shown ., Thus βi are the weights needed to construct the filter with τ3 = 500ms ., As noted in the reservoir computing literature 27 , 48 , 49 output-feedback in general is a very difficult task since it leads to instability ., Therefore the weights βi were not learned online but a method called teacher forcing with noise was applied 27 ., The weights βi were learned in a prior step by using the teacher signal L′ ( t ) = f ∙ oi ∙ −y3 ( t ) ∙ N ( t ) instead of the feedback signal L ( t ) to uncouple the instable learning ., Here y3 ( t ) is the target response for the slowest filter ( Fig 1D ) and N ( t ) is a discrete white noise process that helps to increase the dynamical stability 27 ., The quality of filter construction and the Lyapunov exponent were estimated in a second simulation using the previously learned weights βi for filter construction and the feedback signal L ( t ) ., In the first part of this study we focused on the one-population rate-neuron model previously published by Yamazaki and Tanaka 15 ., While in this previous study the model was used to represent the passage of time , i . e . an internal clock , we now show that it is also possible to use its output to construct exponential filters with various time-constants ., To illustrate the dependence of network stability regime on the amount of feedback we begin by presenting sample impulse responses ( Fig 2 , second row ) for a network ( Fig 2 , top ) with intermediate time constant τw = 50ms and with three values of the recurrent inhibition: w = 0 . 01 , lying in the highly stable region , w = 1 . 4 , close to the edge-of-chaos , and w = 3 , in the chaotic region ., When the weight w was low , ( Fig 2A , w = 0 . 01 ) the network was highly stable to perturbations and showed no long lasting responses ., Close to the edge-of-chaos ( Fig 2B , w = 1 . 4 ) complex , long lasting responses were present ., For larger weights ( Fig 2C , w = 3 ) the network entered a chaotic state in which cells showed random activity without further input modulation ., We further illustrate this dependence in the last two rows of Fig 2 which shows filter constructions ( see Methods ) for three target exponential filters with time constants τi of 10 ms ( Fig 2D1 and 2E1 ) , 100 ms ( Fig 2D2 and 2E2 ) and 500 ms ( Fig 2D3 and 2E3 ) ( chosen to cover the range of performance required for e . g . VOR plant compensation 9; filter construction of intermediate time constants are not shown , but are generally of similar quality ) ., It is clear that in the highly stable regime only fast and intermediate time constant responses could be reconstructed ( dotted light lines ) ., Near the edge-of-chaos acceptable reconstructions were possible at all three time constants ( dark lines ) , and in the chaotic regime reconstruction was always inaccurate and showed oscillatory artifacts ( solid light lines ) ., While Yamazaki and Tanaka 15 argued that this chaotic network state is the preferred network state to implement an internal clock ( compare Fig 2C with Fig 1 from 15 ) these results show that it is disadvantageous when a filter of a continuous signal has to be implemented ( see Discussion ) ., We have noted above ( Methods ) that accurate reconstruction of the impulse response of a linear filter does not imply that the output for other inputs is correct; this requires linearity of the reconstructed filter ., Linearity of the reconstructed filters is investigated in the second row of Fig 2 by comparing their effects on a band-passed noise signal with that of the exact filter ( plotted in black ) , again for time constants τj of 10 ms ( Fig 2D1 ) , 100 ms ( Fig 2D2 ) and 500 ms ( Fig 2D3 ) , It is clear that the reconstruction in the stable regime or the chaotic regime ( light lines ) were much less accurate than in the edge-of-chaos-regime ( dark lines ) ., Note these plots show the response to a test input ( rather than the training input , see Methods ) ., The regularized fitting method used ( LASSO regression , see Methods ) tends to use weights that are as small as possible ., This property is clear in our example , to construct filters from granule cell signals at the edge-of-chaos only a small subset of granule cell responses were necessary ., For the filters with 10 , 100 and 500 ms ( Fig 2D and 2E; w = 1 . 4 ) , the percentage of weights being equal to zero was 90% , 86% and 75% , respectively , and the mean of non-zero weights was 5 . 5 and 11 . 7 and 52 . 6 , respectively ., The high proportion of silent synapses is consistent with experimental findings ( see Discussion ) As discussed previously , the value of w corresponding to the edge-of-chaos can be identified using the Lyapunov exponent ( see Methods ) ., We illustrate this property by investigating the dependence of filter reconstruction accuracy on the Lyapunov exponent ( Fig 3 ) ., Results are shown for three networks with different time constants for the recurrent inhibition: τw = 10ms ( column 1 ) , τw = 50ms ( column, 2 ) Fig 2B and τw = 100ms ( column, 3 ) approximately corresponding to the ranges of membrane and synaptic time constants present in the granular layer ., The top row of Fig 3 shows the Lyapunov exponent of each network plotted against the amount of recurrent inhibition w ., In each case there was a point at which the exponent crossed the zero axis , corresponding to the edge-of-chaos value for that network time constant ., It can be seen that the amount of recurrent inhibition needed decreased as the time constant increased ., The bottom row shows the effect of w on reconstruction accuracy ( measured by R2 goodness-of-fit ) for exponential filters with the three time constants considered previously: τj = 10ms ( blue lines ) , 100ms ( green lines ) and 500ms ( red lines ) for each network ., Performance strongly depended on the weight of the recurrent inhibition ., The goodness-of-fit was best , especially for filters with time constants longer than the internal inhibitory time-constant , for networks close to the edge-of-chaos , just before the transition from stable to chaotic behavior ., Other observations were that while , as expected , the goodness-of-fit for slow filters , e . g . 500ms , increased with the ( inhibitory ) time constant , the performance for fast filters decreased slightly ( Fig 3C2 ) ., Furthermore the performance was best if the inhibitory time constant was equal to the time-constant of the filter ( Fig 3A2 , τ1 = 10ms blue line; Fig 3C2 , τ2 = 100ms green line ) ., Fig 4 investigates the robustness of the properties described above to moderate levels of additive noise and to variability in input signal levels and synaptic weights ., While white noise with amplitude of a = 0 . 01 ( noise amplitude equal to 10% of the input modulation amplitude ) lead to a reduction in goodness-of-fit ( Fig 4A1 ) the principal mechanism of filter construction was not disrupted and the edge-of-chaos was only shifted to larger weights w ( Fig 4A2 ) ., Increasing the between-neuron variability of the mean input excitation to a high value of e . g . vI = 2 ( i . e . 95% of constant input increased to 0–5 from 0 . 8–1 . 2 for default value vI = 0 . 1 ) ( Fig 4B1 solid dark lines ) had almost no benefit for the goodness-of-fit while shifting the edge-of-chaos to larger weight values ., In contrast , imposing larger variability in the inhibitory weight with vw = 2 ( i . e . 95% of weights between 0 and w+4w ) shifted the edge-of-chaos in the opposite direction—towards lower weights ( Fig 4B2 , dotted lines ) , and the quality of filter construction was increased ( Fig 4B1 , dotted lines ) ., This phenomena may be caused by a proportion of input signals or weights being driven to zero due to the positive cut-off which effectively leads to some cells receiving no input and a reduction of connectivity , respectively ., To test the effect of reduced connectivity we examined the direct effect of increased sparseness on reservoir performance ( Fig 4C ) ., Two methods were used to increase sparseness: the first was to decrease the convergence of inhibition to 40 cells ( Fig 4C1 , dotted lines ) by decreasing the network connectivity from a = 0 . 4 to a = 0 . 04 while keeping the network size at Nz = 1k ., The second way was to increase the network size to Nz = 10k while keeping convergence constant at 400 cells ( Fig 4C2 , solid dark lines ) with a = 0 . 04 ., While both cases resulted in an improvement of filter quality , a smaller convergence slightly outperformed an increased network size suggesting that a sampling from less cells is more beneficial since it leads to a higher diversity and variability ., An important requirement for filter construction turned out to be push-pull coding , found for example in the vestibulo-cerebellum , where half of the input signals are inverted ( see Discussion ) ., When the input did not include inverted signals the responses from individual granule cells showed almost no variety in damped oscillations in response to pulse input ( Fig 5A ) ., This consequently lead to an impairment of filter construction performance especially for larger filter time-constants and a shift of the edge-of-chaos to lower weights w ( Fig 5B1 , dark lines ) when compared to the control case ( light lines ) ., Although filter construction performance was only slightly reduced when using regression with positive coefficients only ( see Methods ) ( Fig 5C , light lines ) when push-pull input was present , without push-pull input filter construction quality was heavily reduced ( Fig 5C , dark lines ) ., While the previous model was able to show the principles of filter construction from a simplified model of the granular layer with recurrent inhibition , it did not take into account the fact that inhibition in the granular layer is relayed via a second population of cells , i . e . Golgi cells ., To investigate the effects of this arrangement we extended the one-population model to a two-population model ., The connectivity of the extended model was based on plausible convergence ratios of cw = 4 between Golgi and granule cells and cu = 100 vice versa 50 ., Additional parameters were excitatory time constant τu and the weight of excitation u ( Fig 6 , top ) ., Increasing τu while keeping the inhibitory time constant at τw = 50ms showed that the performance of the two-population model was very similar to the one-population model if the excitation is fast ( Fig 6A1 ) ., However , increasing the excitatory time constant improved the quality of the constructed slow filter ( τ = 500ms ) at the expense of the faster filters ( τ = 10ms and τ = 100ms ) ( Fig 6B1 and 6C1 ) ., Additionally , this leads to a lowered gradient of the Lyapunov exponent ( Fig 6B2 and 6C2 ) ., We therefore focus in the following on the best-case scenario of τu = 1ms and τw = 50ms ., As in the one-population model before ( Fig 2B ) , responses of single granule and Golgi cells in networks close to the edge-of-chaos featured complex but stable , long lasting damped oscillations ( not shown ) ., In Fig 6D we show that increased sparseness , achieved by reducing the convergence onto Golgi cells from cu = 100 to cu = 10 ( light lines ) increased the quality of constructed filters as in the previous model ., However , this time , increasing the granular cell population size to Nz = 10k ( dotted lines ) has almost no beneficial effect , which can be attributed to the bottleneck effect of the small Golgi cell population of Nq = 100 ( compare to Fig 4C , dotted lines ) ., Here , many granule cell responses converge onto a lower dimension of signals , which decreases the fidelity ., On the contrary increasing the granule cell as well as the Golgi cell population size to Nq = Nz = 10k increased the filter construction performance similar to before ( results not shown ) ., Equally , further reducing the Golgi cell population to Nq = 10 for the default case ( Nz = 1k , cu = 100 ) enforced the bottleneck effect and strongly decreased the construction quality of slow filters ( results not shown ) ., The effects of synaptic-weight variability in the two-population model differed for excitatory and inhibitory weights ( Fig 6E ) ., Adding a large variability to excitatory weights vu = 4 increased the goodness-of-fit ( light lines ) just as seen in the model before ., However , adding variability to inhibitory weights vw = 4 decreased the quality of constructed filters ( dotted lines ) ., This can be explained by the low number of connections between Golgi and granule cells of cw = 4 ., Using equal convergence of cw = cu = 20 gave equal effects in increased filter quality with increased variability for excitatory and inhibitory weights ( results | Introduction, Methods, Results, Discussion | Models of the cerebellar microcircuit often assume that input signals from the mossy-fibers are expanded and recoded to provide a foundation from which the Purkinje cells can synthesize output filters to implement specific input-signal transformations ., Details of this process are however unclear ., While previous work has shown that recurrent granule cell inhibition could in principle generate a wide variety of random outputs suitable for coding signal onsets , the more general application for temporally varying signals has yet to be demonstrated ., Here we show for the first time that using a mechanism very similar to reservoir computing enables random neuronal networks in the granule cell layer to provide the necessary signal separation and extension from which Purkinje cells could construct basis filters of various time-constants ., The main requirement for this is that the network operates in a state of criticality close to the edge of random chaotic behavior ., We further show that the lack of recurrent excitation in the granular layer as commonly required in traditional reservoir networks can be circumvented by considering other inherent granular layer features such as inverted input signals or mGluR2 inhibition of Golgi cells ., Other properties that facilitate filter construction are direct mossy fiber excitation of Golgi cells , variability of synaptic weights or input signals and output-feedback via the nucleocortical pathway ., Our findings are well supported by previous experimental and theoretical work and will help to bridge the gap between system-level models and detailed models of the granular layer network . | The cerebellum plays an important role in the learning of precise movements , and in humans holds 80% of all the neurons in the brain , due to numerous small cells called “granule cells” embedded in the granular layer ., It is widely thought that the granular layer receives , transforms and delays input signals coming from many different senses like touch , vision and balance , and that these transformed signals then serve as a basis to generate responses that help to control the muscles of the body ., But how the granular layer carries out this important transformation is still obscure ., While current models can explain how the granular layer network could produce specific outputs for particular reflexes , there is at present no general understanding of how it could generate outputs that were computationally adequate for general movement control ., With the help of a simulated granular layer network we show here that a random recurrent network can in principle generate the necessary signal transformation as long as it operates in a state close to chaotic behavior , also termed the “edge-of-chaos” . | null | null |
journal.pgen.1006765 | 2,017 | Suspected Lynch syndrome associated MSH6 variants: A functional assay to determine their pathogenicity | Lynch syndrome ( LS ) is an autosomal-dominantly inherited predisposition to a variety of malignancies at a young age , mainly colorectal cancer ( CRC ) and endometrial cancer ( EC ) 1 ., It is caused by inactivating germ-line mutations in the DNA mismatch repair ( MMR ) genes MLH1 , MSH2 , MSH6 or PMS2 , or a deletion in the 3’ region of the EPCAM gene that affects MSH2 expression 2–6 ., The DNA MMR system is essential for the fidelity of DNA replication ., Its primary function is the correction of base-base mismatches and insertion-deletion loops that may arise during DNA replication ., Base-base mismatches are recognized by the MSH2-MSH6 heterodimer while MSH2-MSH3 detects loops of unpaired bases ., Following mismatch binding , the MSH heterodimers recruit another heterodimer , MLH1-PMS2 , to coordinate removal and resynthesis of the error-containing strand 7–9 ., A second function of the DNA MMR system is to mediate the toxicity of certain DNA damaging agents such as methylating agents and thiopurines ., These DNA damaging agents create adducts in the genome that give rise to mismatches when replicated ., The DNA MMR system recognizes the mismatches but will remove the incorporated nucleotide rather than the lesion itself , creating a repetitive cycle of nucleotide incorporation and deletion that ultimately leads to DNA breakage and cell death 10 , 11 ., In the absence of MMR , cells tolerate methylation damage , but consequently show high levels of DNA damage-induced mutagenesis on top of a strongly elevated level of spontaneous mutagenesis 12 ., LS patients inherit a functional and a mutant copy of one of the DNA MMR genes ., For cells to become MMR-deficient and develop a mutator phenotype that accelerates carcinogenesis , somatic loss of the wild-type allele is required 13 ., Microsatellite instability ( MSI ) , i . e . , length alterations of repetitive sequences like ( CA ) n or ( A ) n , and loss of immunohistochemical staining ( IHC ) for MMR proteins are considered hallmarks of LS tumors ., Analysis of MSI and IHC on tumor tissue can identify patients who may suffer from LS ., For a definitive LS diagnosis , however , sequence analyses must reveal a pathogenic germline mutation in one of the DNA MMR genes or the 3’ region of EPCAM 14 , 15 ., Many LS-associated sequence variants are nonsense and frameshift mutations that clearly truncate the protein and unambiguously abrogate MMR activity ., Missense mutations that only alter a single amino acid are also frequently identified in suspected-LS patients ., The functional implications of these variants are less clear ., Consequently , the diagnosis of suspected-LS patients carrying missense variants is difficult in the absence of clear segregation and functional data ., As long as the phenotype of these variants of uncertain significance ( VUS ) is unclear , non-carriers cannot safely be discharged from burdensome surveillance programs 16 ., Surveillance programs have proven to significantly reduce morbidity and mortality in LS patients 1 , 17 , 18 , but pose unnecessary psychological and physical stress on carriers of innocent VUS as well as pressure on preventive healthcare ., Therefore , techniques that characterize MMR gene VUS and enable the identification of individuals at risk are urgently needed ., While in the past primarily MSH2 and MLH1 were sequenced to identify LS-causing mutations , in recent years MSH6 has been gained fame for causing LS due to the advancement of DNA sequencing ., However , MSH6 mutation carriers can be difficult to diagnose because they may not entirely fulfill the criteria for LS diagnosis: their age at cancer onset is often later than for MLH1 and MSH2 mutation carriers , and their tumors occasionally stain for MSH6 and have no or low MSI 19–21 ., We therefore extended the applicability of the oligonucleotide-directed mutagenesis screen we recently described for the identification of pathogenic MSH2 variants to MSH6 variants 22 ., The genetic screen uses oligonucleotide-directed gene modification ( oligo targeting ) 23 to introduce variant codons into the endogenous Msh2 gene of mouse embryonic stem cells ( mESCs ) and subsequently identifies pathogenic variants by selecting for cells that are resistant to the thiopurine 6-thioguanine ( 6TG ) ., Here we present the applicability of this screen for the characterization of MSH6 VUS ., The oligonucleotide-directed mutagenesis screen takes a four step approach to the identification of pathogenic MSH6 mutations ( Fig 1 ) :, 1 ) site-directed mutagenesis to introduce the variant of interest into a subset of Msh6+/- mESCs ,, 2 ) selection for cells that consequently lost MMR capacity ,, 3 ) PCR analysis to exclude cells that lost MMR capacity due to loss of the Msh6+ allele ( loss of heterozygosity events ) ,, 4 ) sequence analysis to confirm the presence of the planned mutation in the MMR-deficient cells ., mESCs provide a good study model because the human and mouse MSH6 amino acid sequences share over >86% identity ( S1 Fig ) and mouse models can be made from these cells if VUS need to be studied in vivo ., Msh6+/- mESCs only contain one wild type Msh6 allele ( Msh6+ ) ; the other allele was disrupted by a puromycin-resistance gene and therefore inactivated ( Msh6- ) 24 ., Hence introduction of a specific mutation into the one active Msh6 allele will lead to expression of solely the variant protein and allow immediate investigation of its phenotype ., To achieve this , Msh6 was site-specifically mutated by oligo targeting , a gene modification technique that uses short single-stranded locked-nucleic-acid-modified DNA oligonucleotides ( LMOs ) ( with either sense or antisense orientation ) to substitute a single base pair at a desired location ., LMO-directed base-pair substitution can be achieved at an efficiency of 10−3; thus , about 1 in every 1000 LMO-exposed Msh6+/- mESCs will contain the desired mutation 23 ., To determine whether the substitution abrogated Msh6 activity and this subset of cells consequently lost MMR activity , LMO-exposed mESCs were treated with 6TG ., The thiopurine DNA damaging agent 6TG is highly toxic to MMR-proficient but only moderately toxic to MMR-deficient cells 11 ., Therefore , the appearance of colonies that survived mild 6TG selection is indicative for loss of MMR capacity ., Loss of MMR capacity may arise due to the introduced mutation or due to loss of heterozygosity events that caused loss of the functional Msh6 allele ., To exclude the latter from further investigation , a PCR that detected the presence of both the disrupted and non-disrupted Msh6 alleles was performed 24 ., 6TG-resistant colonies that maintained both Msh6 alleles were sequenced to confirm the presence of the planned mutation ., To demonstrate the ability of the oligonucleotide-directed mutagenesis screen to distinguish pathogenic MSH6 mutations from polymorphisms , a proof of principle study was performed with MSH6 variants G1139S and L1087R that were previously proven to be pathogenic and not pathogenic , respectively 25 , as well as all classified pathogenic and not pathogenic missense variants described in the International Society for Gastrointestinal Hereditary Tumours ( InSiGHT ) colon cancer variant database ( http://insight-group . org/ ) ., This database uses available clinical , in vitro and in silico data to categorize DNA MMR gene sequence variants according to a five-tiered classification scheme as: class 5 , Pathogenic; 4 , Likely pathogenic; 3 , Uncertain; 2 , Likely not pathogenic; and 1 , Not pathogenic 26 ., Msh6+/- mESCs were first exposed to antisense oriented LMOs encoding the desired base-pair substitution ., If subsequent 6TG selection did not reveal resistant colonies encoding the planned mutation , the screen was repeated with sense oriented LMOs ., LMO-mediated introduction of both , pathogenic and not pathogenic variants led to the appearance of 6TG-resistant colonies ., For each LMO , we picked and analyzed 18 colonies ., The vast majority of 6TG-resistant colonies obtained with LMOs encoding polymorphisms had lost the wild-type Msh6 allele by loss of heterozygosity ( LOH ) events , as inferred from allele-specific PCR analysis ., Sequencing of the few 6TG-resistant colonies that had retained both Msh6 alleles ( ±6% ) , did not detect any mutation ( Fig 2A ) ., These background colonies apparently arose from cells that for unknown reasons survived 6TG exposure ., Of the 6TG-resistant colonies that emerged following LMO-mediated introduction of pathogenic mutations , ±40% still contained both Msh6 alleles ., Sequence analysis detected pathogenic mutations in all but one of these 6TG-resistant colonies ( Fig 2B; S2A Fig ) ., Thus , the oligonucleotide-directed mutagenesis screen detected all 4 pathogenic mutations and not one of the 5 non-pathogenic variants , indicating it is capable of distinguishing pathogenic MSH6 mutations from polymorphisms ., We used the oligonucleotide-directed mutagenesis screen to investigate the phenotype of 18 MSH6 VUS described in literature and the InSiGHT database as well as 8 MSH6 VUS detected in suspected-LS patients from the Erasmus Medical Center Rotterdam and the Radboud University Medical Center Nijmegen ( see S1 and S2 Tables for clinical data 27–38; see S3 Fig for location of variants in MSH6 39 , 40 ) ., Of the 26 variants , 18 were not present in 6TG-resistant colonies and hence do not appear to affect MMR activity ., Mutations R510G , A586P , G683D , F703S , L1060R , E1191K , T1217D and T1217I were identified in 6TG-resistant colonies by sequence analysis ( Fig 3A and 3B; S2B Fig ) ., Of note , variants R510G and F703S were detected in only two colonies out of five and four , respectively , that had not resulted from LOH ( Fig 3B ) ., Given the low frequency of LMO-mediated base-pair substitution , we consider the presence of a variant allele in two independent colonies indicative for pathogenicity ., The MMR abrogating effect of all Msh6 variants conferring 6TG-resistance was further characterized by Western blot analyses , MSI assays and methylation-damage-induced mutagenesis assays ., The effect of the identified MMR abrogating mutations on MSH6 and MSH2 protein levels was evaluated by Western blot analyses ( Fig 4 ) ., MSH6 and MSH2 form a heterodimer; consequently , a drop in MSH6 levels is often associated with a slight decrease in MSH2 protein stability ., Protein levels were quantified with respect to Msh6+/- mESCs , which maintain a functional MMR system with about two-third of the MSH6 level observed in Msh6+/+ mESCs 25 ., Known pathogenic mutations V397E , L448P , G1137S and R1332Q reduced MSH6 levels to 7–33% of that seen in Msh6+/- mESCs ., The R1332Q mutation is located in the splice donor site of exon 9 which may explain the appearance of a shorter protein ., The drop in MSH6 levels seen for the known pathogenic mutations was mirrored by variants A586P , G683D , F703S and L1060R that reduced protein levels to 7–24% ., Variants R510G , E1191K , T1217D and T1217I maintained relatively high MSH6 levels of 59–79% ., MSI in MSH6 mutation carriers is largely restricted to mononucleotide markers 41 ., To investigate the effect of the detected Msh6 variants on MSI we used a ( G ) 10-neo slippage reporter ., The neomycin resistance gene ( neo ) in this reporter is rendered out of frame by a preceding ( G ) 10 repeat ., When DNA polymerase slippage errors at the ( G ) 10 repeat such as the deletion of one G or insertion of two Gs remain unnoticed , the neo becomes in frame and generates Geneticin-resistant cells ., Hence the number of Geneticin-resistant colonies is indicative of the frequency of neo-restoring slippage events and the MMR capacity of the cells 42 ., The slippage rates , i . e . , the chance of a slippage event occurring during one cell division , in 6TG-resisant Msh6 VUS expressing mESCs ranged from 5 . 3x10-5 to 5 . 1x10-4; which is around the average rate of 1 . 9x10-4 observed for the known pathogenic mutations and 140 to 1340-fold higher than the slippage rate of 3 . 8x10-7 seen for Msh6+/- MMR-proficient mESCs ( Fig 5 ) ., In addition to increased spontaneous mutagenesis events , MMR-deficient cells also experience increased methylation-damage-induced mutagenesis 43 ., To study the influence of the detected MMR attenuating Msh6 variants on methylation-damage-induced mutagenesis , mESCs were exposed to the methylating DNA damaging agent N-methyl-N’-nitro-N-nitrosoguanidine ( MNNG ) and the number of cells that consequently attained mutations was quantified ., In MMR-proficient cells , DNA replication across MNNG-induced O6-methylguanine lesions is impaired by futile cycles of MMR , ultimately leading to cell death and suppression of methylation-damage-induced mutagenesis ., Under MMR-deficient conditions , however , the MNNG-induced mismatches are not recognized and remain in the genome leading to the accumulation of mutations ., To provide a quick read out for the frequency of mutation accumulation , we measured the number of MNNG-exposed cells that became resistant to a high dose of 6TG for an extended period ., Solely cells that carry an inactivating mutation in Hprt survive stringent 6TG treatment because HPRT is required for 6TG to behave as a DNA damaging agent ., All detected Msh6 variant cell lines showed an elevated MNNG-induced mutator phenotype when compared to the MMR-proficient Msh6+/- mESCs ( Fig 6 ) ., According to literature MSH6-G566R may be pathogenic 29 , 44 , yet our screen did not identify this variant in 6TG-resistant colonies ., Hence we investigated whether the MMR abrogating effect of Msh6-G565R could have been missed by the screen due to technical difficulties ., Rather than applying 6TG selection after oligonucleotide-directed mutagenesis , we purified Msh6G565R/- mESCs using a Q-PCR-based protocol 25 ( S2C Fig ) and subsequently examined their MMR capacity ., Exposure of Msh6G565R/- cells to increasing doses of 6TG revealed that they were equally sensitive to 6TG as Msh6+/- cells ( Fig 7A ) ., In the MSI assay , Msh6G565R/- mESCs did not experience significantly more slippage events than the MMR-proficient control ( Fig 7B ) ., Thus , Msh6-G565R did not attenuate MMR consistent with the oligonucleotide-directed mutagenesis screening result ., The results of our study demonstrate the oligonucleotide-directed mutagenesis screen we previously described for the characterization of MSH2 VUS 22 can be extended to MSH6 VUS ., Combining oligo targeting in Msh6+/- mESCs with 6TG selection and sequence analysis allows pathogenic MSH6 variants to be distinguished from polymorphisms ., The efficacy of the genetic screen was established in a proof of principle study with 4 known pathogenic MSH6 mutations and 5 polymorphisms ., This number was low because of the paucity of MSH6 variants that were classified with 100% certainty ., Not one of the 5 non-pathogenic variants was identified as MMR abrogating ., Also , among the 26 MSH6 VUS we subsequently analyzed , not one of the 4 variants classified as likely not pathogenic was identified as pathogenic by our screen ., Finally , functional assays established that one of the VUS that was not detected as pathogenic by the screen indeed did not influence MMR activity ( G565R ) ., Hence the false positive rate of the screen , i . e . , the chance the screen identified a VUS as MMR abrogating while it was a priori or a posteriori identified as ( likely ) non-pathogenic was <1/10 , giving a specificity >90 . 0% ., The sensitivity of the genetic screen is a measure of the false negative rate; it is the likelihood that a pathogenic mutation is not detected ., All 6 InSiGHT classified pathogenic and likely pathogenic variants as well as the previously proven pathogenic G1139S mutation were recognized as MMR abrogating by the screen , translating to a sensitivity of >85 . 7% ., We used the oligonucleotide-directed mutagenesis screen to investigate the MMR capacity of 26 MSH6 VUS ., Eight of these were found in suspected-LS patients from two medical centers in the Netherlands ., From this clinical cohort , the mouse equivalents of mutations R511G , A587P and F706S were detected by our screen and shown to abrogate MMR ., However , R510G and F703S were detected in only 2/5 and 2/4 6TG-resistant colonies , respectively , that had retained two Msh6 alleles , while the other pathogenic variants were present in virtually all colonies diploid for Msh6 ( Figs 2B , 3A and 3B ) ., The poorer recovery of R510G and F7103S mutants may have been due to a lower success rate of LMO-mediated base-pair substitution ., The pathogenic phenotype observed for these three variants is in line with clinical data: all three variants were detected in patients with MSI-H LS-related tumors and with a family history of LS-related tumors ., In the case of VUS A587P and F706S , relatives with LS-related tumors carried the same mutation ., IHC also demonstrated MSH6 was absent in the patients encoding MSH6-A587P and MSH6-F706S; the IHC data for MSH6-R511G were inconclusive ., The other 5 variants in the clinical cohort , A25S , E221D , G670R , R922Q and c . 3438+6T>C , were not identified as MMR abrogating ., VUS E221D , G670R and R922Q were found in patients who also harbored a second , known pathogenic mutation in one of the DNA MMR genes that was likely causative for the LS phenotype ., E221D was also detected in a second patient who was 83 years old and did not have a family history suspicious for LS ., MSH6-A25S was found in a typical LS tumor , i . e . , a colon tumor showing MSI , loss of heterozygosity of MSH6 , and loss of MSH6 protein expression ., The patient however only had one relative with a colorectal tumor and this tumor was not MSI-high and stained positive for all MMR proteins ., A previous in vitro study also suggested MSH6-A25S is not pathogenic 45; it could be that the tumor arose due to a missed somatic mutation ., VUS c . 3438+6T>C was found in a patient with a family history suspicious of LS ., We however do not know if the relatives with LS-associated cancers also carried this specific MSH6 sequence variant ., IHC failed in the index patient carrying the c . 3438+6T>C variant , therefore we cannot exclude that a somatic mutation or MLH1 hypermethylation caused the MSI in the tumor ., Tumor tissue of one family member was tested and showed no MSI and normal IHC ., It is also possible that the genetic screen was unable to identify c . 3438+6T>C as pathogenic due to differences between the human and mouse MSH6 sequences ., While the MSH6 coding sequence is highly conserved , intron sequences are more variable between species ( S4 Fig shows human and mouse sequence around c . 3438+6 ) ., Hence there is a chance that variant c . 3438+6T>C affects splicing in man but not in mice ., According to several splice site prediction programs ( NNSPLICE , GeneSplicer , Human Splicing Finder ) , however , c . 3438+6T>C does not affect splicing ., The other 18 MSH6 VUS we studied were attained from literature and the InSiGHT database ., The genetic screen found 5 of these variants abrogate MMR: G686D , L1063R , E1193K , T1219D and T1219I ., The detection of G686D and L1063R is in line with their InSiGHT classification , which describes the mutations as likely pathogenic ., Variant E1193K has previously been suggested to cause LS in studies that identified the mutation in patients with ECs that were MSI and did not stain for MSH6 27 , 28 ., Not much clinical data is available for VUS T1219D but Msh6T1217D mice were demonstrated to have increased cancer susceptibility 46 ., VUS T1219I has been described in a CRC patient who had a family history of CRC and a MSI tumor that stained positive for MSH6 , the latter being consistent with the high levels of this variant protein we observed in mESCs ., Both clinical and in vitro data indicate MSH6-T1219I abrogates MMR activity 37 , 45 ., MSH6 VUS R128L , R468H , V509A , Y556F , P623A , S666P , E983Q , R1095C , T1255M and R1304K were not identified as pathogenic in our screen ., These sequence variants were classified as likely not pathogenic by InSiGHT , identified in patients with MLH1 promoter methylation or with MSS and MSH6 positive tumors , or observed in patients for whom little clinical data was available ., VUS S285I , G566R and T1142M were also not detected as MMR attenuating by our screen , yet they seem suspicious for pathogenicity based on available data ., MSH6-T1142M was previously suggested to be probably pathogenic based on clinical data describing the variant in a 27 year old patient with polyps who met the Bethesda guidelines , had a 61 year old mother with polyps , and did not carry pathogenic mutations in any other MMR gene nor showed MLH1 promoter methylation in the tumor 36 ., VUS S285I and G566R were detected in CRC patients with MSI ( low and high , respectively ) tumors that had loss of heterozygosity of MSH6 29 ., Cyr and Heinen 44 investigated the effect of these two mutations on mismatch binding and processing: variant S285I was not found to have a specific MMR attenuating effect but variant G566R was suggested to abrogate MMR by interfering with the ATP-dependent conformational change that must take place to activate downstream repair pathways upon mismatch binding ., We therefore purified Msh6G565R/- mESCs and assessed their MMR capacity ., The Msh6G565R/- cells behaved like MMR-proficient Msh6+/- mESCs , confirming the result of the oligonucleotide-directed mutagenesis screen ., Despite the good performance of our screen and the high amino acid conservation of MSH6 , we cannot exclude Msh6-G565R was not identified as pathogenic due to differences between mice and men ., To fully dissuade this argument we will need to develop the oligonucleotide-directed mutagenesis screen in human cells ., The oligonucleotide-directed mutagenesis screen presented here is a relatively simple tool that can be used to investigate the pathogenic phenotype of many MSH6 VUS in parallel ., While the evolutionary conservation of MMR justifies the use of mouse cells for the majority of VUS , testing of splice-site and intronic mutations necessitates adaptation to human cells ., Also , as long as uncertainty exists about its specificity and sensitivity , functional testing needs to be combined with clinical data and in silico estimations to arrive at a reliable classification of VUS ., Conforming the updated American College of Medical Genetics and Genomics ( ACMG ) standards and guidelines for sequence variant interpretation , we are currently transferring our functional tests to certified Clinical Genetics laboratories and creating an infrastructure where test results are compared and interpreted taking into account all available data ., In this way , LS mutation carriers can be identified with the highest certainty and enrolled in tailored surveillance programs while relatives without the mutation can be excluded from surveillance ., The genetic screen was developed in Msh6+/- mESCs , which contain one active Msh6 allele ( Msh6+ ) and one Msh6 allele that was disrupted by the insertion of a puromycin resistance marker ( Msh6- ) 24 ., The MSH6 variants under investigation were introduced into the Msh6+/- mESCs by oligo targeting using LMOs 23 ., 7x105 Msh6+/- mESCs were seeded in BRL-conditioned medium on gelatin-coated 6 wells and exposed to a mixture of 7 . 5 μl TransIT-siQuest transfection agent ( Mirus ) , 3 μg LMOs and 250 μl serum-free medium the following day ., After 3 days , 1 . 5x106 LMO-exposed cells were transferred to gelatin-coated 10 cm plates and subjected to 6TG ( 250 nM ) ( Sigma-Aldrich ) selection ., After 10 days the 18 largest 6TG-resistant colonies were picked ., Cells that became 6TG-resistant due to loss of heterozygosity events were excluded from further analyses using a PCR specialized to detect the presence of both the disrupted and non-disrupted Msh6 alleles 24 ., 6TG-resistant mESCs that maintained both Msh6 alleles were sequenced to confirm the presence of the planned mutation ., Western blot analyses were performed as described in Wielders et al . 25 ., Rabbit polyclonal antibodies against mMSH2 ( 1:500 ) 47 and mMSH6 ( 1:500 ) 24 as well as mouse polyclonal antibody against γ-Tubulin ( 1:1000; GTU-88 Sigma-Aldrich ) were used as primary antibodies ., Protein bands were visualized using IRDye 800CW goat anti-rabbit IgG and IRDye 800CW goat anti-mouse IgG secondary antibodies ( Li-cor ) and the Odyssey scan ., The infrared fluorescent signals measured by the Odyssey scan are directly proportional to the amount of antigen on the Western blots , allowing quantification of the protein bands ., mESCs were electroporated with the ( G ) 10-neo Rosa26 targeting vector as described in Dekker et al . 48 ., The ( G ) 10-neo Rosa26 targeting vector is composed of a promoterless histidinol resistance gene as well as a neomycin resistance gene ( neo ) that is rendered out of frame by a preceding ( G ) 10-repeat 42 ., Once electroporated , 106 cells were seeded on gelatin-coated 10 cm plates in BRL-conditioned medium and exposed to Histidinol ( 3mM ) ( Sigma-Aldrich ) ., Successful integration of the vector into the Rosa26 locus of the Histidinol-resistant colonies routinely occurs at a frequency of ±95% and was confirmed by Southern blot analyses ., The individual successfully targeted colonies were subsequently expanded to 107 cells and transferred to gelatin-coated 10 cm plates at a density of 105 cells per plate for Geneticin selection ( 600 μg/ml ) ( Life Technologies ) ., After 10 days , the number of Geneticin-resistant colonies was counted and the slippage rate of the variant mESCs calculated using the formula: 0 . 6 x Geneticintotal = N x p x log ( N x p ) , where Geneticintotal is the number of Geneticin-resistant colonies , N the number of cells to which the culture was expanded , and p the number of mutations per cell division ., Experiments were performed in quadruplicate and statistical differences calculated using a one-tailed , unpaired t-test with Welch’s correction ., The MNNG-induced mutagenesis assay was performed as described in Claij and te Riele 43 ., 2 . 5x106 variant mESCs were seeded on an irradiated mouse embryonic fibroblasts feeder layer in 10 cm plates and exposed to 0 or 4μM MNNG ( Sigma-Aldrich ) for 1h the following day ., 40 μM O6-benzylguanine was present in the medium from 1h prior to the MNNG treatment until 6 days after , at which point 1 . 5x106 cells were transferred to gelatin-coated 160 cm2 plates for 6TG selection ( 10 μg/ml ) ., After two weeks of 6TG selection , the number of resistant colonies and hence the frequency of MNNG-induced Hprt mutants could be determined ., Experiments were performed in duplo and the statistical difference between MNNG-treated Msh6+/- mESCs and MNNG-treated variant cell lines calculated using a one-tailed , unpaired t-test with Welch’s correction ., Msh6G565R/- mESCs were made as described by Wielders et al . 25 ., Variant G565R was introduced into Msh6+/- mESCs by oligo targeting and a pure Msh6G565R/- mESC clone was obtained by consecutive rounds of seeding and mutation specific PCR: oligonucleotide-exposed cells were expanded and subsequently seeded on a 96-well plate at a density of 5000 cells per well ., A mutation-specific quantitative PCR was used to identify wells that contained Msh6G565R/- mESCs ., Positive wells were reseeded at lower density and positive wells again identified by Q-PCR ., A pure clone was finally obtained by seeding single cells per well ., Sequence analysis confirmed the creation of Msh6G565R/- mESCs ., The 6TG sensitivity of Msh6G565R/- mESCs was investigated by exposing the variant cell line to increasing doses of 6TG , as described in Wielders et al . 49 ., MMR-deficient Msh6-/- and MMR-proficient Msh6+/- and Msh6+/+ mESCs were taken along for comparison ., We investigated the pathogenic phenotype of MSH6 VUS that were found in suspected-LS patients at the Clinical Genetics departments of the Erasmus Medical Center Rotterdam and Radboud University Medical Center Nijmegen ., We collected tumor characteristics , age at diagnosis , results of molecular diagnostics and germline mutation analysis , and family history from medical records ., MSI analysis was performed with the Bethesda panel 50 or with the Promega pentaplex MSI analysis 51 ., IHC for MLH1 , MSH2 , MSH6 and PMS2 protein was performed as described previously 52 ., Germline mutation analysis of MSH6 was performed by sequencing and multiplex ligation dependent probe amplification ., The in silico prediction model PolyPhen 53 was used to estimate the chance of a variant being deleterious . | Introduction, Results, Discussion, Materials and methods | Lynch syndrome ( LS ) is a hereditary cancer predisposition caused by inactivating mutations in DNA mismatch repair ( MMR ) genes ., Mutations in the MSH6 DNA MMR gene account for approximately 18% of LS cases ., Many LS-associated sequence variants are nonsense and frameshift mutations that clearly abrogate MMR activity ., However , missense mutations whose functional implications are unclear are also frequently seen in suspected-LS patients ., To conclusively diagnose LS and enroll patients in appropriate surveillance programs to reduce morbidity as well as mortality , the functional consequences of these variants of uncertain clinical significance ( VUS ) must be defined ., We present an oligonucleotide-directed mutagenesis screen for the identification of pathogenic MSH6 VUS ., In the screen , the MSH6 variant of interest is introduced into mouse embryonic stem cells by site-directed mutagenesis ., Subsequent selection for MMR-deficient cells using the DNA damaging agent 6-thioguanine ( 6TG ) allows the identification of MMR abrogating VUS because solely MMR-deficient cells survive 6TG exposure ., We demonstrate the efficacy of the genetic screen , investigate the phenotype of 26 MSH6 VUS and compare our screening results to clinical data from suspected-LS patients carrying these variant alleles . | The colorectal and endometrial cancer predisposition Lynch syndrome ( LS ) is caused by an inherited heterozygous defect in one of four DNA mismatch repair ( MMR ) genes ., Deleterious mutations ( e . g . , protein-deleting or -truncating ) in DNA MMR genes unambiguously allow for the clinical diagnosis LS and hence enable appropriate surveillance measures to be taken to reduce cancer risk and ensure early detection of tumors ., However , currently about one-third of detected MMR gene variants are subtle with less clear functional consequences: missense mutations affecting a single amino acid may be innocuous , hence not causing LS , or partially or fully destroy protein function ., As long as uncertainty exists about their pathogenicity , such mutations are labeled ‘variants of uncertain ( clinical ) significance’ ( VUS ) ., VUS hamper genetic counseling and therefore the need for functional testing of VUS is widely recognized ., To functionally annotate MMR gene VUS , we have developed a high content cellular assay in which the VUS is introduced in a cell culture by oligonucleotide-directed gene modification ., Should the VUS be deleterious for MMR , the modified cells survive exposure to the guanine analog 6-thioguanine ( 6TG ) and 6TG-resistant colonies appear ., Should the mutation not affect MMR , no colonies appear ., Here we present the adaptation and application of this protocol to the functional annotation of variants of the MMR gene MSH6 ., Implementation of our assay in clinical genetics laboratories will provide clinicians with information for proper counseling of mutation carriers and treatment of their of tumors . | oligonucleotide-directed mutagenesis, alleles, dna damage, mutation, dna, molecular biology techniques, mutagenesis and gene deletion techniques, frameshift mutation, research and analysis methods, molecular biology, genetic loci, biochemistry, nucleic acids, genetic screens, gene identification and analysis, genetics, mutation detection, biology and life sciences | null |
journal.ppat.1000589 | 2,009 | TbPIF5 Is a Trypanosoma brucei Mitochondrial DNA Helicase Involved in Processing of Minicircle Okazaki Fragments | Trypanosomes and related parasites cause tropical diseases such as sleeping sickness and Chagas disease ., As one of the earliest diverging eukaryotes that contain a mitochondrion 1 , this group of parasites is well known for unusual biological properties ., For example , their mitochondrial genome , known as kinetoplast DNA ( kDNA ) , has an amazing and unprecedented structure , a giant DNA network residing in the cells single mitochondrion 2 , 3 ., The network is a planar structure composed of interlocked DNA rings including several thousand minicircles and a few dozen maxicircles ., Within the mitochondrial matrix the kDNA network is condensed into a disk-shaped structure that is positioned near the flagellar basal body , which resides in the cytoplasm ., The kDNA disk , called the kinetoplast , is actually connected to the basal body by a transmembrane filament system named the tripartite attachment complex ( TAC ) 4 ., Like mitochondrial DNA in other organisms , maxicircles encode ribosomal RNAs and a handful of mitochondrial proteins such as subunits of respiratory complexes ., Many maxicircle transcripts require editing before they can serve as functional mRNAs ., Editing is an unusual RNA processing reaction involving addition or deletion of uridylate residues at specific internal sites of mRNAs ( reviewed in 5 , 6 ) ., In some transcripts , editing occurs on a massive scale , with uridylates introduced by editing constituting more than half of the sequence of the resulting mRNA ., Minicircles encode small guide RNAs that serve as templates for editing , thereby controlling its specificity ., In this paragraph we will briefly discuss the kDNA replication mechanism in T . brucei , focusing on minicircles ., The initial step in replication is the vectorial release of individual minicircles into the space , known as the kinetoflagellar zone ( KFZ ) , between the kDNA disk and the membrane near the flagellar basal body 7 ., Here the free minicircles encounter proteins that assemble and propagate a replication fork , resulting in unidirectional replication as theta structures ., The progeny minicircles are thought to segregate in the KFZ , and then migrate to the antipodal sites , two protein assemblies that flank the kDNA disk and are positioned about 180° apart 8 ., At this time the monomeric minicircle replication products contain either a single continuously synthesized leading strand or they contain unligated Okazaki fragments 9 ., Within the antipodal sites the Okazaki fragments are processed ., Although the detailed processing mechanism is unknown it probably involves several enzymes that localize within the antipodal sites ., These enzymes , which have been studied to varying degrees , include structure-specific endonuclease I 10 , 11 , DNA polymerase β 12 , and DNA ligase kβ 13 ., These enzymes are thought to participate in removal of RNA primers and to fill and close the resulting gaps ., The processed minicircles , containing either the newly synthesized leading strand or lagging strand and still containing at least one gap , are then attached to the network periphery by a topoisomerase II that is also situated in the antipodal sites 14 , 15 ., Since two minicircles are attached for every one removed , the network grows in size ., Only when the minicircle copy number has doubled are their remaining gaps repaired , most likely by DNA polymerase β-PAK 16 and DNA ligase kα 13 , two enzymes that reside within the kDNA disk ., Then the network splits in two and its progeny , each identical to the parent , are pulled into the two daughter cells by their connection ( via TAC ) to the flagellar basal bodies 4 ., Recently we discovered 8 proteins in T . brucei that are related to the Saccharomyces cerevisiae mitochondrial helicase ScPIF1 , and we named them TbPIF1-8 ., Remarkably , six of these are localized at several different positions in the mitochondrion; of the other two , one is nuclear and the other appears to be in the cytoplasm 17 ., We have so far studied only one of the mitochondrial proteins , TbPIF2 , and have found it to be a helicase that is essential for maxicircle replication 17 ., Here we report that TbPIF5 ( Genbank accession No . : XP_847187; GeneDB accession No . : Tb927 . 8 . 3560 ) is a DNA helicase involved in minicircle Okazaki fragment processing , probably by unwinding the hybrid helices between RNA primers and the DNA template ., We previously localized TbPIF5 to the antipodal sites by expressing an ectopic gene encoding a TbPIF5-GFP fusion protein 17 ., To localize TbPIF5 encoded at its endogenous locus , we introduced a sequence encoding a myc epitope at the 3′ end of one endogenous allele of TbPIF5 gene ., This protein would more likely be expressed at its normal level ., Our immunofluorescence studies on this protein confirmed that TbPIF5 localizes within the antipodal sites ( Fig . 1 ) ., Since almost all the cells in an asynchronous log phase culture had this localization , it is likely that this protein does not undergo significant change in its localization during the cell cycle ., To determine whether TbPIF5 is actually a DNA helicase , we expressed it with a His-tag in E . coli and purified it by two steps of chromatography ( Fig . 2A ) ., Recombinant TbPIF5 hydrolyzes ATP in the presence of Mg2+ and M13 ssDNA ( Fig . 2B ) , indicating that it has DNA-dependent ATPase activity ., TbPIF5 also has helicase activity , releasing oligonucleotides that had been annealed to M13 single-stranded circles ( Fig . 2C ) ., As expected , Mg2+ and ATP are required for this reaction ( Fig . 2D ) , and the optimal concentration for both was in the range of 0 . 5 or 1 mM ( Fig . 2D ) ., To determine the polarity of helicase activity , we constructed substrates ( diagrammed in Fig . 2E ) with a short oligonucleotide ( either a or b; 5′ end-labeled with 32Pphosphate ) annealed to either the 5′ or 3′ terminus of oligonucleotide, c . Under conditions in which we observed dissociation of oligonucleotide a from the duplex structure , we could not detect dissociation of oligonucleotide, b . Therefore , as predicted from its homology to the yeast mitochondrial helicase , we conclude that TbPIF5 has a 5′ to 3′ helicase activity ( Fig . 2E ) ., To study the function of TbPIF5 , we first tried RNAi using the pZJM vector 18 ., Although ∼90% of the mRNA was degraded by 2 days after induction of RNAi ( Inset , Fig . S1A ) , there was no effect on cell growth ( Fig . S1A ) ., Use of a stem-loop RNAi vector 18 gave the same result ( data not shown ) ., We then tried to knock out both alleles of TbPIF5 by replacing each allele with a different drug marker ., However , only one allele could be replaced as judged by Southern blot ( Fig . S1B ) ., Because knockout of both alleles may be lethal , we introduced into the cell an ectopic TbPIF5 gene using the vector pLew79-MHTAP 19 ., The ectopic gene expresses TbPIF5 only in the presence of tetracycline , and therefore it should allow deletion of the second genomic allele ., For unknown reasons , this strategy was also unsuccessful using tetracycline concentrations ranging from 2–10 ng/ml ( data not shown ) , and thus we failed to knock out both genomic alleles ., As discussed in the following paragraph , we found unexpectedly that a higher level of tetracycline , which causes overexpression of TbPIF5 , reduces the cells growth rate ., Using the ectopic expression system discussed in the previous paragraph ( except that both endogenous TbPIF5 alleles were still present ) , we found that 2 days of treatment with 1 µg/ml tetracycline caused more than a 15-fold increase ( judged by phosphorimaging ) in TbPIF5 mRNA ( see northern blot inset in Fig . 3A ) ., Furthermore , this treatment reduced the cells growth rate 4 days after tetracycline addition ( Fig . 3A ) , providing evidence that an elevated level of TbPIF5 is deleterious to the cell ., TbPIF5 overexpression also caused shrinkage of kDNA networks as judged by DAPI staining of intact cells ., Fig . 3B shows examples of fluorescence images of wild type cells and those that had undergone 6 days of overexpression ., Fig . 3C shows kinetics of kDNA loss ( determined by visual inspection of fluorescence images like those in panel B ) following induction of overexpression ., At day 6 , only ∼50% of the cells had normal-sized kDNA , 20% had small kDNA , and 30% had no detectable kDNA ., We then used a different approach to evaluate minicircle and maxicircle abundance following induction of TbPIF5 overexpression ., We digested total DNA with HindIII/XbaI , separated the fragments by agarose gel electrophoresis , and then probed a Southern blot for minicircles and maxicircles ( Fig . 3D ) ., After 5 days of overexpression , minicircle abundance decreased by more than half , while there was only a mild effect on the level of maxicircles ( Fig . 3E ) ., These results indicated that TbPIF5 overexpression selectively affects minicircles ., We further examined the isolated kDNA networks by electron microscopy ., The unit-sized network isolated from the uninduced cells has multiple maxicircle loops projecting from the periphery ( arrows in Fig . 4A ) ., In the late stage of replication , maxicircle loops usually concentrate in the central region between the two segregating daughter networks ( arrows in Fig . 4B ) ., After 6 days of TbPIF5 overexpression , some networks have become smaller in size ( Fig . 4C ) , and the structure of some networks is disorganized ( Fig . 4D ) ., As usual , we observed multiple maxicircle loops extending from the edge of the networks in different stage of replication ., However , they do not always concentrate in the central region of the double-sized network that is undergoing segregation ( see an example in Fig . 4E ) ., To investigate whether minicircle loss caused by TbPIF5 overexpression is due to an effect on replication , we fractionated total DNA on an agarose gel in the presence of ethidium bromide and detected free minicircle replication intermediates by probing a Southern blot ( Fig . 5A ) ., After 5 days of overexpression , both covalently-closed and gapped/nicked minicircles decreased by about half , consistent with the decrease in total minicircle abundance ., One unexpected consequence of TbPIF5 overexpression was the appearance of a heterogeneous population of minicircle species migrating as a smear between covalently-closed and gapped/nicked minicircles ., This smear , never before observed and which we call fraction H , is most prominent on days 1 to 4 ., We next fractionated total DNA using neutral/alkaline 2-dimensional gel electrophoresis and analyzed free minicircle species by strand-specific hybridization ( Fig . 5B ) ., In the first dimension , minicircle species were separated in TBE buffer containing ethidium bromide ( conditions identical to those used for the gel in Fig . 5A ) ., In the second dimension , run in 30 mM NaOH , the double-stranded DNA was denatured ., Using 5′-32P -labeled synthetic oligonucleotides , we separately probed for L- ( the leading strand ) and H-strands ( the lagging strand ) ., The probes were complementary to sequences near the 5′ end of the L-strand and within the first Okazaki fragment on the H-strand ., Interpretation of these gels was aided by comparison with our previous 2-D gels of minicircles from the closely-related parasite T . equiperdum 9 as well as from T . brucei 20 ., As mentioned in the Introduction , these and other studies had shown that minicircles replicate unidirectionally via theta structures with the L-strand synthesized continuously and the H-strand discontinuously with ∼100 nucleotide Okazaki fragments ., This mechanism is unusual in that Okazaki fragments are not joined until the θ-structures had segregated into monomeric products; joining is thought to occur within the antipodal sites 9 ., In the control 2-D gel of wild type free minicircle intermediates ( Fig . 5B , upper panels ) , we show a fairly long exposure to reveal the unjoined Okazaki fragments ( OF ) derived from multiply-gapped circles ( MG ) and the diagonal of growing L-strands ranging in size up to ∼1 kb derived from θ-structures ( θ ) ., Joining of most of the Okazaki fragments in a minicircle converts multiply-gapped minicircles to nicked or gapped minicircles ( N/G ) ., Some minor minicircle species previously identified in wild type T . equiperdum and T . brucei such as the knotted minicircle ( K ) , linearized minicircle ( L ) , nicked dimer ( nD ) , and covalently-closed dimer ( ccD ) are not relevant to this paper and not discussed here 9 , 20 , 21 ., Two-dimensional gels of minicircles from cells undergoing TbPIF5 overexpression for 1 day ( Fig . 5B , lower panels ) differed markedly from those from wild type cells ( Fig . 5B , upper panels ) ., We found that fraction H has a ∼1 kb L-strand template and in the next paragraph we will present strong evidence that this strand is circular ., These L-strands form a smear extending from CC to N/G ( H , left lower panel in Fig . 5B ) ., The reason for smearing is that prior to denaturation they had been paired with H strands varying in size ., The latter molecules form a diagonal , never observed previously , in the size range of 0 . 1 to near 1 kb ( H , right lower panel in Fig . 5B ) ., Thus , fraction H likely consists of a circular L-strand paired with a family of growing H strands ., Since the probe detects only the first Okazaki fragment to be synthesized , the H-strand fragments in the diagonal must include the first and form a family of ligated contiguous Okazaki fragments ., Strand-specific hybridization also suggested a decrease in level of growing L-strands on θ-structures ( compare L-strand diagonal , designated θ , in left upper panel in Fig . 5B with corresponding area of left lower panel ) , although since different exposures were used it is not possible to make a firm conclusion on this point ., To further characterize fraction H , we purified free minicircles by sucrose gradient centrifugation ( Fig . 5C ) and treated these molecules with T4 DNA polymerase ( plus all four dNTPs ) , T4 DNA ligase ( plus ATP ) , or both together ( Fig . 5D ) ., DNA polymerase alone converts fraction H to the position of gapped/nicked minicircles , but DNA ligase alone barely affects the mobility of fraction H . However , both enzymes together convert a substantial portion of fraction H to covalently-closed minicircles ., This experiment not only indicates that fraction H is a gapped molecule with ligated Okazaki fragments but also provides evidence that the L-strand of fraction H is a circle ., If TbPIF5 is involved in primer removal , it is possible that its overexpression might reduce the number or length of primers on either free minicircles or those linked to the network ., We previously reported that in T . brucei there are no ribonucleotides on the 5′ end of either the newly synthesized L-strand or the first Okazaki fragment on minicircles that were linked to the network 11 ., However , we never had searched for primers on free minicircles ., Using the strategy we developed previously 11 , we investigated whether primers were present before and after TbPIF5 overexpression ( Fig . 6 ) ., We isolated kDNA networks and free minicircle intermediates ( from both uninduced and 1 day overexpression cells ) , digested them with TaqI , and fractionated the products on a denaturing 9% polyacrylamide gel ., We then probed a Southern blot for the first Okazaki fragment ., This fragment , containing ∼73 nucleotides but with a slightly heterogeneous 3′ end , had been converted by TaqI to a slightly smaller fragment ( 66 nucleotides ) with a homogeneous 3′ end ( Fig . 6A ) ., This species , whether derived from free minicircles or network minicircles , was not altered by alkali treatment , indicating that there are no ribonucleotides on its 5′ end or anywhere else within the molecule ., Using a similar strategy , we searched for ribonucleotides at the 5′ terminus of the continuously-synthesized L-strand ., We cleaved the minicircles with HpyCH4V , which release a 69 nucleotide terminal L-strand fragment ( Fig . 6B ) ., Again , there is no ribonucleotide attached at the 5′ end of the newly-synthesized L-strand ., In a recent search for T . brucei mitochondrial DNA helicases , we found that the genome encodes 8 proteins related to ScPIF1 , a mitochondrial helicase of S . cerevisiae ., Remarkably , 6 of the T . brucei PIF1-related gene products are mitochondrial 17 ., Here we report the properties of one of these enzymes , TbPIF5 , which is localized in the antipodal sites ( Fig . 1 ) ., As shown in Fig . 2 , we found that a recombinant protein had helicase activity , with a 5′ to 3′ polarity , similar to that of the yeast homolog 22 ., We did not observe a phenotype following RNAi of TbPIF5 , even though ∼90% of the mRNA was depleted within 2 days ( Fig . S1 ) ., We could knock out one , but not both alleles of TbPIF5 , raising the possibility that the gene is essential ., Surprisingly , the genome of a related kinetoplastid , Leishmania major , encodes only 7 PIF1-like helicase genes , and the counterpart of TbPIF5 gene is apparently absent 17 ., Although this fact might support an argument that TbPIF5 could be dispensable we cannot rule out the possibility that other PIFs may take over TbPIF5s functions in L . major ., Like T . brucei , the T . cruzi genome contains 8 genes related to ScPIF1 ., Although RNAi and single allele knockouts did not affect cell growth or kDNA size as determined by DAPI staining , we did observe a striking effect of TbPIF5 overexpression on the replication of minicircles ., Not only was there a slowing of growth and loss of kDNA minicircles ( Fig . 3 ) , but there was an alteration in joining of Okazaki fragments ( Fig . 5 ) ., Before we discuss these new data , we will review what is known about primer removal and other processing reactions of minicircle Okazaki fragments ., We will also review Okazaki fragment joining in the nucleus of other eukaryotes ., There is a fundamental difference between processing of trypanosome minicircle Okazaki fragments with that in other cells ., In either prokaryotes or eukaryotes , Okazaki fragment primers are generally removed and fragments are ligated immediately after their synthesis 23 ., In trypanosome mitochondria , on the other hand , minicircle Okazaki fragments are not joined until after the progeny minicircles have segregated ., In T . brucei , theta-type replication apparently occurs in the KFZ , and then the segregated progeny are thought to migrate to the antipodal sites ( probably with one sister minicircle going to each antipodal site 2 ) ., At this stage the progeny molecules with a newly-synthesized H-strand are designated multiply-gapped circles , and the gaps are positioned between the ∼100 nucleotide Okazaki fragments 9 , 24 ., The presence in the antipodal sites of multiply-gapped minicircles ( with a 3′ OH terminus on each Okazaki fragment ) explains the intense in situ labeling of these sites by terminal deoxynucleotidyl transferase and a fluorescent dNTP 25 ., The antipodal sites also contain enzymes that likely function in primer removal and gap repair ., These include structure-specific endonuclease I ( SSE-1 , homologous to the 5′ exonuclease domain of bacterial DNA polymerase I ) 26 ., RNAi of SSE-1 confirms its involvement in primer removal 11 ., It is likely that following primer removal all but one of the gaps are repaired by DNA polymerase β and DNA ligase kβ , both of which are positioned in the antipodal sites 13 , 16 ., Following repair of most but not all gaps , these minicircles , together with their sister minicircles ( also containing a single gap adjacent to or overlapping the L strand start site ) are reattached to the network periphery by a topoisomerase II that is also positioned in the antipodal sites 14 , 15 ., Neither free minicircles nor network minicircles from procyclic T . brucei contain 5′ ribonucleotides derived from primers ( Fig . 6 and 11 ) , suggesting that in these cells primer removal is efficient ., In contrast , we found one or two ribonucleotides on network minicircles ( both on the leading strand and at least on the first Okazaki fragment ) , in cells that had undergone RNAi knockdown of SSE-1 11 ., However , the newly-synthesized L-strands on network minicircles in T . equiperdum bloodstream forms have one or two 5′ ribonucleotides 27 and in the related parasite C . fasiculata has up to six 28 ., No residual RNA primer was found associated with the minicircle H strand fragments in T . equiperdum 29 ., Finally , we do not know for any of these parasites the initial length of the primer or all of the enzymes involved in their removal ., C . fasciculata has a mitochondrial RNase H1 30 and a comparable enzyme is found in T . brucei 31; this enzyme may also contribute to primer removal ., To understand processing of minicircle Okazaki fragments it is essential to consider the enzymology of this complex pathway in nuclei of other eukaryotes ., Proteins involved in this process include flap endonuclease 1 ( FEN1 ) , RNase H , Dna2p , replication protein A ( RPA ) , DNA polymerase δ , and DNA ligase I 23 ., RNase H removes the primer one nucleotide upstream of RNA-DNA junction 32 , and the remaining ribonucleotide is then cleaved by FEN1 33 ., S . cerevisiae also contains an RNase H-independent pathway in which DNA polymerase δ can strand-displace the RNA primer , forming a flap intermediate ., Most flap intermediates are short and can be cleaved by FEN1 itself 34–37 ., However , long flaps ( >30 bases ) may also be generated by DNA polymerase δ ., The long flap is then coated by the single-strand binding protein RPA , which recruits Dna2p , a protein with both 5′ to 3′ helicase and nuclease activities ., Dna2p cleaves the long flap into a shorter flap that is subsequently removed by FEN1 ., Finally the resulting gap is repaired by polymerase δ and ligase I 38 , 39 ., Recent studies in yeast have uncovered a role for PIF1 helicase in these reactions ( ScPIF1 is found in both the mitochondria and the nucleus ) 40–42 ., The genetic interaction between PIF1 , DNA2 and a subunit of pol δ ( POL32 ) , together with the biochemical studies 43 , 44 , indicate that Pif1p may assist pol δ in generating the flap , which is processed subsequently by Dna2p 45 ., The mechanism by which Pif1p functions in this process is still unclear ., Here we found that TbPIF5 plays an important role in minicircle Okazaki fragment maturation ., Our most significant finding was that overexpression of TbPIF5 causes accumulation of fraction H , which is a minicircle species that contains a growing lagging strand ( ranging from 0 . 1 kb to 1 kb ) on the 1 kb L-strand templates ., We now propose a model explaining how TbPIF5 overexpression causes accumulation of fraction H ( Fig . 7B ) ., As discussed above ( and diagramed in Fig . 7A ) , Okazaki fragment joining in wild type cells does not occur until after minicircle progeny have segregated and migrated to the antipodal sites ., TbPIF5 ( alone or together with other proteins ) likely unwinds RNA primers , generating flaps that are subsequently degraded ., The gaps are filled and repaired probably by DNA polymerase β and DNA ligase kβ ., To prevent pre-maturation of Okazaki fragments , cells must tightly control the recruitment of some key enzymes such as TbPIF5 ., For example , TbPIF5 may bind to the minicircle progeny only after their segregation and migration to the antipodal sites ., It would not be surprising that overexpression of TbPIF5 perturbs the timing and location of Okazaki fragment processing ., Excess TbPIF5 could bind to minicircle θ-structures , triggering premature removal of primers ( Fig . 7B ) and permitting joining of Okazaki fragments ., If TbPIF5 also removes RNA primers that are not yet extended by a DNA polymerase , then further extension of the H-stand would be effectively blocked ., L-strand synthesis would proceed to completion , allowing segregation of a sister with a full length newly-synthesized L-strand and another with a truncated H strand in which the Okazaki fragments had been joined ., The latter molecules , with a heterogeneously-sized H-strand , form fraction H . Topoisomerase II might not recognize these molecules and therefore fail to reattach them to the network ., Thus , fraction H gradually accumulates , presumably within the antipodal sites ., This defect in minicircle attachment could explain the shrinking and eventual loss of kDNA that occurs following overexpression of TbPIF5 ., Further studies are needed on this helicase and other proteins involved in primer removal to fully understand the mechanism of minicircle Okazaki fragment processing ., Procyclic strain 29-13 ( from G . Cross , Rockefeller University ) was used for RNAi ., Procyclic strain 927 was used for the localization experiment ., Conditions for cell culture and transfection were described previously 18 , 46 ., The first 500 bp of the TbPIF5 coding sequence were PCR-amplified using genomic DNA isolated from procyclic strain 427 and inserted into the pJZM and stem-loop vectors 18 ., RNAi methodologies were described previously 18 ., DNA and RNA purification , gel electrophoresis , Southern blotting , Northern blotting , and sucrose gradient sedimentation were performed as described previously 20 ., The TbPIF5 knockout was conducted as described previously 47 ., Electron microscopy of isolated kDNA networks was done as described 48 ., Fragments of the 3′-end of TbPIF5 coding region ( 500 bp ) and its neighboring 3′ untranslated region ( 500 bp ) were PCR amplified using primers a–d: a , 5′GACCGGTACCCGTCTCACGCGCTTACCTATTG 3′; b , 5′ GCAGCTCGAGTTCTTCCACTTCCCCTTCATACTCCCC 3′; c , 5′ GCGGGGATCCCCGAGAGCGATGAGCGAAAAAG 3′; d , 5′ GCATCGGGGCGGCCGCACTCTCTCTCTCTCCATCTATGAATGC 3′ ., PCR products were inserted into pMOTag33M 49 ., After digestion with Acc65I and NotI , the DNA fragments were transfected into procyclic strain 927 ., The coding sequence ( minus the first 49 amino acids which constitute a predicted mitochondrial targeting signal ) was amplified by PCR , cloned into pET28a ( Novagen ) , and transformed into the E . coli Rosetta™ ( DE3 ) pLysS strain ( Novagen ) ., The cells were inoculated into 500 ml of LB medium ( containing 34 µg/ml chloramphenicol and 30 µg/ml kanamycin ) and grown at 37°C to an OD600 nm of 0 . 6 ., After addition of 1 mM IPTG , the culture was incubated for another 3 h at 25°C ., Cells were harvested by centrifugation ( 8000 g , 10 min ) and the cell pellet was resuspended in 20 ml buffer A ( 50 mM sodium phosphate , 300 mM NaCl , 10 mM imidazole , pH 8 . 0 ) ., After lysis by sonication , the suspension was centrifuged ( 10000 g , 30 min ) and the supernatant was mixed gently with 2 ml Ni-NTA slurry ( Qiagen ) ( 1 h , 4°C ) ., The Ni-NTA beads were then washed 4 times with 2 ml buffer B ( 50 mM sodium phosphate , 300 mM NaCl , 20 mM imidazole , pH 8 . 0 ) ., Proteins were eluted 3 times with 0 . 5 ml buffer C ( 50 mM sodium phosphate , 300 mM NaCl , 250 mM imidazole , pH 8 . 0 ) ., The eluates were dialyzed overnight at 4°C against buffer D ( 25 mM Tris-HCl , 300 mM NaCl , 1 mM DTT , pH 7 . 5 ) ., The samples were loaded onto a 0 . 5 ml heparin-Sepharose FF ( Bioscience Healthcare ) column equilibrated with the same buffer ., Recombinant protein was eluted at 0 . 8 M NaCl and dialyzed against buffer E ( 25 mM Tris-HCl , 100 mM NaCl , 1 mM DTT , pH 7 . 5 ) ., Recombinant TbPIF5 is very unstable and it was freshly prepared for the activity assays ., For ATPase assay , recombinant TbPIF5 ( 10 , 20 , and 50 ng ) was incubated ( 20 µl reaction , 10 min , 37°C ) with 8 . 25 nM γ-32P ATP ( 6000 Ci/mmol ) , 150 µM non-radioactive ATP , 50 mM Tris-HCl , pH 8 . 5 , 50 mM NaCl , 2 mM DTT , 2 mM MgCl2 , 0 . 25 mg/ml bovine serum albumin , and 50 ng M13mp18 ssDNA ., Samples ( 1 µl ) were spotted onto a polyethyleneimine-cellulose plate ( J . T . Baker , USA ) and developed in 1 . 0 M formic acid/0 . 5 M LiCl followed by autoradiography ., For helicase assays , the M13-based substrate was constructed as described 50 and the substrates for polarity assay were made as described 51 ., Assays ( 20 µl each ) contained various amounts of TbPIF5 , 50 mM Tris-HCl , pH 8 . 5 , 50 mM NaCl , 2 mM DTT , 2 mM MgCl2 , 2 mM ATP , 0 . 25 mg/ml bovine serum albumin , and the substrate ( 15 fmol ) ., Reactions were incubated at 37°C for 10 min and subjected to electrophoresis with a 12% polyacrylamide gel in 0 . 5×TBE ( 150 V , 1 h ) ., The gel was dried and autoradiographed . | Introduction, Results, Discussion, Materials and Methods | Trypanosoma bruceis mitochondrial genome , kinetoplast DNA ( kDNA ) , is a giant network of catenated DNA rings ., The network consists of a few thousand 1 kb minicircles and several dozen 23 kb maxicircles ., Here we report that TbPIF5 , one of T . bruceis six mitochondrial proteins related to Saccharomyces cerevisiae mitochondrial DNA helicase ScPIF1 , is involved in minicircle lagging strand synthesis ., Like its yeast homolog , TbPIF5 is a 5′ to 3′ DNA helicase ., Together with other enzymes thought to be involved in Okazaki fragment processing , TbPIF5 localizes in vivo to the antipodal sites flanking the kDNA ., Minicircles in wild type cells replicate unidirectionally as theta-structures and are unusual in that Okazaki fragments are not joined until after the progeny minicircles have segregated ., We now report that overexpression of TbPIF5 causes premature removal of RNA primers and joining of Okazaki fragments on theta structures ., Further elongation of the lagging strand is blocked , but the leading strand is completed and the minicircle progeny , one with a truncated H strand ( ranging from 0 . 1 to 1 kb ) , are segregated ., The minicircles with a truncated H strand electrophorese on an agarose gel as a smear ., This replication defect is associated with kinetoplast shrinkage and eventual slowing of cell growth ., We propose that TbPIF5 unwinds RNA primers after lagging strand synthesis , thus facilitating processing of Okazaki fragments . | Trypanosoma brucei is a protozoan parasite that causes human sleeping sickness in sub-Saharan Africa ., Trypanosomes are primitive eukaryotes and they have many unusual biological features ., One prominent example is their mitochondrial genome , known as kinetoplast DNA or kDNA ., kDNA , with a structure unique in nature , is a giant network of interlocked DNA rings known as minicircles and maxicircles ., kDNA superficially resembles chain mail in medieval armor ., The network structure dictates an extremely complex mechanism for replication , the process by which two progeny networks , each identical to their parent , are formed ., These progeny networks then segregate into the daughter cells during cell division ., One feature of this replication pathway , in which discontinuously synthesized strands of minicircles are joined together in a reaction involving an enzyme known as a helicase , is the subject of this paper ., Since there is nothing resembling kDNA in human or animal cells , and since kDNA is required for viability of the parasite , enzymes involved in this pathway are promising targets for chemotherapy . | molecular biology/dna replication, microbiology/parasitology | null |
journal.pcbi.1006355 | 2,019 | Gap junctions set the speed and nucleation rate of stage I retinal waves | Spontaneous activity spreads through neuronal systems of many different mammal species during development ., Crucial roles are attributed to this spontaneous activity 1 ., Among the most prominent roles is the synaptic refinement in the retina , where spatio-temporally correlated bursts of activity are observed , and it was found that blocking these waves disrupts eye-specific segregation into the visual thalamus 2 , 3 ., Therefore , much effort has been devoted in recent years ( e . g . 4–7 ) to understand the mechanisms responsible of retinal waves ., The observed patterns of spontaneous activity in the developing retina are remarkably similar across many species 1 ., These patterns have been characterized as spatially correlated bursts of activity in the ganglion cell ( GC ) layer , which are followed by periods of silence 8–10 ., So far , three different stages of retinal waves have been described in rodents , ( for review see e . g . 1 ) ., These different stages are characterized by their underlying circuits , which mature subsequently in development ., In stage I , bursts of activity spread between retinal ganglion cells ., In this stage , few synapses are identifiable and waves are mediated by gap junctions ( GJs ) and adenosine 11 ., Stage II begins with the onset of synaptogenesis and ends with the maturation of glutamatergic circuits while stage III waves end with eyeopening and the onset of vision 12 , 13 ., Here , we exclusively focus on the earliest developmental stage ( stage I ) ., This stage is prior to the emergence of functional chemical synapses in the retina ., Waves show random initiation sites , no directional bias , and a propagation speed of about 450 μm/s ., Via patch-clamp recordings , stage I retinal waves were found to be initiated and propagated in the GC layer 11 ., In this work we develop a theoretical model of the retina and limit ourselves to a GC layer of bursting neurons which are coupled by GJs ., These electrical synapses are formed between each of the major neuron types in the vertebrate retina 14–18 and play a major role in signal processing and transmission of visual information ( for a review , see 18 ) ., GJs are formed by two apposed hemichannels , each one formed by an hexameric array of proteins know as connexins ., In mammals , connexin-36 and connexin-45 were clearly identified in neurons located in the inner retina 15 , 19 ., Both types of connexins follow a distinct expression pattern during retinal development 20 ., GJ coupling between neurons has been addressed in various theoretical studies ( see e . g . 21 , 22 ) and has received particular attention in the context of large-scale brain rhythms ( e . g . 23 , 24 ) and traveling wave dynamics ( see e . g . 25 , 26 ) ., However , their involvement in the maturation process of the retina is not yet fully understood 27 ., GJs have been proposed as the responsible mediator of stage I retinal waves but not yet been used in a model of such waves 5 , which is the problem that we intend to solve with this study ., From a physical perspective , GJs act with integration times of the order of milliseconds and were thus argued not to be the mediator of stage I waves 5 , 9 , which are much slower compared to this time-scale ., In this work , we present a model of stage I retinal waves , formed by a network of bursting cells ., The cells are coupled by the Ohmic currents through GJs which corresponds to the discretized version of a diffusive coupling ( see e . g . 28 for a recent example of complex pattern generation with such a coupling ) ; for recent studies of wave propagation using the alternative spatially extended coupling by an integral kernel , see e . g . 29 , 30 ., For our model , we show that under certain conditions , the wave propagation can be sufficiently slow to be the responsible mediator for stage I retinal waves ., We discuss analytical estimations of the propagation velocities and compare them to extensive numerical simulations of networks of up to 12 , 000 neurons ., Our analytical work , based on diffusively coupled bursting neurons , applies methods from nonlinear dynamics and pattern formation to differential equations with discontinuous resettings ., Furthermore , we study the repetitive nucleation of waves caused by noisy input currents and discuss the dependence of the nucleation rates on the noise intensities ., We use the phenomenological Izhikevich neuron model 31 , 32 , known for displaying biologically plausible dynamics ., Due to its discontinuous fire and reset mechanism , it is a computationally efficient model of a bursting neuron ., Comparable dynamics can be obtained from two-dimensional excitable models such at the Morris-Lecar model , under incorporation of an additional third dimension ,, e . g ., a calcium-dependent potassium current , cf ., Sec . 5 . 2 in 33 The model can be regarded as a quadratic integrate-and-fire neuron for the membrane voltage Vi ( t ) of the ith neuron with an additional slow recovery variable ui ( t ) , also referred to as gating variable ( cf ., Fig 1, ( a ) for the nullclines of the system ) :, τ V d V i d t= a ( V i - V rest ) ( V i - V crit ) - u i + R I i , ( 1 ), τ u d u i d t= b V i - u i , ( 2 ), if : V i ≥ V peak → { V i = V reset , u i = u i + d ., ( 3 ), The membrane recovery variable provides negative feedback to the voltage ( cf ., Fig 1, ( b ) and 1, ( c ) top ) ., The parameters a , b , d as well as Vrest , Vcrit , Vreset , and Vpeak determine the spiking regime of the neuron , with Vrest < Vcrit < Vpeak ., The time-scales of the voltage and gating variable are defined by τV and τu , respectively ., For u ( t ) ≡ 0 and I ( t ) ≡ 0 , Vrest and Vcrit are the stable and the unstable fixed points of the dynamics , respectively ., If Vi ≥ Vpeak , the membrane potential is reset to Vreset , the kth spike time , ti , k , is registered , and the recovery variable is increased by the constant value, d . We choose parameters such that the burst characteristics of our model neuron illustrated in Fig 1 roughly agree with experimental measurements from Syed et al . 11 ., Specifically , we aim at a burst duration of about 1 − 2 seconds ( cf ., Fig 1, ( c ) bottom ) and a spike frequency during bursts of about 5 − 15 Hz ., We find those characteristics reasonably met for: a = 0 . 1 , b = 0 . 3 , d = 1 . 2 , τV = 100 msec , τu = 0 . 0003−1 msec , Vrest = −76 mV , Vcrit = −48 mV , Vpeak = 30 mV , Vreset = −50 mV ., The bursting mechanism is illustrated in Fig 1 ., The chosen time-scale of the gating variable u is comparatively large , but not uncommon for cortical neurons 34 ., The total current RIi = RIgap , i + Inoise , i is a superposition of the intrinsic noise current and GJ currents from neighboring cells ( see below ) ., The intrinsic noise originates from fluctuations of the various channel populations ( sodium , calcium , and different potassium channels , see, e . g ., 35 ) and is approximated by white Gaussian noise:, R I noise , i = τ V 2 D ξ i ( t ) , ( 4 ), with 〈ξi ( t ) 〉 = 0 and 〈ξi ( t ) ξj ( t′ ) 〉 = δij δ ( t − t′ ) and D is the noise intensity ., We perform simulations at discrete times with a time step of Δt = 0 . 1 msec according to an Euler-Maruyama integration scheme , see supporting information S1 Text ., Ganglion cells are distributed within the ganglion cell layer with a decreasing density towards the outer regions of the retina ., For instance , the density in rabbits covers a range from 5000 cells/mm2 down to 200 cells/mm2 ( the mean value is 800 ) 36 ., In a previous study of retinal waves observed in rats , Butts et al . 4 used a ganglion cell density of ∼ 4000 cells/mm2 ., In their simulations they placed neurons in a regular triangular lattice for which the given density translates to a lattice spacing of 17 μm ., Because we focus on the rabbit retina , we assume a triangular lattice with a different lattice spacing of 38 μm , reflecting the lower cell density ( 800 cells/mm2 ) for this system ., The reported experimental observations on characteristics of stage I retinal wave were obtained from retina patches of roughly 3 × 5 mm ., A mean cell density of 800 cells/mm2 translates to a total cell number estimate of 12 , 000 cells in the studied system ., For comparability , we use a similar number of cells for simulations ( i . e . 12 , 100 = 110 × 110 ) ., The triangular lattice structure can be seen in Fig 2, ( a ) ., Here , we ignore for simplicity the inhomogeneous and irregular structure of the ganglion cell layer ., We place N = n × n single ganglion cells in a rectangular domain on a triangular lattice such that every cell is connected with GJs to six nearest neighbors , ( the lattice structure is illustrated in Fig 2, ( a ) ) ., For illustrative purposes , we will also consider a one-dimensional chain , in which each neuron has only two neighbors ., Because we are interested only in stage I waves , prior to synaptogenesis , these cells are not connected to any other cells , i . e . bipolar and amacrine cells are not part of our model ., We choose a common approach ( e . g . 21 ) to model the GJ current as diffusive and instantaneous coupling by, R I gap , i = G ∑ n = neighbor ( V n - V i ) , ( 5 ), where G is the rescaled dimensionless GJ coupling , i . e . G = R/Rgap ., The membrane resistance R of retinal ganglion cells can experimentally be measured and is in the range of 100-500 MΩ ,, e . g ., 37 ., Rgap is the GJ resistance between neighboring ganglion cells in the retina , which depends on the connexin type and the transjunctional voltage difference and is roughly Rgap ≈ 1GΩ 38 , 39 ., The values of R and Rgap imply a physiological range for our parameter of G ∈ 0 . 1 , 0 . 5 ., Because the time course of the action potential produced by our neuron model is only a coarse approximation of the electrophysiological shape of a spike , the GJ coupling may be stronger or weaker than assumed here ., This gives additional justification for choosing a wider range of G . For the two-dimensional setup , we apply two different boundary conditions ., For estimating the noise dependence of propagation velocities and nucleation rates , we perform small system simulations ( N∼50-260 ) with periodic boundary conditions in both directions ( system on a torus ) in order to avoid strong finite-size effects ., Simulations of the full system with N∼12 , 000 are carried out with two additional layers of neurons on the boundary , that are not exposed to intrinsic noise ( cells on the system boundary have fewer neighbors , between 2 and 5 instead of 6 ) ., Neurons in the two outer layers of the large simulations are discarded from all statistical evaluations ., Single propagating waves running through the network can be captured by the population activity 40, A ( t ) = 1 N Δ t A ∑ i = 1 N ∑ k ∫ t t + Δ t A d t δ ( t - t i , k ) , ( 6 ), where the index k runs over the spike times of the ith neuron ., Hence , A ( t ) is the firing rate , averaged over the network and the time bin ΔtA ., We use ΔtA = 0 . 5 seconds , which is comparatively large and covers multiple spikes when the cells are bursting ., If we couple cells in a chain ( as indicated in Fig 2, ( a ) —1D ) and initiate a burst in one of them , we see a propagation of the burst along the chain ( cf ., Fig 2, ( b ) ) ; similar voltage traces have also been seen in simulation of computational models of cortex slices , e . g . 41 ., A higher propagation speed can be achieved by increasing the GJ conductance parameter G Fig 2, ( c ) ., The picture is similar in our two-dimensional setup , for which snapshots are shown in Fig 2, ( d ) ., In this case , the wave has been evoked by enforcing a burst in the lower left corner ., It propagates as a circularly shaped wave front , which is a consequence of the regularity and rotational symmetry of the system ., The gating variable u ( lower row in Fig 2, ( d ) ) can be associated with the experimentally accessible calcium dynamics and resembles calcium fluorescences images 11 ., Compared to the membrane potential ( top row ) , the wavefront of the gating variable lags behind , as it slowly builds up during the burst ., In both , one-dimensional and two-dimensional simulations in Fig 2 , we have set the intrinsic noise intensity to zero in order to illustrate that wave propagation does not hinge on the presence of fluctuations ., We note already here , that the propagation speed in the two-dimensional system matches the order of magnitude of biologically observed values ., To determine the speed of the waves from simulation such as shown in Fig 2, ( c ) , we approximate the wave’s shape as circular with a fixed center ., We define a wavefront as the group of neurons that spike within the same time bin of Δt = 0 . 1 seconds ( see left illustration in Fig 3, ( a ) ) and measure the front’s mean distance from the center and its mean time instance of occurrence ., From the differences of these distances and times , we determine the mean velocity , which we find to be weakly distance dependent , but saturating at about 350 μm from the origin of the wave , cf ., Fig 3, ( b ) ., In the following , all velocity values are averaged over measurements for the range of distances 350 − 650 μm ( shaded area in Fig 3, ( b ) ) from the point of initiation and we refer to this measuring method as concentric method ., The velocities are shown in Fig 3, ( c ) as a function of the GJ parameter for the physiologically relevant range of G ( see Methods ) ., We obtain velocities that are in the range of values observed in the rabbit retina 11 , cf ., the shaded area in Fig 3, ( b ) ., The experimental mean value of about 450 μm/sec is attained for G ≈ 0 . 4 ., The propagation and its speed can be theoretically understood as follows ., Assuming a steep wave profile , the speed of the wave is given by the inverse of the time it takes a bursting neuron to excite its neighbors , times the displacement of the corresponding wave fronts ., We refer to this time as burst onset time difference ( BOTD ) ., For simplicity , we neglect noise and consider in the following a one-dimensional setup consisting of three neurons: one initially quiescent neuron ( i ) is connected to a bursting neuron ( i − 1 ) on one side and to a quiescent neuron ( i + 1 ) on the other side ., They are separated by the lattice spacing ℓ = 38 μm , hence the velocity is defined as v1D = ℓ/TB ., Therein , TB denotes the analytical approximation of the BOTD for this one-dimensional case ., The approximation TB for the BOTD between neighboring neurons can be derived using three assumptions ( details in S1 Text ) ., First , we assume a constant gating variable ( u ( t ) ≈ ur = const ) , which is reasonable on a short time-scale , because τu ≫ τV ., Second , we replace the voltage variable of the bursting neuron Vi−1 ( t ) by its temporal average V ¯ b = const , that can be analytically calculated ( see S1 Text ) and for our standard parameters is V ¯ b = - 34 mV ., Third , we replace the voltage of the quiescent neuron that is not directly connected to the bursting neuron by the resting potential , Vi+1 = Vr ., Consequently , the GJ current seen by the driven neuron reads R I gap , i = G ( V i - 1 + V i + 1 - 2 V i ) ≈ G ( V ¯ b + V r - 2 V i ( t ) ) , and the resulting dynamics until the voltage Vi reaches the peak potential for the first time is effectively one-dimensional and can be recast to the form ( cf . details in S1 Text ) :, τ V d V i d t ≈ a ( V i - V rest ) ( V i - V crit ) - u r + G ( V ¯ b + V r - 2 V i ) ., ( 7 ), This first order ordinary differential equation can be solved via separation of variables to find t ( V ) ., We obtain it by first calculating the difference of the times from the voltage being at its peak potential and its resting potential ., However , the driven neuron is already exposed to the driving GJ current while the voltage of the bursting neuron travels to its first spike time ( cf . Fig . A of S1 Text ) ., Therefore , for simplicity we subtract the first inter-spike interval TISI from the beforehand calculated time difference:, T B ( G ) =t ( V peak ) - t ( V r ) - T ISI ., ( 8 ), The explicit expression is lengthy and derived in S1 Text , resulting in Eq ., O of S1 Text ., Comparing TB to simulations of a one-dimensional chain shows a reasonable agreement ( cf . Fig . A of S1 Text ) , although the theory overestimates the simulated values , in particular , for larger values of G . For comparison we also discuss a corresponding result for the wave velocity in the continuum limit in S1 Text ., In the two-dimensional setup at larger times , the wave attains a planar shape as indicated in Fig 3, ( a ) , where red circles represent bursting neurons and blue and yellow circles represent driven and quiescent neurons ., Now , we assume that the wave front is perfectly flat and all neurons shown in the same color share an identical voltage ., In that case , the propagation mechanism simplifies to two bursting neurons exciting one quiescent neuron , whose membrane potential is further affected by two quiescent neurons ., Hence , we can mimic the quasi one-dimensional situation by doubling the value of G and additionally taking into account the modification of the effective length , i . e . ℓeff = ( 3/4 ) 1/2 ℓ , see Fig 3, ( a ) ., Consequently , we can approximate the velocity in the two-dimensional system as, v 2 D ( G ) =3 / 4 · ℓ T B ( 2 G ) ., ( 9 ), Calculated velocities v2D ( G ) are shown in Fig 3, ( c ) by the blue line , underestimating the true velocity ( circles ) but providing a correct order-of-magnitude estimate ., Note that so far we restricted the considerations to a purely deterministic setup ., Our simulations with noise indicate that moderate fluctuations have only little impact on the mean velocities ., In the stochastic version of our system , we observe spontaneous waves that resemble those found in experiments 11 ., Experimentally , it was observed by Syed et al . 11 that the spontaneously nucleated waves appear with a mean inter-wave interval TIWI of 36 seconds ., In our model , waves are initiated by noise , since neurons are set in the excitable regime and cannot generate periodic spiking or bursting without external input ., We expect that the nucleation rate per neuron depends strongly on the noise intensity D . To characterize this dependence , we simulate small systems ( N∼50-260 , see Methods ) with periodic boundary conditions for two different values of GJ coupling and different noise intensities , cf ., Fig 4 ., With the understanding that every neuron has the same chance to trigger a wave , the global nucleation rate should be linear with N to a first approximation ., Thus we measure the nucleation rate per neuron as r = 1/ ( TIWI N ) ., As demonstrated in Fig 4 by the linear dependence of the rate’s logarithm on the inverse noise intensity , we obtain an Arrhenius rate, r=r 0 exp ( - Δ U / D ) ., ( 10 ), The effective potential barrier ΔU depends on G and the system size N and saturates for sufficiently large systems ( inset ) for both values of G . The increase of the potential barrier with G can be understood to first approximation by the effective change of the current-voltage relation in the single neuron ., The GJ coupling term Eq ( 5 ) leads to an effective increase in the leak current that stabilizes the resting potential and makes it harder to initiate a burst ., This mechanism is dominant in comparison to the influence of other coupling effects and the stochasticity of the neighbors on the nucleation rate ( supported by additional simulations , see Fig . B of S1 Text ) ., The more subtle dependence of ΔU on the system size can be explained as follows: Coupling stochastic neurons in small systems with periodic boundary conditions leads to spatial correlations and thus effectively to stronger noise ., This effect can be neglected for large system sizes or weak coupling , but has a measurable effect otherwise ( cf . Fig 4 and Fig 4 inset ) ., Our results so far can be used to predict the mean inter-wave interval and the propagation speed of retinal waves for a system size N = 12 , 100 that roughly corresponds to the experimentally studied patch size in Ref ., 11 ., Vice versa , we can infer an approximate value of the noise intensity D that leads to the experimentally observed value of TIWI = 36 seconds and test this in numerical simulations of the full system ., For our estimation of the rough value of the noise intensity in a large system , we have to take into account that the single neuron undergoes a substantial refractory period of Tref ≈ 14 seconds after bursting ( estimated from small-system simulations investigating the minimal mean inter-wave interval for various noise intensities ) ., The mean inter-wave interval is then given by TIWI = Tref + 1/N ⋅ r ( D ) and the estimated value of the noise intensity follows from the Arrhenius law , Eq ( 10 ) , as, D *=Δ U / ln N ( T IWI - T ref ) r 0 ≈ 0 ., 050 ( 11 ), ( for G = 0 . 4 , and r0 = 6 and ΔU = 0 . 71 , fit parameters from Fig 4 , solid line with N = 256 ) ., The estimated parameters , G = 0 . 4 and D = 0 . 050 , can now be used in a large-scale simulation ., In Fig 5, ( a ) , we show snapshots of the full system’s gating variable ( a proxy for the experimentally accessible calcium concentration ) ., The wave front seen in the experimentally observable area ( box in Fig 5, ( a ) ) looks similar to experimental measurements , cf ., Ref ., 11 ., From Fig 5, ( b ) , it becomes evident that the mean inter-wave interval becomes much shorter for a slight increase in D . The mean inter-wave interval at these parameter values is not exactly 36 seconds , but somewhat larger: these statistics depend very sensitively on the value of the noise intensity ( i . e . on the second leading digit , cf ., Fig 5, ( c ) middle ) ., This is seen in the global population activity , that reveals a wave going through the system as a single peak vs . time ., The dependence of crucial neural statistics on the noise intensity is illustrated in Fig 5, ( c ) ., In contrast to the mean inter-wave interval , the mean velocity of the wave does not depend strongly on the noise ( Fig 5, ( c ) , top ) but stays close to the experimentally observed mean value ( dashed line ) ., This is due to the fact , that the wave , once it is initiated , is largely determined by the deterministic propagation mechanism explained above ., The fine tuning of the noise intensity shows that the experimental value of 〈TIWI , exp〉 = 36 seconds is attained for a noise level of D = 0 . 052 , slightly larger than D* ( estimated above ) ., How realistic is this noise level ?, To address this question , we show at the bottom of Fig 5, ( c ) the standard deviation of the subthreshold voltage fluctuations , σV , as a function of the noise intensity D . σV increases only slightly with D and attains values around 1 . 6 mV ., To our knowledge , there are no detailed investigations of intrinsic noise sources in retinal ganglion cells at embryonic age ., Because in this developmental stage there are no chemical synapses present 42 , the synaptic background fluctuations can be excluded for our system: in the recurrent networks of the cortex , fluctuations stem mainly from the many synaptic interactions among the neurons , resulting in the famous asynchronous irregular state 43 that can be highly variable 44–46 ., In our system , one likely source of variability is channel noise that typically leads to small membrane potential fluctuations with a standard deviation σV below 0 . 6 mV 47 , 48 ., The noise intensity that is required for the experimentally observed inter-wave interval results in sub-threshold voltage fluctuations that are three times bigger , cf ., Fig 5, ( c ) bottom , suggesting that besides ion channel noise there are additional sources of fluctuations present ., These could result from stochasticity of GJs itself but also indirectly from GJs via differences in individual resting potentials ( for the heterogeneity of the resting potential in similarly sized cells , pyramidal cells in the cortex , see 49 ) ., In any case , the apparent voltage fluctuations of about 1 . 6 mV are well within the range of experimentally observed voltage noise in embryonic ganglion cells ( cf . Fig . 1 in Ref . 11 ) ., The investigations presented in this paper propose a GJ-based model of stage I waves in the developing retina ., Starting with a neuron model that roughly reproduces the spiking properties of a burst of one single retinal ganglion cell , we incorporated GJ coupling of physiologically plausible strength and temporally uncorrelated fluctuations ., This allowed us to reproduce the characteristics of wave nucleation and slow wave propagation in the early retina ., Earlier it was believed that GJs can play a role in fast neural transmissions only 5 , 9 , since the current in electrical synapses responds much quicker than neurotransmitters in chemical synapses ., As shown in our paper , however , it is possible to obtain a limited transmission speed in a simple Ohmic model of the GJ coupling ., Furthermore , although stochastic fluctuations are strong enough to ignite bursts with the correct nucleation rate , they do not distort the propagating fronts very much , i . e . the wave propagation is still a reliable process ., The reason for the slow transmission we observe can be found in the nonlinear dynamics of the single neuron ., The Izhikevich model that we use for the ganglion cell is essentially a quadratic integrate-and-fire neuron model with a slow adaptation variable ., This model is the normal form of a saddle-node bifurcation and has a pronounced latency if close to this bifurcation , i . e . the spike response to a current step ( in our case provided by a neighboring bursting cell ) is considerably delayed because the system experiences the “ghost of the former fixed point” , see Ref ., 50 ., The presence of weak noise modifies this picture only slightly 51 ., Although our model accounts for the most important features of wave nucleation and propagation for stage I retinal waves , it cannot explain the strong variability of the experimentally measured statistics ( error of velocity ±91 μm/sec 11 ) ., This is due to a number of model simplifications , which we now concludingly discuss ., Firstly , the real system is much more heterogeneous than in our model , both with respect to the lattice structure as well as with respect to the local coupling between cells; secondly , GJs may couple more than next neighbors and their conductivity may be noisy and voltage gated; thirdly , the detailed dynamics of ganglion cells is certainly more complex than can be captured by the Izhikevich model; last but not least , the white Gaussian noise in our model is a rather coarse approximation of the channel noise and other fluctuations in the system ., In our model , we arranged the neurons on a highly regular lattice with a cellular spacing according to an experimentally determined mean value of cell density , neglecting the strong heterogeneities in the distribution 36 ., On this lattice , each cell is connected to exactly six nearest neighbors ., Given the aforementioned heterogeneity , the numbers and distances between neighbors will be more broadly distributed than in our model ., Incorporating these heterogeneous features in the simulations would likely broaden the range of observed velocities and thus better reflect the considerable variability found in experimentally measured values ., The soma size of ( rabbit ) retinal ganglion cells ( < 30μm , e . g . Ref 36 ) is smaller than our employed lattice spacing , implying GJ coupling between dendrites rather than soma-soma coupling only ., The size of the dendritic arbor of retinal ganglion cells is ∼ 100 − 130μm , thus suggesting direct communication between cells that are up to the threefold of the lattice spacing apart ., In our simulations with only next-neighbor coupling , we could reproduce the experimentally observed velocity with a comparatively large coupling constant of G = 0 . 4 ( physiological range was G ∈ 0 . 1 , 0 . 5 , see Methods ) ., It is conceivable , that this large G value is an effective description of a system with larger effective GJ neighborhood but with a smaller ( and possibly distance-dependent ) coupling value G . Put differently , we expect similar results for the wave speed in a system with extended coupling neighborhood but reduced coupling strength per connection ( with the latter still being within the physiological range ) ., Regarding the neuron model and the incorporation of noise , we note that for developed retinal ganglion cells detailed multi-compartment conductance-based models with stochastic ion channels exist 35 ., With more electrophysiological data available , it will certainly be possible to develop biophysically more realistic models of the bursting ganglion cell at the early stage ., Furthermore important for our problem will be the incorporation of stochastic models of GJs 52 with voltage-dependent kinetics 53 , 54 and the heterogeneity of physiological parameters such as the resting potential ., Such detailed models are certainly difficult to simulate for large networks but could be employed to estimate the total noise intensity in the system and to identify the dominant noise source , cf ., similar approaches in Refs ., 35 , 55 , 56 . | Introduction, Methods, Results and discussion | Spontaneous waves in the developing retina are essential in the formation of the retinotopic mapping in the visual system ., From experiments in rabbits , it is known that the earliest type of retinal waves ( stage I ) is nucleated spontaneously , propagates at a speed of 451±91 μm/sec and relies on gap junction coupling between ganglion cells ., Because gap junctions ( electrical synapses ) have short integration times , it has been argued that they cannot set the low speed of stage I retinal waves ., Here , we present a theoretical study of a two-dimensional neural network of the ganglion cell layer with gap junction coupling and intrinsic noise ., We demonstrate that this model can explain observed nucleation rates as well as the comparatively slow propagation speed of the waves ., From the interaction between two coupled neurons , we estimate the wave speed in the model network ., Furthermore , using simulations of small networks of neurons ( N≤260 ) , we estimate the nucleation rate in the form of an Arrhenius escape rate ., These results allow for informed simulations of a realistically sized network , yielding values of the gap junction coupling and the intrinsic noise level that are in a physiologically plausible range . | Retinal waves are a prominent example of spontaneous activity that is observed in neuronal systems of many different species during development ., Spatio-temporally correlated bursts travel across the retina at a few hundred μm/sec to facilitate the maturation of the underlying neuronal circuits ., Even at the earliest stage , in which the network merely consists of ganglion cells coupled by electric synapses ( gap junctions ) , it is unclear which mechanisms are responsible for wave nucleation and transmission speed ., We propose a model of gap junction coupled noisy neurons , in which waves emerge from rare stochastic fluctuations in single cells and the wave’s transmission speed is set by the latency of the burst onset in response to gap junction currents between neighboring cells . | cell physiology, medicine and health sciences, action potentials, nervous system, membrane potential, ocular anatomy, condensed matter physics, junctional complexes, electrophysiology, neuroscience, gap junctions, ganglion cells, waves, animal cells, physics, cellular neuroscience, wave propagation, retina, cell biology, anatomy, synapses, physiology, neurons, biology and life sciences, cellular types, physical sciences, ocular system, afferent neurons, retinal ganglion cells, neurophysiology, nucleation | null |
journal.pbio.1001518 | 2,013 | Selecting One of Several Mating Types through Gene Segment Joining and Deletion in Tetrahymena thermophila | Unicellular eukaryotes reproduce asexually , but most also have a sexual stage to their life cycle that increases genotypic variability ., Sexual partners are usually morphologically indistinguishable and mating types , as part of a self/non-self recognition system , foster outbreeding ., Mating types were first discovered by Sonneborn in the ciliate Paramecium aurelia 1 ., This discovery initiated the field of microbial genetics , as mating types were subsequently found in bacteria and a diversity of microbial eukaryotes ., The number of mating types and the mechanisms of mating type determination vary widely among unicellular eukaryotes 2–6 ., T . thermophila is a ciliate that segregates germline and somatic functions into two nuclei with distinct genome structures: the diploid micronucleus ( germline ) and the polyploid macronucleus ( somatic ) ., Starvation induces mating ( conjugation ) between two cells of different mating types ., During conjugation ( Figure S1 ) the parental somatic nucleus is destroyed while new somatic and germline nuclei are differentiated from a zygote nucleus ., This differentiation includes extensive site-specific genome rearrangements , including fragmentation of the germline chromosomes , de novo telomere addition , and deletion of thousands of internal eliminated sequences ( IESs ) 7 ., Mating type is also determined at this stage 8 ., The T . thermophila germline mat locus was first described by Nanney et al . in 1953 9 and remains the only locus known to control mating type specificity in this organism ., These authors reported that the mat locus determines a spectrum of seven mating types ( I–VII ) , one of which is stochastically and irreversibly expressed in each new somatic nucleus ., Extensive field collections have revealed no additional mating types 10 ., Two classes of germline mat alleles are known 10–13 ., The mat-1-like alleles encode mating types I , II , III , V , and VI , while mat-2-like alleles encode mating types II , III , IV , V , VI , and VII 9 ., All the strains used in this work are homozygous for the mat-2 allele of inbred strain B . Alternative DNA deletions , rather than epigenetic gene silencing , were proposed to be responsible for mating type determination 14 ., The work reported here , made possible by the molecular identification of the mating type genes , has revealed a type of programmed DNA rearrangement in the somatic nucleus that assembles a gene pair of one mating type and deletes the rest ., The genetically mapped mat locus 15–17 was assigned to a roughly 300-kb segment of a somatic chromosome sequence assembly ( Figure S2 ) ., As cells must be starved to mate , we assumed that a candidate mating type gene would be expressed in a mating type-specific manner during starvation and not expressed during growth ., In a previous whole-transcriptome RNA-seq study 18 , mRNA was prepared and sequenced from starved SB4217 ( mating type V or mt V ) cells as well as from starved and growing SB4220 ( mt VI ) cells ( Table S1 ) ., To identify mating type candidate genes , we mapped the RNA-seq reads to the 300-kb segment of the mt VI somatic reference genome 15 , 19 ., Two adjacent genes in this region showed mating type-specific expression in starved cells and no expression during growth ( Figure 1A ) making them good mating type gene candidates ., We named these genes MTA and MTB ., A transcript for each gene was assembled primarily from reads that mapped to the mt VI reference genome ., Reads from mt VI covered both genes except for one small gap in MTA , which was filled in by cDNA sequencing ( unpublished data ) ., Northern blot analysis ( Figures 2 and S3 ) confirmed a single transcript for each mt VI gene ., Only the terminal exons of MTA and MTB could be assembled from the mt V reads that mapped to the mt VI reference genome ( Figure 1B ) ., In addition , a partial transcript was assembled de novo from the mt V RNA-seq reads ( Text S1 ) ., Two thirds of this partial transcript has 99 . 9% identity with the terminal exon of mt VI MTA gene but the remainder is absent from the mt VI somatic reference genome and could encode a mating type-specific segment ., The MTA and MTB genes identified above are arranged head to head , are divergently transcribed ( Figure 1B ) and are predicted to code for unique proteins ., The MTA gene ( TTHERM_01087810 , KC405257 ) is predicted to encode a 161-kD protein while the MTB gene ( TTHERM_01087820 , KC405257 ) is predicted to encode a 194-kD protein ., Each terminal exon is unique in the somatic mt VI genome sequence and both are predicted to encode transmembrane ( TM ) helices ., TM domain proteins that can localize to the cell surface could play a role in self/non-self recognition , since cell-cell contact is required to stimulate cells to mate 20–22 ., If the MTA and MTB genes determine mating type , they may also be essential for mating ., This was addressed by removing the entire somatic gene pair of mt VI ( SB210 ) by homologous gene replacement ( Figure S4A and S4B ) 23 ., The gene pair knockout ( MT– ) abolished the cells ability to pair or produce progeny when mixed with starved wild-type ( wt ) cells of a different mating type or with cells of the same mating type ., Identical results were obtained with three independent knockout strains ., In contrast , control assays of mating between two wt strains of different mating types showed high levels of pair formation and produced abundant ( >85% ) progeny ., Each gene of the mt VI gene pair was deleted independently to investigate the functional relationship between the two genes ., For both single knockouts , RT-PCR showed that removal of one gene did not abolish expression of the remaining gene ( Figure S4C ) ., Three independent MTB knockouts ( MTB– ) gave the same results as the gene pair knockout ., No progeny were produced when MTB– cells were mixed with wt cells of a different mating type ., The MTA knockout ( MTA– ) retained mating specificity but very little mating competence ., It paired extremely poorly and rarely produced progeny ( 0 . 16% on average ) when mated with wt cells of a different mating type ., No pairs or progeny were detected when it was mated to cells of the same ( mt VI ) mating type ., Identical results were obtained with three independent knockout clones ., To determine whether other mating types express genes containing the TM exons shared by mts V and VI , we isolated RNA from starved , mature strains of each mating type ( Table S1 ) ., Northern blot analysis revealed that cells of every mating type have MTA– and MTB-like transcripts ( Figures 2 and S3 ) ., The length of the transcripts is similar to the lengths of the RNA-seq assembled transcripts , 4 . 8 kb for MTA6 ( mt VI MTA ) and 5 . 7 kb for MTB6 ( mt VI MTB ) ., These results , in combination with the RNA-seq results , support the hypothesis that all mating types have MTA and MTB genes consisting of two segments: one encoding a highly conserved TM segment found in all mating types and the other encoding a larger mating type-specific segment ., To identify the genes of the germline mat locus , we used the mt VI MTA6 and MTB6 gene pair sequence as query in a BLAST search of the SB210 germline genome sequence ( Tetrahymena Comparative Sequencing Project , Broad Institute of Harvard and MIT , http://www . broadinstitute . org/ ) ., Multiple matching discontiguous segments were observed over a 91-kb region of the germline ., The mating type-specific segments of MTA6 and MTB6 matched once in the middle of this region ., Additional matches were due to the conserved TM exons of MTA6 and MTB6 , each of which matched six times within this region ., This led us to identify five additional gene pairs containing sequences homologous to those of the TM exons of MTA6 on the left and MTB6 on the right ., The genes are arranged in a tandem array of six similarly oriented gene pairs , the number of mating types encoded by the mat-2 allele ( Figure 3 ) ., Sequence immediately flanking the mat locus is identical in the germline and somatic genomes ., Before carrying out detailed analysis of the mat locus , we filled all sequence gaps in this region and corrected sequence errors ( Tables S2 and S3 ) ., In the finished sequence we found that each gene pair consists of an MTA- and an MTB-like gene ., These are composed of a unique mating type-specific segment , and a terminal TM exon segment that is highly conserved among the MTA ( or MTB ) genes ., The germline mat locus lacks a complete gene pair ., The mat locus array begins and ends with the only complete genes within the array , later shown to be MTA2 and MTB3 , respectively ( Figure 3 ) ., The TM exons of all the other mating type genes are truncated , indicated by the use of lower case “tm” ( for example , MTA-tm or tm ) ., Assembly of a somatic mating type gene pair requires joining of mating type-specific segments to the full-length copies of the MTA2– and MTB3-TM exons located at the ends of the array ., A mating type was assigned to each germline gene pair segment by Southern blot analysis using probes from unique regions of each germline gene pair ., Each probe was found to be mating type-specific , hybridizing to a single band from the somatic nucleus of one mating type ( Figure 4 ) ., This result clearly shows that only one mating type gene pair remains in the somatic nucleus ., The order of the mating type gene pairs in the germline was identified as II – V – VI – IV – VII – III ( Figure 3 ) ., Using the above information , the somatic mat locus of each mating type was sequenced from mature mating type strains ( Tables S1 and S3 ) derived from a mating between strains SB210 mt VI and SB1969 mt II ., The entire germline mat locus from SB1969 mt II was sequenced and found to be identical to that of SB210 ( Table S3 ) ., In the mature mating type strains every somatic gene pair has full-length MTA– and MTB-TM exons joined to a mating type-specific segment , an arrangement identical to that of the somatic mt VI gene pair ., The TM exons of the other mating types revealed several single nucleotide polymorphisms when compared to the mt VI gene pair ( see below ) , but otherwise are identical ., The mating type genes represent two gene families ., Predicted proteins within the MTA family are of similar size ( 1423–1494 aa ) ., Clustal Omega alignment 24 , 25 of the six predicted MTA proteins reveals their TM exons share 99 . 6% amino acid identity ( Text S2A ) ., Mating type-specific regions were compared by means of all-by-all pairwise alignments of every MTA mating type-specific amino acid sequence using BLASTP ., On average , the alignments covered 98% ( range 92%–100% ) of the sequences , and showed 42% ( range 38%–47% ) sequence identity and 60% ( range 58%–65% ) sequence similarity ( identical and conservative substitutions ) ; expected values ranged from 1E-162 to less than 1E-200 ., Predicted proteins within the MTB family are also of similar size ( 1 , 733–1 , 749 aa ) ., Clustal Omega alignment of the six MTB proteins shows their TM exons share 99 . 4% amino acid identity ( Text S2B ) ., Analogous pairwise alignments of every MTB mating type-specific amino acid sequence on average covered 99% ( range 97%–100% ) of the sequences , and showed 43% ( range 41%–46% ) sequence identity and 62% ( range 60%–64% ) sequence similarity; expected values were all less than 1E-200 ., The two protein families were compared by all-by-all BLASTP alignments of MTA versus MTB predicted amino acid sequences; in every case , the only significant match ( expected value around 1E-08 ) was restricted to a ∼80 amino-acid cysteine-rich segment containing furin-like repeats , starting about 50 amino acids into the TM exon-encoded sequence ., Clustal Omega alignment of the furin-like repeats within the 12 TM exons is shown in Text S3 ., Cysteines at 12 positions and other amino acids at 14 positions are absolutely conserved among the furin-like repeats of the 12 TM exons ., The function of cysteine rich , furin-like repeat domains is not known , but they are found in some endoproteases and cell surface receptors 26 ., The mating type-specific segments of the germline gene pairs differ in size by up to 8 . 5 kb ., This variation is due to the presence of IESs , germline-specific sequences that interrupt a contiguous region of somatic-destined sequence , within the array ., By comparing somatic sequences to the germline genome sequence , we identified six IESs , ( Figure 3; Table S4 ) ., Each was confirmed by cloning and sequencing PCR products from the germline and somatic nuclei ( unpublished data ) ., The IESs lie within introns in mating type-specific segments or in an intergenic region; they range in size from 299 to 5 , 989 bp ., No other differences were found between the germline and somatic sequences in the mating type-specific segments ., Additional germline-limited sequence separates adjacent mating type gene pairs in the germline array ( Table S4 ) ., We identified homologs of the MTA and MTB genes in the somatic genome sequence of several additional species ( Figure 5 ) ., Somatic genome sequence is available for two Tetrahymena species that are within the same subgroup 27 as T . thermophila ( T . malaccensis and T . elliotti ) and two more distantly related species ( T . borealis and T . pyriformis ) ( T . malaccensis , T . elliotti , T . borealis at the Broad Institute website , T . pyriformis strain GL by W . Miao , unpublished data ) ., T . malaccensis and T . borealis have systems with six and seven mating types , respectively , and like T . thermophila , mating type determination is stochastic , without influence of the parental mating types 28 ., The mating type system of T . elliotti is unknown ., The same is true of T . pyriformis , where the GL strain is sole representative of this species ., This strain also lacks a germline nucleus and thus would be sterile if it could mate ., Nucleotide and protein BLASTN and TBLASTN searches using the sequence of the conserved TM exons led us to identify single-copy , head-to-head MTA and MTB homologs of approximately the same length for all four related species ( Text S4 ) ., The results of a phylogenetic analysis ( Figure, 5 ) and Clustal Omega alignment ( Text S5 ) showed the mating type of the sequenced strain of T . elliotti to be most closely related to T . thermophila mating type III ., Similarly , the mating type of the sequenced strain of T . malaccensis is most closely related to mt IV ., Alignments of the predicted amino acid sequences are shown in Text S5 ., For the remaining species , specific mating type relationships could not be recognized either because they carry a homolog of the mt I gene of the T . thermophila mat1 allele , which has not yet been sequenced , or the sequence divergence is too great ., Neither T . thermophila MTA nor MTB protein show similarity to any of the other ciliate mating type protein deposited in GenBank , a total of 19 distinct proteins from four Euplotes species and one Blepharisma japonicum protein , as determined by BLASTP with expected value threshold\u200a=\u200a10 ., The six MTA TM exon segments of the germline SB210 mat locus were aligned , delineating the position at which each germline tm segment is truncated ( Figure, 6 ) and revealing 59 polymorphic sites ( Table S5 ) ., The MTB TM exon segments were similarly examined and 52 polymorphic sites were found ( Table S6 ) ., With only one exception , none of the polymorphic nucleotides generate stop codons or reading frame shifts and most are unique to a particular gene pair ., Unique polymorphic nucleotides within the germline TM exon segments allow us to deduce the germline origin of somatic MTA-TM and MTB-TM exon DNA ., During differentiation of a new somatic nucleus a pair of intact MTA and MTB genes must be assembled from the germline genes ., One possibility is that joining occurs between the ends of the mating type-specific segment and the start of the MTA2 and MTB3 full-length TM exons ., If this were the case , all the progeny would have full-length TM exons identical to those of the MTA2 and MTB3 germline genes ., Alternatively , joining could occur at internal locations within the germline TM exons ., In this case , the somatic TM exons would contain novel combinations of the unique polymorphic nucleotides found in the germline tm segments ., Somatic mating type gene pair sequences from the mature strains mentioned above , and the SB210 shotgun macronuclear genome sequence , were found to contain novel combinations of these polymorphic nucleotides ( see below ) , suggesting that joining can occur within the germline TM exons ., To determine more precisely where the incomplete gene pairs are joined to full-length germline TM exons , we compared the sequences of TM exons from newly differentiated somatic nuclei to those of germline TM exon segments ., We sequenced individual somatic TM exons from progeny that had not yet undergone the first cell division ( exconjugants , Figure S1 stage 3 ) ., We constructed “collapsed alignments” to concisely represent all the polymorphisms in the somatic and germline nuclei ( Texts S6–S8 ) ., Schematic representations of the complete set of sequenced exons are shown in Figure 6 ., Somatic MTA2 and MTB3 genes , which are already complete in the germline , showed no evidence of any joining event ., The TM exons of every other somatic mating type gene showed polymorphic nucleotide combinations not present in the germline genome ( Figure 6; Texts S7 and S8 ) ., A single , simple joining event connecting a truncated germline tm segment to the full-length germline MTA2-TM exon explains 98% of the somatic MTA-TM exons present in early progeny ( Table S7A ) ., The MTA join sites were mapped to a 269-bp segment near the start of the MTA2-TM exon ., A single , simple event explains joining to the full-length germline MTB3-TM exon in 74% of the sequenced exons ( Table S7A ) ., This percentage varies from 42% for somatic MTB2 to 89% for somatic MTB6 ., The MTB join sites mapped to intervals distributed throughout the germline TM exon sequence ., The number of distinct join sites may have been exaggerated if PCR template switching 29 reshuffled nucleotide diversity in these sequenced TM exons ( Text S9 ) ., These data confirm that many if not all of the joining events occur within the TM exon rather than exclusively between the TM exon and the mating type-specific segment ., The frequency of novel nucleotides ( not present in the germline ) is less than one in 50 , 000 sequenced base pairs ( Figure 6 legend ) , showing that the joining events are highly precise ., Analysis of the TM exons of twenty 120-fission strains ( Table S7; Texts S10 and S11 ) shows that MTB-TM exons undergo additional recombination after the resumption of vegetative multiplication ., Highly significant differences between 0- and 120-fission cells are observed for the MTB-TM exons , whether one compares the number of haplotypes explained by a single joining event , or by recombination events involving more than two germline genes , or by gene conversions ( Table S7 ) ., PCR template switching is excluded as a spurious source of recombination in these results ( Text S9 ) ., We believe these events largely represent intragenic secondary recombination , distinct from the single , simple recombination events responsible for mating type determination ., Our findings suggest that mating type determination in T . thermophila involves a remarkable type of programmed genome rearrangement ., We have identified a pair of mating type genes that are arranged head-to-head ., Each mating type is characterized by a similarly organized pair of somatic genes and each gene of the pair encodes a TM domain shared by all mating types ., Starvation is required for mating and induces transcription of both genes ., Both genes are required for wt levels of pair formation and progeny production ., The germline genome contains an array of incomplete gene pairs , one for each mating type ., During development of the somatic nucleus in progeny cells , the germline array undergoes rearrangement to assemble one complete gene pair and delete all others in the somatic chromosome ., Thus , mating type determination occurs by deletion rather than by an epigenetic gene silencing mechanism ., These findings account for the irreversibility of mating type determination ., The mating type locus can be thought of as a multi-state developmental switch where the switch is stochastically and permanently set to one state in the somatic genome ., The removal of either or both genes caused a significant inhibition of pairing between cells of different mating types , suggesting the MTA and MTB genes are both fundamental for recognition of cells of a different mating type ( allorecognition ) ., This inhibition of pairing suggests that the gene products may be functioning cooperatively for allorecognition ., In addition to allorecognition , the gene products could be distinguishing self to prevent homotypic pairing ., If this were the case , homotypic pairing would be observed in the absence of one or both genes ., This does not appear to be a function of the MTA and MTB genes because pairing between starved cells of the same mating type was not observed in our knockouts ., At least two events are required to assemble a complete somatic mating type gene pair from the mat germline array ( see model shown in Figure 7 ) ., At the left end of the gene pair , the MTA-tm segment must be joined to the single copy , full-length MTA2-TM exon located at the far left end of the array ., At the right end of the same gene pair , the MTB-tm segment must join to the single copy , full-length MTB3-TM exon located at the far right end of the array ., The breakage and rejoining mechanism is highly precise ., Since both joining events occur within translated exons segments , without this precision mating competence could be lost ., Possible mechanisms include homologous recombination and precise nonhomologous end joining ., The mechanism will become clearer once we experimentally determine which of the observed recombination events are essential to mating type determination and which are unrelated to this process ., Regardless of the mechanism , an interesting question is how joining at opposite ends is coordinated to result in the assembly of a somatic gene pair ., A stochastically selected germline gene pair may be epigenetically marked , its two ends cut , and full length TM exons joined coordinately ., Alternatively , each end could be processed independently resulting in the deletion of one or more gene pairs from either end , until only one complete gene pair remains ., Additional knowledge of the mechanism will be needed to understand how mating type frequencies are influenced by environmental conditions , such as temperature and nutritional state 30 , 31 ., In addition to the single , simple recombination events associated with mating type determination , we have observed secondary recombination events in somatic TM exons , especially MTB TM exons ., These events are particularly frequent in the MTB TM exons of mature cell lines ( Table S7 ) ., As explained in Text S9 , artifacts of PCR template-switching are excluded in these results ., Since the majority of joined TM exons from 24-h exconjugants show no evidence of secondary recombination , these events are probably unrelated to mating type determination ., Presumably they chiefly reflect recombination between multiple somatic chromosome copies carrying independently differentiated TM exons prior to the purification brought about by assortment during vegetative multiplication ( Figure S1 ) ., A number of recombination events , most simply interpreted as gene conversions , have also been detected among MTB exon haplotypes ., We believe that these MTB gene conversions are also due to the secondary recombination described above and are unrelated to Tetrahymena mating type determination , in part because gene conversions are found in only a small minority of the sequenced TM exons in 24-h exconjugants ., In addition , gene conversion per se cannot result in the loss of intervening mating type gene pairs ., Gene conversion is responsible for mating type switching in yeast , but no intervening DNA is lost in yeast mating type switching 32 ., Programmed somatic DNA rearrangements are well known among the ciliates 33 , 34 ., In T . thermophila , approximately 6 , 000 IESs in the germline genome are excised during differentiation of a new somatic nucleus 35 ., The deletions that join TM exons to mating type-specific segments differ in several important ways ., IES excision is imprecise; precision is not required , as nearly all IES are found in intergenic regions or within introns 36 ., In contrast , the deletions involved in mating type determination are highly precise and occur within the coding segment of the TM exon ., Furthermore , IES excision is maternally controlled; only sequences absent from the parental somatic genome are targeted for elimination 37 , 38 ., Mating type , on the other hand , is stochastically inherited; determination of mating type in each progeny cell occurs autonomously during the differentiation of the new somatic nucleus ., Mating type-specific sequences absent from the parental somatic nucleus escape deletion by the IES excision mechanism and are retained in progeny somatic nuclei ., Finally , preliminary experiments ( unpublished data ) indicate that mating type determination occurs several hours after excision of IES within the mat locus ., All these considerations lead us to conclude that these two processes , which occur in the differentiating somatic nucleus , proceed by different mechanisms ., In mating type determination , DNA breakage and rejoining occurs physically independently and precisely at both ends of one gene pair ., This leads to the assembly of one complete gene pair and the excision of the other germline gene pairs from the somatic chromosome ., To our knowledge , this type of programmed genome rearrangement is novel , at least in ciliate molecular biology ., The modular organization of the T . thermophila germline mat locus ( Figure 3 ) in combination with rare unequal meiotic crossing-over between homologous germline TM/tm domains could facilitate rapid evolutionary change in the number of available mating types ., This hypothesis is consistent with the existence of two T . thermophila germline mat allele classes specifying different numbers of mating types ( five for mat-1 and six for mat-2 ) ., mat-1-like alleles carry mt I but are missing mts IV and VII ., mat-2-like alleles are the opposite , carrying mts IV and VII in adjoining gene pairs while missing mt I . Using somatic genome sequence data we assigned a mating type to the sequenced strains of two other Tetrahymena species by virtue of their similarities to T . thermophila mating types ., This suggests that a similar mating type system is conserved in multiple Tetrahymena species ., If so , the mechanism proposed above could also explain the finding that the number of mating types described in species of the genus Tetrahymena is dynamic , ranging from 3 to 9 ( reviewed in 39 ) ., Using the strong sequence conservation observed at the TM exons , it may be possible to isolate and sequence mating type genes from many species of the genus Tetrahymena to investigate the evolution of their mating type system ., T . thermophila is a model organism for eukaryotic biology 15 ., Future research of this mating type system should advance our knowledge in several areas of biology ., The biochemical functions of the MTA and MTB gene products are of interest for understanding the principle of self/non-self discrimination ., The study of genomic rearrangements employed for mating type determination can inform mechanisms of genome dynamics in other systems ., All of the T . thermophila strains used here have the inbred strain B genetic background 40 ., As such , they are mat2/mat2 homozygotes and can be of any one of mating types II–VII ., The somatic and germline genomes have been sequenced from strain SB210 41 ., A panel of mature strains of different mating types , F1s of a cross of SB210×SB1969 , was obtained by propagating F1 cells for ∼120 fissions , subcloning and determining their mating type , all using established methods 42 ., The germline mat locus alleles of SB210 and SB1969 are identical to the nucleotide , as determined by sequencing SB1969 ( Table S3 ) ., This identity should extend to the germline of all members of the F1 panel ., Strains are listed in Table S1 and are available through the National Tetrahymena Stock Center ( http://tetrahymena . vet . cornell . edu ) ., An RNA-seq-based whole transcriptome analysis was done using strains SB4217 ( mt V ) and SB4220 ( mt VI ) ; detailed information of RNA extraction , library construction , and deep RNA sequencing can be found in our previous work 18 ., For the work here , we compared three conditions: starved mating type VI cells , growing mating type VI cells , and starved mating type V cells ., To look for transcription differences between mating type V ( SB4217 ) and mating type VI ( SB4220 ) , RNA-seq reads were first mapped to the ∼300-kb region of SB210 ( mt VI ) that includes the mat locus ( Figure S2 ) using TopHat 43 ., Mapped reads were assembled as transcript fragments using Cufflinks 44; the command lines of this mapping-then-assembly pipeline were described previously 18 ., Gbrowse genome viewer ( http://gmod . org/wiki/GBrowse ) was setup to visually check the transcription differences using as input the TopHat mapping results and the Cufflinks assemblies ., Necessary data format transformations were performed using SAMTools 45 and ad hoc Perl scripts ., The ∼300-kb region was then manually examined in Gbrowse for significant mating type-specific transcription differences ., We de novo assembled a mating type V partial transcript using the Trinity transcriptome assembler ( 2011-08-20 release version ) 46 ., The command line used was: Trinity . pl –seqType fq –output output –left left . fq –right right . fq –run_butterfly –bflyHeapSpace 20000M ., To identify de novo assembled mating type V partial transcripts related to the mating type locus , similarity searches ( BLASTN ) were performed against the de novo transcript assemblies using as query the sequences of genes TTHERM_ 01087810 and TTHERM_ 01087820 ., All identified mating type V transcripts were aligned with mating type VI transcripts of the two genes to discern conserved and mating type-specific sequences ., Table S2 lists primers used to PCR amplify gaps in the SB210 germline sequence in the region of the mat locus , the mat locus from the SB1969 germline sequence , regions flanking IESs in the somatic genome , segments for knockout constructs , and the TM exons from the somatic nuclei of mature strains ., DNA was prepared as described 47 ., PCR products were amplified with Finnzymes Phusion High-Fidelity DNA Polymerase and cloned using the Zero Blunt TOPO PCR Cloning Kit for Sequencing ( Invitrogen ) ., Primers for sequencing were chosen using the SB210 genome sequence ., Sequencing by Sanger dideoxy sequencing was carried out at Eton Bioscience Inc ., RNA was extracted using the Qiagen RNeasy kit from 10-ml cultures at 2×105 cells/ml that had been starved for 3 h at 30°C in 10 mM Tris ( pH 7 . 4 ) ., 15 µg of RNA was loaded per lane on a 1% gel , subject to electrophoresis for ∼2 h at 120 V in formaldehyde buffer , and set up for downward transfer in denaturation buffer to a charged nylon membrane ., To prepare hybridization probe , 150 ng of PCR product was labeled with dATP32P by random primer labeling for 72 h at room temperature , followed by removal of unincorporated dATP32P nucleotides using the QIAquick nucleotide removal Kit ( Qiagen ) ., Pre-hybridization and hybridization with ULTRAhyb solution ( Ambion ) was at 45°C for 2 and ∼16 h , respectively ., Blo | Introduction, Results, Discussion, Materials and Methods | The unicellular eukaryote Tetrahymena thermophila has seven mating types ., Cells can mate only when they recognize cells of a different mating type as non-self ., As a ciliate , Tetrahymena separates its germline and soma into two nuclei ., During growth the somatic nucleus is responsible for all gene transcription while the germline nucleus remains silent ., During mating , a new somatic nucleus is differentiated from a germline nucleus and mating type is decided by a stochastic process ., We report here that the somatic mating type locus contains a pair of genes arranged head-to-head ., Each gene encodes a mating type-specific segment and a transmembrane domain that is shared by all mating types ., Somatic gene knockouts showed both genes are required for efficient non-self recognition and successful mating , as assessed by pair formation and progeny production ., The germline mating type locus consists of a tandem array of incomplete gene pairs representing each potential mating type ., During mating , a complete new gene pair is assembled at the somatic mating type locus; the incomplete genes of one gene pair are completed by joining to gene segments at each end of germline array ., All other germline gene pairs are deleted in the process ., These programmed DNA rearrangements make this a fascinating system of mating type determination . | Tetrahymena thermophila is a single-celled organism with seven sexes ., After two cells of different sexes mate , the progeny cells can be of any one of the seven sexes ., In this article we show how this sex decision is made ., Every cell has two genomes , each contained within a separate nucleus ., The germline genome is analogous to that in our ovaries or testes , containing all the genetic information for the sexual progeny; the somatic or working genome controls the operation of the cell ( including its sex ) ., We show that the germline genome contains a tandem array of similarly organized but incomplete gene pairs , one for each sex ., Sex is chosen after fertilization when a new somatic genome is generated by rearrangement of a copy of the germline genome ., One complete sex gene pair is assembled when the cell joins DNA segments at opposite ends of the array to each end of one incomplete gene pair; this gene pair is thus completed and becomes fully functional , while the remaining sex gene pairs are excised and lost ., The process involves programmed , site-specific genome rearrangements , and the physically independent rearrangements that occur at opposite ends of the selected gene pair happen with high reliability and precision . | developmental biology, model organisms, molecular cell biology, genetics, biology, genomics, evolutionary biology, microbiology, genetics and genomics | In Tetrahymena, a multi-sexed single-celled organism, the sex of the progeny is randomly determined by site-specific recombination events that assemble one complete gene pair and delete all others. |
journal.pgen.1006443 | 2,016 | Loss of C9orf72 Enhances Autophagic Activity via Deregulated mTOR and TFEB Signaling | Amyotrophic lateral sclerosis ( ALS ) is a fatal neurodegenerative disease characterized by the progressive degeneration of motor neurons ., Frontotemporal dementia ( FTD ) is the second most common type of dementia in people younger than 65 and is characterized by degeneration of the frontal and temporal lobes of the brain ., A hexanucleotide repeat expansion ( HRE ) , ( GGGGCC ) n , in the promoter or intron of the uncharacterized gene , chromosome 9 open reading frame 72 ( C9orf72 ) , has been found to be the most common cause of both ALS and FTD 1 , 2 and has been linked to a number of other neurological disorders ., How the C9orf72 HRE leads to neurodegeneration remains to be determined , although both gain-of-toxicity and loss-of-function mechanisms have been proposed ., The gain-of-toxicity mechanisms involve both RNA and protein products generated from the expanded hexanucleotide repeats ., For example , RNAs containing the expanded repeats can interfere with the functions of specific RNA-binding proteins 3–5 , and toxic repeat polypeptides can be generated through repeat-associated non-ATG-dependent translation 6–10 ., However , the HRE could be pathogenic through loss-of-function mechanisms when the expression of the C9orf72 gene is disrupted ., Multiple studies have demonstrated that C9orf72 RNA and protein levels are reduced in patient cells and brains 11–15 ., Although partial knockdown of C9orf72 in the brain or its neural-specific deletion does not affect survival in mice 16 , 17 , loss of C9orf72 orthologs in zebrafish and C . elegans has deleterious effects 18 , 19 ., Studies of these loss-of-function mechanisms are hampered by a lack of knowledge about the physiological function of the C9orf72 protein ., Bioinformatic analysis suggested that C9orf72 is a DENN-like protein 20 , 21 , which is a family of proteins that regulate small GTPases and membrane trafficking ., DENN domain-containing proteins have also been implicated in autophagy and in the mammalian target of rapamycin ( mTOR ) signaling pathways 22 ., Although a recent study has reported that C9orf72 regulates autophagy and endosomal trafficking 23 , the function of the C9orf72 protein remains largely unknown ., Here we report the findings in mice and human cells that loss of C9orf72 inhibits mTOR signaling and leads to a profound upregulation of transcription factor EB ( TFEB ) and enhanced autophagy flux ., We further show that C9orf72 interacts with another DENN-like protein Smith-Magenis syndrome chromosome region candidate 8 ( SMCR8 ) , which also regulates mTOR signaling and autophagy ., The results suggest that a deficiency in the function of C9orf72 may contribute to the pathogenesis of relevant neurodegenerative diseases ., To study the physiological functions of C9orf72 in mammals , we generated a knockout ( KO ) mouse model lacking the protein ., Human C9orf72 has one orthologous gene in the mouse , 3110043O21Rik , which is located on chromosome 4 ., For convenience , we refer to the mouse gene as C9orf72 hereafter ., The mouse C9orf72 gene is predicted to produce seven transcripts , three of which are protein-coding , as compared to the human C9orf72 gene , which produces three transcripts and two protein isoforms ., The mouse C9orf72 proteins share 98% identity with their human C9orf72 counterparts ( S1 Fig ) ., We generated C9orf72 KO mice by using a mouse embryonic stem ( ES ) cell line that contains a heterozygous allele of a 7754 base pair deletion in the C9orf72 gene ., This deletion results in the removal of exons 2–6 and is predicted to produce nonfunctional truncated protein products from all three protein-coding transcripts of the mouse C9orf72 gene ( Fig 1A ) ., We further removed the neomycin cassette by crossing the C9orf72 KO male mice carrying the original targeted allele with SOX2-Cre transgenic females ( Fig 1A ) ., Western blotting of brain homogenates from C9orf72 wild-type and KO littermates , using an antibody predicted to detect all mouse C9orf72 isoforms , showed a protein band at 55 kDa ( corresponding to mouse isoform 1 ) , not present in the C9orf72-/- samples ( Fig 1B ) , confirming that our KO mice lack C9orf72 in brain ., We were unable to detect the other two mouse C9orf72 isoforms , suggesting that mouse isoform 1 is the major isoform in the mouse brain ., The homozygous C9orf72 KO mice showed a decrease in survival compared with littermates , with more than 50% dead in 600 days ( Fig 1C ) ., This decrease in survival was also observed in heterozygous C9orf72+/- animals to a lesser degree with only about 20% dead in 600 days ., Both C9orf72 homozygous and heterozygous knockout mice developed normally before exhibiting rapidly progressive lethargy before death ., The stage of lethargy could last for days up to a month ., At the end stage , the animals showed a lack of excitability or response to external stimuli ( S1 Movie ) ., In post-mortem examination , consistent with recent reports of immune dysregulation in C9orf72 knockout mice 24–27 , we observed splenomegaly in the C9orf72-/- mice ., The spleen was generally increased in length from ~3/4 inches to 1–1 . 25 inches ., In addition , we frequently observed potential tumors in the thymus or in the regions of the abdomen ., There was no obvious neuronal cell death in brain or spinal cord , but functional deficits of the nervous system could not be excluded ., The exact cause of death for these C9orf72 knockout mice remains to be determined ., Although we observed no obvious neuronal defects in C9orf72 KO mice , it is possible that C9orf72 has functions in the nervous system in response to stresses ., Thus , we asked if mTOR signaling is altered when C9orf72 is absent , since mTOR signaling is a central signaling pathway that senses the stresses related to nutrient availability , oxygen , and energy levels 28 ., Also , DENN-like proteins have been implicated in mTOR signaling and nutrient sensing 29–31 and C9orf72 contains DENN domains ., We monitored mTOR activity by assessing the phosphorylation of its downstream target ribosomal protein S6 kinase B1 ( S6K1 ) ., Cells were starved for amino acids for 50 minutes before amino acids were added back to induce the phosphorylation of S6K1 ., Interestingly , knockdown of C9orf72 in HEK293T cells resulted in a decrease in the phosphorylation of S6K1 within 10 to 20 minutes after addition of amino acids , as compared with control cells transfected with scrambled control shRNAs ( Fig 2A and 2B ) ., These results suggest that the loss of C9orf72 decreases mTOR activation after amino acid stimulation ., To study the molecular defect in the complete absence of C9orf72 protein , we generated mouse embryonic fibroblasts ( MEFs ) from C9orf72 wild-type and KO littermates ., And we assessed the phosphorylation of S6K1 in the C9orf72-/- MEF lines ., Phosphorylation of S6K1 was decreased in C9orf72-/- MEF lines compared with lines derived from wild-type littermates ( Fig 2C ) , suggesting that mTOR activation after amino acid stimulation is diminished in the absence of C9orf72 ., Subsequently , we asked whether the observed reduction of mTOR activation in the absence of C9orf72 impacts the function of TFEB , a transcription factor that is a master regulator of lysosome biogenesis and autophagy-related genes , and a substrate of mTOR 32 ., In an autoregulatory loop , nuclear translocation of TFEB leads to increased expression of itself ., Phosphorylation of TFEB by mTOR prevents its translocation to the nucleus and causes down-regulation of TFEB ., We transfected GFP-TFEB into HEK293T cells and observed that knockdown of C9orf72 resulted in a significant increase in GFP-TFEB levels ( Fig 3A and 3B ) , consistent with the decrease in mTOR activity ., Moreover , imaging analysis indicated that nuclear localization of GFP-TFEB was significantly increased upon knockdown of C9orf72 as compared with cells treated with control shRNAs ( Fig 3C and 3D ) ., Western blotting of the nuclear and cytoplasmic fractions further confirmed that GFP-TFEB was enriched in the nucleus upon knockdown of C9orf72 ( Fig 3E ) ., Next , we validated these results in C9orf72-/- and wild-type MEF cells ., Consistently , the complete absence of C9orf72 led to a significant increase in the nucleus to cytoplasm ratio of GFP-TFEB signals ( Fig 3H and 3F ) ., Furthermore , consistent with the notion that TFEB promotes the biogenesis and activity of lysosomes 32 , we observed a significant increase in the number of LysoTracker-stained acidic vesicles in the C9orf72-/- MEF cells , confirming functional consequences on lysosomes of enhanced nuclear TFEB ( Fig 3H and 3G ) ., We then questioned whether our results held true in vivo ., Analysis of brain homogenates by western blotting from all examined C9orf72 KO mice showed a dramatic increase in endogenous TFEB levels compared with wild-type controls ( Fig 3I ) , consistent with our results in tissue culture ., We next asked whether downstream targets of TFEB were also increased by loss of C9orf72 ., Indeed , western blot analysis of lysosome-associated membrane glycoprotein 1 ( LAMP1 ) , which is a transcriptional target of TFEB 33 , indicated that LAMP1 was profoundly increased in the C9orf72 KO mouse brains ( Fig 3I ) ., Another related lysosomal protein LAMP2 was also markedly increased in the absence of C9orf72 in the KO mouse brains ., Taken together , these results suggest that , consistent with the inhibition of mTOR signaling , loss of C9orf72 increases TFEB activity ., Since we observed a function of C9orf72 in mTOR signaling and mTOR is known to negatively regulate autophagy , we assessed the levels of the autophagy marker LC3 by immunoblotting in these cells ., During autophagy , LC3I is processed to LC3II via lipidation , which allows for insertion of the LC3 protein into the autophagosome membrane ., Our results show a significant increase of LC3I in C9orf72-/- MEFs when compared with wild-type MEFs , indicating that basal autophagy is altered in these cells ( S2A and S2B Fig ) ., To test the role of C9orf72 in neurally differentiated cells , we generated embryonic stem cells from C9orf72 KO mice and littermate controls and differentiated them into the motor neuron precursors that further grew into mature motor neurons ( ~40% of the culture ) plus astrocytes and oligodendrocytes ., We assessed the level of LC3 by immunoblotting and found that the C9orf72-/- cells enriched with motor neurons showed a substantial accumulation in LC3I ( S2C Fig ) , in line with what was observed in C9orf72-/- MEFs ., The decreases in LC3II/LC3I ratio observed in our western blots can indicate a defect in lipidation or an increase in degradation via the lysosome ., To distinguish between these two possibilities , we assessed LC3 levels after nutrient deprivation-induced autophagy in the absence and presence of the lysosomal inhibitor Bafilomycin in C9orf72-/- and wild-type MEF cells ., We found that the Bafilomycin-induced accumulation of LC3II was significantly enhanced in C9orf72-/- MEFs compared with wild-type MEFs ( Fig 4A and 4B ) , indicative of an enhanced autophagic flux in C9orf72-depleted cells ., To further examine the status of autophagic flux , we analyzed the numbers of LC3-positive autophagic vesicles and the colocalization between LC3 vesicles and Rab7 , a late endosome-/lysosome-associated GTPase that marks mature autophagolysosomes 34–36 ( Fig 4C–4F ) ., Quantification of LC3-positive vesicles in the absence and presence of Bafilomycin , demonstrates that , consistent with western blot results , the number of LC3-positive autophagic vesicles was significantly increased in C9orf72-/- MEFs ( Fig 4D ) ., An autophagic flux index , defined as the difference in the volumes of LC3-positive vesicles before and after Bafilomycin treatment , was quantified , further confirming the increased autophagic flux capacity in C9orf72-/- MEFs ( Fig 4E ) ., Similarly , we observed an increase in the colocalization of LC3-positive autophagic vesicles with Rab7-positive vesicles ( Fig 4F ) , confirming enhanced autophagolysosome formation ., In addition to induced autophagy , we also examined basal autophagic flux under fully supplemented nutrient conditions in the absence of C9orf72 ( S3 Fig ) ., Despite of relatively low level of signals , the LC3/Rab7 vesicle colocalization assay indicated a trend that there were more LC3-positive vesicles and more colocalized LC3/Rab7 vesicles in Bafilomycin-treated C9orf72-/- MEFs than in wild-type control cells ( S3A Fig ) ., This result is consistent with the western analysis of LC3 protein levels , in which LC3II accumulated robustly in Bafilomycin-treated C9orf72-/- MEFs ( S3B Fig ) ., In addition to MEF cells , we observed a similar result in HEK293T cells for autophagic flux after knockdown of C9orf72 ., Under nutrient deprivation , in cells treated with C9orf72 shRNA , despite a decrease in LC3II/LC3I ratio before Bafilomycin treatment , lysosomal inhibition induced a robust accumulation of LC3II ( S3C Fig ) , suggesting that the total autophagic flux was increased ., Taken together , the observed increases in autophagic activity are consistent with the impairment of mTOR signaling and the profound increase of TFEB as results of loss of C9orf72 ., Of note , our results do not rule out the possibility that C9orf72 functions in other aspects of autophagy ., For example , we examined the activity of ATG4B , which catalyzes the cleavage of proLC3 to produce LC3I and also removes LC3II from the autophagosome membrane after it fuses with the lysosome ( S4A Fig ) ., Knockdown of C9orf72 in HEK293T cells resulted in a significant decrease in the signal of an ATG4B activity luciferase reporter when compared with control cells ( S4B Fig ) , suggesting that ATG4B activity is impaired under basal conditions ., However , no change was detected in ATG4B protein levels by western blotting upon knockdown of C9orf72 ( S4C Fig and S4D Fig ) , indicating that the reduction in ATG4B activity was not due to a decrease in its protein level ., The unchanged level of ATG4B protein could presumably make it readily available to support the enhanced autophagic flux observed in nutrient deprivation-treated C9orf72 deficient cells ., We next investigated whether the absence of C9orf72 alters the markers of autophagy in vivo ., Since mTOR signaling senses nutrient stresses and autophagy induction is a natural response to nutrient stresses through mTOR , we asked whether the absence of C9orf72 affects the autophagic response under these stress conditions ., We applied amino acid withdrawal by feeding mice a low-protein diet that is well-tolerated in young animals 37 ., Beginning at 4 months of age , gender-matched wild-type and C9orf72 KO littermates were fed either normal or amino acid-deficient chow for four weeks before tissues were harvested for analysis ( Fig 5A ) ., We first examined autophagy in the brain of C9orf72 KO mice ., Because of the low levels of LC3 conversion during starvation in the brain 38 , we assessed the levels of the autophagy marker protein p62 ., Western blotting of brain homogenates showed a slight decrease in p62 levels in C9orf72 KO mice when compared with wild-type littermates , a defect that became more pronounced when the mice were on the low-protein diet ( Fig 5C and 5D ) ., The decrease of p62 was not due to change in its solubility since no insoluble p62 was detected in western analysis or histological examinations ( Fig 5B ) ., The lack of accumulation of p62 in the brains of C9orf72 KO mice suggests an increased autophagy activity ., Consistent with the results in the mouse brain , we also observed a decrease in p62 levels in C9orf72-/- MEF lines compared with wild-type cells ( S5A Fig and S5B Fig ) ., We next examined the liver , a common tissue type used to study autophagy , harvested from the wild-type and C9orf72 KO littermates , for changes in LC3 ., We observed a decrease in the level of LC3II protein or relative increase of LC3I protein in C9orf72-/- livers relative to the wild-type controls under the low protein diet condition ( S5C Fig and S5D Fig ) ., To gain molecular insight into the function of C9orf72 , we performed a quantitative proteomic screen for protein interactors of the C9orf72 protein using stable isotope labeling by amino acids in cell culture ( SILAC ) mass spectrometry ( Fig 6A ) ., Human C9orf72 Isoform A with a C-terminal Flag tag was expressed in HEK293T cells metabolically labeled with 13C , 15N L-Arginine and L-Lysine and immunoprecipitated using Flag-tag beads ., A parallel immunoprecipitation was performed using unlabeled mock-transfected cells as a control to identify proteins that bound to the Flag-tagged beads alone ., The resulting immunoprecipitates were pooled and analyzed via mass spectrometry to identify proteins that were enriched by the C9orf72 bait ., We identified SMCR8 as the top C9orf72 interactor since it had the highest SILAC ratio or enrichment ( S6A and S6B Fig and S1 Table ) ., Notably , SMCR8 , although uncharacterized , is also a DENN-like protein 20 , 21 ., We validated this interaction by co-immunoprecipitation , with Flag-tagged C9orf72 pulling down endogenous SMCR8 in HEK293T cells ( Fig 6B ) ., Conversely , reciprocal immunoprecipitation experiments demonstrated that an anti-SMCR8 antibody pulled down Flag-tagged C9orf72 , confirming their interaction ( Fig 6C ) ., The interaction was further validated by co-immunoprecipitation of co-expressed Flag-SMCR8 and C9orf72-V5 proteins ( S6C Fig ) ., Consistently , GFP-tagged C9orf72 and mCherry-tagged SMCR8 both localized to the nucleus and the cytoplasm in HEK293T cells ( S6D Fig ) ., Since we identified SMCR8 as the most abundant protein interactor of C9orf72 , we asked whether C9orf72 influences the level of SMCR8 protein ., While examining the brain lysates from the C9orf72 KO mice , we observed a dramatic reduction in the level of SMCR8 protein ., Although present in wild-type brains , SMCR8 was not detected in C9orf72-/- brain homogenates by western blotting ( Fig 6D ) ., Examination of SMCR8 transcripts by qPCR showed no reduction in its mRNA levels , supporting that C9orf72 influences SMCR8 at the protein level ( S7A Fig ) ., Notably , WD repeat-containing protein 41 ( WDR41 ) , another protein identified in our proteomic screen ( S1 Table ) and recently confirmed to be an interactor of the C9orf72/SMCR8 complex 28 , 29 , was not decreased in C9orf72-/- brain samples ( S7C Fig ) ., In addition , overexpression of C9orf72 in HEK293T cells increases SMCR8 , suggesting that C9orf72 regulates SMCR8 protein levels ( Fig 6F and 6H ) ., To further study the function of SMCR8 , we obtained a CRISPR/Cas-9 generated SMCR8 KO cell line ., This cell line contains a frameshift mutation in the first exon of SMCR8 resulting in the loss of the full-length protein product ( S8A–S8C Fig ) ., Since we observed that C9orf72 regulates SMCR8 protein levels , we asked whether SMCR8 reciprocally influences the levels of C9orf72 ., By examining the lysates from the SMCR8 KO cells by western blotting , we observed a dramatic reduction in the level of C9orf72 protein ( Fig 6E ) ., We observed the same effect on the C9orf72 protein when we treated HEK293T cells with validated SMCR8 shRNA compared with control cells transfected with a scrambled shRNA control ( S8D and S8E Fig ) ., Examination of C9orf72 transcripts in SMCR8 KO cells by qPCR showed an increase in its mRNA levels ( S7B Fig ) , suggesting that the loss of SMCR8 decreased the C9orf72 protein level not by reducing its RNAs ., Next we studied how SMCR8 regulates C9ORF72 protein levels ., We first asked whether the regulation occurs due to changes in protein stability or turnover ., Since the level of C9ORF72 was too low to allow for chase experiments to probe their turnover in the SMCR8 KO cells , we overexpressed C-terminal-V5 tagged C9orf72 and N-terminal-mCherry tagged SMCR8 , or an mCherry only control , into HEK293T cells and studied their protein levels ., Compared with the mCherry control , the expression of mCherry-SMCR8 substantially increased the level of C9orf72-V5 ( Fig 6G ) ., Importantly , under a 12-hr chase condition after treatment of the cells with the translation inhibitor cycloheximide , mCherry-SMCR8 dramatically stabilized the co-expressed C9orf72-V5 as compared with the mCherry control ( Fig 6G ) ., We also confirmed that the C9orf72-V5 protein was degraded through both proteosomal and lysosomal pathways , since inhibition of proteasomal degradation by MG132 treatment or inhibition of lysosomal degradation by Bafilomycin treatment stabilized the C9orf72-V5 protein ( S7E Fig ) ., These data indicate that C9orf72 and SMCR8 form a stable cognate protein complex that protects C9orf72 from degradation ., Given the connections between C9orf72 and SMCR8 , we asked whether loss of SMCR8 plays a role in mTOR signaling similar to that of C9orf72 ., In accordance with the results from C9orf72-/- MEF cells , knockout of SMCR8 led to a similar defect ., In SMCR8 KO HAP1 cells , the phosphorylation of S6K1 after amino acid treatment was significantly decreased when compared with control cells ( Fig 7A and 7B ) ., Next , we investigated if loss of SMCR8 also affected autophagy ., First , we examined LC3 levels after shRNA-mediated knockdown of SMCR8 in HEK293T cells by immunoblotting ., As observed in C9orf72-/- MEFs and C9orf72 shRNA treated HEK293T cells , knockdown of SMCR8 led to a decrease in the ratio of LC3II to LC3I , when compared with cells treated with scrambled shRNA ( Fig 7C and 7D ) ., Additionally , Bafilomycin treatment of the cells under starvation showed a similar accumulation of LC3II with the SMCR8 knockdown as that of the control cells ( Fig 7E ) ., Thus , the autophagic flux appears to be intact in the absence of SMCR8 in this cell line ., In the present study , we have identified a function of C9orf72 in regulating mTOR signaling and autophagy ., Loss of C9orf72 leads to deficiency in the phosphorylation of S6K1 and increase of TFEB protein levels and nuclear activity , demonstrating a regulatory role of C9orf72 in the mTOR signaling pathway upstream of autophagy ., We identified the major interacting partner of C9orf72 protein as SMCR8 ., The most structurally homologous proteins to SMCR8 and C9orf72 in the human proteome are folliculin ( FLCN ) and folliculin-interacting proteins ( FNIP1 or 2 ) , respectively 20 , 21 ., Like SMCR8 and C9orf72 , FNIP and FLCN are DENN domain-containing proteins 20 , 21 that interact with each other in a protein complex 29 , that have also been shown to regulate autophagy and mTOR signaling 30 , 31 ., Since the FNIP and FLCN complex was shown to function as either GAP or GEF for the Rag GTPases in the mTORC1 pathway , we speculate that the C9orf72-SMCR8 complex may function in a similar fashion in autophagy and mTOR signaling ., Our results demonstrate that loss of C9orf72 can alter the dynamics of autophagy ., We observed a relative increase in LC3I levels upon loss of C9orf72 ( S2 Fig ) , in consistence with a recent report for LC3 levels in C9orf72 KO mouse liver and spleen tissues 39 , which we interpret as an increase in autophagosome turnover instead of a decrease in LC3II formation ., In support of this model , we did not observe a decrease in LC3II levels after Bafilomycin treatment under full nutrient conditions , suggesting that the formation of LC3II is intact ( S3 Fig ) ., Moreover , we observed increased autophagic flux in response to nutrient deprivation in C9orf72-/- cells ( Fig 4 ) ., Consistent with our model of increased autophagic flux , we observed a loss of mTOR activity after loss of C9orf72 , which is classically associated with increases in the autophagic pathway ., In support of our finding , a recent study showed decreased mTOR signaling in C9orf72-depleted HeLa cells 40 ., Importantly , we observed a substantial increase of TFEB and its lysosomal targets in C9orf72 knockout mice ( Fig 3 ) ., As a master regulator of lysosome biogenesis , TFEB is known to promote cellular lysosomal capacity and autophagy 32 ., Consistent with our findings , we also observed a decrease in levels of the autophagy receptor p62 in brain tissues from C9orf72 KO mice and observed a similar decrease in the C9orf72 KO MEFs ., Interestingly , it was recently reported that loss of the SMCR8 homologue folliculin similarly results in decreased mTOR signaling and a TFEB-mediated enhancement of the lysosomal compartment 31 ., There have been recent reports describing C9orf72’s functions in autophagy 39 , 41–43 , including a decrease in autophagy initiation as a result of knockdown of C9orf72 41 , 42 ., These observations are not necessarily mutually exclusive to our present study ., C9orf72 might play a multifunctional role in different steps of the autophagic pathways ., While C9orf72 may influence the function of the FIP200/ULK1 autophagy initiation complex 41 , 42 , it could also regulate mTOR signaling and TFEB and thus promote autophagic flux , as observed in the present study ., Furthermore , the manifestation of the phenotypes could be influenced by the dynamic nature and condition-dependent activity levels of autophagy pathways ., Due to the reduced state of mTOR signaling in C9orf72-depleted cells , the increased autophagic flux of these cells could be more readily revealed under nutrient deprivation conditions , as employed in the present study ., Notably , the autophagy receptor p62 is both a substrate of autophagy and a transcriptional target of TFEB 44 , therefore it is subject to opposing regulation by upregulation of TFEB ., Taken together , our study provides evidence that long-term loss of C9orf72 leads to physiological changes that are characterized by reduced mTOR activity , in consistence with increased TFEB signaling leading to enhanced cellular lysosomal capacity and autophagic flux ., Since multiple studies have reported that the hexanucleotide repeat expansion led to reduced expression of C9orf72 mRNAs and proteins in patient cells and brains 11–15 , the defects associated with loss of C9orf72 protein function could contribute to the pathogenesis of relevant neurodegenerative diseases ., Several studies have reported that neither mice lacking C9orf72 protein nor those expressing the human C9orf72 gene containing the HRE mutation exhibited major neuronal loss 17 , 45–47 , with the exception of one study reporting neurodegeneration in transgenic mice expressing HRE-containing C9orf72 48 ., Our observation that C9orf72 ablation changes LC3 levels in motor neuron cultures suggests that loss of C9orf72 might affect neuronal functions ., Autophagy and nutrient sensing are essential for neuronal health and their alteration is an increasingly recognized feature in aging-related neurodegenerative diseases 49 , 50 ., Of note , several autophagy-related genes , including p62 , optineurin , and TBK1 , have been linked to ALS 51–53 ., Proteinaceous inclusions positive for p62 are a pathologic feature in brains from patients carrying the C9orf72 HRE mutation 54 ., Taken together , our findings suggest that C9orf72 protein has a function in the metabolic processes of the cell and reduction in its function may contribute to related age-dependent neurodegenerative diseases ., The animal protocol ( MO15M165 ) was approved by the Johns Hopkins Animal Care and Use Committee following the National Research Council’s guide to the Care and Use of Laboratory Animals ., C9orf72 cDNA ( HsCD00398737 ) was obtained from Arizona State University and SMCR8 cDNA ( HsCD00347993 ) from Harvard Plasmid Repositories ., The C9orf72 constructs were generated using the Gateway cloning system ( ThermoFisher , Waltham , MA ) with a C-terminal 3xFlag or V5 tag ., The SMCR8 constructs were generated with an N-terminal Flag or mCherry tag using Gateway or classical cloning methods , respectively ., All shRNAs were cloned into the pRFP-C-RS vector ( Origene ) , which was modified to remove the RFP coding sequence via digestion with MluI and BglII followed by blunting and religation ., The following shRNA sequences were used: 5’ctgtgttacctcctgaccagtcagattga 3’ ( SMCR8 ) ; 5’cttccacagacagaacttagtttctacct 3’ ( C9orf72 ) ., The autophagy luciferase assay plasmids were kindly provided by Brian Seed ( Harvard ) and the normalization plasmid pCMV-SEAP was from Addgene ( 24595 , Alan Cochrane , University of Toronto ) ., GFP-TFEB was obtained from Addgene ( 38119 , Shawn Ferguson , Yale University ) ., GFP-TFEB used for MEF experiments was described before 55 ., For GFP-LC3 , human LC3 was cloned into pEGFP-C1 ., RFP-Rab7 was generated from EGFP-Rab7 ( a kind gift from Bo van Deurs at University of Copenhagen ) by exchanging EGFP into RFP ., Mouse ES cell lines containing a heterozygous allele of 3110043O21Riktm1 . 1 ( KOMP ) Mbp were obtained from the KOMP repository ., The ES cells with a strain background of C57BL/6N-Atm1Brd were microinjected into blastocysts , and the germline-transmitted allele was maintained on the C57BL/6 background ., Male mice bearing the original targeting allele were crossed with SOX2-Cre recombinase transgenic female mice ( Jackson Laboratory , 008454 ) to remove the LoxP-flanked neomycin selection cassette ., The resulting allele was bred to heterozygotes and homozygotes that were used in this study ., The genotyping primers were the following: gaatggagatcggagcacttatgg ( wild-type , forward ) , gccttagtaactaagcttgctgccc ( wild-type , reverse ) , gcacaagctatgttcatttgg ( KO , forward ) , gactaacagaagaacccgttgtg ( KO , reverse ) ., For the low-protein diet assay , 16 week old , gender-matched littermates were fed a low-protein diet ( Test Diet 5767 , 5% protein ) or standard chow for 4 weeks prior to tissue collection ., Mouse tissue was lysed in modified RIPA buffer ( 50 mM Tris pH 6 . 8 , 150 mM NaCl , 0 . 5% SDS , 0 . 5% Sarkosyl , 0 . 5% NP40 , 20 mM EDTA , Roche protease inhibitors ) using a Dounce homogenizer , sonicated , and used for further analysis ., For the survival analysis , Kaplan-Meyer curves were generated using GraphPad Prism software ., All cells were maintained in DMEM supplemented with 10% FBS unless otherwise noted ., The SMCR8 knockout HAP1 cells ( HZGHC003606c011 ) were created at Horizon Genomics ( Vienna , Austria ) by using CRISPR/Cas9 and maintained in IMDM supplemented with 10% FBS ., All cell lines were cultured in 95% O2/5% CO2 ., Cell lines were transfected using Lipofectamine 2000 ( ThermoFisher ) according to the manufacturer’s instructions ., Mouse embryonic fibroblasts were isolated from Day 13 embryos by trypsin digestion and their genotypes confirmed by PCR ., The lines were immortalized by transfecting cells with the SV40-T antigen-expressing plasmid pSG5 Large T using Lipofectamine 2000 ., The cells were passaged at least 5x to ensure the homogeneity of the cell population before use in experiments ., To isolate embryonic stem cells , 14-week old C9orf72 heterozygous females were treated with Pregnant Mare Serum Gonadotropin via intraperitoneal injection followed by injection 24 hours later with human chorionic gonadotrophin to induce superovulation prior to mating with C9orf72 heterozygous males ., Embryos were collected 48 hours after the second injection at the transgenic core facility at Johns Hopkins University and the genotypes confirmed by PCR ., Wild type and C9ORF72-/- ES cells were cultured on 0 . 1% gelatin coated plates in 2i media consisting of half of DMEM/F12 and half of Neurobasal media containing N2-supplement ( ThermoFisher Scientific 17502048 ) , B-27 supplement ( ThermoFisher Scientific 17504044 ) , 0 . 05% BSA ( ThermoFisher Scientific 15260037 ) , 50 units Penicillin-Streptomycin , 1 μM PD03259010 ( Stemgent 04–0006 ) , 3 μM CHIR99021 ( Stemgent 04–0004 ) , 2 mM Glutamine , 150 μM Monothioglycerol ( Sigma M6145 ) and 1 , 000 U/ml LIF ., Motorneuron differentiation protocol was modified from a previously reported induction protocol using retinoic acid and Smoothened agonist ( SAG , Millipore ) 56 ., Briefly , 1 X 106 ES cells were harvested by dis | Introduction, Results, Discussion, Materials and Methods | The most common cause of the neurodegenerative diseases amyotrophic lateral sclerosis and frontotemporal dementia is a hexanucleotide repeat expansion in C9orf72 ., Here we report a study of the C9orf72 protein by examining the consequences of loss of C9orf72 functions ., Deletion of one or both alleles of the C9orf72 gene in mice causes age-dependent lethality phenotypes ., We demonstrate that C9orf72 regulates nutrient sensing as the loss of C9orf72 decreases phosphorylation of the mTOR substrate S6K1 ., The transcription factor EB ( TFEB ) , a master regulator of lysosomal and autophagy genes , which is negatively regulated by mTOR , is substantially up-regulated in C9orf72 loss-of-function animal and cellular models ., Consistent with reduced mTOR activity and increased TFEB levels , loss of C9orf72 enhances autophagic flux , suggesting that C9orf72 is a negative regulator of autophagy ., We identified a protein complex consisting of C9orf72 and SMCR8 , both of which are homologous to DENN-like proteins ., The depletion of C9orf72 or SMCR8 leads to significant down-regulation of each other’s protein level ., Loss of SMCR8 alters mTOR signaling and autophagy ., These results demonstrate that the C9orf72-SMCR8 protein complex functions in the regulation of metabolism and provide evidence that loss of C9orf72 function may contribute to the pathogenesis of relevant diseases . | C9orf72 is one of many uncharacterized genes in the human genome ., The presence of repeated nucleotides in the non-coding region of the C9orf72 gene ( GGGGCC ) has been linked to the neurodegenerative diseases Amyotrophic Lateral Sclerosis ( ALS ) and Frontotemporal dementia ( FTD ) ., However , how the presence of these repeats in the gene leads to neurodegeneration is unknown ., One possible explanation is that the repeats lead to a reduced expression of the C9orf72 gene and loss of function of the C9orf72 protein ., Although C9orf72 is well-conserved among multi-cellular organisms , its protein function remains to be determined ., In this study , we demonstrated that loss of C9orf72 reduces mTOR signaling and enhances autophagy ., mTOR signaling and autophagy are important for the cellular maintenance of metabolic balances , especially under stress conditions ., C9orf72 protein exists in a complex with another DENN-like protein , SMCR8 , which also regulates mTOR signaling and autophagy ., We generated mice lacking C9orf72 , which died prematurely and showed dramatic upregulation of TFEB , a crucial transcriptional regulator of autophagy and lysosomal genes , that integrates mTOR activity state and autophagic capacity ., We propose that C9orf72 function is important for metabolic control and its deficiency can contribute to the development of neurodegenerative diseases . | cell death, autophagic cell death, lysosomes, ecology and environmental sciences, vesicles, molecular probe techniques, cell processes, immunoblotting, animal models, model organisms, immunoprecipitation, molecular biology techniques, cellular structures and organelles, research and analysis methods, immunoblot analysis, mouse models, molecular biology, precipitation techniques, community ecology, cell biology, ecology, trophic interactions, biology and life sciences | null |
journal.pcbi.1000042 | 2,008 | Shaping Embodied Neural Networks for Adaptive Goal-directed Behavior | One of the most important features of the brain is the ability to adapt or learn to achieve a specific goal , which requires continuous sensory feedback about the success of its motor output in a specific context ., We developed tools 1–3 for closing the sensory-motor loop between a cultured network and a robot or an artificial animal ( animat ) 4 in order to study learning directly through behavior of the artificial body and its interaction with its environment ., Compared to animal models , the cultured network is a simpler and more controllable system to investigate basic network computations; confounding factors such as sensory inputs , attention , and behavioral drives are absent , while diverse and complex activity patterns remain 5–9 ., Previously , an embodied cultured networks ability to control an animat or a mobile robot was demonstrated without a specifically defined goal 2 , 10 ., In another case , animats were designed to avoid obstacles 11 or follow objects 12 , but deterministically and without learning ., By using a lamprey brainstem to control a mobile robot , Mussa-Ivaldi et al . demonstrated the embodied in vitro networks tendency to compensate the sensory imbalance caused by artificially altering the sensitivity of the sensors at one side of the robot ., Without a pre-defined goal and external training stimulation , long-term changes in behavior in response to the sensory imbalance were found in embodied lamprey brainstems 13 , however , the changes were unpredictable 14 ., In order to further understand the learning capability of an embodied cultured network for goal-directed behavior , we need to investigate how the network can be shaped and rewired , and how to direct this change ., Previous studies have demonstrated the potential for disembodied cultured networks to achieve functional plasticity ., This neural plasticity provides a potential learning capability to cultured networks ., Jimbo et al . 15 used a localized tetanic stimulus to induce long-lasting changes in the network responses that could be either potentiated or depressed depending on the electrode used to evoke the responses ., Moreover , we and others previously found that such tetanus-induced plasticity was spatially localized and asymmetrically distributed 16 , 17 ., By delivering two different tetanic stimulation patterns , Ruaro et al . trained a cultured network to discriminate the spatial profiles of the stimuli ., These results suggest that different stimulation patterns can shape diverse functional connectivity in cultured networks ., By incorporating closed-loop feedback , Shahaf and Marom 18 showed unidirectional learning: to induce an electrode-specific increase in response ., This simple form of learning was achieved by a binary training: to stop a periodic stimulation at one electrode when the desired response level at the target electrode was obtained ., In order to scale to more complex behavior , we need to create more structured training stimuli and detailed activity metrics to investigate whether an embodied cultured network can learn multiple tasks simultaneously ., Unlike in vivo systems , the sensory-motor mapping and training algorithm in an embodied cultured network are defined by the experimenters ., In order to efficiently find an effective closed-loop design among infinite potential mappings , we first embodied a biologically-inspired simulated network to study an adaptive goal-directed behavior in an animat: learning to move toward and stay within a user-defined area in a 2-D plane ., The simulated network of 1000 leaky integrate-and-fire neurons expressed spontaneous and evoked activity patterns similar to that of the dissociated cortical cultures 19 ., Furthermore , a similar but larger simulated network showed that localized coherent input resulted in shifts of receptive and projective fields similar to those observed in vivo 20 ., Thus simulated networks show promise for analyzing biological adaptation with various closed-loop designs ., The closed-loop design we discuss here consists of four unique elements: Here , we demonstrate adaptive goal-directed behavior in the simulated network , where multiple tasks were learned simultaneously ., The desired behavior could only be achieved with proper selection of stimuli to encode sensory inputs and a variety of training stimuli with adaptive selection contingent on the animats behavior ., While lacking the characteristic layered structure of in vivo cortical tissue , the biologically-inspired simulated network still could be functionally shaped , and showed meaningful behavior , demonstrating that these neural networks have an innate ability to process information ., The proposed design is not restricted to a particular sensory-motor mapping , and could be applied with different and more complex goal-directed behaviors , which may provide a useful in vitro model for studying sensory-motor mappings , learning , and memory in the nervous system ., We used three networks with different connectivity , each with 5 different sets of CPSs ( randomly selected CPSQ1–CPSQ4 ) ., These 15 setups with different network connectivity and sensory-motor mappings were used for the following simulation experiments:, In order to validate the use of RBS to maintain desired behavior , the animat was run with RBS between context-control probing sequences ( CPSs ) without training ( no PTS ) , and the results were compared to the animats performance without RBS ( CPSs only ) ., An example of the time course of the animats distance from the origin is shown in Figure 2A ., The motor mapping was transformed ( by , see Figure 1B ) to obtain desired movements before the simulation ., Therefore , in the beginning of both simulations with RBS and without RBS , the animat moved in desired directions in each quadrant and stayed within the inner circle ., The animat maintained this desired behavior for the entire hour over 90% of the time when RBS was applied , whereas it moved outward after 10 minutes when no RBS was applied ., The mutual information between the movement angle and the sensory input is shown in Figure 2B ., When the animat started moving outward in an undesired direction , the mutual information decreased significantly ., This indicates decreasing stability of the animats movement under the same sensory input ., The mutual information during the last 10 minutes ( P2 period in Figure 2B ) was compared to the mutual information during the first 10 minutes ( P1 ) in the 15 simulations ( 3 networks , 5 different selections of CPSs each ) ( Figure 2C ) ., With RBS , the mutual information in P2 was 1 . 42±0 . 15 bits ( mean±SEM , n\u200a=\u200a1800 measures , 15 networks , 120 measures in 10 min per network ) , which was comparable to 1 . 53±0 . 09 bits in P1 ( p\u200a=\u200a0 . 77 , Wilcoxon signed-rank test ) ., Without RBS , the mutual information in P2 was 0 . 14±0 . 10 bits , which was significantly lower than 1 . 40±0 . 24 bits in P1 ( p<1e-4 ) ., This indicates that RBS with an aggregate frequency of 3 Hz maintained stability of the network input-output function , validating the use of RBS to maintain desired behavior in the animat ., Furthermore , the results also suggested that repetitive non-training stimuli ( CPSs and RBS ) were unable to induce enough plasticity to systematically alter the animats behavior ., We investigated the networks ability to learn a user-defined goal behavior by “switching” the sensory mapping ., A motor mapping was created ( through transformations ) to obtain desired movements before the experiment began ( Figure 1B ) ., The animats performance was observed for 10 minutes , demonstrating robust goal-directed behavior ( Figures 3 and 4 ) ., Then the sensory mapping was suddenly and drastically altered , so that the animats behavior was no longer correct ., Specifically , a CPS appropriate for evoking movement toward the center from Q1 was now delivered when the animat was in Q3 , and vice versa ., Learning was then quantified by the animats ability to adapt to the new , fixed sensory mapping and exhibit goal-seeking behavior ., Ten simulations , out of 15 , showed successful adaptation to the switch ., One successful simulation is shown in Figure 3A , and the corresponding movie is shown in Supplemental Material Movie S1 ., Immediately after the switch , as expected , the animat moved outward in the quadrants where the sensory mapping switch was performed ( Q1 and Q3 ) ., Patterned training stimulation ( PTS ) , paired stimulation designed to induce STDP throughout any shared activation pathways in the network , began to shape the network synaptic weights , and the desired behavior was restored under the switched mapping ., An unsuccessful simulation is shown in Figure 3B ., In 5 unsuccessful simulations , the animat kept moving outward and was repeatedly put back into the inner circle whenever it reached the outer circle ., The training was unable to restore the desired behavior throughout a 4-hr simulation ., In Figure 3B , only the first 90 minutes are shown for clarity ., Distance plots for all 15 simulations are shown in Figure 4 ., For successful simulations , the average time for the adaptation was 88 . 6±12 . 2 minutes ( mean±SEM , n\u200a=\u200a10 successful-learning simulations ) ., Two different types of unsuccessful learning are also indicated ( Type I and Type II failures , see below ) ., One-third of the simulations showed unsuccessful learning but were nevertheless informative ( see Figure 4 ) ., Two types of failures were observed in these following 5 unsuccessful experiments ., In order to verify that the successful adaptation in the overall system was contributed by learning in the network , and not solely by the adaptive process in the artificial training algorithm , we repeated the original successful-learning simulations with the STDP algorithm turned off ., We found that the desired behavior could not be restored without the STDP algorithm , or long-term plasticity , in the network ., This also rules out frequency-dependent synaptic depression as the adaptation mechanism , since that algorithm was left turned on ., The comparison of the animats movement in one successful-learning simulation and its corresponding simulation without STDP is shown in Figure 7 , and the comparison of learning curves is shown in Figure 7B ., Among all original successful-learning simulations , the average probability of successful behavior before the switch was 63 . 3±3 . 5% ( n\u200a=\u200a10 successful-learning simulations ) , dropped significantly to 9 . 8±1 . 1% after the switch ( p<5e-4 , Wilcoxon signed-rank test ) , and increased significantly back to 53 . 6±3 . 5% after 88 . 6±12 . 2 minutes when the desired behavior was restored ( p<5e-4 ) ( Figure 7C ) ., The probability of successful behavior after the switch was comparable to that before the switch ( p\u200a=\u200a0 . 09 ) ., For all corresponding simulation without STDP algorithm , the probability of successful behavior before the switch was 68 . 4±4 . 6% ( n\u200a=\u200a10 simulations without STDP ) , dropped significantly to 6 . 2±0 . 8% after the switch ( p<5e-4 ) , but showed no significant increase at the end of the simulation ( 6 . 4±0 . 9% ) ( p\u200a=\u200a0 . 91 ) ( Figure 7C ) ., This indicated that network long-term plasticity was essential for successful learning in the closed-loop system ., Different PTSs were delivered at different times before the desired behavior was restored ., The training history from the same successful-learning example shown in Figure 7 is shown in Figure 8A ., We hypothesized that the same PTS might have different effects at different points in time because the network would be in different states ., Therefore , successful adaptations would require application of PTSs in a certain sequence ., In order to test this hypothesis , we ran 10 additional simulations with only one PTS pattern available for training in each quadrant , instead of a pool of 660 PTSs as in the original stimulations ( see Methods ) ., These were the four most often used PTSs in the original simulations , one for each quadrant ., For the example shown in Figure 8A , only PTS #575 was delivered in the new simulation when training was required due to unsuccessful movement in Q1 ., We compared the original simulation and the corresponding new simulation by their learning curves ( one example is shown in Figure 8B ) ., The probability of successful behavior generally kept increasing after the switch for the original successful-learning simulation where multiple PTS patterns were available for training ( gray curve ) , but not for the new simulation where only a single PTS pattern was available ( blue curve ) ., A significant increase of the probability of successful behavior after the sensory mapping switch was found in the original successful-learning simulations ( p<5e-4 ) ( Figure 8D , and also Figure 7C ) ., However , all 10 new simulations with only the four most frequent PTSs available showed no significant increase of the probability of successful behavior from immediately after the switch ( 9 . 2±1 . 0% ) to the end of the simulation ( 10 . 1±3 . 7% ) ( p\u200a=\u200a0 . 61 , Wilcoxon signed-rank test ) ( Figure 8D ) ., This shows that not only one PTS , but a sequence of different PTSs was needed in order to restore the desired behavior ., We have demonstrated that successful adaptations to altered sensory mappings required a sequence of different PTSs , which was determined by the real-time feedback contingent on the animats performance ., In order to investigate the importance of behavior-contingent training for successful learning , we recorded the whole stimulation sequence ( PTS and RBS ) for each successfully adapted case and replayed it into the same network with the same initial state and same sensory-motor mapping ., Different random seeds for fluctuations in neurons membrane potentials and synaptic currents were used between the successful-learning simulations and the replayed training simulations ., This difference would lead to different network responses , and thus different movement trajectories and different CPS sequences ., However , the effect of non-training stimuli ( CPSs and RBS ) on shaping the network was insignificant , as shown in Figure 2 ., Therefore , whether the network could adapt to the new sensory mapping solely depended on the effect of training stimulation ., The replayed training stimulation was no longer contingent on whether or not desired movement occurred ., In 10 stimulation-replay experiments , the animat was unable to show successful adaptation to the sensory mapping switch ( shown as “non-contingent” in the example of Figure 9A ) , which had been successful with behavior-contingent training ( shown as “contingent” ) ., A comparison of the learning curves for this example is shown in Figure 9B ., With contingent training , a significant increase of the probability of successful behavior after the sensory mapping switch was found ( p<5e-4 ) ( Figure 9C , and also Figure 7C ) ., However , with replayed training stimulation , the average probability of successful behavior in the last 10 minutes of the simulations was 11 . 6±2 . 2% , which is comparable to 9 . 2±1 . 8% measured within 10 minutes after the switch ( p\u200a=\u200a0 . 47 ) ( Figure 9C ) ., In order to understand how successful ( closed-loop ) and replayed ( open-loop ) training stimulation shaped the network differently , we visualized the changes in weights of all synapses by using Principal Components Analysis ( PCA ) ., The first three components ( PC1 to PC3 ) of the network synaptic weights for the contingent training simulation and the non-contingent training simulation example shown in Figure 9A are plotted over time ( Figure 9D ) ., Starting from the same initial synaptic weights , the network diverged to different synaptic weights distributions as the training became progressively less contingent on the network activity and the animats performance ., We have demonstrated that two different sets of network synaptic weights that were responsible for the desired behavior under two different sensory mappings ( Pre and Post-contingent in Figure 9D ) ., We then further investigated whether under a specific sensory mapping , the desired behavior could only be exhibited by a specific set of network synaptic weights ., After the network adapted to the switched sensory mapping , we switched the sensory mapping back to the original sensory mapping to see whether the network could re-adapt to the original mapping ( Figure 10 ) ., After the switch-back , the behavior-contingent patterned training stimulation was able to restore the desired behavior under the original sensory mapping ( Figure 10A ) , but with a different set of network synaptic weights ( Figure 10B ) ., This indicates that multiple synaptic configurations , or “solutions” , existed for the desired behavior ., RBS was hypothesized to negate “attractors” in network synaptic weight distributions caused by spontaneous activity ( mainly network-wide synchronized bursts of activity called barrages ) , and to prevent network synaptic weights from drifting to such attractors after inducing plasticity with electrical stimulation 19 ., RBS with an aggregate frequency of 1 Hz reduced the occurrence of spontaneous barrages by at least 10 times in the simulated network and dissociated cortical cultures 19 ., By reducing the occurrence of spontaneous barrages , the network synaptic weights were mainly affected by activity evoked by RBS ., Since RBS was random spatially and temporally , the evoked activity had an unbiased randomizing effect on changing network synaptic weights ., In a different approach , a barrage-control stimulation protocol consisting of a group of electrodes cyclically stimulated with an aggregated frequency of 50 Hz was found to completely eliminate spontaneous barrages 32 ., Similar to RBS , the barrage-control stimulation stabilized tetanus-induced plasticity in dissociated cortical cultures ( Madhavan R , Chao ZC , Potter SM , unpublished data ) ., However , different mechanisms might be involved ., RBS evoked network-wide responses with unbiased spatiotemporal structure , while the barrage-control stimulation desynchronized spontaneous activity into spatially localized and temporally dispersed responses ., In this study , the aggregate stimulation frequency of RBS was increased from 1 to 3 Hz so that the amount of stimulation in RBS and PTS were comparable ., RBS did stabilize network synaptic weights ( the network synaptic weights were clustered in Pre period in Figure 9D ) and also stabilized the network input-output function ( see Figure 2 ) ., Even though sharing the same network connectivity and the same PTS pools , some simulations showed successful learning and others were unsuccessful ., Therefore , we concluded that the selection of CPSs for sensory encoding , which was the only remaining difference , was crucial for determining the success of adaptation ., We found that the stimulations used to encode sensory inputs should evoke neither overly localized nor largely overlapped responses ., Too much localization reduced the possibility to improve movement directions in switched quadrants , and too much overlap caused unwanted changes in un-switched quadrants ., These results suggest a certain level of independence is required between responses to stimulations used to encode different sensory inputs , which could be achieved by using smaller and distinct recording areas to determine movement , or by offsetting the CA through the motor mapping transformation so that the probability of a CA to point in different directions is more uniform ., Furthermore , correlated changes in responses to different sensory inputs could also be avoided by using training stimulation that only causes localized plastic changes ., These findings could instruct the designs of implant electrode geometries and feedback stimulation patterns in prosthetics to achieve a more efficient and effective adaptation ., We showed that long-term plasticity in the network ( STDP ) was essential for the adaptation in the overall system ( see Figure 7 ) ., Short-term plasticity ( frequency-dependent synaptic depression , see Methods and Supplemental Material Text S1 ) alone was not able to achieve successful adaptation ( Figure 7 ) ., Furthermore , learning curves indicate that fewer training stimuli were required to maintain the desired behavior after the system had adapted ( see Figure 7B and Figure 8B ) ., These suggest that the improved performance was not due to short-term elastic responses to the stimulation ., Elastic change was observed in dissociated cultures where the neurons responsiveness adapted to very low frequency stimulation but relaxed back within minutes after stimulation was removed 33 , 34 ., Using paired pulses with different stimulation electrodes and different inter-pulse intervals was one possible design for training ., More optimal training algorithms likely exist ., Using stimulation sequences with more than two stimuli could help shape the network synaptic weights to a desired state , since they might evoke a greater variety of response patterns and produce different behaviors ., However , the tradeoff is that a larger pool of possible training stimuli could lead to a longer training duration before successful adaptation ., Furthermore , a different algorithm to adaptively update the probability of selecting PTSs might better find appropriate PTSs and remove unhelpful ones in the pool ., The simulated network was used to explore many different possible sensory-motor mappings and training algorithms ( not described here ) because of savings in preparation time and an ability to monitor all synaptic weights ., The described algorithm successfully demonstrated adaptive goal-directed behavior with multiple sensory-motor mappings ., This closed-loop algorithm is not restricted to a particular type or a particular number of sensory-motor mappings ., Integrate-and-fire networks have been used previously for demonstrating goal-directed learning 35 , 36 ., In this work , we constructed a simulated network , specifically to mimic living MEA cultures , in order to find a closed-loop design that might be applicable to show goal-directed learning living cultures ., In another study , we tested our closed-loop algorithm in a cortical network cultured over an MEA , where we successfully avoid Type I and Type II failure to train a living network to control the movement of an animat in a desired direction ( Chao ZC , Bakkum DJ , Potter SM , unpublished data ) ., Studying neural networks basic computational properties , such as parallel signal processing and learning , by working with simulated/living in vitro networks could lead to direct development of more advanced artificial neural networks , more robust computing methods , and even the use of neurally controlled animats themselves as biologically-based control systems . | Introduction, Methods, Results, Discussion | The acts of learning and memory are thought to emerge from the modifications of synaptic connections between neurons , as guided by sensory feedback during behavior ., However , much is unknown about how such synaptic processes can sculpt and are sculpted by neuronal population dynamics and an interaction with the environment ., Here , we embodied a simulated network , inspired by dissociated cortical neuronal cultures , with an artificial animal ( an animat ) through a sensory-motor loop consisting of structured stimuli , detailed activity metrics incorporating spatial information , and an adaptive training algorithm that takes advantage of spike timing dependent plasticity ., By using our design , we demonstrated that the network was capable of learning associations between multiple sensory inputs and motor outputs , and the animat was able to adapt to a new sensory mapping to restore its goal behavior: move toward and stay within a user-defined area ., We further showed that successful learning required proper selections of stimuli to encode sensory inputs and a variety of training stimuli with adaptive selection contingent on the animats behavior ., We also found that an individual network had the flexibility to achieve different multi-task goals , and the same goal behavior could be exhibited with different sets of network synaptic strengths ., While lacking the characteristic layered structure of in vivo cortical tissue , the biologically inspired simulated networks could tune their activity in behaviorally relevant manners , demonstrating that leaky integrate-and-fire neural networks have an innate ability to process information ., This closed-loop hybrid system is a useful tool to study the network properties intermediating synaptic plasticity and behavioral adaptation ., The training algorithm provides a stepping stone towards designing future control systems , whether with artificial neural networks or biological animats themselves . | The ability of a brain to learn has been studied at various levels ., However , a large gap exists between behavioral studies of learning and memory and studies of cellular plasticity ., In particular , much remains unknown about how cellular plasticity scales to affect network population dynamics ., In previous studies , we have addressed this by growing mammalian brain cells in culture and creating a long-term , two-way interface between a cultured network and a robot or an artificial animal ., Behavior and learning could now be observed in concert with the detailed and long-term electrophysiology ., In this work , we used modeling/simulation of living cortical cultures to investigate the networks capability to learn goal-directed behavior ., A biologically inspired simulated network was used to determine an effective closed-loop training algorithm , and the system successfully exhibited multi-task goal-directed adaptive behavior ., The results suggest that even though lacking the characteristic layered structure of a brain , the network still could be functionally shaped and showed meaningful behavior ., Knowledge gained from working with such closed-loop systems could influence the design of future artificial neural networks , more effective neuroprosthetics , and even the use of living networks themselves as a biologically based control system . | neuroscience/behavioral neuroscience, neuroscience, computational biology/computational neuroscience | null |
journal.pcbi.1006606 | 2,018 | Bayesian adaptive dual control of deep brain stimulation in a computational model of Parkinson’s disease | Deep brain stimulation ( DBS ) is an effective therapy for treating the motor symptoms of Parkinson’s disease ( PD ) , and is often used to complement dopamine replacement therapy in patients who have progressed to severe stages of PD 1 ., The clinical success of DBS relies on selecting stimulation parameters that both relieve symptoms and avoid persistent stimulation-induced side effects ., Identifying clinically optimized stimulation settings , or in other words programming the pulse generator , is conducted by a movement disorders specialist through a laborious trial-and-error process ., The process involves parsing through several free parameters including electrode configuration , stimulation amplitude , pulse frequency , and pulse width ., However , because the programming process is both time-intensive and exhausting for the patient 2 , 3 , most clinical programming visits focus on a truncated set of four monopolar electrode configurations in which stimulation amplitude is increased for each setting to the point of inducing persistent side effects ., Recent advances in DBS technology have rendered the programming process even more challenging ., For instance , directional DBS leads with eight 4 or as many as thirty-two contacts 5 are emerging for clinical use , and new stimulation algorithms are increasing the dimensionality of the programming process , adding additional free parameters 6–12 ., As these new technologies become more widely available , programming next-generation DBS systems will no longer be feasible with current trial-and-error approaches 13 ., Implantable DBS systems have been designed to deliver stimuli and record the resulting neural responses , thus providing a framework for implementing closed-loop DBS algorithms 14 that can intelligently select the optimal stimulation parameters for each patient at any point in time ., Key to the development of a closed-loop DBS strategy is defining a biomarker as feedback for a controller; the biomarker must correlate well with PD symptoms , although it need not be causal ., Synchronous activity in the beta range ( 12-35 Hz ) of local field potentials ( LFPs ) is one possible candidate ., While the precise role of beta oscillations in the basal ganglia are under debate , increased beta band activity within the basal ganglia has been associated with anti-kinetic symptoms of PD 15 ., Specifically , elevated beta power has been observed in the dorsolateral portion of the subthalamic nucleus ( STN ) in human patients 16–18 as well as the globus pallidus ( GP ) , but to lesser extent 19–21 ., There is also evidence that a reduction in beta power , either by medication 22–24 or DBS 25 , correlates with improved UPDRS scores ., Two separate types of beta-based feedback stimulation policies have been proposed: power or amplitude feedback and phase feedback ., In the former implementation , an amplitude-responsive adaptive STN-DBS algorithm initiated stimulation only when the amplitude in the beta band of STN LFPs exceeded a manually set threshold 7 , 8 ., This approach resulted in significant reduction in parkinsonian motor signs and overall reduction in stimulation on-time compared to conventional , isochronal DBS ( cDBS ) ., In the latter case , stimulation was triggered off of the phase of the beta oscillation , delivering phase-locked bursts to optimally disrupt beta oscillations for PD 9 , 10 or low frequency oscillations for tremor 11 , 12 ., However , while both stimulation policies are closed-loop , neither is autonomous; each requires manually setting yet another free parameter ., A visualization of these two differing stimulation policies are show in Fig 1 , as well as a combined phase and power feedback stimulation policy ., We will use the term “power” here , as opposed to “amplitude” , to disambiguate this parameter from other stimulation parameters ., As one can readily convert between power and amplitude , the terms are essentially interchangeable ., In this study , we designed and tested a Bayesian adaptive dual control algorithm that can efficiently and autonomously learn the parameters of both phase and power feedback stimulation , as well as other stimulation parameters ., We evaluated the algorithm in a computational mean-field model of the basal ganglia-thalamacortical system that simulated beta rhythms and response to electrical stimulation , and we compared the algorithms performance to other optimization strategies ., In order to develop and test the adaptive dual control algorithm , we used a physiologically realistic mean-field model of the basal ganglia-thalamocortical system ( BGTCS ) , developed by van Albada and Robinson 26 , 27 ., The BGTCS modeled the mean firing rate and voltage of nine cortical and subcortical structures with second-order differential equations , the structure of which is shown below in Fig 2 ., The model was capable of simulating both the naïve state , as well as a dopamine-depleted ( DD ) state , with a strong beta rhythm ., In this study we tested the Bayesian adaptive dual controller in the dopamine-depleted state of the model to suppress its beta oscillation ., For a detailed description of the equations governing the model and how parameters were set , see van Albada and Robinson , 2009 26 , 27 ., The BGTCS model produced LFP signals generally comparable in spectral content to those measured in humans with Parkinson’s disease undergoing DBS surgery ., In order to simulate the effects of DBS within the model , stimuli were incorporated as a direct current injection into the target structure ., As the integration timestep of the model ( 1 ms ) was much greater than the duration of the first phase of a typical DBS pulse ( 60 μs–240 μs ) , the stimulus pulse was integrated to obtain the total charge , which was then divided by the membrane capacitance to yield the change in voltage due to a single DBS pulse ., The resultant ΔV was added directly to the voltage of the target structure ., Fig 3 shows example voltage traces from the GPi of the BGTCS in the naïve , DD , and DD with cDBS states , as well as the power spectrum from each trace ., The power spectra revealed several salient features of the BGTCS ., First , it produced an oscillation in the beta range ( at 29 Hz ) , and the power of that oscillation increased in the DD state ., Second , simulated conventional DBS at 130Hz ( cDBS ) , similar to what has been used clinically , reduced the power of the 29 Hz oscillation ., Thus , the model of the dopamine-depleted state 1 ) produced oscillations with a pronounced beta peak , and 2 ) responded to cDBS in a realistic manner ., This model was then used to design , test , and evaluate the Bayesian adaptive dual controller ., The tuning of stimulation parameters for DBS was formulated as a control problem: We have a system ( the patient ) whose symptoms we wish to control ( i . e . reduce ) with stimulation ., However , unlike normal control problems , here we have dual goals: We wish to control the patient’s symptoms as well as possible using the best known stimulation parameters , but also must explore the parameter space to identify new parameters that may be better than the current best , thus allowing for better control in the future ., This balance between control and information gathering , or exploitation and exploration leads to the concept of dual control 28 ., In order to accomplish these conflicting goals , we implemented an adaptive dual controller ( ADC ) for DBS , which is composed of two components: ( 1 ) an inner parameterized stimulator and ( 2 ) an outer parameter adjustment loop ., The inner loop can be any stimulator with parameters to tune , from a traditional cDBS system to new closed-loop DBS algorithms , and may or may not incorporate feedback from the patient ., For example , a power-based DBS algorithm would turn stimulation on or off based upon the power of an oscillation measured from the patient ., Conversely , cDBS would not measure any feedback signals ., The outer parameter adjustment loop acts to tune the parameters of the stimulator , and operates on a relatively long timescale ., The outer loop is given a specification , or goal , which it attempts to meet through an iterative process: selecting a parameter value ( or values ) , observing the effect of that value on its goal , estimating the effects of new values , and then selecting the next value ., For example , with an power-based DBS algorithm , the outer loop would begin by selecting a power threshold for the inner loop ., The inner loop would then execute stimulation with that parameter value for some pre-determined amount of time , after which the outer loop would observe the effects of that value on some biomarker and select a new value ., The general structure of an adaptive dual controller for DBS is shown in Fig 4 ., Traditional cDBS can be viewed as a simplistic ADC , where an isochronal stimulator takes the place of the parameterized stimulator , and the clinician acts as the parameter adjuster ., The clinician’s specification is to improve the patients quality of life ., During a clinic visit , they select stimulation parameters and observe the effects ., The clinician uses his or her experience to build a mental estimation of the relationship between parameters and quality of life , and uses this map to intelligently determine which parameter combinations to try ., At the end of the visit , however , the loop is broken and the patient is sent home with the clinician-optimized settings ., Here , we designed a Bayesian ADC with two components: an inner phase/power feedback stimulator , and an outer Bayesian optimization parameter adjustment loop ., We first describe the components individually , and then describe the combined Bayesian ADC ., The inner feedback stimulator had three parameters: ( 1 ) oscillation phase trigger , ( 2 ) oscillation power threshold , and ( 3 ) stimulus amplitude ., In order to implement phase and power feedback stimulation , a real-time method of accurately estimating both the phase and power of an oscillation was paramount ., Previously , phasic stimulation had been accomplished by band-pass filtering the signal and then using the time since the preceding zero crossing to approximate phase 12 ., Power-based stimulation had been achieved by rectifying and smoothing the band-passed signal for 400 ms 7 , 8 ., The Hilbert transform is often used to extract the phase and power of a signal ., However , the Hilbert transform is acausal , making it impossible to implement in real time ., We recently developed a novel sliding Fourier transform , called the Sliding Windowed Infinite Fourier Transform ( SWIFT ), X n ( ω ) = e - 1 / τ e j ω X n - 1 ( ω ) + x n , ( 1 ), along with the αSWIFT ,, X n ( ω ) α = X n ( ω ) s l o w - X n ( ω ) f a s t , ( 2 ), described in 29 ., Unlike other methods of phase/power estimation , the SWIFT directly and efficiently calculates the Fourier transform of the signal in real time , centered on ω = 2πf/fs and windowed with an infinite length , causal exponential window ., In fact , the SWIFT is a causal approximation of the Hilbert transform ., The αSWIFT employs the α window ( the difference between two exponentials with different time constants ) , and has improved frequency resolution ., Here , we used the αSWIFT to calculate the phase and power of the beta oscillation in real time ., The SWIFT has two parameters which control its behavior: the center frequency ω , and the time constant τ ( or two time constants , τslow and τfast for the αSWIFT ) ., The center frequency , ω was set to match the center frequency of the beta peak in the model ., The ( slow ) time constant controls the time-frequency tradeoff of the SWIFT: a shorter time constant leads to higher temporal resolution , but lower frequency resolution ( wider frequency response ) ., To balance this tradeoff , we matched the width of the SWIFT’s frequency response to the width of the model’s beta peak at -6 dB ( or 50% power reduction ) ., The model’s beta peak had a width of ±1 . 15 Hz at -6 dB , and so we set τslow = 0 . 240 s to match , which can be readily calculated from the Fourier transform of the SWIFT’s exponential window ., τfast was set to τslow/5 , which smooths the output without significantly altering the SWIFT’s frequency response ., Fig 1 shows the phase/power feedback stimulation algorithm operating on an example LFP , extracting phase/power using the αSWIFT , and triggering stimulation off phase when the power is above threshold ., In this context , the SWIFTs parameters are selected to filter the signal around the oscillation produced by the BGTCS ., The SWIFT parameters for a physiological signal could be selected in a similar manner: The center frequency and width can be estimated from the power spectral density measured from a sample signal ., A concern is that a physiological signal’s center frequency may wander more than the BGTCS model; this could be addressed by periodically re-estimating the SWIFT parameters from the raw signal ., Alternatively , Jackson et al , 2016 described a method of estimating the real time phase of a frequency-modulated signal by combining three real time Fourier transforms ( RTFT ) operating at neighboring frequencies , which produces a flat frequency response over the frequency band of interest ., Their method could easily be augmented to use the SWIFT in the place of the RTFT 30 ., While many optimization algorithms could be used for the outer loop , the problem of creating an ADC for DBS has several constraints which make Bayesian optimization ( BayesOpt ) ideal ., The goal of BayesOpt is to find the minimum of the objective function with as few evaluations as possible 31–34 , and indeed is among the most efficient algorithms at doing so 32 , 35–38 ., BayesOpt also provides a framework for explicitly balancing exploration and exploitation in order to efficiently find the global minimum ., To reduce the number of function evaluations , BayesOpt only approximates the objective function accurately in regions where it is profitable to do so , and samples coarsely everywhere else 39 ., This is ideal for tuning stimulation parameters as the patient is likely to have little tolerance for exploration , and so we wish to find their optimal settings with as few steps as possible ., The power and efficiency of BayesOpt stems from the incorporation of prior belief about the objective function with available evidence ( through Bayes theorem ) to build a model of the objective function ,, P ( M | E ) ∝ P ( E | M ) P ( M ) ., ( 3 ), That is , the posterior probability of a model M given some evidence E , is proportional to the likelihood of E given M multiplied by the prior probability of M . BayesOpt then uses this model to direct sampling and trade off exploration and exploitation 40 ., BayesOpt consists of three steps ., First , a prior distribution is defined over the objective function ., Second , a set of N previously gathered measurements , D 1 : N , are combined with the prior through Bayes rule to obtain a posterior distribution ., Finally , the acquisition function , which is a function of the posterior distribution that predicts the utility of sampling , is used to determine where next to sample to maximize the utility ., Putting the above components together , we constructed a Bayesian adaptive dual controller ., The controller had two components: an inner feedback stimulation loop , which applied stimulation based on the phase/power of the beta oscillation , and an outer Bayesian parameter adjustment loop which optimized the parameters of the inner feedback stimulator to maximally suppress the beta oscillation ., In order for the Bayesian ADC to find an optimal parameter set , there must exist at least one minimum over the feedback stimulator’s parameter space ., We swept the space on a 643 grid ( oscillation phase trigger , oscillation power threshold , and stimulus amplitude ) , and measured the average beta power over the last 50 s of a 100 s simulation ., Fig 7 shows three 2D slices through the parameter space ( with the third parameter held constant at its global minimum ) ., We see that the model’s beta power responded to all three parameters , and that a minimum existed ., The sweep also revealed a complex underlying landscape with flat regions , nonlinearities , and local minima , which may prove difficult for optimization algorithms to navigate ., Fig 7 only shows three 2D slices through 3D volumetric data; there are other complex interactions which are not seen in these planes ., Most importantly , however , the parameter sweep revealed that the BGTCS has a global minimum ., Next , we tested the Bayesian ADC’s ability to efficiently locate the global minimum in this complex landscape ., After having verified the existence of a global minima , we ran the Bayesian ADC in all 7 parameter combinations ., In the 1 and 2D cases , the variable ( s ) not being optimized over were fixed at their global minimum ., Fig 8 shows a 1D example of the Bayesian ADC optimizing the stimulus phase trigger , with stimulus amplitude and power threshold held constant ., By the 6th function evaluation , the Bayesian ADC was already sampling near the optimal stimulus phase ., The Bayesian ADC was able to build an accurate representation of the BGTCS’ response to stimulation in relatively few function evaluations ., The ADC took few exploratory steps , and did so to optimally cover the space and gather information about the underlying function ., In this example , we see that at function evaluation 16 , the Bayesian ADC chose to explore near −π , before returning to the optimal region around 3π/4 ., Finally , we empirically analyzed the Bayesian ADC , and compared the BayesOpt outer loop’s performance to other optimization strategies ., We chose to compare to two types of algorithms: gradient-approximating algorithms , such as the Nelder-Mead ( NM ) simplex 43 , and global algorithms such as DIRect 44 ., Each algorithm was bounded on the same interval , and initial conditions were selected uniformly at random ., We selected NM because we expected it to outperform most other gradient-approximating algorithms , most of which are not robust to noise ., The NM approximates the gradient using a simplex , whose vertices are often far enough apart to return the correct search direction , even in the presence of noise ., We selected DIRect due to its ability to quickly blanket the search space ., Through examination of the optimization landscape ( Fig 7 ) , we can draw several key insights regarding the nature of closed-loop stimulation in the context of a biological system , and relate these results to other studies in the field ., The Bayesian ADC’s key advantages stem from fitting a Gaussian process ( GP ) to data , and then using the GP to intelligently sample , explicitly balancing exploration and exploitation to find the global minimum ., Through fitting the GP , BayesOpt is able to learn and account for both the length scale of the parameters as well as the noise level , and is among the most efficient algorithms in terms of number of function evaluations required to find the minimum ., The Bayesian ADC is also able to balance exploration and exploitation: it is able to find the minimum quickly and exploit it , but continues to explore intelligently to ensure that it has arrived at the global minimum ., Gradient-approximating methods ( such as NM ) can quickly descend towards a minimum , but are unable to explore globally , are sensitive to initial conditions , and are vulnerable to becoming trapped in local minima or wandering around flat regions ., Global exploration methods ( such as DIRect ) , do not rely on gradients and can quickly find the global minimum ., However , such algorithms are often purely exploratory , and never transition to exploitation ., When trying to optimize stimulation parameter settings , balancing exploration and exploitation is critical ., We need to approach the minimum as quickly as possible , but also avoid local minima while preventing unnecessary exploration , as the patient is likely to have little tolerance for wildly varying stimulation parameters ., BayesOpt provides a framework for balancing exploration and exploitation in a way that most other algorithms do not ., Of course , selecting the optimal balance between exploration and exploitation is not a trivial task ., All acquisition functions have a hyperparameter which controls the exploration/exploitation tradeoff ., The GP-UCB algorithm we emplyed is no regret in the limit as N → ∞ with ν = 1 ., However , we are less interested in achieving 0 regret than we are with achieving a low regret quickly ., Thus , a smaller ν should be chosen , such as ν = 0 . 25 , to encourage exploitation ., Furthermore , this hyperparameter could be adapted over time: if the patient feels like too many exploratory settings are being chosen , ν could be decreased ., The Bayesian ADC is not without its limitations ., First and foremost , because Bayesian optimization relies on calculating and inverting the covariance matrix of the inputs , the complexity grows as O ( n3 ) , where n is the number of observations ., Therefore , the time it takes to compute the next parameter set increases as the cube of the number of samples ., However , in this case , and many clinical applications , the time it takes to assess the effects of a single parameter set is relatively long ( seconds in this model , minutes in the clinic , or hours or even days at home ) ., Therefore , as long as it takes less time to compute the next parameter set than it does to evaluate a parameter set , this will not be an issue ., Additionally , the Bayesian ADC assumes the existence of a static response surface , although this need not be the case ., If the patient’s response to stimulation is changing over the course of the measurements , the Bayesian ADC will not converge ., However , if the time-course of this change is long relative to the time-course of the measurements , this could be overcome ., Furthermore , instead of using all previous observations , we could limit the algorithm to use only the most recent N , thereby allowing the algorithm to “forget” , forcing it to re-explore changing areas ., This “forgetting” strategy could be used to solve the aforementioned complexity problem as well ., Finally , neural networks ( NNs ) could be used to estimate the GP , which would address both the scalability problem ( becoming linear in n , instead of cubic ) , and the stationarity problem ( as NNs naturally “forget” training data far in the past ) 54 , 55 ., The Bayesian ADC framework we present here has broad applicability for tuning stimulation parameters across diseases and devices ., At its heart , the Bayesian ADC framework is simply a method for efficiently optimizing the parameters of a controller using some feedback signal ., Both the inner loop and the feedback signal can be designed to fit the problem at hand ., In this paper , we present a Bayesian adaptive dual controller for the suppression of pathological oscillations ., The Bayesian ADC was shown to perform well in a computational model of Parkinson’s disease for selecting the optimal parameters to reduce the oscillation power ., The Bayesian ADC was composed of two parts , an inner feedback stimulator , and an outer BayesOpt parameter tuning loop ., As compared to other algorithms , BayesOpt was able to efficiently tune stimulation parameters , explicitly balancing exploration and exploitation to find the optimal settings in as few function evaluations as possible ., Finally , the Bayesian ADC is generalizable , both across diseases and stimulator designs . | Introduction, Methods, Results, Discussion | In this paper , we present a novel Bayesian adaptive dual controller ( ADC ) for autonomously programming deep brain stimulation devices ., We evaluated the Bayesian ADC’s performance in the context of reducing beta power in a computational model of Parkinson’s disease , in which it was tasked with finding the set of stimulation parameters which optimally reduced beta power as fast as possible ., Here , the Bayesian ADC has dual goals:, ( a ) to minimize beta power by exploiting the best parameters found so far , and, ( b ) to explore the space to find better parameters , thus allowing for better control in the future ., The Bayesian ADC is composed of two parts: an inner parameterized feedback stimulator and an outer parameter adjustment loop ., The inner loop operates on a short time scale , delivering stimulus based upon the phase and power of the beta oscillation ., The outer loop operates on a long time scale , observing the effects of the stimulation parameters and using Bayesian optimization to intelligently select new parameters to minimize the beta power ., We show that the Bayesian ADC can efficiently optimize stimulation parameters , and is superior to other optimization algorithms ., The Bayesian ADC provides a robust and general framework for tuning stimulation parameters , can be adapted to use any feedback signal , and is applicable across diseases and stimulator designs . | Deep brain stimulation ( DBS ) is an effective therapy for treating motor symptoms of Parkinson’s disease ., However , the clinical success of DBS relies on selecting stimulation parameters that both relieve symptoms while avoiding side effects ., Currently , DBS devices are programmed using a laborious trial-and-error process , requiring multiple clinic visits over the course of months ., As DBS leads and algorithms become more complex , it will become impossible to select optimal DBS parameters manually ., There is a clear need for an intelligent , automated approach to parameter tuning ., We present a novel Bayesian adaptive dual controller ( ADC ) , which can autonomously tune stimulation parameters ., It uses a feedback signal measured from the patient to quantify the efficacy of a set of stimulation parameters , and uses this information to intelligently find the parameters which work best for each individual patient ., The Bayesian ADC has the potential to improve DBS efficacy and reduce clinic visits by efficiently finding the best stimulation parameters . | medicine and health sciences, neurochemistry, chemical compounds, neurodegenerative diseases, applied mathematics, brain electrophysiology, social sciences, electrophysiology, biomarkers, neuroscience, organic compounds, surgical and invasive medical procedures, hormones, simulation and modeling, algorithms, deep-brain stimulation, optimization, cognitive psychology, mathematics, functional electrical stimulation, brain mapping, bioassays and physiological analysis, amines, neurotransmitters, catecholamines, intelligence, dopamine, research and analysis methods, chemistry, electrophysiological techniques, movement disorders, biochemistry, psychology, organic chemistry, physiology, neurology, parkinson disease, biogenic amines, biology and life sciences, physical sciences, cognitive science, neurophysiology | null |
journal.pgen.1002170 | 2,011 | Genome-Wide Association Study Identifies HLA-DP as a Susceptibility Gene for Pediatric Asthma in Asian Populations | Asthma is the most common chronic disorder in children , and asthma exacerbation is an important cause of childhood morbidity and hospitalization ., The prevalence of childhood asthma in Japan is 5 . 0% among school children in 2006 1 , and an estimated 300 million people worldwide have asthma 2 ., Asthma is characterized by airway hyperresponsiveness and inflammation , tissue remodeling , and airflow obstruction ., Infiltration of lymphocytes , mast cells , and eosinophils in the airways cause airway inflammation , and T helper ( Th ) type 2 cytokines play crucial roles in orchestrating the inflammatory responses; thus , asthma is considered a Th2-type immune disease ., Previously conducted genome-wide association studies ( GWAS ) for asthma identified association with the loci on chromosomes 17q21 ( ORMDL3 for Caucasian pediatric asthma , odds ratio ( OR ) =\u200a1 . 45 , P\u200a=\u200a1×10−10 ) 3 , 5q21 ( PDE4D for pediatric asthma , OR\u200a=\u200a0 . 6 , P\u200a=\u200a4 . 7×10−7 ) 4 , 9q21 . 31 ( TLE4 for Hispanic pediatric asthma , OR\u200a=\u200a0 . 6 , P\u200a=\u200a6 . 8×10−7 ) 5 , and 1q31 ( DENND1B for Europeans and African ancestries 6 , OR\u200a=\u200a0 . 77 and 1 . 41 , respectively; combined P\u200a=\u200a1 . 7×10−13 ) ., A GWAS for severe asthma identified association with the region between RAD50 and IL5 on chromosome 5q ( OR\u200a=\u200a1 . 64 , P\u200a=\u200a3 . 0×10−7 ) and HLA-DR/DQ ( OR\u200a=\u200a0 . 68 , P\u200a=\u200a9 . 6×10−6 ) , but they did not include a replication dataset 7 ., Recently , Moffatt et al . conducted a large-scale GWAS in Caucasian populations and identified 6 loci ( IL18R1 , HLA-DQ , IL33 , SMAD3 , GSDMB/GSDMA , and IL2RB ) associated with asthma 8 ., In the present study , we conducted the first GWAS in Asian population for pediatric asthma by using Illumina HumanHap550/610-Quad BeadChip ( Illumina , San Diego , USA ) ., The GWAS flow chart is shown in Figure 1 ., We analyzed 450 , 326 SNPs in 938 cases and 2 , 376 controls , using standard quality control practices ( Table S1 ) ., The genotypes in cases and controls were compared using the Cochran–Armitage trend test ( Figure 2 ) ., There was only minor inflation of the genome-wide statistical results owing to population stratification ( genomic control ( λGC ) =\u200a1 . 048; Figure 3 ) ., Five SNPs ( rs3019885 , rs987870 , rs2281389 , rs2064478 , and rs3117230 ) showed strong association with pediatric asthma with P<1×10−8 ., Of these , rs2064478 and rs3117230 were in complete linkage disequilibrium ( LD ) ( r2\u200a=\u200a1 ) with rs2281389 ., In order to validate the results of the GWAS , we tested the remaining 3 SNPs ( rs3019885 , rs987870 , and rs2281389 ) in 2 independent replication cohorts comprising Asians ( Japanese and Koreans ) , considering P<0 . 05 as significant replication ., Of these 3 SNPs , significant associations were noted at rs987870 in both cohorts ( Table 1 ) ., To merge the findings of these studies , we conducted meta-analysis with a fixed-effects model by using the Mantel–Haenszel method ., As shown in Table 1 , the Mantel–Haenszel P value of 2 . 3×10−10 was noted for rs987870 ( OR\u200a=\u200a1 . 40 , confidence interval ( CI ) =\u200a1 . 26–1 . 55 ) ., The rs987870 is located between HLA-DPA1 and HLA-DPB1 ., Genotype imputation using MACH 9 revealed association between asthma and the SNPs that were in strong LD with rs987870 ( Figure 4 , Table S2 ) ., Moreover , rs987870 C allele was in complete LD with DPA1*0201 ( r2\u200a=\u200a1 ) ., We determined HLA-DPA1 genotypes by using direct sequencing and MACH imputation of the data from 1135 cases and 2376 controls and found that DPA1*0201 was strongly associated with pediatric asthma ( P\u200a=\u200a5 . 2×10−10 , OR\u200a=\u200a1 . 52 , Table 2 ) ., Then , we determined the HLA-DPB1 genotypes in 1135 cases and 2296 controls and found that DPB1*0901 was associated with pediatric asthma ( P\u200a=\u200a2 . 0×10−7 , OR\u200a=\u200a1 . 49 , Table 3 ) ., DPB1*0901 was in strong LD with DPA1*0201 and rs987870 C allele ( D prime\u200a=\u200a0 . 93 ) ., Because more than 90% of pediatric asthma patients were allergic to house dust mites , it is possible that the association was due to IgE reactivity ( sensitization ) against mites ., We performed an association study for mite sensitization using independent adult subjects without allergic respiratory diseases such as asthma and perennial allergic rhinitis ( 367 subjects with house dust mite-specific IgE and 1633 subjects without mite-specific IgE ) ., Subjects with house dust mite-specific IgE were non-allergic in terms of symptoms but possessed mite-specific IgE ., Subjects without mite-specific IgE did not exhibit allergic symptoms ., We did not find an association between rs987870 and mite sensitization ( P\u200a=\u200a0 . 54 , OR\u200a=\u200a1 . 07 , Table S3 ) ., Our GWAS in Asian populations found HLA-DP as susceptibility gene for pediatric asthma ., Majority of pediatric asthmas are atopic ( i . e . , familial tendency to produce IgE antibodies against common environmental allergens ) and possess specific IgE against the house dust mite ., Mite sensitization is more prevalent in Asia than in Europe and is observed in 39% of the general adult population in Japan 10 ., High prevalence of mite sensitization in asthmatic children has also been reported in Taiwan , where 94 . 2% of children with asthma are sensitized against Dermatophagoides pteronyssinus 11 ., However , only a small subset of subjects with house dust allergy develop asthma 12 ., We performed an independent association study for mite sensitization in adult subjects without allergic respiratory diseases and did not find an association between rs987870 and mite sensitization without symptoms ., If the relative risk for mite sensitization in the individuals carrying a putative risk allele was 1 . 4 and the allele frequency was 0 . 15 compared to that in individuals without the allele , the statistical power of the sample size for mite sensitization study was 0 . 92 at an alpha level of 0 . 05 ., These results suggested that DPA1*0201 and DPB1*0901 may be associated with asthma rather than IgE production against house dust mite ., The association signal was stretched in the region of HLA-DPB2 , collagen , type XI , alpha 2 ( COL11A2 ) , and Retinoid X receptor beta ( RXRB ) ( Figure 4 ) ., Because of LD in this region , it is difficult to specifically identify causative variants ., HLA-DPB2 is a pseudogene ., COL11A2 encodes a component of type XI collagen called the pro-alpha2 ( XI ) chain ., Mutations in COL11A2 have been associated with non-syndromic deafness , otospondylomegaepiphyseal dysplasia , Weissenbacher-Zweymüller syndrome , and Stickler syndrome ( OMIM ID *120290 ) ., RXRB belongs to the RXR family and is involved in mediating the effects of retinoic acid ., RXRB forms a heterodimer with the retinoic acid receptor and thus preferentially increases its DNA binding and transcriptional activity at promoters containing retinoic acid 13 ., All SNPs showing strong association with asthma ( P<1×10−10 ) were located in introns or intergenic regions ., LD of these associated SNPs with rs987870 was not strong; therefore , it is likely that the functional effect is due to DPA1*0201 and DPB1*0901 ., In HLA-DP , Caraballo et al . reported that DPB1*0401 is significantly decreased in patients with allergic asthma in Mulatto population ( an admixture population of European and African ancestries ) 14 ., Apart from the study of Caraballo et al . , the association between HLA-DP alleles and asthma was restricted to occupational 15 or aspirin-induced asthma 16 ., Howell et al . reported associations between HLA-DR genotype and asthma and between HLA-DPA1*0201 and IgE specific to grass pollen mix and the pollen allergen Phl p 5 17 ., Grass pollen allergy is not a major cause of asthma in Japan 18; therefore , the HLA-DPA1*0201 association in the present study was less likely to be due to sensitization to grass pollen ., DPA1*0201 has also been reported to be positively associated with lower levels of rubella-induced antibodies 19 , cytokine immune responses against measles vaccine 20 , and ulcerative colitis 21 , and negatively associated with type 1 diabetes 22 ., DPB1*0901 was shown to be associated with systemic sclerosis 23 , non-permissive mismatches for hematologic stem cell transplantation 24 , ulcerative colitis 21 , and Takayasus arteritis 25 ., HLA-DP molecules present short peptides of largely exogenous origin to CD4-positive helper T cells and other T cells , leading to subsequent immunological responses ., T cells recognize complex formation between a specific HLA type and a particular antigen-derived epitope ., Therefore , HLA molecules capable of binding a particular epitope can restrict T cell induced-immune responses , leading to association between particular HLA types and immune-related diseases ., Type 1 diabetes is a Th-1 type immune disease ., Varney et al . studied 1 , 771 type 1 diabetes multiplex families , analyzing them by the affected family-based control method 26 , and found that DPA1*0201 has a protective effect on the development of type 1 diabetes ( adjusted P\u200a=\u200a5×10−4 , OR 0 . 7 ) 22 ., Epidemiologic studies have associated type 1 diabetes with lower prevalence of asthma and other allergic diseases 26 , 27 ., Also , the previous GWAS of rheumatoid arthritis , other Th-1 type immune disease , has shown that rs987870 C allele confers protection against rheumatoid arthritis 28 ., These findings suggest that HLA-DPA1*0201 could determine Th1/Th2 dominance and could partially explain the inverse relationship between asthma and Th-1 type immune diseases ., Previous GWAS involving European , Mexican , and African populations showed association of asthma with SNPs located in several newly discovered genes ., Our GWAS dataset supported an association between identical SNPs reported in ORMDL3/GSDMB/GSDMA , IL5/RAD50/IL13 , HLA-DR/DQ , and SMAD3 and pediatric asthma ( P<0 . 05 , Table S4 ) ., Two asthma GWA studies revealed an association of HLA-DQ with pediatric/adult asthma in Caucasians 7 , 8 ., HLA-DQ , like HLA-DP , is an αβ heterodimer of the MHC Class II type ., Like HLA-DP , HLA-DQ recognizes and presents foreign antigens , but is also involved in recognizing common self-antigens and presenting those antigens to the immune system ., We failed to replicate the top SNPs of PDE4D , TLE4 , DENND1B , IL18R1 , and IL2RB that were reported in the original articles , but several SNPs in the regions surrounding PDE4D and IL2RB showed significant association when we set the significance level at P\u200a=\u200a0 . 05 ( Table S4 ) ., The different LD patterns/allele frequencies observed in PDE4D and IL2RB in Asians and Caucasians may explain the different SNP associations observed in different ethnic populations ., rs1342326 in IL33 was not polymorphic in the Asian population ., There were several limitations of the present GWAS ., The controls for the GWAS and 1st replication samples were from adult populations ., Information regarding history of asthma in early childhood or other asthma-related information ( i . e . , status of allergic sensitization and lung function ) was not collected for these controls ., Therefore , we cannot exclude the possibility that our control samples may include subjects who outgrew asthma ., The prevalence of pediatric asthma in Japan is around 5%; therefore , our GWAS samples have reduced power compared with that of selected controls ., In the 1st replication Japanese controls , subjects with present and past history of allergic diseases were excluded , and Korean controls in the 2nd replication were non-allergic pediatric controls ( Table S5 ) ., The genomic control value in the present study was 1 . 053 , indicating minor population stratification ., The Japanese population comprises 2 clusters ( Hondo and Ryukyu; Hondo is the mainland of Japan and Ryukyu is the name of the island south of Japan ) ., We performed principal component analysis using EIGENSTRAT software 29 to identify subjects belonging to Ryukyu ., Because 2nd or 3rd generation Chinese live in Japan , and the genetic population structure in Chinese differs from that in Japanese , we also performed principal component analysis to exclude Chinese subjects ., Although hidden population stratification may exist , its influence on the final results is not expected to be significant ., rs3019885 is located in intron 2 of solute carrier family 30 ( SLC30A8 ) , and showed strong association in the GWAS population ., The association was replicated in the independent Japanese samples , but not in the Korean population ., SLC30A8 is a zinc efflux transporter expressed at high levels only in the pancreas; the GWAS revealed that variants of SLC30A8 are associated with type 2 diabetes 30 ., Japanese and Koreans are genetically close but we cannot exclude the possibility that the association of rs3019885 with pediatric asthma is population specific ., In conclusion , we performed the first GWAS in Asian population for pediatric asthma and found that DPA*0201/DPB1*0901 is strongly associated with pediatric asthma ., The association with the HLA-DP locus emphasizes the importance of the HLA-class II molecules on the biological pathways involved in the etiology of pediatric asthma , and suggests that HLA-DP can be a therapeutic target for asthma ., The study was approved by the institutional review board and the ethics committee of each institution ., Written informed consent was obtained from each participant in accordance with institutional requirements and the Declaration of Helsinki Principles ., Characteristics of pediatric asthma cases and controls are summarized in Table S5 ., Genotyping for GWAS was performed using the Illumina HumanHap550v3/610-Quad Genotyping BeadChip ( Illumina ) , as per manufacturers instruction ., In replication analyses , genotyping of each individual was performed with the TaqMan genotyping system ( Applied Biosystems ) on an ABI PRISM 7900HT Sequence Detection System ( Applied Biosystems ) ., PCR was performed on a 384-well format , and automatic allele calling was performed using ABI PRISM 7900HT data collection and analysis software , version 2 . 2 . 2 ( Applied Biosystems ) ., HLA-DPB1 genotyping of 1135 cases , 794 controls ( control, 1 ) and 1475 controls ( control, 2 ) were performed with the WAKFlow HLA typing kit ( Wakunaga , Hiroshima , Japan ) , as per manufacturers instruction ., First , the target DNA was amplified by polymerase chain reaction ( PCR ) with biotinylated primers specifically designed for each HLA-DPB1 locus ., Then , the PCR product was denatured and hybridized to complementary oligonucleotide probes immobilized on fluorescent-coded microsphere beads ., Concurrently , the biotinylated PCR product was labeled with phycoerythrin-conjugated streptavidin and immediately examined with the Luminex 100 system ( Luminex , Austin , TX ) ., Genotype determination and data analysis were performed with the WAKFlow typing software ( Wakunaga ) ., HLA-DPA1 genotyping was performed with direct sequencing of exon 2 with forward primer 5′-TCAGGATGCCCAGACTTTCAA-3′ and reverse primer 5′-CAGGGGGCACTTAGGCTTCC-3′ , and with the sequencing primer 5′-TCAGGATGCCCAGACTTTCAA-3′ using the BigDye Terminator v . 1 . 1 Cycle Sequencing Kit ( Applied Biosystems ) on an ABI PRISM 3130 Genetic Analyzer ( Applied Biosystems ) ., In the GWAS , we examined the potential genetic relatedness on the basis of pairwise identity by state for all of the successfully genotyped samples using the EIGENSTRAT software 29 ., In the GWAS , we genotyped 978 cases with pediatric asthma and 2402 controls using Illumina HumanHap550v3/610-Quad Genotyping BeadChip ( Illumina , San Diego , USA ) ., Samples of duplicated ( identical individual or monozygotic twin ) , first- , second- , and third-degree pairs were detected , and the individual with a lower call rate was excluded from further analysis ., PCA was performed , and the results were combined with those obtained for our in-house Ryukyu and Han Chinese reference samples ., Yamaguchi-Kabata et al . characterized the Japanese population structure using the genotypes for 140 , 387 SNPs in 7003 Japanese individuals , along with 60 European , 60 African , and 90 East-Asian individuals , in the HapMap project and found that the Japanese population is composed of 2 clusters ( Hondo and Ryukyu ) 36 ., Hondo is the biggest island of Japan , and the island of Ryukyu is located in southern Japan ., Also , we have 2nd or 3rd generation Chinese living in Japan , and Chinese present a different genetic population structure from Japanese ., Therefore , we excluded samples belonging to Han Chinese or Ryukyu , and 938 cases and 2376 controls were considered for further analysis ., Cluster plots of SNPs were checked by visual inspection and SNPs with ambiguous calls were excluded ., We excluded SNPs with a low genotyping rate ( <90% ) , minor allele frequency less than 0 . 01 in either pediatric asthma cases or controls , or with Hardy-Weinberg equilibrium P value<10−4 in controls ., Finally , 450 , 326 SNPs were used for the GWAS ., Details regarding the exact number of remaining SNPs after applying each quality control criterion are available in Table S1 ., The genomic control inflation factor ( λGC ) , defined as the median association test statistic across all SNPs divided by its expected value , was calculated by the method proposed by Devlin et al . 37 ., GWAS and replication analyses were performed using the Cochran–Armitage trend test and χ2 test ., The meta-analysis was performed with the Mantel–Haenszel approach as a fixed-effects model 38 ., All statistical findings were reported without correction ., The results of GWAS were plotted with GWAS GUI v0 . 0 . 2 39 ., HLA-DP region was plotted with LocusZoom 40 ., The power calculation was performed with Genetic Power Calculator 41 ., Quantile-quantile ( Q-Q ) plot was plotted with ggplot2 package 42 in R version 2 . 10 . 0 ( http://www . r-project . org/ ) ., Imputation of genotypes in the DP region was performed with MACH version 1 . 0 9 with 1000 Genome Project data ( 1000G 2010-6 release , http://www . sph . umich . edu/csg/yli/mach/download/1000G-2010-06 . html ) ., The HLA-DP region was in strong linkage disequilibrium and some DPB1 alleles were known to be linked with particular DPA1 alleles ., First , we imputed HLA-DPA1 alleles by using the actual genotype data of samples obtained from Illumina HumanHap550v3/610-Quad ( Illumina ) and 1000 Genome Project data of Asian origin ( JPT+CHB ) ( http://www . sph . umich . edu/csg/abecasis/MaCH/download/1000G-2010-06 . html ) ., The accuracy of the imputated data was confirmed by direct sequencing ., The error rate of imputation was 1/352 ( 0 . 003 ) . | Introduction, Results, Discussion, Materials and Methods | Asthma is a complex phenotype influenced by genetic and environmental factors ., We conducted a genome-wide association study ( GWAS ) with 938 Japanese pediatric asthma patients and 2 , 376 controls ., Single-nucleotide polymorphisms ( SNPs ) showing strong associations ( P<1×10−8 ) in GWAS were further genotyped in an independent Japanese samples ( 818 cases and 1 , 032 controls ) and in Korean samples ( 835 cases and 421 controls ) ., SNP rs987870 , located between HLA-DPA1 and HLA-DPB1 , was consistently associated with pediatric asthma in 3 independent populations ( Pcombined\u200a=\u200a2 . 3×10−10 , odds ratio OR\u200a=\u200a1 . 40 ) ., HLA-DP allele analysis showed that DPA1*0201 and DPB1*0901 , which were in strong linkage disequilibrium , were strongly associated with pediatric asthma ( DPA1*0201: P\u200a=\u200a5 . 5×10−10 , OR\u200a=\u200a1 . 52 , and DPB1*0901: P\u200a=\u200a2 . 0×10−7 , OR\u200a=\u200a1 . 49 ) ., Our findings show that genetic variants in the HLA-DP locus are associated with the risk of pediatric asthma in Asian populations . | Asthma is the most common chronic disorder in children , and asthma exacerbation is an important cause of childhood morbidity and hospitalization ., Here , taking advantage of recent technological advances in human genetics , we performed a genome-wide association study and follow-up validation studies to identify genetic variants for asthma ., By examining 6 , 428 Asians , we found rs987870 and HLA-DPA1*0201/DPB1*0901 were associated with pediatric asthma ., The association signal was stretched in the region of HLA-DPB2 , collagen , type XI , alpha 2 ( COL11A2 ) , and Retinoid X receptor beta ( RXRB ) , but strong linkage disequilibrium in this region made it difficult to specifically identify causative variants ., Interestingly , the SNP ( or the HLA-DP allele ) associated with pediatric asthma ( Th-2 type immune diseases ) in the present study confers protection against Th-1 type immune diseases , such as type 1 diabetes and rheumatoid arthritis ., Therefore , the association results obtained in the present study could partially explain the inverse relationship between asthma and Th-1 type immune diseases and may lead to better understanding of Th-1/Th-2 immune diseases . | medicine, genetics of the immune system, clinical immunology, genetics, genetics and genomics, biology, human genetics, immunology, respiratory medicine, pulmonology, asthma | null |
journal.pgen.1003671 | 2,013 | Integrated Model of De Novo and Inherited Genetic Variants Yields Greater Power to Identify Risk Genes | The genetic architecture of autism spectrum disorders ( ASD ) is complex and thought to involve the action of at least hundreds of genes ., Yet , despite this complexity , four recent studies 1–4 identified five novel genes affecting the risk for ASD from whole-exome sequencing ( WES ) of 932 ASD probands ., The studies made these discoveries by also sequencing the parents of the probands and thereby discovering a multiplicity of independent Loss-of-Function ( LoF ) mutations in each of these five genes ., The multiplicity is key: due to the rarity of de novo LoF events , two or more independent recurrent events in a sample of this size generate more evidence for association than would two LoF variants found in a comparable case and control sample ., Thus , even though de novo events are rare , these observations provide an excellent signal-to-noise ratio , have proven valuable in the pursuit of reliable signals for genes affecting the ASD risk , and are likely to form the foundation for many studies targeting gene discovery in the future 5 ., Note , however , that the multiplicity test is using only a small fraction of all the information collected by a WES study ., Many other de novo events occur , beyond LoF , and these are ignored ., Moreover it completely ignores inherited rare variants within families ., And , of course , delineation of rare variants into inherited and de novo is challenging or impossible for case-control studies ., We conjecture that the distribution of variation , whether inherited , de novo and from case-control , can be leveraged , in combination with the de novo mutations , to maximize the statistical power to detect risk genes ., We propose an integrated model of de novo mutations and transmitted variation to address these challenges ., We demonstrate that both the number of de novo mutations and the numbers of different types of transmitted variations in family trios ( father , mother and an affected child ) , follow simple distributions dependent on a set of common parameters: mutation rates , relative risks of mutations and population frequency of the variants ., This model readily incorporates additional data from case-control studies ., The statistical framework of our model enables us to rigorously analyze the genetic architecture of a complex disease , conduct power and sample size analysis , and identify risk genes with higher sensitivity ., Through simulations we show that the power of our novel statistical test , called TADA for “transmission and de novo association” , is substantially higher than competing tests ., Our simulations also provide guidance in planning future studies targeting discovery of genes involved in the risks of complex diseases , henceforth , risk genes ., We demonstrate the benefits of TADA through an extensive study of ASD using published WES data from 932 ASD trios as well as nearly 1000 ASD subjects and matched control subjects from the ARRA Autism Sequencing Consortium ( AASC ) study 6 , 7 ., Using the model underlying TADA , we estimate there are approximately 1000 genes that play a role in risk for ASD , with an average relative risk of approximately 20 due to LoF in one of these genes ., Finally , we identify several potential novel ASD risk genes ( genes whose mutations affect the risk of ASD ) using TADA and the ASD data ., For concreteness we start by reviewing the multiplicity test to detect risk genes by evaluating the independent recurrence of de novo mutations in the same gene ., The multiplicity test classifies a gene as affecting risk if it sustains or more recurrent de novo LoF mutations in a sample of families ., Based on computations of expected rates of de novo events as a function of a genes exonic length and base pair composition 2 , a recent study 1 found that LoF events for is significant evidence to declare a gene as a risk gene ( , genomewide ) ., Applying this threshold to data from four ASD family studies 1–4 led to the discovery of five novel genes affecting ASD risk ., A weakness of the multiplicity test is that it produces a single threshold for the entire genome , regardless of the heterogeneity amongst genes in their sizes and base pair composition , and its threshold is a function of sample size , so that the threshold for is inadequate when the sample increases to ., To illustrate the power of the Multiplicity Test and its properties , we performed some simulations using genetic parameters that are described and estimated in the next section ., As demonstrated previously 1 , the power for detecting a gene increases monotonically with increasing sample size and it depends strongly on the genes mutation rate ( Figure 1A ) ., Although the per gene power is relatively low , for a disorder like ASD , more than 60 genes are expected to contain at least two LoF mutations with families ( Figure 1B ) ., The corresponding false discovery rate ( FDR ) is less than 5% for and well below 10% for as large as 5 , 000; switching to a threshold of to diminish false discoveries leads to a significant loss in power ( Figure 1B ) ., The original treatment of the multiplicity test as requiring a single threshold is simple to adjust ., Instead one can compute the p-value for each gene using a Poisson model for the probability of observing or more recurrent de novo events based on the genes mutation rate ., We will call such a test the De Novo Test ., This test automatically incorporates the number of families and a gene specific mutation rate to determine the likelihood of recurrent de novo events ., TADA model is formulated for sequence data from individual genes ., Data for the model can come from sequences of trios ( unaffected parents and an affected child ) and from cases and controls ., Given the information from a gene , namely the pattern of de novo mutations and inherited , damaging variants in the affected progeny , the goal is to relate the data with the underlying genetic parameters such as the relative risk of the mutations ., In the model , we restrict the class of variation to rare and deleterious mutations acting dominantly and assume subjects can be classified as carrying one of two “alleles” , those with a deleterious mutation of this type ( ) and those without ( ) ., We put alleles in quotes because , for example , we treat all LoF events in the same gene as a single LoF “allele” ., Because severe mutations are generally present at very low frequencies in the population ( typically ) , there are effectively two possible genotypes per gene , and ., If we let denote the allele frequency of , then the frequencies of the genotypes and in the population are approximately and , respectively ., For a trio consisting of unaffected parents and an affected child , there are four likely genotype combinations ( Figure 2 ) , of which only three are informative: if both parents are homozygous , a heterozygous child results from a de novo mutation; and if one parent is heterozygous , the allele is either transmitted or not ., Based on the de novo and transmitted alleles , we formulate a likelihood model for the observed data ., Let denote the rate of mutation for the gene being analyzed per generation and chromosome; let denote the genotype relative risk for the genotype ; and let and denote the penetrance of and , respectively ., Let , and be the counts of each of the three outcomes ( de novo , transmitted and nontransmitted , respectively ) , from a sample consisting of families ., These counts approximately follow Poisson distributions ( see Text S1 for derivation ) : , , and ., For case-control data , counts of genotype in cases and controls follow a Poisson distribution with approximate rate parameters and , respectively ( see Text S1 ) ., From this structure it is apparent that the transmitted counts can be viewed as a type of case-control data with sample size ., Combining data , let be the total number of in the controls plus the number of transmitted variants , and let be the total number of in the cases plus the number of transmitted variants ., It follows that ( 1 ) for which and ., The resulting probability model has three parameters ( ) per gene ., For each gene , the mutation rate per gene ( ) can be estimated from its exonic length and nucleotide content 1 and hence this quantity can be treated as known ., The statistical problem for each gene is to estimate and then test if ., We conjecture that a more powerful strategy to discover risk genes from family data is to combine the information on de novo and inherited mutations into an unified statistical framework , such as the one we just proposed , which forms the basis for TADA ., TADA tests the hypothesis against the alternative ., A traditional likelihood ratio test will not work well in this setting because one or more of the counts will be zero for many genes , leading to poor maximum likelihood estimates for and ., To circumvent this problem we cast TADA in a Hierarchical Bayes ( HB ) framework , thereby improving estimates of and by pooling information across all genes , but still modeling rates as gene-specific ., The underlying assumption is that LoF and severe missense mutations are rare in all genes and hence we can learn about the frequency distribution in a given gene by looking at the distribution across all genes ., Likewise , we can learn about how mutations in one gene affect risk by examining the range and distribution of risks across all disease-related genes ., The HB model assumes a fraction of the genes are associated with the disorder ( model ) ; the remaining fraction follow the null model ( model ) ., Under , the relative risk is constrained ( ) , but under , is assumed to follow a distribution across risk genes ., For both models , the frequency of severe mutations per gene , , is assumed to vary by gene , with some commonality across the genome ., The distributions of and under both models are specified by prior parameters , and we estimate the values of these parameters by maximizing the marginal likelihood of the data ( this is known as the Empirical Bayes method , see Methods ) ., Once the prior parameters are estimated , we compute the evidence for and for each gene ., Specifically , for the i-th gene , let be its data , the evidence for is defined as: ( 2 ) where is given by Equation 1 , and represent the prior distributions ., Unlike the likelihood-based test , the evidence for is not based on point estimates of and ; instead it integrates out the two parameters ., The model evidence of can be defined similarly , except that is fixed at 1 ., The Bayes factor of any gene is the ratio of to ., The statistical significance of the Bayes factor is given by its p-value , determined empirically by simulating data under the model assuming ( see Text S1 ) ., Some insights into the relationship to a likelihood-ratio test ( LRT ) can be gained by examining an approximation of , the Bayes factor: ( 3 ) where the parameters are estimated by Bayesian mean posterior estimators ., These parameter estimates are a weighted average of the maximum-likelihood estimate for the i-th gene and the mean of the prior distributions ., For example , is interpolated between the allele frequency derived from all genes and the gene-specific estimate ( Figure S1 ) ., Thus the Bayes factor is similar to the LRT except that we utilize a refined estimator of the allele frequency ., The model just described is designed for a single type of mutation ( say LoF ) , but it can incorporate multiple types ., For different types of mutations , such as LoF and damaging missense mutations , the distributions of and are likely to be different , so we model each type of mutation and estimate the prior parameters separately using the HB framework ., Then the total Bayes factor of a gene is the product of the Bayes factor from each type of mutation , and the p-value can be computed similarly from simulations ., In practice , we note that the damaging missense mutations predicted by bioinformatic tools likely contain a number of mutations having no effect on the gene function , thus we introduce an additional model to account for this feature , downweighting the evidence from missense mutations ( see Methods ) ., The TADA method we described can also be used for de novo data alone ., Basically , we ignore inherited and standing variants , but allow multiple types of de novo mutations ., The details are not repeated here , but are provided in our supporting Website ( see Methods ) ., We call this simplified model , TADA-Denovo , and it is particularly useful for genes with multiple de novo events in different categories ( e . g . some nonsense and some missense mutations ) ., We use the proposed model to estimate the number of ASD risk genes ( ) , their average relative risk ( ) , and the distribution of the population frequency of the mutations ., These estimates yield insight into the genetics of ASD and pave the way for realistic simulations to study the power of statistical tests ., Our overall strategy is first to use de novo mutations to estimate an approximate range of the parameter values , then use the HB method to refine these estimates using both family and the case-control data ., Consider the de novo LoF mutations in families 1–4 ., These data reveal a total of de novo LoF mutations across all genes , and multiple-hit genes ( at least 2 independent de novo LoF events per gene ) ., Our goal is to find values of and that best predict the observed counts and ( Text S1 ) ., We assume that the relative risk of an ASD risk gene varies across genes , with the average relative risk of the LoF mutations equal to ., The mathematics of TADA reveal there is an inverse relationship between and ( Figure 3A , see Equation 27 in Text S1 ) ., For an alternative and more intuitive explanation of why these parameters have an inverse relationship , see the arguments in 2 ., For any given value of , we can compute the expected number of multiple-hit genes; matching the expected with the observed value of , we estimate the the number of ASD risk genes is between 550 to 1000 ( Figure 3B ) ., In the next step , we use the HB model to estimate the most likely value of within this range , and the result is ASD risk genes , with the corresponding relative risk ( see Text S1 ) ., These estimates are similar to published results using somewhat different methods 1 , 2 ., We examine evidence for the hypothesis that the population frequency of LoF mutations for ASD risk genes ( ) is lower than that for non-risk genes ( ) because mutations in ASD risk genes are under stronger negative selection than the average gene ., These frequencies are of interest because they have a major influence on the power of association test 8 ., We estimate based on the number of LoF variants in the case-control data from the AASC 7 and the transmitted/nontransmitted data from 641 families ( the transmission data are only available for a subset of the 932 families ) ., To obtain the empirical distribution of across all genes we first count the frequency of the LoF mutations in each gene ( Figure 3C ) ; we find a substantial number of genes with 0 LoFs ., We next estimate the prior distributions of under the null and alternative models , respectively , using the HB model and find they provide a good fit to the observed data ( Figure 3C , Figure S1 ) ., From these analyses the mean of under , i . e . the average for ASD risk genes , is about , significantly smaller than that of non-risk genes , ( see Text S1 for a description of how the HB model uses a mixture model to permit estimation of parameters specific to ASD risk genes without actually classifying genes as such . ) Notably , while the empirical estimate of for most genes is 0 ( thus not useful for inference ) , the value of from the HB model is never equal to 0 due to smoothing ., Using the same procedures we also estimated these parameters for missense mutations that are probably damaging according to the PolyPhen prediction 9 ( denoted as Mis3 mutations ) ., Estimates reveal lower risk for these mutations , as expected , and lower for ASD risk genes compared with non-ASD genes ( Table S1 ) ., Equipped with estimates of the genetic parameters , we can simulate genetic data under the model and assess the performance of statistical methods ., We compare performance of three tests: De Novo , as described in Section 2 . 1; TADA , described in Section 2 . 3; and a “Meta test” , which combines two tests , one based on de novo events and the other on inherited variants , via meta analysis ., For the meta test we compute the p-value from data on inherited variants using a Fisher exact test , treating transmitted/untransmitted events as case-control data; and compute a p-value for de novo events using the De Novo test ., Then these p-values are combined using Fishers method ., In all the simulations , different parameters are used to generate the data , yet TADA always uses the same set of parameters derived from the real data , as described previously ., Thus these results establish the robustness of TADA under different parameter settings and thus , to some extent , how it should behave for real data ., Because TADA is a novel method , data were first simulated under the null hypothesis of no association to obtain the distribution of the TADA test statistic and its associated p-values ., The results show that the test is well calibrated and type I error is properly controlled ( Figure S2 ) ., Next , data were simulated under the alternative model , using different sample sizes and different combinations of the parameters and , within the range of plausible values estimated in the previous section ., This comprehensive simulation showed TADA has superior power relative to the other two tests ( Figure S3 ) ., In Figure 4 , we show a selected portion of the simulation results under the most likely scenarios , reflecting the trade-off between relative risks and allele frequencies , i . e . mutations with high risks are likely to exist in lower frequencies in the population ., For a gene with typical parameter values ( Figure 4B ) , the power of the TADA test , at , was about fivefold larger than that of the other two tests ., To assess the performance of the tests from a genome-wide analysis , we generated realistic simulated counts based on the estimated genetic parameters for ASD , namely average relative risk of 20 and risk genes , among a total of 18 , 000 genes sequenced ., We focus on false discovery rate ( FDR ) , calibrating the empirical FDR to control at 10% , and estimated power as the number of true discoveries ., Results confirmed the advantage of TADA ( Figure S4A ) ., For example , at , TADA identified more than 200 ASD risk genes at FDR below 10% , while the De Novo and Meta tests identify about 50 and 70 genes at this level of FDR , respectively ( cf Figure 1 ) ., We performed additional simulations with somewhat different procedures to demonstrate the robustness of these findings ., In one experiment , we simulated data under the average relative risk of 10 , instead of 20 , while TADA still uses the relative risk of 20 ., The power of all methods was significantly reduced , as expected , yet TADA still performed better than both de novo test and the simple meta-analysis ( Figure S4B ) ., In another experiment , the simulation procedure incorporated the possible dependency between the LoF frequency of a gene ( ) and its relative risk ( ) , based on simple mutation-selection balance: the two were not sampled independently , but rather the frequency was inversely proportional to the risk ( see Methods ) ., Despite this change of simulation model , the results were virtually identical to those from earlier simulations ( Figure S4C ) ., The data we used were all reported de novo mutations from 932 ASD families 1–4; transmitted mutations from 641 of these families; and case-control data from the AASC , consisting of 935 ASD subjects and 870 controls 7 ., Each missense mutation was classified into a category of damage to the protein based on its predicted effect on the coding sequence using PolyPhen2 9: benign ( Mis1 ) ; possibly damaging ( Mis2 ) ; and probably damaging ( Mis3 ) ., Note that de novo LoF mutations occurred at about two-fold enriched rate in the probands relative to the unaffected siblings ( Figure 5A , Table S2 ) ., The rate for de novo Mis3 was also higher in probands than siblings , but the difference was not as striking ., There is essentially no difference in probands and siblings for other types of mutations ., We thus applied the TADA method to the LoF and Mis3 mutations ., The overall inflation of the results due to population stratification is negligible: a modified 7 genomic control factor 10 ( see Text S1 ) ., There is significant enrichment of genes with low p-values compared with random expectation ( Figure 5B ) : 244 genes have , 64 more than expected under the null model ., There is an intriguing coincidence in the excess of small p-values - namely that it is very similar to the excess number of genes with single-hit de novo LoF events in ASD subjects compared to their unaffected siblings 1 ., Notably the large tail in the QQ plot is largely driven by the de novo LoF events , and appears to reflect true signal instead of inflation ., We control for the multiple hypothesis testing using the Benjamini-Hochberg procedure 11 ., Fifteen genes meet the criteria of a False Discovery Rate less than 20% ( Table 1 , see Table S3 for the complete results ) ., The list includes all five genes with two de novo LoF mutations , as well as several novel genes that are promising candidates for ASD based on existing evidence ., For the novel predictions , the p-values from the de novo data alone are far from achieving genome-wide significance ( the column in Table 1 ) and would be impossible to identify without combining the de novo , transmitted and case-control data ., The results of TADA generally depend on the estimates of the mutation rates of the genes , as well as the Bayesian prior parameters of the model ., We perform additional analyses to study how sensitive the results are to these parameters ., Based on our findings , we choose several genes from Table 1 for this investigation ., Although the error of mutation rate estimation is likely small 1 , we vary the mutation rate of each gene: from 1/2 of the estimated rate to twice the rate ., As expected , the p-value increases as the mutation rate increases , although overall the impact is modest ( Figure S5A ) ., Next we vary the Bayesian prior parameter , , which represents the average relative risk over all risk genes , from 10 to 20 ., The p-values from TADA are even less sensitive to this parameter ( Figure S5B ) ., For disorders like ASD , recent results show that detection of de novo LoF events can be a powerful means of discovering novel risk genes 1–4 ., Yet de novo events are relatively rare , roughly one per exome , and de novo LoF events even more so , and thus many families must be assessed to identify multiple de novo LoF events in the same gene ., To make the most of this experimental design , we develop a new statistical approach , TADA , that utilizes both transmitted and de novo variants from nuclear families and case-control data to determine genetic association ., TADA builds on the simple multiplicity test , which relies on recurrent de novo events , but it creates a full analytical framework to incorporate all of the information on the distribution of rare variation ., The result is a test with greater power ., Our test achieves its good performance properties by providing an analytic framework that links the observed pattern of de novo mutations with the underlying genetic parameters , such as the relative risk conveyed by such mutations ., In addition to analyzing data for novel gene discovery , this framework can be used to analyze the power of a test and predict the required sample size to attain sufficient power for future investigations ., Moreover , by using empirical Bayes methods , TADA refines estimates of allele frequencies of the damaging mutations by using the full genome to estimate these quantities ., This approach increases the information in the transmitted variants in each gene considerably and yet maintains good control of false discoveries ., Association studies evaluating cases and controls have been a common design for identifying variation affecting risk for complex diseases ., It has proven successful for identifying common variation affecting risk , after sufficient samples had been amassed to ensure variation having modest impact on risk could be detected 12 ., Common variants surely play a role in ASD 13 , 14 , but the effect sizes are small 15 and it will be challenging to detect individually-significant SNPs ., Indeed virtually every discovery for ASD risk genes traces to rare and de novo variants 1–4 , 16–20 ., As the cost of sequencing drops , genetic research increasingly focused on the role of rare variants in complex diseases such as ASD , but the sample size has been limited and so has the yield of such studies ., For a sample of nearly 1000 ASD case and well matched controls the ARRA ASD sequencing consortium ( AASC ) found no significant associations 7 , except for variation acting recessively 6 ., These results comport with studies of other disorders and suggest that large sample sizes will be required to achieve good power in rare variant association studies 21 ., Arguably a fundamental difficulty is that most of the mutations with large effects tend to be under strong negative selection , existing at very low frequencies in the population 22 ., Variants that occur with greater frequency often have smaller effect on the phenotype , reducing the power of gene-based test statistics ., Our analysis provides insight into some advantages of de novo over case-control studies , especially for LoF events ., The de novo test gains power because the mutation rate for genes can be estimated accurately from supplementary sources , and need not be estimated as part of the statistical procedure ., Because of the low mutation rate , the number of de novo LoF events expected by chance is very small , and thus we could attach high statistical significance to any gene with more than one independent LoF mutation ., While a single de novo LoF event is certainly not definitive evidence , it can put a gene on the short list as a risk gene – for ASD , it is more likely than not an ASD risk gene ., In contrast , for case-control data , we require an estimate of the allele frequency under the null hypothesis ., When the mutant allele is very rare ( as for ASD risk genes ) , a very large sample is required to ensure that this frequency is indeed small ., Another feature of observed de novo mutations is that they have not been subject to the force of purifying selection , which plays a key role in shaping the pattern of standing variation ., Therefore it is likely that de novo mutations , especially LoF mutations , have stochastically larger effect sizes than rare variation transmitted for generations , because selection tends to drive down allele frequencies of variants having large effects on reproductive success ., Moreover , allele frequency is inversely tied to power , critical for any experimental design ., Therefore studies utilizing de novo variation can have distinct advantages , in terms of power , relative to those that do not ., By simulations we demonstrate that the power of TADA is higher than tests based solely on de novo events or standard meta-analysis that combines p-values from de novo and inherited data ( transmission or case/control ) ., There are two explanations for this gain of power ., First , TADAs hierarchical model uses the information in the case-control ( or transmission ) data more efficiently than the standard hypergeometric or trend test ., One important property of LoF mutations , compared to less severe functional variants , is their rarity in the population ( Figure 3C ) ., TADA , which is similar in spirit to a Poisson test of rare events , is able to exploit the rarity of these damaging events by estimating the distribution of LoF alleles across the exome ( see Figure S1B ) , whereas the other methods cannot ., Second , because damaging de novo mutations are rare , most genes will not harbor them even when thousands of cases have been sequenced ., For such genes , using Fishers method to combine the de novo p-value , which will be close to 1 , with the p-value from the case-control data penalizes the overall test statistic ., In contrast , the Bayesian approach uses de novo events when they are informative and disregards the de novo data when they are uninformative; the Bayes factor from de novo in such cases would be close to 1 , making little contribution to the genes total Bayes factor ., We estimate that there are about 1 , 000 ASD risk genes with average relative risk about 20 ., In a recent paper using the same de novo data , the number of ASD risk genes ( ) was estimated at 370 4 ., In that paper , the expected number of genes with recurrent LoF events was derived as a function of , and equating it to 5 ( the observed number ) , produced the solution that ., The analysis made the implicit assumption that all ASD risk genes are equally likely to sustain multiple de novo LoF events ., In Text S1 we show , using Jensens Inequality , that the non-uniform distribution of the mutation rates and the relative risks among the ASD risk genes leads to a significant under-estimation of , explaining the discrepancy between our results and those of Iossifov et al . 4 ., When applied to ASD data , TADA predicts a number of novel ASD risk genes ( Table 1 ) , as well as supporting results for known ASD risk genes ., For some of the newly implicated genes it is straightforward to garner other supporting evidence for their role in ASD ., S100G is a downstream target of CHD8 , a key transcriptional regulator often disrupted in ASD subjects 23 ., CUL3 plays a critical role in neurodevelopment 24 , 25 and in particular regulates synaptic functions 26 ., A recent study identified an additional de novo protein-changing mutation in CUL3 in ASD probands 27 , replicating our finding here ., COL25A1 , a brain-specific collagen , was implicated in risks for Alzheimers disease 28 and antisocial personality disorder 29 ., Inspection of other genes slightly below our chosen FDR threshold reveals several more interesting genes that likely play some role in ASD ( all ranked among the top 25 , see Table S3 ) ., TBR1 , a transcription factor critical in brain development , regulates several known ASD risk genes 30 ., A recent study has identified recurrent de novo disruptive mutations in TBR1 in ASD subjects 23 ., MED13L , a component of the Mediator Complex , is intriguing because of its role in Rb/E2F control of cell growth 31 and the fact that RB/E2F plays a key role in neurogenesis 32 and neuronal migration 33 ., Recently MED13L has been associated with risk for schizophrenia 34 ., NFIA is a member of the NFI transcription factor family , thought to have a neuroprotective role 35 , and NFIA-knockout mice display profound defects in brain development 36 ., Genotyping/sequencing errors can introduce biases in data analyses , especially those for family data 37 , 38 and for combining data across multiple heterogeneous studies 39 ., Our analyses are likely robust to these possible biases because the variant calls were all carefully evaluated:, ( i ) all de novo mutations described previously 1–4 and analyzed here , a total of 122 LoF and 314 damaging missense mutations , have been validated by previous studies;, ( ii ) the case-control data have been carefully harmonized to minimize batch effects by using stringent quality control filters 7; and, ( iii ) for the case-control data , all variant calls in two genes ( CHD8 and SCN2A ) have been evaluated by Sanger sequencing and 20 out of 20 validate , further supporting the quality of the variant ca | Introduction, Results, Discussion, Methods | De novo mutations affect risk for many diseases and disorders , especially those with early-onset ., An example is autism spectrum disorders ( ASD ) ., Four recent whole-exome sequencing ( WES ) studies of ASD families revealed a handful of novel risk genes , based on independent de novo loss-of-function ( LoF ) mutations falling in the same gene , and found that de novo LoF mutations occurred at a twofold higher rate than expected by chance ., However successful these studies were , they used only a small fraction of the data , excluding other types of de novo mutations and inherited rare variants ., Moreover , such analyses cannot readily incorporate data from case-control studies ., An important research challenge in gene discovery , therefore , is to develop statistical methods that accommodate a broader class of rare variation ., We develop methods that can incorporate WES data regarding de novo mutations , inherited variants present , and variants identified within cases and controls ., TADA , for Transmission And De novo Association , integrates these data by a gene-based likelihood model involving parameters for allele frequencies and gene-specific penetrances ., Inference is based on a Hierarchical Bayes strategy that borrows information across all genes to infer parameters that would be difficult to estimate for individual genes ., In addition to theoretical development we validated TADA using realistic simulations mimicking rare , large-effect mutations affecting risk for ASD and show it has dramatically better power than other common methods of analysis ., Thus TADAs integration of various kinds of WES data can be a highly effective means of identifying novel risk genes ., Indeed , application of TADA to WES data from subjects with ASD and their families , as well as from a study of ASD subjects and controls , revealed several novel and promising ASD candidate genes with strong statistical support . | The genetic underpinnings of autism spectrum disorder ( ASD ) have proven difficult to determine , despite a wealth of evidence for genetic causes and ongoing effort to identify genes ., Recently investigators sequenced the coding regions of the genomes from ASD children along with their unaffected parents ( ASD trios ) and identified numerous new candidate genes by pinpointing spontaneously occurring ( de novo ) mutations in the affected offspring ., A gene with a severe ( de novo ) mutation observed in more than one individual is immediately implicated in ASD; however , the majority of severe mutations are observed only once per gene ., These genes create a short list of candidates , and our results suggest about 50% are true risk genes ., To strengthen our inferences , we develop a novel statistical method ( TADA ) that utilizes inherited variation transmitted to affected offspring in conjunction with ( de novo ) mutations to identify risk genes ., Through simulations we show that TADA dramatically increases power ., We apply this approach to nearly 1000 ASD trios and 2000 subjects from a case-control study and identify several promising genes ., Through simulations and application we show that TADAs integration of sequencing data can be a highly effective means of identifying risk genes . | medicine, disease mapping, neuropsychiatric disorders, mathematics, mental health, child psychiatry, epidemiology, statistics, genetics, biology, human genetics, genetics of disease, statistical methods, psychiatry | null |
journal.pgen.1007999 | 2,019 | Transcription-dependent spreading of the Dal80 yeast GATA factor across the body of highly expressed genes | In eukaryotes , gene transcription by RNA polymerase II ( Pol II ) is initiated by the binding of specific transcription factors to double-stranded DNA ., The yeast transcription factors target regulatory regions called UAS or URS ( for Upstream Activating/Repressing Sequences ) , generally directly adjacent to the core promoter ., The generated regulatory signals converge at the core promoter where they permit the regulation of Pol II recruitment via the ‘TATA box-binding protein’ and associated general transcription factors 1 , 2 ., The transcription factor binding sites are usually short sequences ranging from 8 to 20 bp 3 ., They are most often similar but generally not identical , differing by some nucleotides from one another 3 , making it sometimes difficult to predict whether a given UAS will function as such in vivo ., GATA factors constitute a family of transcription factors highly conserved among eukaryotes and characterized by the presence of one or two DNA binding domains which consists of four cysteines ( fitting the consensus sequence CX2CX17-18CX2C ) coordinating a zinc ion followed by a basic carboxy-terminal tail 4 ., While vertebrate GATA factors possess two adjacent homologous zinc fingers , fungal ones contain only one single zinc finger , being most closely related to the C-terminal vertebrate zinc finger 5 , 6 , which is the one responsible for determining the binding specificity of GATA-1 , the founding member of the GATA factor family 7 ., The specificity of GATA factor binding has been thoroughly characterized in yeast 8–10 and metazoans 11–18 ., In addition , structure determinations of protein-DNA complexes , first for GATA-1 4 , then for its fungal orthologue AreA 19 , allowed for the identification of the subtle determinants of DNA specificity for GATA factors ., Notably , the conserved DNA binding domain of GATA factors was reported to bind to consensus sequences ( corresponding to GATAA ( G ) or GATTAG for the yeast GATA factors described hereafter ) , as shown in various organisms using direct or indirect methods 4 , 19–22 ., These consensus sequences are accordingly referred to as GATA motifs ., Since its discovery 40 years ago in chicken cells , the family of GATA factors was extended in human cells and represents master regulators of hematopoiesis and cancer 23 ., However , although approximately 7 million GATA motifs can be found in the human genome , the GATA factors occupy only 0 . 1–1% of them ., Conversely , other regions are occupied by GATA factors despite lacking the consensus motif 24 , 25 ., Consistently , even if most GATA factors bind to core GATA sequences , peculiar specificities have been reported for the flanking bases as well as for the fourth base of the GATA core element 26–29 ., These studies revealed an elevated flexibility in the recognition sites for vertebrate and fungal GATA factors , much greater than previously anticipated , making the search for GATA sites and their enrichment in GATA-regulated genes tedious and unproductive ., In addition , GATA factors can swap among them for the same motif and switch from active or repressive transcriptional activity ., All these observations developed the main paradigm shift of how GATA factors are recruited and reside on the chromatin 30 , 31 ., In yeast , the family of GATA transcription factors contains over 10 members 32 ., Four of them are implicated in the regulation of Nitrogen Catabolite Repression ( NCR ) -sensitive genes , the expression of which is repressed in the presence of a preferred nitrogen source ( glutamine , asparagine , ammonia ) and derepressed when only poor nitrogen sources ( e . g . proline , leucine , urea ) are available 10 ., The key GATA factors involved in NCR signaling are two activators ( Gln3 and Gat1/Nil1 ) and two repressors ( Gzf3/Nil2/Deh1 and Dal80/Uga43 ) 33–38 ., In a perfect feedback loop , the expression of DAL80 and GAT1 is also NCR-sensitive , which implies cross- and autogenous regulations of the GATA factors in the NCR mechanisms 38–41 ., Under nitrogen limitation , expression of DAL80 is highly induced 35 , and Dal80 enters the nucleus where it competes with the two GATA activators for the same binding sites 20 , 39 , 42 ., Although initially described as being active under nitrogen abundance 37 , 38 , the Gzf3 repressor also localizes to NCR-sensitive promoters in conditions of activation 40 ., The sequence conservation among the four yeast NCR GATA factors is remarkable and the residues involved in contacts with the DNA , thus specificity determination , are 100% conserved ., In this respect , the binding sites of Dal80 on target DNA are likely to be recognized also by Gln3 , Gat1 and Gzf3 28 ., In vitro , the Gln3 and Gat1 activators bind to single GATA sequences , presumably as monomers 43 , like their orthologous vertebrate counterparts , while Dal80 was found to bind to two GATA sequences , 15–35 bp apart , in a preferred tail-to-tail orientation or to a lower extent in a head-to-tail configuration 9 , 20 , 39 , 44 ., In vivo , GATA factor binding site recognition also appears to require repeated GATA motifs within promoters , as shown for the NCR-sensitive DAL5 promoter 45–47 ., This led to the actual fuzzy definition of UASNTR , consisting in two GATA sites located close to one another to present a binding platform for GATA factors 45–47 ., Finally , in some cases , the existence of auxiliary promoter sequences was shown to compensate single GATA site , allowing for transcriptional activation 48 , although this was never as efficient as additional GATA sites 49 ., The antagonistic role of Dal80 also requires multiple GATA sites 39 , 42 , and inactivation of one of the four GATA sites of the UGA4 promoter results in the loss of the Dal80-repressive activity while affecting moderately Gln3- and Gat1- activation capacity 20 ., In summary , although NCR-sensitive genes are recognized to contain at least one GATA site , and often more , a precise definition of the minimal element required for binding and transcriptional regulation is still lacking ., In yeast , genome-wide ChIP analyses have allowed gaining insights into the GATA factor gene network through the identification of direct targets 50–53 ., However , these studies were not performed in activating conditions , when all GATA factors are expressed , localized in the nucleus and active , so that the current list of GATA factor targets are likely to be underestimated ., On another hand , bioinformatic analyses have shown that , since GATA sequences are short , they can be found almost everywhere throughout the genome ., Therefore , based on the sole criteria of the presence of repeated GATA sequences in yeast promoters , a third of the yeast genes could hypothetically be NCR regulator targets 54 ., However , such GATA motif repetitions have been found in the promoter of 91 genes , inducible by GATA activators in absence of a good nitrogen source , supposed to be directly targeted by the GATA activators 55 ., Nevertheless , the functionality of these hypothetical UAS still needs to be directly demonstrated in vivo 1 ., Here , we provide the first genome-wide identification of Dal80 targets in yeast , in physiological conditions where Dal80 is fully expressed and active ., Using a ChIP-Seq approach combined to a bioinformatic peak-calling procedure , we defined the exhaustive set of Dal80-bound promoters , which turned out to be much larger than anticipated ., Our data indicate that at some promoters , Dal80 recruitment occurs independently of GATA sites ., Strikingly , Dal80 was also detected across the body of a subset of genes bound at the promoter , globally correlating with high and Dal80-sensitive expression ., Mechanistic single-gene experiments confirmed the Dal80 binding profiles , further indicating that Dal80 spreading across gene bodies requires active transcription ., Finally , co-immunoprecipitation experiments revealed that Dal80 physically interacts with active form of Pol II ., In order to determine the genome-wide occupancy of a GATA factor in yeast , our rationale was to choose Dal80 as it is known to be highly expressed in derepressing conditions and forms chromosome foci when tagged by GFP 56 ., We grew yeast cells in proline-containing medium and performed a ChIP-Seq analysis using a Dal80-Myc13-tagged strain and the isogenic untagged strain , as a control ( Fig 1A ) , after ensuring that the Myc13-tagged form of Dal80 was functional ( S1A Fig ) ., Dal80-bound regions were then identified using a peak-calling algorithm ( see Material & Methods ) ., A promoter was defined as bound by Dal80 on the basis of a >75% overlap of the -100 to -350 region ( relative to the downstream ORF start site ) by a peak ( Fig 1B ) ., We chose to use as the reference coordinate the translation initiation codon rather than the transcription start site ( TSS ) since the latter has not been accurately defined for all genes ., Then , our arbitrary definition of the promoter as the -350 to -100 region relative to the ATG codon was based on the distribution of the TSS-ATG distance for genes with an annotated TSS ( median and average distance = 58 and 107 bp , respectively; see S1B Fig ) ., Strikingly , Dal80 was found to bind to 1269 gene promoters ( Fig 1C and 1D and S1 Table ) ., This number , corresponding to 22% of all protein-coding gene promoters , is much higher than anticipated given the roughly hundred target genes generally cited for the GATA transcriptional activators Gat1 and Gln3 55 , 57 , presumably sharing binding sites with Dal80 ., However , we noted that some peaks ( 221 ) overlapped several promoters ( 471 ) , mainly of divergent genes ( 442 ) , as shown in Fig 1E for an illustrative example ., Despite it is possible that in such cases , only one of the two divergent promoters is targeted by Dal80 , the number of in vivo Dal80 target sites we identified here has been extensively extended from what was acknowledged so far ., Among the genes showing Dal80 binding at their promoter , we noticed a significant enrichment for cytoplasmic translation genes , as well as genes involved in small molecule biosyntheses , including amino acids ( S2 Table ) ., Before our work , very few studies have investigated the transcriptional targets of Dal80 in vivo in conditions of nitrogen deprivation ., One of them , based on mini-arrays 58 , identified 19 Dal80-regulated genes , all of which have been isolated in our ChIP-Seq analysis ( highlighted in orange in column B of S3 Table ) ., As expected given the similarity between binding sites of Dal80 and the other nitrogen-regulated GATA factors , other genes related to previous nitrogen regulation screens 55 , 57–64 are also significantly enriched within our list: 103 of the 205 previously identified nitrogen-regulated genes have been identified in our ChIP-Seq analysis using Dal80 as the bait , which is much more than expected by chance ( P<0 . 001 , Chi-square test; S3 Table , column B ) ., Surprisingly , analysis of GATA site occurrence over Dal80-bound and unbound promoters revealed no difference between the two classes , 48 . 2% and 51 . 3% of Dal80-bound and unbound promoters containing at least two GATA sites , respectively ( Fig 1F ) ., Likewise , we observed no major difference between the Dal80-bound and unbound promoters in respect of the GATA sites spacing ( S1C Fig ) and orientation ( S1D Fig ) preferences defined in vitro for Dal80 binding 9 ., Intriguingly , 20% of Dal80-bound promoters do not contain any GATA site ( Fig 1F ) , indicating that Dal80 recruitment can also occur independently of the presence of consensus GATA sites ( see S1B Fig for visualization of Dal80 recruitment to a GATA-less promoter ) ., In summary , our ChIP-Seq analysis revealed that Dal80 binds to a set of promoters larger than previously expected , targeting biosynthetic functions and protein synthesis in addition to nitrogen catabolite repression ., We asked whether Dal80-binding to promoters could be associated to regulation of gene expression by the nitrogen source and/or Dal80 ., We therefore performed RNA-seq in wild-type cells grown in glutamine- and proline-containing medium , and in dal80Δ cells grown in proline-containing medium ., Firstly , we identified 1682 ( 30% ) genes differentially expressed ( fold-change ≥2 or ≤0 . 5 , P ≤0 . 01 ) in wild-type cells according to the nitrogen source provided ( Fig 2A ) , including 754 genes upregulated ( NCR-sensitive ) and 928 downregulated ( revNCR-sensitive ) in proline-containing medium ( see lists in S4 Table ) ., Consistent with previous reports , DAL80 was found in our set of NCR-sensitive genes ( S4 Table ) , showing very low expression in glutamine-containing medium and strong derepression in proline ( S2A Fig ) ., More globally , 97 of the 205 genes previously identified as NCR-sensitive were also found in our list ( P<0 . 0001 , Chi-square test; S4 Table ) ., In parallel , we identified 546 genes showing significantly altered expression ( fold-change ≥2 or ≤0 . 5 , P ≤0 . 01 ) in proline-grown dal80Δ cells compared to wild type ( Fig 2B; S5 Table ) ., In agreement with the previously described repressive activity of Dal80 35 , 232 genes are indeed negatively regulated by Dal80 ( up in dal80Δ; red dots in Fig 2B ) ., Unexpectedly , 314 genes are positively regulated by Dal80 ( down in dal80Δ; blue dots in Fig 2B ) ., This is the first in vivo global indication suggesting a positive function for Dal80 in gene expression ., The Dal80-repressed group was enriched for genes involved in small molecule catabolic processes ( S6 Table ) , while the Dal80-activated genes were mostly involved in amino acid biosynthesis ( S7 Table ) ., Again , we noticed an overlap between Dal80-regulated genes and nitrogen regulated genes that were identified in other screens: 86 of the 205 previously identified nitrogen-regulated genes have been identified as Dal80-regulated , which is much more than expected by chance ( P<0 . 0001 , Chi-square test; column D of S3 Table ) ., Globally , we observed a significant correlation between Dal80-sensivity and regulation by the nitrogen source ( P<0 . 00001 , Chi-square test; Fig 2C; see also S2B Fig ) ., Indeed , there are more NCR-sensitive Dal80-activated and Dal80–repressed genes than expected in case of independence ( Fig 2C; see also S2B Fig ) ., Similarly , the number of revNCR-sensitive Dal80-repressed genes is also significantly higher than expected by chance ( Fig 2C; see also S2B Fig ) ., In contrast , the number of revNCR-sensitive Dal80-activated genes is significantly lower than expected by chance ( Fig 2C; see also S2B Fig ) , indicating a negative correlation in this case ., This observation is consistent with the DAL80 gene itself being NCR-sensitive , so that the Dal80-activated genes can only be activated when DAL80 is expressed ., More importantly , Dal80 recruitment to promoters significantly correlated with nitrogen- and Dal80-sensitivity ., In fact , nitrogen-regulated expression and Dal80-binding are not independent , as NCR-sensitive ( 212 ) and especially revNCR-sensitive ( 325 ) genes are significantly enriched in Dal80-bound genes ( P<0 . 00001 , Chi-square test; Fig 2D; see also S2C Fig ) ., We also observed a significant correlation between Dal80-sensitive gene expression and Dal80 recruitment at the promoter: 211/546 of Dal80-regulated genes were bound by Dal80 , including 120/314 Dal80-activated and 91 Dal80-repressed genes , which again is much more than expected by chance ( P<0 . 00001 , Chi-square test; Fig 2E; see also S2D Fig ) ., Fig 2F shows an illustrative example of an NCR-sensitive , Dal80-activated gene ( UGA3 ) , the promoter of which is bound by Dal80 ( Fig 1E ) ., S3A Fig shows the RNA-Seq signals for another NCR-sensitive , Dal80-repressed and Dal80-bound gene ( MEP2 ) , correlating with Pol II occupancy levels ( S3B Fig ) ., In summary , there is a significant correlation between Dal80 recruitment to the promoter of genes and a regulation by the nitrogen source and/or Dal80 at the RNA level , indicating that Dal80 recruitment to promoters is physiologically relevant ., More specifically , we identified a subset of 211 Dal80-bound genes that are regulated by Dal80 ( S3 Table ) , and that are therefore a robust class of direct Dal80 targets ., The metagene analysis described above revealed that the genes bound by Dal80 at the promoter also display a signal along the gene body , although this intragenic signal remains globally lower than in the promoter-proximal region ( Fig 1D ) ., This observation prompted us to investigate the possibility that Dal80 also occupies the gene body , at least for a subset of genes ., We identified 189 genes showing Dal80 intragenic occupancy , according to a >75% overlap of the ORF by a Dal80-Myc13 peak ( Fig 3A and 3B ) ., Among them , 144 ( 76% ) were also bound at the promoter ( Fig 3B ) ., On the other hand , 45 genes showing Dal80 intragenic binding were not bound at the promoter ( Fig 3B ) ., Hence , we distinguished four classes of genes ( S8 Table ) :, ( i ) those bound by Dal80 at the promoter only ( “P” class; Fig 3C; S8 Table , column C ) ,, ( ii ) those showing both promoter and intragenic binding ( “P&O” class; Fig 3D; S8 Table , column E ) ,, ( iii ) those bound across the ORF only ( “O” class; Fig 3E; S8 Table , column D ) ,, ( iv ) the unbound genes ( Fig 3F ) ., Interestingly , we noted that the global Dal80-Myc13 signal at the promoter was higher for the “P&O” class in comparison to the “P” class ( Fig 3C and 3D ) ., Most of the genes of the “O” class are not Dal80-sensitive ( 40/45; S8 Table , column J ) ., Furthermore , a substantial fraction of them correspond to small dubious ORFs , close to or even overlapping an adjacent Dal80-bound gene promoter ., In these cases , the limited resolution of the ChIP-Seq technique , combined to the small size of these genes , might have allowed them to pass the filters we used to identify Dal80 intragenic binding ., Overall , these observations suggest that the existence of the “O” class is likely to be physiologically irrelevant ., Therefore , this class will not be further considered in our study ., In conclusion , we identified a subset of genes showing intragenic Dal80 occupancy , in most cases correlating with a strong Dal80 recruitment at the promoter ., We asked whether Dal80 occupancy across gene bodies correlates with nitrogen-regulated gene expression and Dal80-sensitivity ., We observed that nitrogen-regulated genes ( NCR and revNCR; Fig 4A; see also S4A Fig ) and Dal80-regulated genes ( Dal80-activated and -repressed; Fig 4B; see also S4B Fig ) were significantly more represented in the P&O class compared to the Dal80-unbound class ., Strikingly , we also observed that the genes of the P&O class are more expressed than the unbound genes ( P < 2 . 2e-16 , Wilcoxon rank-sum test; Fig 4C ) but also than the P-bound genes ( P = 1 . 3e-14 , Wilcoxon rank-sum test; Fig 4C ) ., However , it should be noted that a fraction of P-bound and unbound genes are expressed to higher levels than genes of the “P&O” class ( S4C and S4D Fig ) , indicating that high expression does not always imply intragenic Dal80 occupancy ., Together with the observation that genes of the “P&O” class globally showed higher Dal80-Myc13 ChIP-Seq signal at the promoter than those of the “P” class ( Fig 3C and 3D ) , our results indicate that Dal80 occupancy across gene bodies correlates with a stronger recruitment at the promoter and higher expression in proline-containing medium ., This raises the question of the specificity of the intragenic signal observed by ChIP-Seq ., Indeed , for several proteins , unspecific ChIP signals have been detected across the body of a subset of highly expressed Pol II- and Pol III-dependent genes , referred to as ‘hyper-ChIPable’ loci 65–67 ., We asked whether genes of our P&O class have been previously identified as ‘hyper-ChIPable’ ( S9 Table , column G ) ., This comparison indicated that 48/1125 of the P-bound genes and 27/144 of the P&O genes match with hyper-ChIPable loci ( S4E and S4F Fig; see also S9 Table , columns H-I ) , suggesting that for a minority of cases , the intragenic Dal80 signal could be due to the ‘hyper-ChIPability’ of the locus and therefore be non-specific ., However , since these ‘hyper-ChIPable’ loci were defined under growth conditions that are different from those used in our study ( growth in rich medium vs proline-containing synthetic medium ) , we aimed to get a more robust control for the specificity of Dal80 within gene bodies ., Our rationale was to evaluate how similar and/or specific two close GATA factors could share/distinguish this “so called” artefactual hyper-ChIPability property ., We performed a similar ChIP-Seq analysis using another GATA factor , the Gat1 activator 68 , using the same conditions and following the same experimental procedure as described above ( Figs 1A , 1B & 3A ) ., Interestingly , 83 . 2% ( 936/1125 ) of the promoters bound by Dal80 were also bound by Gat1 ( S4G Fig; S9 Table , column E ) , reinforcing the accuracy of the extended list of novel GATA-bound genes in yeast ., Strikingly , the proportion of common targets among the P&O class dramatically decreased , 55% ( 79/144 ) of the genes bound by Dal80 at the promoter and across the gene body also showing promoter and intragenic binding for Gat1 ( S4H Fig; S9 Table , column F ) ., Importantly however , 65/144 P&O for Dal80 do not display intragenic binding for Gat1 ( S4H Fig; S9 Table , column F ) , although Gat1 is recruited to the promoter of 57 of them ., Thus , we can define a subset of 57 genes showing a specific intragenic occupancy of Dal80 , while both Dal80 and Gat1 are recruited to their promoters similarly ., As an illustrative striking example , Fig 4D shows a snapshot of the ChIP-Seq signals across MEP2 , a well-characterized NCR-sensitive gene , the promoter of which is bound by the two GATA factors , but only Dal80 is found within the gene body ., To summarize , Dal80 occupancy across the gene body correlates with high expression levels ., In a substantial proportion of cases , intragenic occupancy was found to be specific for Dal80 , as another GATA factor also recruited to the promoter in the same experimental conditions was not detected within the gene body ., In order to validate our genome-wide observations and get additional mechanistic insights into the molecular bases of Dal80 occupancy across the body of highly expressed genes , we characterized the binding profile of Dal80 along the ammonium permease-coding gene MEP2 , an NCR-sensitive gene of the “P&O” class ( see Fig 4D ) ., ChIP experiments followed by qPCR confirmed that Dal80 binds not only the promoter , but also across the coding region of MEP2 in proline-grown cells ( Fig 5A and 5B ) ., No signal was observed in glutamine-grown cells ( Fig 5B ) , indicating that Dal80 recruitment only occurs when it is expressed ( S2A Fig ) ., To determine whether Dal80 intragenic occupancy is mediated by nascent RNA binding during transcription , we performed a similar ChIP experiment on the MEP2 gene , treating the chromatin with RNase before the immunoprecipitation ., Our results show no significant change of the Dal80-Myc13 signal across MEP2 upon RNAse treatment of the chromatin extracts before the immunoprecipitation ( Fig 5C ) , indicating that Dal80 occupancy across the gene body does not depend on RNA ., Since genes of the Dal80 “P&O” class are globally highly expressed , we asked whether active transcription is a prerequisite for Dal80 binding across the ORF ., Our strategy was to select an NCR gene for which Dal80 is bound at the promoter when repressed and then monitor Dal80 occupancy once the gene is activated ., Our RNA- and ChIP-Seq data allowed us to isolate the UGA4 locus , another well-characterized NCR-sensitive gene , bound by Dal80 at the promoter ( Fig 6A; see snapshot in S5A Fig ) ., UGA4 expression is induced by GABA ( γ-aminobutyric acid ) and is strongly repressed by Dal80 in the absence of the inducer 69 ., To derepress UGA4 without inducer , a Dal80-specific deletion in the C-terminal leucine zipper domain was generated , impairing Dal80 repressive activity without affecting its binding capacity 34 , 44 ., Indeed , in the Dal80ΔLZ-Myc13 strain ( Fig 6B ) , the steady-state level of UGA4 mRNA ( S5B Fig ) and Pol II occupancy ( S5C Fig ) both increased to derepressed levels in non-inducing conditions , like in a dal80Δ strain ., Strikingly , in these conditions , full-length Dal80-Myc13 binding was restricted to the UGA4 promoter ( Fig 6A; see also S5A Fig ) , while Dal80ΔLZ-Myc13 binding was detected at the promoter and across the body of UGA4 ( Fig 6A ) ., Interestingly , the leucine zipper of Dal80 and consequently , its dimerization , needed for UGA4 repression , were not required for its localization across the UGA4 gene body ., Importantly , these results confirm that promoter binding is not sufficient to confer intragenic binding , but suggest that transcription activation is required ., Altogether , these observations prompted the important mechanistic question of how Dal80 can be localized to gene bodies upon transcription activation ., In order to test if the presence of an NCR-sensitive promoter could confer intragenic Dal80 binding across the body of a non-NCR-sensitive gene , we placed the URA3 ORF under the control of different promoters bound or not by Dal80: the MEP2 and TDH3 promoters as P&O representative , the ALD6 promoter for the P class and the VMA1 promoter , which is not bound by Dal80 ( Fig 7A ) ., When driven by PMEP2 , the expression of URA3 becomes NCR-sensitive and followed wild-type MEP2 expression ( S6 Fig ) , correlating with Pol II recruitment over the URA3 ORF ( Fig 7B ) ., In these conditions , we observed Dal80-Myc13 binding at the promoter of MEP2 and also across URA3 ( Fig 7C ) ., Similarly for PTDH3-URA3 construct , Dal80 also was relocalized within the URA3 ORF , although to a lesser extent ., Importantly , Dal80 binding was not detected across URA3 when it was expressed from its native locus , under the control of its promoter ( Fig 7C ) or under the control of the Dal80-bound PALD6 or unbound PVMA1 ( Fig 7C ) , reinforcing the idea that those promoters fail to carry sufficient information for Dal80 to occupy the URA3 ORF ., Among the obvious characteristics , we noticed that Pol II occupancy is higher within those P&O URA3 genes than the P only , suggesting that transcription strength might be a key determinant for Dal80 localization across the ORF ., Interestingly , among the P&O fusions ( MEP2 and TDH3 ) , we noted a difference in Dal80 binding levels to the adjacent URA3 ORF , while those of Pol II remain similar across the two coding regions , suggesting that Pol II level might not be the only factor that control Dal80 occupancy ., In conclusion , these results show that for the same URA3 sequence , the Dal80 occupancy displays distinct features depending only on the promoter characteristics to be classified as P , P&O or unbound , reflecting transcriptional strength ., We propose that Dal80 presence within the ORF could be attributed to a spreading mechanism , controlled by Pol II complex and Dal80-promoter recognition capacity ., These results exclude strongly DNA motif ( s ) as a main determinant for Dal80 spreading into ORF but rather raise the question of the direct implication of Pol II itself ., To test the hypothesis that the active Pol II complex could be responsible for Dal80 spreading beyond Dal80-bound promoters , we assessed the effect of rapid inactivation of Pol II using the thermosensitive rpb1-1 strain 70 , 71 ., We analyzed Dal80-Myc13 binding along MEP2 in WT and rpb1-1 cells ., When rpb1-1 cells were shifted at 37°C for 1h , MEP2 mRNA and Pol II levels showed a 2-fold ( S7A Fig ) and >10-fold decrease ( S7B Fig ) , respectively , reflecting the expected transcription shut-down when rpb1-1 cells are shifted in non-permissive conditions ., In the same conditions , we observed a significant >5-fold reduction of Dal80-Myc13 levels across the MEP2 ORF , while the binding at the promoter was not affected ( Fig 8A ) ., This result reinforces the idea that Dal80 spreading across the body of NCR-sensitive genes is strongly correlated to an active Pol II ., To get insights into the mechanism by which Dal80 associates to actively transcribed gene bodies , we tested whether it physically interacts with the transcriptionally engaged form of Pol II ( Fig 8B ) ., Total protein extracts from Dal80-Myc13 cells were immunoprecipitated with antibodies directed against the Pol II CTD and its phospho-forms Ser2P and Ser5P , respectively characteristic of elongating and initiating Pol II forms ., All three antibodies enabled effective immunoprecipitation , whereas no antibody and nonspecific antibody controls generated a lower or no signal at all ., Thus , Dal80 would physically interact with phosphoforms of the Pol III , suggesting a strong association with Pol II engaged in active transcription from initiating to elongating polymerase ., Together , our data indicate that Dal80 spreading across the body of NCR-sensitive genes depends on active transcription and that Dal80 interacts with the transcriptionally active forms of Pol II , supporting a model where Dal80 spreading across the body of highly expressed , NCR-sensitive genes might be the result of Dal80-Pol II association at post-initiation transcription phases ., Eukaryotic GATA factors belong to an important family of DNA binding proteins involved in development and response to environmental changes in multicellular and unicellular organisms , respectively ., In yeast , four GATA factors are involved in Nitrogen Catabolite Repression ( NCR ) , controlling gene expression in response to nitrogen source availability ., One of them , the Dal80 repressor , itself NCR-sensitive , acts to modulate the intensity of NCR responses ., Over the past decade , a number of studies have screened the genome aiming at gathering an inventory of genes regulated by the nitrogen source ., Although >500 genes have been shown to be differentially expressed upon change of the nitrogen source 57 , 64 , the list of NCR-sensitive genes was reduced to about 100 , based on their sensitivity to GATA factors 55 , 57 , 60 , 63 , suggesting that the number of Dal80 targets would be situated in that range ., Here , using ChIP-Seq , we identified 1269 Dal80-bound promoters , which considerably extends the list of potential Dal80 targets ., In fact , the number of Dal80-bound promoters could even have been greater ., Indeed , the GATA consensus binding site is rather simple and short , so that in yeast , a total number of 10 , 000 putative binding sites can be found in all protein-coding gene promoters , 2930 promoters having at least two GATA sites , which is thought to be a prerequisite for in vivo binding and function of the GATA factors ., The difference between the number of promoters with ≥2 GATA sites and the number of Dal80-bound promoters suggests the existence of a selectivity for Dal80 recruitment ., This selectivity could rely on promoter architecture and/or chromatin structure , conditioning the requirement for auxiliary DNA binding factors that would stabilize Dal80 at some promoters ., Moreover , although we observed a significant correlation between Dal80 binding and regulation , the expression of most of the Dal80-bound genes was not affected in a dal80Δ mutant strain ., Again , Dal80-dependence for transcribing these genes , as well as their NCR sensitivity , could require the presence of yet unknown cofactors which are not produced or inactive under the tested growth conditions ., In mammals , GATA factors also display an extraordinary complexity in the relationships between binding and expression regulation ., Like Dal80 , GATA-1 and GATA-2 only occupy a small subset of their abundant binding motif throughout the genome , and the presence of the conserved binding site is insufficient to cause GATA-dependent regulation in most instances 72 ., GAT | Introduction, Results, Discussion, Materials and methods | GATA transcription factors are highly conserved among eukaryotes and play roles in transcription of genes implicated in cancer progression and hematopoiesis ., However , although their consensus binding sites have been well defined in vitro , the in vivo selectivity for recognition by GATA factors remains poorly characterized ., Using ChIP-Seq , we identified the Dal80 GATA factor targets in yeast ., Our data reveal Dal80 binding to a large set of promoters , sometimes independently of GATA sites , correlating with nitrogen- and/or Dal80-sensitive gene expression ., Strikingly , Dal80 was also detected across the body of promoter-bound genes , correlating with high expression ., Mechanistic single-gene experiments showed that Dal80 spreading across gene bodies requires active transcription ., Consistently , Dal80 co-immunoprecipitated with the initiating and post-initiation forms of RNA Polymerase II ., Our work suggests that GATA factors could play dual , synergistic roles during transcription initiation and post-initiation steps , promoting efficient remodeling of the gene expression program in response to environmental changes . | GATA transcription factors are highly conserved among eukaryotes and play key roles in cancer progression and hematopoiesis ., In budding yeast , four GATA transcription factors are involved in the response to the quality of nitrogen supply ., Here , we have determined the whole genome binding profile of the Dal80 GATA factor , and revealed that it also associates with the body of promoter-bound genes ., The observation that intragenic spreading correlates with high expression levels and exquisite Dal80 sensitivity suggests that GATA factors could play other , unexpected roles at post-initiation stages in eukaryotes . | statistics, gene regulation, regulatory proteins, dna-binding proteins, dna transcription, chi square tests, mathematics, genome analysis, transcription factors, sequence motif analysis, research and analysis methods, sequence analysis, bioinformatics, proteins, mathematical and statistical techniques, gene expression, statistical methods, genetic loci, biochemistry, statistical hypothesis testing, database and informatics methods, genetics, biology and life sciences, physical sciences, genomics, gene prediction, computational biology | null |
journal.pcbi.1003024 | 2,013 | Reinforcement Learning Using a Continuous Time Actor-Critic Framework with Spiking Neurons | Many instances of animal behavior learning such as path finding in foraging , or – a more artificial example – navigating the Morris water-maze , can be interpreted as exploration and trial-and-error learning ., In both examples , the behavior eventually learned by the animal is the one that led to high reward ., These can be appetite rewards ( i . e . , food ) or more indirect rewards , such as the relief of finding the platform in the water-maze ., Important progress has been made in understanding how learning of such behaviors takes place in the mammalian brain ., On one hand , the framework of reinforcement learning 1 provides a theory and algorithms for learning with sparse rewarding events ., A particularly attractive formulation of reinforcement learning is temporal difference ( TD ) learning 2 ., In the standard setting , this theory assumes that an agent moves between states in its environment by choosing appropriate actions in discrete time steps ., Rewards are given in certain conjunctions of states and actions , and the agents aim is to choose its actions so as to maximize the amount of reward it receives ., Several algorithms have been developed to solve this standard formulation of the problem , and some of these have been used with spiking neural systems ., These include REINFORCE 3 , 4 and partially observable Markov decision processes 5 , 6 , in case the agent has incomplete knowledge of its state ., On the other hand , experiments show that dopamine , a neurotransmitter associated with pleasure , is released in the brain when reward , or a reward-predicting event , occurs 7 ., Dopamine has been shown to modulate the induction of plasticity in timing non-specific protocols 8–11 ., Dopamine has also recently been shown to modulate spike-timing-dependent plasticity ( STDP ) , although the exact spike-timing and dopamine requirements for induction of long-term potentiation ( LTP ) and long-term depression ( LTD ) are still unclear 12–14 ., A crucial problem in linking biological neural networks and reinforcement learning is that typical formulations of reinforcement learning rely on discrete descriptions of states , actions and time , while spiking neurons evolve naturally in continuous time and biologically plausible “time-steps” are difficult to envision ., Earlier studies suggested that an external reset 15 or theta oscillations 16 might be involved , but no evidence exists to support this and it is not clear why evolution would favor slower decision steps over a continuous decision mechanism ., Indeed biological decision making is often modeled by an integrative process in continuous time 17 , where the actual decision is triggered when the integrated value reaches a threshold ., In this study , we propose a way to narrow the conceptual gap between reinforcement learning models and the family of spike-timing-dependent synaptic learning rules by using continuous representations of state , actions and time , and by deriving biologically plausible synaptic learning rules ., More precisely , we use a variation of the Actor-Critic architecture 1 , 18 for TD learning ., Starting from the continuous TD formulation by Doya 19 , we derive reward-modulated STDP learning rules which enable a network of model spiking neurons to efficiently solve navigation and motor control tasks , with continuous state , action and time representations ., This can be seen as an extension of earlier works 20 , 21 to continuous actions , continuous time and spiking neurons ., We show that such a system has a performance on par with that of real animals and that it offers new insight into synaptic plasticity under the influence of neuromodulators such as dopamine ., The goal of a reinforcement learning agent is to maximize its future rewards ., Following Doya 10 , we define the continuous-time value function as ( 3 ) where the brackets represent the expectation over all future trajectories and future action choices , dependent on the policy ., The parameter represents the reward discount time constant , analogous to the discount factor of discrete reinforcement learning ., Its effect is to make rewards in the near future more attractive than distant ones ., Typical values of for a task such as the water-maze task would be on the order of a few seconds ., Eq ., 3 represents the total quantity of discounted reward that an agent in position at time and following policy can expect ., The policy should be chosen such that is maximized for all locations ., Taking the derivative of Eq ., 3 with respect to time yields the self-consistency equation 19 ( 4 ) Calculating requires knowledge of the reward function and of the environment dynamics ( Eq 1 ) ., These are , however , unknown to the agent ., Typically , the best an agent can do is to maintain a parametric estimator of the “true” value function ., This estimator being imperfect , it is not guaranteed to satisfy Eq ., 4 . Instead , the temporal difference error is defined as the mismatch in the self-consistency , ( 5 ) This is analog to the discrete TD error 1 , 19 ( 6 ) where the reward discount factor plays a role similar to the reward discount time constant ., More precisely , for short steps , 19 ., An estimator can be said to be a good approximation to if the TD error is close to zero for all ., This suggests a simple way to learn a value function estimator: by a gradient descent on the squared TD error in the following way ( 7 ) where is a learning rate parameter and is the set of parameters ( synaptic weights ) that control the estimator of the value function ., This approach , dubbed residual gradient 19 , 25 , 26 , yields a learning rule that is formally correct , but in our case suffers from a noise bias , as shown in Models ., Instead , we use a different learning rule , suggested for the discrete case by Sutton and Barto 1 ., Translated in a continuous framework , the aim of their optimization approach is that the value function approximation should match the true value function ., This is equivalent to minimizing an objective function ( 8 ) A gradient descent learning rule on yields ( 9 ) Of course , because is unknown , this is not a particularly useful learning rule ., On the other hand , using Eq ., 4 , this becomes ( 10 ) where we merged into the learning rate without loss of generality ., In the last step , we replaced the real value function derivative with its estimate , i . e . , , and then used the definition of from Eq ., 5 . The substitution of by in Eq ., 10 is an approximation , and there is in general no guarantee that the two values are similar ., However the form of the resulting learning rule suggests it goes in the direction of reducing the TD error ., For example , if is positive at time , updating the parameters in the direction suggested by Eq ., 10 , will increase the value of , and thus decrease ., In 19 , a heuristic shortcut was used to go directly from the residual gradient ( Eq . 7 ) to Eq ., 10 ., As noted by Doya 19 , the form of the learning rule in Eq ., 10 is a continuous version of the discrete 1 , 27 with function approximation ( here with ) ., This has been shown to converge with probability 1 28 , 29 , even in the case of infinite ( but countable ) state space ., This must be the case also for arbitrarily small time steps ( such as the finite steps usually used in computer simulations of a continuous system 19 ) , and thus it seems reasonable to expect that the continuous version also converges under reasonable assumptions , even though to date no proof exists ., An important problem in reinforcement learning is the concept of temporal credit assignment , i . e . , how to propagate information about rewards back in time ., In the framework of TD learning , this means propagating the TD error at time so that the value function at earlier times is updated in consequence ., The learning rule Eq ., 10 does not by itself offer a solution to this problem , because the expression of explicitly refers only to and at time ., Therefore does not convey information about other times and minimizing does not a priori affect values and ., This is in contrast to the discrete version of the TD error ( Eq . 6 ) , where the expression of explicitly links to and thus the TD error is back-propagated during subsequent learning trials ., If , however , one assumes that the value function is continuous and continuously differentiable , changing the values of and implies changing the values of these functions in a finite vicinity of ., This is in particular the case if one uses a parametric form for , in the form of a weighted mixture of smooth kernels ( as we do here , see next section ) ., Therefore , the conjunction of a function approximation of the value function in the form of a linear combination of smooth kernels ensures that the TD error is propagated in time in the continuous case , allowing the temporal credit assignment problem to be solved ., We now take the above derivation a step further by assuming that the value function estimation is performed by a spiking neuron with firing rate ., A natural way of doing this is ( 11 ) where is the value corresponding to no spiking activity and is a scaling factor with units of reward units×s ., A choice of enables negative values , despite the fact that the rate is always positive ., We call this neuron a critic neuron , because its role is to maintain an estimate of the value function ., Several aspects should be discussed at this point ., Firstly , since the value function in Eq ., 11 must depend on the state of the agent , we must assume that the neuron receives some meaningful synaptic input about the state of the agent ., In the following we make the assumption that this input is feed-forward from the place cells to the ( spiking ) critic neuron ., Secondly , while the value function is in theory a function only of the state at time , a spiking neuron implementation ( such as the simplified model we use here , see Models ) will reflect the recent past , in a manner determined by the shape of the excitatory postsynaptic potentials ( EPSP ) it receives ., This is a limitation shared by all neural circuits processing sensory input with finite synaptic delays ., In the rest of this study , we assume that the evolution of the state of the agent is slow compared to the width of an EPSP ., In that limit , the firing rate of a critic neuron at time actually reflects the position of the agent at that time ., Thirdly , the firing rate of a single spike-firing neuron is itself a vague concept and multiple definitions are possible ., Lets start from its spike train ( where is the set of the neurons spike times and is the Dirac delta , not to be confused with the TD signal ) ., The expectation is a statistical average of the neurons firing over many repetitions ., It is the theoretically favored definition of the firing rate , but in practice it is not available in single trials in a biologically plausible setting ., Instead , a common workaround is to use a temporal average , for example by filtering the spike train with a kernel ( 12 ) Essentially , this amounts to a trade-off between temporal accuracy and smoothness of the rate function , of which extreme cases are respectively the spike train ( extreme temporal accuracy ) and a simple spike count over a long time window with smooth borders ( no temporal information , extreme smoothness ) ., In choosing a kernel , it should hold that , so that each spike is counted once , and one often wishes the kernel to be causal ( ) , so that the current firing rate is fully determined by past spike times and independent of future spikes ., Another common approximation for the firing rate of a neuron consists in replacing the statistical average by a population average , over many neurons encoding the same value ., Provided they are statistically independent of each other ( for example if the neurons are not directly connected ) , averaging their responses over a single trial is equivalent to averaging the responses of a single neuron over the same number of trials ., Here we combine temporal and population averaging , redefining the value function as an average firing rate of neurons ( 13 ) where the instantaneous firing rate of neuron is defined by Eq ., 12 , using its spike train and a kernel defined by ( 14 ) This kernel rises with a time constant and decays to 0 with time constant ., One advantage of the definition of Eq ., 12 is that the derivative of the firing rate of neuron with respect to time is simply ( 15 ) so that computing the derivative of the firing rate is simply a matter of filtering the spike train with the derivative of the kernel ., This way , the TD error of Eq ., 5 can be expressed as ( 16 ) where , again , denotes the spike train of neuron in the pool of critic neurons ., Suppose that feed-forward weights lead from a state-representation neuron to neuron in the population of critic neurons ., Can the critic neurons learn to approximate the value function by changing the synaptic weights ?, An answer to this question is obtained by combining Eq ., 10 with Eqs 13 and 16 , leading to a weights update ( 17 ) where is the time course of an EPSP and is the spike train of the presynaptic neuron , restricted to the spikes posterior to the last spike time of postsynaptic neuron ., For simplicity , we merged all constants into a new learning rate ., A more formal derivation can be found in Models ., Let us now have a closer look at the shape of the learning rule suggested by Eq ., 17 ., The effective learning rate is given by a parameter ., The rest of the learning rule consists of a product of two terms ., The first one is the TD error term , which is the same for all synapses , and can thus be considered as a global factor , possibly transmitted by one or more neuromodulators ( Figure 1 ) ., This neuromodulator broadcasts information about inconsistency between the reward and the value function encoded by the population of critic neurons to all neurons in the network ., The second term is synapse-specific and reflects the coincidence of EPSPs caused by presynaptic spikes of neuron with the postsynaptic spikes of neuron ., The postsynaptic term is a consequence of the exponential non-linearity used in the neuron model ( see Models ) ., This coincidence , “Hebbian” term is in turn filtered through the kernel which corresponds to the effect of a postsynaptic spike on ., It reflects the responsibility of the synapse in the recent value function ., Together these two terms form a three-factor rule , where the pre- and postsynaptic activities combine with the global signal to modify synaptic strengths ( Figure 2A , top ) ., Because it has , roughly , the form of “TD error signalHebbian LTP” , we call this learning rule TD-LTP ., We would like to point out the similarity of the TD-LTP learning rule to a reward-modulated spike-timing-dependent plasticity rule we call R-STDP 6 , 16 , 30–32 ., In R-STDP , the effects of classic STDP 33–36 are stored into an exponentially decaying , medium term ( time constant ) , synapse-specific memory , called an eligibility trace ., This trace is only imprinted into the actual synaptic weights when a global , neuromodulatory success signal is sent to the synapses ., In R-STDP , the neuromodulatory signal is the reward minus a baseline , i . e . , ., It was shown 32 that for R-STDP to maximize reward , the baseline must precisely match the mean ( or expected ) reward ., In this sense , is a reward prediction error signal; a system to compute this signal is needed ., Since the TD error is also a reward prediction error signal , it seems natural to use instead of ., This turns the reward-modulated learning rule R-STDP into a TD error-modulated TD-STDP rule ( Figure 2A , bottom ) ., In this form , TD-STDP is very similar to TD-LTP ., The major difference between the two is the influence of post-before-pre spike pairings on the learning rule: while these are ignored in TD-LTP , they cause a negative contribution to the coincidence detection in TD-STDP ., The filtering kernel , which was introduced to filter the spike trains into differentiable firing rates serves a role similar to the eligibility trace in R-STDP , and also in the discrete TD ( ) 1 ., As noted in the previous section , this is the consequence of the combination of a smooth parametric function approximation of the value function ( each critic spike contributes a shape to ) and the form of the learning rule from Eq ., 10 ., The filtering kernel is crucial to back-propagation of the TD error , and thus to the solving of the temporal credit assignment problem ., Having shown how spiking neurons can represent and learn the value function , we next test these results through simulations ., However , in the actor-critic framework , the actor and the critic learn in collaboration , making it hard to disentangle the effects of learning in either of the two ., To isolate learning by the critic and disregard potential problems of the actor , we temporarily sidestep this difficulty by using a forced action setup ., We transform the water-maze into a linear track , and “clamp” the action choice to a value which leads the agent straight to the reward ., In other words , the actor neurons are not simulated , see Figure 2B , and the agent simply “runs” to the goal ., Upon reaching it at time , a reward is delivered and the trial ends ., Figure 2C shows the value function over color-coded trials ( from blue to red ) as learned by a critic using the learning rule we described above ., On the first run ( dark blue trace ) , the critic neurons are naive about the reward and therefore represent a ( noisy version of a ) zero value function ., Upon reaching the goal , the TD error ( Figure 2D ) matches the reward time course , ., According to the learning rule in Eq ., 17 , this causes strengthening of those synapses that underwent pre-post activity recently before the reward ( with “recent” defined by the kernel ) ., This is visible already at the second trial , when the value just before reward becomes positive ., In the next trials , this effect repeats , until the TD error vanishes ., Suppose that , in a specific trials , reward starts at the time when the agent has reached the goal ., According to the definition of the TD error , for all times the -value is self consistent only if — or equivalently ., The gray dashed line in Figure 2C shows the time course of the theoretical value function; over many repetitions the colored traces , representing the value function in the different trials , move closer and closer to the theoretical value ., The black line in Figure 2C represents the average value function over 20 late trials , after learning has converged: it nicely matches the theoretical value ., An interesting point that appears in Figure 2C is the clearly visible back-propagation of information about the reward expressed in the shape of the value function ., In the first trials , the value function rises only for a short time just prior to the reward time ., This causes , in the following trial , a TD error at earlier times ., As trials proceed , synaptic weights corresponding to even earlier times increase ., After trials in Figure 2C , the value function roughly matches the theoretical value just prior to , but not earlier ., In subsequent trials , the point of mismatch is pushed back in time ., This back-propagation phenomenon is a signature of TD learning algorithms ., Two things should be noted here ., Firstly , the speed with which the back-propagation occurs is governed by the shape of the kernel in the Hebbian part of the learning rule ., It plays a role equivalent to the eligibility trace in reinforcement learning: it “flags” a synapse after it underwent pre-before-post activity with a decaying trace , a trace that is only consolidated into a weight change when a global confirmation signal arrives ., This “eligibility trace” role of is distinct from its original role in the term , where it is used to smooth the spiking activity of the critic neurons ( Eq . 12 ) ., As such , one might be tempted to change the decay time constant of the term in the learning rule so as to control back-propagation speed , while keeping the “other” of the signal fixed ., In separate simulations ( not shown ) , we found that such an ad-hoc approach did not lead to a gain in learning performance ., Secondly , we know by construction that this back-propagation of the reward information is driven by the TD error signal ., However , visual inspection of Figure 2D , which shows the traces corresponding to the experiment in Figure 2C , does not reveal any clear back-propagation of the TD error ., For , a large peak mirroring the reward signal ( gray dashed line ) is visible in the early traces ( blue lines ) and recedes quickly as the value function correctly learns to expect the reward ., For , the is dominated by fast noise , masking any back-propagation of the error signal , even though the fact that the value function is learned properly shows it is indeed present and effective ., One might speculate that if a biological system was using such a TD error learning system with spiking neuron , and if an experimenter was to record a handful of critic neurons he would be at great pain to measure any significant TD error back-propagation ., This is a possible explanation for the fact that no back-propagation signal has been observed in experiments ., We have already discussed the structural similarity of a TD-modulated version of the R-STDP rule 6 , 30 , 31 with TD-LTP ., Simulations of the linear track experiment with the TD-STDP rule show that it behaves similarly to our learning rule ( data not shown ) , i . e . , the difference between the two rules ( the post-before-pre part of the coincidence detection window , see Figure 2A ) does not appear to play a crucial role in this case ., We have seen above that spiking neurons in the “critic” population can learn to represent the expected rewards ., We next ask how a spiking neuron agent chooses its actions so as to maximize the reward ., In the classical description of reinforcement learning , actions , like states and time , are discrete ., While discrete actions can occur , for example when a laboratory animal has to choose which lever to press , most motor actions , such as hand reaching or locomotion in space , are more naturally described by continuous variables ., Even though an animal only has a finite number of neurons , neural coding schemes such as population vector coding 37 allow a discrete number of neurons to code for a continuum of actions ., We follow the population coding approach and define the actor as a group of spiking neurons ( Figure 3A ) , each coding for a different direction of motion ., Like the critic neurons , these actor neurons receive connections from place cells , representing the current position of the agent ., The spike trains generated by these neurons are filtered to produce a smooth firing rate , which is then multiplied by each neurons preferred direction ( see Models for all calculation details ) ., We finally sum these vectors to obtain the actual agent action at that particular time ., To ensure a clear choice of actions , we use a -winner-take-all lateral connectivity scheme: each neuron excites the neurons with similar tuning and inhibits all other neurons ( Figure 3B ) ., We manually adjusted the connection strength so that there was always a single “bump” of neurons active ., An example of the activity in the pool of actor neurons and the corresponding action readout over a ( successful ) trial is given in Figure 3C ., The corresponding maze trajectory is shown in Figure 3D ., In reinforcement learning , a successful agent has to balance exploration of unvisited states and actions in the search for new rewards , and exploitation of previously successful strategies ., In our network , the exploration/exploitation balance is the result of the bump dynamics ., To see this , let us consider a naive agent , characterized by uniform connections from the place cells to the actor neurons ., For this agent , the bump first forms at random and then drifts without preference in the action space ., This corresponds to random action choices , or full exploration ., After the agent has been rewarded for reaching the goal , synaptic weights linking particular place cells to a particular action will be strengthened ., This will increase the probability that the bump forms for that action the next time over ., Thus the action choice will become more deterministic , and the agent will exploit the knowledge it has acquired over previous trials ., Here , we propose to use the same learning rule for the actor neurons synapses as for those of the critic neurons ., The reason is the following ., Let us look at the case where : the critic is signaling that the recent sequence of actions taken by the agent has caused an unexpected reward ., This means that the association between the action neurons that have recently been active and the state neurons whose input they have received should be strengthened so that the same action is more likely to be taken again in the next occurrence of that state ., In the contrary case of a negative reinforcement signal , the connectivity to recently active action neurons should be weakened so that recently taken action are less likely to be taken again , leaving the way to , hopefully , better alternatives ., This is similar to the way in which the synapses from the state input to the critic neurons should be strengthened or weakened , depending on their pre- and postsynaptic activities ., This suggests that the action neurons should use the same synaptic learning rule as the one in Eq ., 17 , with now denoting the activity of the action neurons , but the signal still driven by the critic activity ., This is biologically plausible and consistent with our assumption that is communicated by a neuromodulator , which broadcasts information over a large fraction of the brain ., There are two critical effects of our -winner-take-all lateral connectivity scheme ., Firstly , it ensures that only neurons coding for similar actions can be active at the same time ., Because of the Hebbian part of the learning rule , this means that only those which are directly responsible for the action choice are subject to reinforcement , positive or negative ., Secondly , by forcing the activity of the action neurons to take the shape of a group of similarly tuned neurons , it effectively causes generalization across actions: neurons coding for actions similar to the one chosen will also be active , and thus will also be given credit for the outcome of the action 16 ., This is similar to the way the actor learns in non-neural actor-critic algorithms 18 , 19 , where only actions actually taken are credited by the learning rule ., Thus , although an infinite number of actions are possible at each position , the agent does not have to explore every single one of them ( an infinitely long task ! ) to learn the right strategy ., The fact that both the actor and the critic use the same learning rule is in contrast with the original formulation of the actor-critic network of Barto et al . 18 , where the critic learning rule is of the form “TD error×presynaptic activity” ., As discussed above , the “TD error×Hebbian LTP” form of the critic learning rule Eq ., 17 used here is a result of the exponential non-linearity used in the neuron model ., Using the same learning rule for the critic and the actor has the interesting property that a single biological plasticity mechanism has to be postulated to explain learning in both structures ., In the Morris water-maze , a rat or a mouse swims in an opaque-water pool , in search of a submerged platform ., It is assumed that the animal is mildly inconvenienced by the water , and is actively seeking refuge on the platform , the reaching of which it experiences as a positive ( rewarding ) event ., In our simulated navigation task , the learning agent ( modeling the animal ) is randomly placed at one out of four possible starting locations and moves in the two-dimensional space representing the pool ( Figure 4A ) ., Its goal is to reach the goal area ( of the total area ) which triggers the delivery of a reward signal and the end of the trial ., Because the attractor dynamics in the pool of actor neurons make it natural for the agent to follow a straight line , we made the problem harder by surrounding the goal with a U-shaped obstacle so that from three out of four starting positions , the agent has to turn at least once to reach the target ., Obstacles in the maze cause punishment ( negative reward ) when touched ., Similar to what is customary in animal experiments , unsuccessful trials were interrupted ( without reward delivery ) when they exceeded a maximum duration ., During a trial , the synapses continually update their efficacies according to the learning rule , Eq ., 17 ., When a trial ends , we simulate the animal being picked up from the pool by suppressing all place cell activity ., This results in a quick fading away of all neural activity , causing the filtered Hebbian term in the learning rule to vanish and learning to effectively stop ., After an inter-trial interval of 3s , the agent was positioned in a new random position , starting a new trial ., Figure 4B shows color-coded trajectories for a typical simulated agent ., The naive agent spends most of the early trials ( blue traces ) learning to avoid walls and obstacles ., The agent then encounters the goal , first at random through exploration , then repeatedly through reinforcement of the successful trajectories ., Later trials ( yellow to red traces ) show that the agent mostly exploits stereotypical trajectories it has learned to reach the target ., We can get interesting insight into what was learned during the trials shown in Figure 4B by examining the weight of the synapses from the place cells to actor or critic neurons ., Figure 4C shows the input strength to critic neurons as a color map for every possible position of the agent ., This is in effect a “value map”: the value the agent attributes to each position in the maze ., In the same graph , the synaptic weights to the actor neurons are illustrated by a vector field representing a “policy preference map” ., It is only a preference map , not a real policy map because the input from the place cells ( represented by the arrows ) compete with the lateral dynamics of the actor network , which is history-dependent ( not represented ) ., The value and policy maps that were learned are experience-dependent and unique to each agent: the agent shown in Figure 4B and C first discovered how to reach the target from the “north” ( N ) starting position ., It then discovered how to get to the N position from starting positions E and W , and finally to get to W from S . It has not however discovered the way from S to E . For that reason the value it attributes to the SE quarter is lower than to the symmetrically equivalent quarter SW ., Similarly the policy in the SE quarter is essentially undefined , whereas the policy in the SW quarter clearly points in the correct direction ., Figure 4D shows the distribution of latency – the time it takes to reach the goal – as a function of trials , for 100 agents ., Trials of naive agents end after an average of ( trials were interrupted after ) ., This value quickly decreases for agents using the TD-LTP learning rule ( green ) , as they learn to reach the reward reliably in about trials ., We previously remarked that the TD-LTP rule of Eq ., 17 is similar to TD-STDP , the TD-modulated version of the R-STDP rule 6 , 30 , 31 , at least in form ., To see whether they are also similar in effect , in our context , we simulated agents using the TD-STDP learning rule ( for both critic and actor synapses ) ., The blue line in Figure 4D show that the performance was only slightly worse tha | Introduction, Results, Discussion, Models | Animals repeat rewarded behaviors , but the physiological basis of reward-based learning has only been partially elucidated ., On one hand , experimental evidence shows that the neuromodulator dopamine carries information about rewards and affects synaptic plasticity ., On the other hand , the theory of reinforcement learning provides a framework for reward-based learning ., Recent models of reward-modulated spike-timing-dependent plasticity have made first steps towards bridging the gap between the two approaches , but faced two problems ., First , reinforcement learning is typically formulated in a discrete framework , ill-adapted to the description of natural situations ., Second , biologically plausible models of reward-modulated spike-timing-dependent plasticity require precise calculation of the reward prediction error , yet it remains to be shown how this can be computed by neurons ., Here we propose a solution to these problems by extending the continuous temporal difference ( TD ) learning of Doya ( 2000 ) to the case of spiking neurons in an actor-critic network operating in continuous time , and with continuous state and action representations ., In our model , the critic learns to predict expected future rewards in real time ., Its activity , together with actual rewards , conditions the delivery of a neuromodulatory TD signal to itself and to the actor , which is responsible for action choice ., In simulations , we show that such an architecture can solve a Morris water-maze-like navigation task , in a number of trials consistent with reported animal performance ., We also use our model to solve the acrobot and the cartpole problems , two complex motor control tasks ., Our model provides a plausible way of computing reward prediction error in the brain ., Moreover , the analytically derived learning rule is consistent with experimental evidence for dopamine-modulated spike-timing-dependent plasticity . | As every dog owner knows , animals repeat behaviors that earn them rewards ., But what is the brain machinery that underlies this reward-based learning ?, Experimental research points to plasticity of the synaptic connections between neurons , with an important role played by the neuromodulator dopamine , but the exact way synaptic activity and neuromodulation interact during learning is not precisely understood ., Here we propose a model explaining how reward signals might interplay with synaptic plasticity , and use the model to solve a simulated maze navigation task ., Our model extends an idea from the theory of reinforcement learning: one group of neurons form an “actor , ” responsible for choosing the direction of motion of the animal ., Another group of neurons , the “critic , ” whose role is to predict the rewards the actor will gain , uses the mismatch between actual and expected reward to teach the synapses feeding both groups ., Our learning agent learns to reliably navigate its maze to find the reward ., Remarkably , the synaptic learning rule that we derive from theoretical considerations is similar to previous rules based on experimental evidence . | computational neuroscience, biology, neuroscience, learning and memory | null |
journal.pcbi.1004370 | 2,015 | Multiscale Mechanical Model of the Pacinian Corpuscle Shows Depth and Anisotropy Contribute to the Receptor’s Characteristic Response to Indentation | Mechanoreceptors , a major component of the somatosensory system , detect specific physical stimuli and produce neural signals that give rise to sensations such as touch and pain 1 ., Cutaneous mechanoreceptors respond to mechanical stimuli and consist of afferent nerve fibers surrounded by specialized end organs that collectively encode a wide range of different touch sensations 2 , 3 ., The Pacinian corpuscle ( PC ) is a cutaneous mechanoreceptor that responds primarily to vibratory stimuli in the frequency range of 20–1000 Hz 4 , 5 ., The PC has low spatial sensitivity across the surface of the skin , and the receptive field of a single PC may span an entire hand 2 ., The PC is located in the dermis of glabrous skin 6 , 7 ., The PC has been widely studied because of its relatively large size ( Fig 1 ) ., The PC is about 1 mm in length by 0 . 67 mm in width and has an ovoid shape with a single myelinated nerve fiber located along the long axis of the receptor 8 , 9 ., The nerve fiber is surrounded by three main zones: an inner core , which contains bilaterally arranged cytoplasmic lamellae; an intermediate growth zone; and an outer core , which consists of 30 or more concentrically aligned collagenous lamellae 9 ., The lamellae are believed to act collectively as a high-pass filter that shields the nerve fiber at the receptor center from low frequency , high amplitude stimuli 2 , 10 , 11 ., The PC is a difficult organ to understand because its function involves a complex , interrelated set of biological , chemical , mechanical , and electrical phenomena ., Much recent work has focused on the development of the PC and identification of relevant transcription factors , such as c-Maf and ER-81 12 , 13 ., Release of GABA from the inner core has also been suggested to play a role in the rapid adaptivity of the PC 14 ., In the electrical and neural engineering arena , Lesniak and Gerling 15 have recently put forward computational models of the tactile mechanosensory system ., None of these studies , however , address the fundamental mechanics of the PC and the role of its structure in determining how skin displacement is transmitted to the PC neurite ., The work of Güçlü et al . 7 was an important exploration of PC mechanics ., They used finite-element modeling to investigate the role of the PC’s geometry in its mechanical response to static indentation ., Experimental data in which PCs were indented by cylindrical contactors with step waveforms of various amplitudes were compared to the computational models ., Semi-infinite plane and ovoid models produced similar displacements within the PC in response to static indentation , and neither model matched the localization of strain near the contractor seen in the experiment ., The purpose of this study was to test two specific hypotheses about the biomechanics of the PC ., First , Güçlü et al . rejected the hypothesis that receptor shape leads to the observed mechanical behavior of the PC , leaving open the question of how the strain concentrates near the indentation site; herein , we tested the hypothesis that mechanical anisotropy contributes to the strain localization ., Second , having concluded that a structural model of the PC is mechanically acceptable , we used that model to test the hypothesis that deep embedding within the skin contributes to the low spatial sensitivity and large receptive field of the PC ., A multiscale scheme 16–18 was used to model the response of the PC to indentation ., The method is summarized here and described in detail elsewhere 16–18 ., A finite-element model at the macroscopic level was coupled with representative volume elements ( RVE ) , each comprised of a fiber network , at the microscopic scale ( Fig 2 ) ., Each finite element contained eight Gauss points , each with an associated RVE ., Each RVE contained a network of 500–700 fibers in a constrained mixture ( cf . 19 , 20 ) with a nearly incompressible neo-Hookean matrix ., In this study , each element received a unique set of fiber networks depending upon its location within the mesh ., Macroscopic-level deformations were passed down to the microscopic level , the networks within each RVE were thus stretched , and the force exerted by each fiber , F , was calculated using the fiber constitutive equation, F=AB ( exp ( BEf ) −1 ), ( 1 ), where A is a measure of fiber stiffness , B is a measure of fiber nonlinearity , and Ef is the fiber Green strain computed from the fiber stretch , λf ,, Ef=0 . 5 ( λf2−1 ), ( 2 ), From the fiber forces on the RVE boundaries , the volume-averaged Cauchy stress at each Gauss point within the macroscopic element was calculated as, Sijmacro=1V∫VSijmicro\xa0dV=∑bcxiFj, ( 3 ), where V is the RVE volume , Sijmicro is the microscale stress , bc refers to summation over all network nodes on the RVE boundary , xi is the boundary fiber cross-link i-coordinate , and Fj is the force acting on the boundary fiber cross-link by the fiber in the j-direction ., The averaged stress balance was given as 17, ( 4 ), where uk is the RVE boundary displacement and nk is the normal vector to the RVE boundary ., The displacements were updated until the stress balance Eq ( 4 ) had equilibrated ., Simulations were run on 64 cores at the Minnesota Supercomputing Institute ., All finite-element meshes were generated in ABAQUS ., A mesh convergence study on the isolated PC problem gave average errors of 5% in nodal displacement between the coarsest mesh and the finest mesh ( which was used in the study ) ., Based on this result and our previous studies 16 , we expect our numerical results to have errors of at most 5% ., Delaunay networks were created to populate the RVEs within the multiscale model ., To capture the anisotropy of the collagenous lamellae within the PC , the networks for each finite element were aligned with the PC’s surface ( Fig 3A ) ., In the embedded models , the networks populating skin elements were made transverse orthotropic and aligned with the surface of the skin to reflect dermal collagen organization 21 ., Material constants A and B for eq ( 2 ) were set at 114 μN and 10 respectively , obtained from fitting the multiscale model to data from a published study on dermal mechanics 22 ., The Poisson’s ratio for the neo-Hookean matrix was set at ν = 0 . 47 for the simulations to model a nearly-incompressible matrix ., The matrix shear modulus was set at G = 4 . 2 kPa 16 ., The properties of the Pacinian corpuscle lamellae are not known , and estimates of the modulus of the lamellar layers have ranged from 1 KPa 23 to 0 . 5 MPa 11 ., Neither of those bounds was based on a published mechanical measurement of the properties of collagen fibers within the PC; the small value was estimated based on assumed high compliance of the basement membrane , and is consistent with an anecdotal reference from the literature ( 24 cited by 10 , but no published data in 24 ) , and the latter was based on arterial wall ., Recent work 23 has found that using a low modulus for the lamellar stiffness gives more accurate predictions of PC response in the high-frequency range , but collagen fibers have been found to have moduli in the MPa range 25 , as have basement membranes from the renal tubule 26 and ocular lens 27–29 ., It is thus clear that a better theoretical and structural description is needed , and it is likely that the choice of modulus will depend on the structure of that model ., Since the emphasis of this work was on the effect of anisotropy , not on differences in stiffness between the corpuscle and the surrounding skin ( which could be a very important factor and which should be explored in future studies ) , we chose to give the fibers in our PC model the same properties in as the fibers in the skin ., The isolated Pacinian corpuscle was meshed as a half-ellipsoid , with a major axis of length 1 mm and a minor axis of length 0 . 5 mm ., The PC model contained 2984 hexagonal elements ., The isolated PC model was subjected to 10 μm indentation by a 250 μm diameter indenter to simulate the experiments performed by Güçlü et al . and the associated finite-element model 7 ., The indenter displaced nodes vertically , as shown in Fig 3A ., For consistency with the Güçlü experiments , the nodal displacements in the isolated PC model were analyzed ., The displacement of nodes located along the top 100 μm of the z-axis was calculated after 10 μm indentation ., Strain ( εyy ) along the long axis of the PC was calculated for comparison to the embedded model ., The isolated PC was subjected to a 10 μm indentation with an indenter of diameter 250 μm to mimic the experiment and simulation of Güçlü et al . The multiscale model captured the nonlinear trend in displacement seen in the experimental data and not predicted by an isotropic linear elastic model ( Fig 4A ) ., The multiscale model populated with isotropic Delaunay networks rather than circumferentially-aligned networks also produced results similar to Güçlü et al . ’s isotropic linear elastic model ., Fig 4B shows the displacement of nodes along a cross section through the x-z plane of the PC at y = 0 ., As seen in Fig 4B , the multiscale model predicted a nonlinear spacing between nodes , with a greater nodal gap occurring with increasing depth ., The Von Mises stress was calculated at each element in the PC after 10 μm indentation for the cases of isotropic networks and circumferentially-aligned networks ., In Fig 5 , the isotropic case shows stress of approximately 2x higher magnitude around the indenter than those shown for the circumferentially-aligned case ., The strain along the long axis of the PC was calculated over 25 steps of 1 μm indentation ., As seen in Fig 6 , the strain along the long axis of the axon increased with indentation into the PC ., The strain calculated for the isotropic network case was approximately 10x higher than that calculated for the circumferentially-aligned network case ., The long-axis strain along the PC resulting from indentation at various nodes along the surface was also compared for the epidermis and dermis models ( Fig 7 ) ., The epidermal PC model showed large strain in response to loading directly above the PC that dropped off quickly as the indenter moved away from the PC ., This drop-off implies more spatial sensitivity and thus a smaller receptive field ., The dermal PC model shows less indenter position dependence and thus a larger receptive field ., The Von Mises strain was calculated within every element in the dermal- and epidermal-embedded cases after 10 μm indentation for indentations above the center of the PC and 750 μm down its long and short axes ., As seen in Fig 8 , the Von Mises strain in the PC in the dermis case shows little variation when the structure is indented at different locations ., In both cases , PC strain is less than that of the immediately surrounding tissue because of its greater degree of anisotropy ., The strain in the PC in the epidermis case shows greater variations with indenter location ., The long-axis strain along the PC resulting from 10 μm indentation at various nodes along the surface of the skin was compared for horizontally-aligned and vertically-aligned PCs embedded within a dermal mesh ( Fig 9 ) ., The horizontal PC model showed positive strains or no strain resulting from indentation ., The vertical PC model always showed negative strains in response to indentation ., This result shows that indentation within the receptive field of a horizontally-aligned PC always results in positive axial stretch of the neurite ., Indentation of the vertical model does not result in neurite stretch ., This study used a multiscale finite-element model to determine that the structure of the PC is an important contributor to the nonlinear behavior of the receptor ., In addition , it showed that the deep dermal location of the PC provides it with lower spatial sensitivity ., Several factors must be considered when interpreting our results ., First , the mechanical stimulus was a fixed indentation into the PC with no transient effects ., As such , this study addresses the location and magnitude of the stimulus but it does not take stimulus frequency into account when determining the PC mechanical response , as others have 11 ., A static model was chosen as suitable for comparison with experiment 7 , but it would not be appropriate for simulating the vibrotactile response ., While Hubbard 31 also investigated PC mechanics , the results from the current paper cannot be directly compared to that study , which placed the PC within a hinged apparatus rather than stimulating with a vertical indenter ., Therefore , the experiments performed by Güçlü et al . were used to validate the current model ., Second , the PC was treated as incompressible , and no fluid movement was allowed within the PC even though such flow is known to be important 11; thus , our model must be interpreted as the instantaneous response of the PC ., The time-dependent response of the PC is necessary to address , as it is crucial to the PC’s role as a high-pass filter to vibration 5 , 11 , so insights from the current study should center on instantaneous response ., Also , the experimental PC literature 4 , 5 , 32 focuses appropriately on the use of directly applied sinusoidal displacements to elicit the response of the PC to vibratory stimuli , providing a rich data set on the dynamical response of the PC ., A model of PC mechanics should include a dynamical component ( fluid flow , viscoelasticity , or both ) to account for the phase difference that can occur between skin and PC stimulation , and also between PC stimulation and receptor response ., The mechanical model used in this study also simplified the structure of the PC to account only for anisotropy within the receptor and not for its specific components and detailed structure ., The receptor capsule is composed of concentrically-arranged collagenous lamellae through which mechanical forces are transduced ., The lamellae consist of condensed cell layers separated by layers of pressurized fluid 33 ., The structure of the capsule is believed to play an important role in determining which mechanical forces are transmitted to the axon 34 ., Each lamella contains a single layer of flat squamous epithelial type cells , with interlamellar spacing increasing with distance from the inner core 10 ., Tight junctions between cells within each lamella prevent fluid flow across lamellae , but flow within the fluid layer between adjacent lamellae is possible and is significant at low frequencies ., Loewenstein and Skalak first proposed that the role of the PC capsule’s lamellar structure is that of a series of mechanical high-pass filters to shield the nerve fiber at the center of the receptor from low frequency , high amplitude stimuli 11 ., To place this model in the broader context , it is more advanced than that of Güçlü , which is isotropic and linear elastic , but does not provide the single-lamellar-level description of Lowenstein and Skalak or of subsequent variations thereon 23 , 35 ., The multi-scale approach of the current model , in which RVE’s are introduced with position-dependent fiber orientations , could be extended to more complex microstructures as greater structural information becomes available ., It is also notable that the lamella-based models 11 , 23 , 35 can account for the apparent viscoelasticity of the tissue by incorporating interlamellar flow; our model would not be able to do so but could incorporate a continuous viscous contribution similar to that derived previously 33 in a homogenized model of the PC ., Clearly , there is need for a more detailed microstructural model that can address other aspects of PC behavior and perhaps can resolve the disconnect between the high stiffness typical of collagenous tissues and the low stiffness reported experimentally 24 and used to describe vibrotactile mechanics of the PC 23 ., The mechanical model presented in this study simplified the neurophysiology of the PC action potential generation into axial stretch of its central nerve fiber ., The solid black line in Fig 7 indicates the division between the positive and negative stretch ., Because the exact mechanism of axon excitation is unknown , it is possible that the neurite could also be stimulated during compression along the long axis ., Long-axis compression could , for example , be experienced as positive stretch in other directions due to neurite incompressibility ., Thus , while it is possible that long-axis compression could lead to axon excitation , only long-axis stretch was considered in this study ., It was initially proposed by Gray & Ritchie 30 that sensory receptors respond to mechanical stimulation resulting from nerve stretch ., After PC compression studies performed by Hubbard in 1958 were unable to measure a change in axon length within the error of measurement , other possible mechanical mechanisms for transduction were proposed 31 ., Specifically , Hubbard proposed that a change in the ratio between the major and minor axes of the cross section of the nerve fiber leads to a change in surface area and thus membrane stretch ., Fig 10 shows a comparison between the model of axon stretch along the long axis used in this study and the area change proposed by Hubbard ., Both long-axis strain and area strain increase monotonically with indentation into the PC ., The area change of the PC is approximately six times the long-axis strain after 25 μm of indentation ., The orientation of the PC with respect to the surface of the skin was analyzed in this study ., The results presented in Fig 9 show that indentation within the receptive field of a horizontally-aligned PC always results in either axial stretch of the neurite or no neurite stretch ., Indentation within the receptive field of a vertically-aligned PC always results in neurite compression ., The assumption used in this study that PC action potential generation is the result of axial stretch of the neurite and not axial compression means that the vertically-aligned PC was never activated by static compression ., This study only shows the results of two PC orientations to static compression ., A vibratory stimulus would involve both application and removal of an indenter due to oscillations , creating a more complex field that would likely include long-axis stretch in both cases ., Thus , it is expected that different results on PC orientation would be obtained from models of vibration ., It has previously been shown 36 that the electrophysiological response of an intact PC or its decapsulated nerve terminal changes polarity as it is rotated 90 degrees along its long axis ., The same study also showed that a decapsulated terminal reversed the polarity of its neural response when compressed horizontally to the nerve or vertically to the nerve and proposed that the bilateral arrangement of cells around the terminal is responsible for changing how compression is transmitted to the neurite in these different orientations ., It has also been shown 37 that compression of an intact PC along its long axis requires a much stronger stimulus to cause depolarization than that required in compression along the short axis ., The current study models the transmission of mechanical stimuli through the lamellae to the neurite , and showed differences with PC orientation , but does not address the electrophysiological effects reported by others 36 , 37 ., A combined model ( cf . 33 , 38 ) could lend greater insight into the interaction between mechanical and electrophysiological events ., Incorporation of isotropic Delaunay networks into the PC multiscale model rather than circumferentially-aligned networks resulted in a lack of shape dependence similar to that observed in Güçlü’s finite-element study 7 ., The current study thus confirms the finding that the ellipsoidal shape of the PC is not per se responsible for the observed mechanical behavior ., In addition , the use of isotropic Delaunay networks further confirmed Güçlü‘s result that a homogeneous isotropic model of the PC cannot predict the experimentally observed displacement pattern in response to indentation ., This study showed that the internal anisotropic structure is an important factor leading to the nonlinear displacements through the PC ., The nonlinear reduction in displacement of lamellae located closer to the central core is in agreement with the hypothesis that the lamellar structure can help protect the nerve from large deformations under large skin surface loads ., The current model bases the mechanical properties of the networks representing the skin on data from uniaxial mechanical tests on dermis 22 ., There are many factors that can influence the mechanics of skin , which can vary with anatomical location , proximity to blood vessels , thickness , body weight , hair-cycle stage , skin disease , and experimental conditions such as humidity 39–42 ., Specifically , mechanical and structural properties such as viscoelasticity and anisotropy of skin can vary with age and anatomical location 42 ., The mechanical properties of a skin sample are influenced by structural components such as the collageneous fiber network and the presence of different layers , which exhibit different mechanical properties , within the skin 35 , 40 , 41 , 43 ., Selecting different data for fitting of the skin element mechanical properties used in this study would likely change the quantitative results of the current study due to differences in the aforementioned factors ., The overall qualitative results of this study , however , are not expected to change ., The mechanics of skin can also vary depending on the type of load applied , as skin behaves differently under compression and tension 39 , 40 ., In vivo skin can also be in different amounts of tension depending on anatomical location , the body position , and the individual 42 ., All of the aforementioned factors could be considered in future models , with the current model functioning as a basis for subsequent studies and as a low-order model of skin behavior ., Past experiments have shown that the isolated PC is a highly sensitive mechanoreceptor with nanometer sensitivity 2 ., The simulations performed in this study model the sensitivity of an isolated and embedded PC to micrometer-scale indentations ., Sensitivity thresholds of the PC have been reported previously as 3 nm applied directly to the PC and 10 nm applied to the surface of the skin 2 ., In the current study , the smallest indentation tested on the isolated PC was 1 μm , which corresponds to an amplitude of approximately 7 μm applied to the skin after comparison of the strain along the long axis of the PC in the isolated and dermis-embedded models ., This ratio could change for a more anatomically detailed model , in which multiple receptors rather than just one are located within the dermis ., There are currently no published experimental data on the mechanics on a PC embedded in skin ., While this work focused on the PC , its results can also provide insight into other cutaneous mechanoreceptors ., The embedded PC simulations showed that a PC located in the dermis of the skin was able to replicate the low spatial sensitivity of a PC in vivo ., The PC embedded in the epidermis had a higher spatial sensitivity within the receptive field tested in the simulations ., Receptors located closer to the surface of the skin , such as the Meissner corpuscle and Merkel cell-neurite complex , show decreased receptive fields and thus increased spatial sensitivity to mechanical events on the skin surface 44 ., The current study could also be expanded to include the different geometries and cutaneous locations of other mechanoreceptors . | Introduction, Methods, Results, Discussion | Cutaneous mechanoreceptors transduce different tactile stimuli into neural signals that produce distinct sensations of touch ., The Pacinian corpuscle ( PC ) , a cutaneous mechanoreceptor located deep within the dermis of the skin , detects high frequency vibrations that occur within its large receptive field ., The PC is comprised of lamellae that surround the nerve fiber at its core ., We hypothesized that a layered , anisotropic structure , embedded deep within the skin , would produce the nonlinear strain transmission and low spatial sensitivity characteristic of the PC ., A multiscale finite-element model was used to model the equilibrium response of the PC to indentation ., The first simulation considered an isolated PC with fiber networks aligned with the PC’s surface ., The PC was subjected to a 10 μm indentation by a 250 μm diameter indenter ., The multiscale model captured the nonlinear strain transmission through the PC , predicting decreased compressive strain with proximity to the receptor’s core , as seen experimentally by others ., The second set of simulations considered a single PC embedded epidermally ( shallow ) or dermally ( deep ) to model the PC’s location within the skin ., The embedded models were subjected to 10 μm indentations at a series of locations on the surface of the skin ., Strain along the long axis of the PC was calculated after indentation to simulate stretch along the nerve fiber at the center of the PC ., Receptive fields for the epidermis and dermis models were constructed by mapping the long-axis strain after indentation at each point on the surface of the skin mesh ., The dermis model resulted in a larger receptive field , as the calculated strain showed less indenter location dependence than in the epidermis model . | We performed computer simulations of the mechanical behavior of the Pacinian corpuscle ( PC ) , a sensory receptor in the skin that helps detect short-term contact and high-frequency vibration ., The PC is composed of a series of tissue layers , and we found that this characteristic structure may explain the response of the PC in indentation experiments ., We also found that the deep placement of the PC within the skin allows it to detect stimuli over a wide area of the skin but not to distinguish the specific location of the stimulus ., These findings are a step but still an early step towards the broad goal of understanding how mechanical stimuli to the skin are translated into neural signals to the spinal cord and brain , and many open questions still remain about how the different sensors of the nervous system work together to create our sense of touch . | null | null |
journal.pntd.0007183 | 2,019 | A novel diagnostic algorithm equipped on an automated hematology analyzer to differentiate between common causes of febrile illness in Southeast Asia | Arboviruses and bacterial infections such as salmonellosis , leptospirosis , and rickettsiosis are common causes of acute febrile illness in tropical and subtropical countries 1–3 ., Discriminating between these infections is of great importance to triage patients in need of antibiotics or monitoring for dengue complications ., In daily practice , dengue and bacterial infections are often diagnosed on clinical grounds and many patients are prescribed antibiotics without laboratory confirmation of a bacterial infection ., Confirmatory microbiological tests , including blood cultures , serology , molecular tests , and antigen- or antibody-based rapid tests are frequently unavailable and suffer from important diagnostic limitations ., An alternative for pathogen-specific diagnostic tests is the assessment of the host immune response , using biomarkers such as C-reactive protein ( CRP ) or procalcitonin ( PCT ) 4 , 5 ., Disease-specific changes in circulating blood cells may also be helpful , for example , leukopenia and thrombocytopenia support a diagnosis of dengue 6 ., The discriminatory performance of cell numbers alone is , however , insufficient for clinical decision-making ., A promising development is the ability to measure phenotypic changes in blood cells by automated hematology analyzers ., For example , activated leukocytes contain more lipid rafts in their cell membrane and altered intracellular DNA/RNA levels 7 which can be quantified using specific reagents and distinct fluorescence patterns 8 , 9 ., Based on the principle that different infections evoke different patterns in blood cell number and phenotype , a diagnostic algorithm called the Infection Manager System ( IMS ) , was developed for use on Sysmex hematology analyzers ., The IMS indicates whether an inflammatory response is present and whether an arboviral , bacterial , or malarial origin is suspected ., The aim of our present study was to enroll adult patients with common causes of undifferentiated fever in Southeast Asia in order to train and evaluate the diagnostic performance of the IMS for these infections , as well as to compare the diagnostic performance against CRP and PCT ., A prospective cohort study was conducted between July 2014 and February 2016 in three hospitals ( Hasan Sadikin University Hospital , Salamun General Hospital , and Cibabat General Hospital ) and two primary care outpatient clinics , all located in Greater Bandung , the capital of the West Java province in Indonesia ., Patients aged 14 years and above presenting an acute febrile illness and clinical suspicion of an arboviral infection , salmonellosis , leptospirosis , rickettsiosis , or any other common bacterial infection were enrolled ., Exclusion criteria included pregnancy and the suspicion of a chronic infection , such as tuberculosis or HIV , and severe concomitant conditions like dialysis , autoimmune diseases , or malignancies ., The sample size of 600 individuals was based on the assumption that a proven or probable bacterial or arboviral infection could be diagnosed in 50% of enrolled patients and that enteric fever , leptospirosis , or rickettsiosis could be diagnosed in approximately 20% ( n = 30 ) of subjects with a proven or probable bacterial infection ., To test how often the IMS flags an inflammatory response in healthy adults , the trained IMS was also tested in a cohort of healthy Dutch adults , derived from a well-established prospective population-based study , incorporating 13 , 432 individuals from the north of the Netherlands ( www . lifelines . nl ) ., The first selection of patients was done by treating physicians at the participating health facilities on the basis of clinical features and routine additional examinations ., Demographic data , medical history , physical examination , results of laboratory and radiology tests , and suspected diagnosis were recorded in a standardized electronic study case report form ., All admitted patients were followed up three days after enrolment to evaluate the clinical picture and perform additional diagnostic tests on indication ., A policlinic visit was planned with the same purpose between days 7–14 after enrolment day ., Non-admitted patients were followed up twice: first appointment between 2–7 days after enrolment , a second appointment within one week thereafter ., Fig 1 summarizes the study flow and diagnostic procedures ., Blood was drawn at inclusion in all patients for immediate hemocytometry and microbiological testing ., EDTA plasma , serum , and whole blood were stored at -80°C for additional microbiological tests ., Initial microbiological tests were performed at the discretion of the treating physician ., These included the performance of blood cultures in patients with a suspected bacterial sepsis or enteric fever , pus cultures in case of an abscess , and dengue NS1 rapid test or serological tests for suspected dengue , enteric fever , or leptospirosis ., Radiological examinations such as a chest X-ray were performed on indication ., Next , stored blood of all enrolled subjects was tested using the following diagnostic algorithm: dengue diagnostics were performed using a dengue NS1 antigen rapid diagnostic test ( RDT ) , and if negative , paired dengue IgM and IgG serology and dengue PCR ., Furthermore , RDTs or serology were done on all samples for chikungunya IgM , Salmonella IgM ( Tubex® ) , and Leptospira IgM ( Panbio® ) ., In case of a positive chikungunya IgM , Salmonella IgM score ≥4 or a positive Leptospira IgM , specific serum or whole blood PCRs for these pathogens were performed ., The remaining cases without a proven diagnosis were tested for Rickettsia typhi IgM and IgG , followed by a specific R . typhi real-time PCR in case of a positive result ., The following case definitions were used: a proven dengue virus infection was defined as:, i ) positive result of NS1 RDT or dengue PCR , or, ii ) seroconversion of anti-dengue IgM and/or IgG , or, iii ) fourfold or greater increase of anti-dengue IgG titers in convalescent serum ., Chikungunya or Leptospirosis were proven when the PCR was positive ., Salmonellosis was proven when Salmonella spp ., were isolated from blood culture or when the whole blood Salmonella PCR was positive ., Murine typhus was proven when there was seroconversion or a four-fold increase in IgM or IgG R . typhi titer or a positive PCR on the buffy coat ., A proven cosmopolitan bacterial infection was defined as isolation of a pathogenic pathogen from blood culture or other sterile location , or by a combination of clinical features and results of radiology , for example in case of pneumonia ., Malaria was proven if Plasmodium parasites were detected on a blood smear ., In case no proven diagnosis was obtained , two experienced clinicians ( AvdV and QdM ) graded the remainder of the cases as probable or possible arboviral or bacterial infection without any further sub-classification or as fever from unknown origin ., Grading was done using all clinical data and additional investigations , but without results of IMS and CRP or PCT ., Hemocytometry was done on EDTA blood within 4 hours using Sysmex XN-1000 , Sysmex XN-550 , and a regular Sysmex XE-5000 analyzer ., Details of the performed microbiological tests and the CRP and PCT measurements are given in S1 Table ., The IMS is based on novel techniques that quantify cellular activation and cell membrane composition using distinct fluorescence and surfactant reagents that target RNA , DNA , and bioactive lipids , respectively 8–10 ., The IMS algorithm is given in S1 Fig . The IMS first flags whether an inflammatory response is detected and if so , whether it fits a bacterial , ( arbo ) viral , or malarial origin or cannot be classified and designated as an unspecified inflammatory response ., When no inflammatory reaction is noticed , no message is given ., The sponsor was not involved in data acquisition , including results of hemocytometry or microbiological assays ., Employees of the sponsor were involved in the training of the IMS algorithm using the first 200 enrolled cases with the goal to further optimize the IMS performance ., For this training , the sponsor had access to clinical information , results from microbiology and radiology examinations , and the tentative cause of the febrile illness as classified by the clinical study team ., Results of PCRs and CRP/PCT were not yet available at that time ., Next , the final version of the IMS was tested on all evaluable cases with employees of the sponsor classifying all enrolled patients into: no sign of inflammation , or suspected arboviral , bacterial , malarial , or unspecified inflammation ., For this classification , the sponsor was blinded to all clinical data , results from additional tests and the final classification by the study team of the cause of the febrile illness ., Whereas the IMS classification was performed by the sponsor in this feasibility study , the intention is to create an analyzer that directly reports the IMS classification after measurement of the blood sample without requiring data to be sent to another site for analysis ., For CRP and PCT the following cut-off levels were evaluated in predicting a bacterial etiology of fever: for CRP >20 mg/L and >40 mg/L and for PCT >0 . 5 ng/mL and >2 . 0 ng/mL plasma levels upon admission , respectively 2 ., For additional analyses , a special group named ‘antibiotics’ , was created , containing individuals who were flagged as either bacterial or unspecified inflammation by the IMS , as antibiotics may be indicated in these cases ., Patients with a proven concomitant arboviral-bacterial infection were also classified as bacterial infection ., Descriptive statistics were conducted for all variables ., Differences in hematology parameters between groups were analyzed using Wilcoxon rank sum test in case of two groups and Kruskal-Wallis test in case of more than two groups ., All statistical analyses were performed using R ( R Core Team ( 2016 ) ) ., All procedures followed were in accordance with the ethical standards of the Helsinki Declaration ., All study participants provided written informed consent ., In patients aged 14–18 years , a parent or guardian provided informed consent with written assent by the child ., The study protocol was approved by the Ethics Committee of Hasan Sadikin General Hospital ( LB . 02 . 01/C02/515/I/2015 , LB . 02 . 01/C02/2352/II/2016 ) ., A total of 600 patients were enrolled ., A total number of 137 patients were subsequently excluded because of missing data , mostly because of insufficient follow-up while no proven diagnosis was made ., From the remaining 463 subjects , 342 patients could be classified as having a proven , probable , or possible arboviral , bacterial , combined arboviral-bacterial , or malaria infection ( Fig 1 ) ., A total number of 89 individuals had a proven arboviral infection: 81 cases with dengue , based on a positive result of a dengue NS1 antigen test ( n = 68 ) , IgM dengue seroconversion ( n = 9 ) , or dengue PCR ( n = 4 ) and eight cases with chikungunya ., Three patients with IgM dengue seroconversion also had a bacteremia ( two Salmonella spp . and one Staphylococcus aureus ) ., A total of 94 patients had a proven bacterial infection: murine typhus ( n = 26 ) , salmonellosis ( n = 16 ) , leptospirosis ( n = 6 ) and cosmopolitan bacterial infections , including bacteremia ( n = 13 ) , community-acquired pneumonia ( n = 15 ) , skin or soft tissue infection ( n = 11 ) , urinary tract infection ( n = 5 ) and single cases of puerperal infection and peritonitis ., A total number of 121 patients were classified as unknown origin of infection ., Baseline characteristics of participants with a proven infection are summarized in Table 1; characteristics of participants with proven or probable infections are given in S2 Table ., In total , 82% of the enrolled patients were hospitalized and ten patients died during hospitalization , all from the proven bacterial group ., Fig 2 and Fig 3 show the results of a selection of novel leukocyte parameters per infection or aggregated in arboviral or bacterial infections ., Whereas there was a large overlap in the number of activated neutrophils ( Neut-RI ) and monocytes ( Re-Mono ) across the different infections , dengue was characterized by a marked increase in AS-Lymph and Re-Lymph , which are considered to represent plasma cells and reactive lymphocytes , respectively ., In contrast , chikungunya was not associated with increased AS-Lymph or Re-Lymph ., Participants with the intracellular bacterial infections salmonellosis and murine typhus also had significantly higher Re-Lymph than those with other bacterial infections ( salmonellosis vs . leptospirosis P = 0 . 006; salmonellosis vs . cosmopolitan bacterial infection P< 0 . 0001; murine typhus vs . leptospirosis P = 0 . 007; murine typhus vs . cosmopolitan bacterial infections P< 0 . 0001 ) ., Table 2 summarizes the diagnostic performance of the IMS ., An inflammatory response was flagged in all but two cases; one case of dengue in whom the dengue diagnosis was based on IgM seroconversion , and one patient with salmonellosis ., Overall , the sensitivity , specificity , positive predictive value ( PPV ) , and negative predictive value ( NPV ) of the IMS for arboviral infections were 69 . 7% , 97 . 9% , 96 . 9% and 77 . 3% , respectively , and for bacterial infections 77 . 7% , 93 . 3% , 92 . 4% and 79 . 8% ., Inflammation remained unclassified in 19 . 1% and 22 . 5% of patients with a proven bacterial or arboviral infection , respectively ., Importantly , six out of seven ( 86% ) cases with proven chikungunya were classified as unspecified inflammation ., Similarly , a relatively high proportion of cases with murine typhus were either classified as unspecified inflammation ( 27% ) or arboviral inflammation ( 8% ) ., None of the other proven or probable bacterial infections were classified as arboviral ., The three cases with a combined arboviral-bacterial infection were all flagged as bacterial infection ., One of four malaria cases was not correctly flagged as being malaria ., Fig 4A shows CRP and PCT plasma levels at study enrolment per infection , and Fig 4B provides these levels for cases aggregated in proven or proven/probable bacterial or arboviral etiology ., In the proven cases , a bacterial etiology was associated with significantly higher CRP and PCT levels than a proven arboviral etiology with median ( IQR ) CRP levels of 110mg/L ( 52-192mg/L ) vs . 11mg/L ( 5-23mg/L; P<0 . 0001 ) and PCT levels of 2 . 6ng/mL ( 0 . 8–7 . 5ng/mL ) and 0 . 4ng/mL ( 0 . 2–0 . 7ng/mL; P<0 . 0001 ) , respectively ( Table 1 and Fig 4A ) ., Table 3 summarizes the diagnostic performance of the IMS compared with CRP and PCT ., A special category , named ‘antibiotics’ , was created for the IMS result , containing individuals who were flagged as either bacterial or unspecified inflammation by the IMS , as antibiotics may be indicated in these ., In total , 88% and 84% of bacterial cases had CRP levels above the pre-defined cut-offs of >20mg/L or >40mg/L , respectively , whereas 81% and 54% had PCT levels >0 . 5ng/mL or >2 . 0ng/mL , respectively ., For the arboviral group , 72% and 91% of cases had CRP levels below these cut-offs and 55% and 93% PCT levels below these cut-offs , respectively ., The optimal CRP plasma level cut-off to distinguish between a bacterial and viral etiology was 36 . 6 mg/L ( sensitivity 85 . 1% with specificity 91 . 0%; area under the receiver operating characteristic ( ROC ) curve 0 . 92 ) and for PCT 0 . 96ng/mL ( sensitivity 72 . 3%; specificity 83 . 1%; area under the ROC curve 0 . 81 ) ., Overall , CRP with a cut-off of 40mg/L had a somewhat higher sensitivity for bacterial infections than the IMS with a somewhat lower specificity ., Using the ‘antibiotics’ classification in IMS shifted the balance to a higher sensitivity and higher NPV , but lower specificity compared with CRP ., PCT performed less well than either the IMS or CRP ., Finally , we determined how frequently the IMS flags an inflammatory response in healthy individuals ., A total of 13 , 432 Dutch subjects were available from the lifelines cohort that had no sign or symptoms of illness or abnormality on routine laboratory examination and in whom IMS data were accessible as well ., The IMS indicated an unspecified inflammatory response in five participants ., The main finding of the present study is that a novel diagnostic algorithm operating on an automated Sysmex hematology analyzer , called the IMS , is capable of confirming the presence of an infection in Indonesian adults presenting with an acute febrile illness and discriminate arboviral from bacterial infections ., The IMS is based on the principle that pathogens induce specific changes in the number and phenotype of circulating blood cells and that these changes can differentiate viral from bacterial infections ., The idea that algorithms incorporating novel blood count parameters may be used as decision tools for antibiotic therapy is supported by recent studies in febrile children 9 and ICU patients 11 , 12 ., In resource-limited countries , costly and expertise-reliant diagnostic assays cannot be performed routinely ., The IMS has the advantage that it operates on a standard hematology analyzer with results being available within a few minutes at an affordable price ., In health facilities with a hematology analyzer , the IMS holds promise as an alternative for pathogen-specific RDTs or host biomarker tests , and as a tool for a more targeted use of pathogen-specific diagnostic assays ., In addition , in patients with dengue , daily hemocytometry is advised to monitor platelet and leukocyte counts ., This offers a unique opportunity to combine diagnostics with clinical monitoring ., The arboviral group in our study mainly comprised of dengue cases ., Dengue is the most common arboviral infection with more than one third of the worlds population living in areas at risk for infection 13 ., Dengue was characterized by increases in antibody synthesizing ( AS-Lymph ) and reactive lymphocytes ( Re-Lymph ) , in combination with thrombocytopenia and a high immature platelet fraction ., Polyclonal plasmacytosis has previously been reported to be a feature of dengue infections 14 , 15 ., In chikungunya cases , elevations in AS-Lymph and Re-Lymph were not observed and 86% of chikungunya infections were classified as ‘unspecified inflammation’ ., The diagnostic performance of the IMS for viral infections other than dengue , including common respiratory infections and other arboviruses such as Zika , therefore awaits to be determined ., Bacterial infections were also aggregated into one group because of relatively low numbers per group ., Interestingly , Salmonella spp ., and R . typhi are intracellular growing bacteria and infections with these pathogens elicited a distinct pattern with a significantly higher Re-Lymph ., Therefore , our data suggest that the IMS also has the potential to differentiate among specific subtypes of bacterial infections ., IMS classified a substantial number of infections as ‘unspecified’ inflammation ., Because antimicrobial therapy may still be warranted in conditions flagged as unspecified inflammation , a category ‘antibiotics’ was created ., The NPV of the IMS for the ‘antibiotics’ category was high ( 95 . 5% ) , suggesting that the IMS holds promise to improve the correct use of antibiotics as well as antimicrobial stewardship in these settings ., Dengue-bacterial co-infections are probably underestimated and withholding antibiotics may have severe consequences 16 ., Fortunately , in the three patients with a proven double infection in our study , the IMS scored all as bacterial infections ., The IMS can also provide an indication on the presence of malaria , but novel techniques using laser technologies and reagents specifically designed for malaria detection using Sysmex analyzers are currently under clinical evaluation ( ClinicalTrials . gov Identifier: NCT02669823 ) ., Overall , the trained IMS performed comparable to CRP with the latter having a slightly higher sensitivity but lower specificity to diagnose bacterial infections ., Including cases with unclassified inflammation in the bacterial etiology group ( ‘antibiotics’ category ) , the balance shifted to a higher sensitivity , but lower specificity ., Cut-offs for clinical decision making depend on the clinical setting ., So far , only a few studies have reported CRP or PCT levels in tropical infections 2 , 17 ., Our findings are comparable to those by Wangrangsimakul et al . who also found a CRP level of 36mg/L as the optimal cut-off level to distinguish between bacterial and viral causes of undifferentiated fever in Thailand 2 ., We enrolled patients suspected of having specific infections that are very common throughout much of Southeast Asia ( e . g . dengue , enteric fever , leptospirosis , murine typhus ) and our findings are therefore most likely applicable to areas outside Indonesia ., The performance of the IMS in areas with a different infection epidemiology is currently unknown ., Results of a diagnostic study investigating the performance of the IMS in Sub-Saharan Africa are expected in the coming year ( ClinicalTrials . gov , NCT02669823 ) ., The IMS software operates on routine hematology analyzers ( Sysmex XN series ) and results are provided within one minute ., The costs associated with the assay are expected to be in the range of a regular full blood count ., A full blood count is among the most commonly performed laboratory tests–also in resource-poor areas in Asia–and introduction of the IMS algorithm is especially promising for the workup of febrile patients in larger healthcare facilities where hemocytometry analyzers are already in routine use , but which lack facilities for more specialized microbiological assays ., Limitations of the present study are that proof of infection , using microbiology or imaging studies , was obtained in only 35% of cases ., Our results do not however differ very much from other similar studies in low-income settings 18 , 19 ., Secondly , we used stringent microbiological criteria ., Despite our efforts to include as much ‘tropical’ infections as possible , the total number of proven tropical bacterial infections remained limited ., In line with other studies , we also found that murine typhus is an important and often unrecognized infection 2 , 20 , 21 ., Thirdly , our study did not include consecutive febrile patients , but limited selection to those patients suspected of having a specific type of infection in order to train the IMS algorithm ., This , together with the stringent microbiological criteria , may have led to selection bias , e . g . dengue patients of whom the majority had a positive NS1 antigen test ., Confirmatory validation studies enrolling consecutive febrile patients are therefore required ., Lastly , a cohort of healthy Dutch instead of Indonesian individuals was used to test how frequently the trained IMS indicates inflammation in absence of an infection ., Inclusion of a large control population from the same demography would have been preferred , because factors such as ethnicity and living conditions may influence hematological reference ranges ., Nonetheless , earlier data showed that reference ranges on Sysmex analyzers in a Dutch and Asian ( Indian ) population of healthy adults were fairly similar 22 , 23 , suggesting that important differences in IMS performance are not expected ., Age-related differences in reference ranges are bigger , especially between children below the age of six years and adults ., Our study did not include children and it is important to emphasize that the IMS first needs validation in children as well as other healthy and patient populations in other areas before it can be introduced on a routine basis ., In conclusion , the IMS is a promising novel diagnostic algorithm that can be equipped on a standard hematology analyzer and can be used to triage patients in need of antibiotics or monitoring for dengue complications . | Introduction, Methods, Results, Discussion | Distinguishing arboviral infections from bacterial causes of febrile illness is of great importance for clinical management ., The Infection Manager System ( IMS ) is a novel diagnostic algorithm equipped on a Sysmex hematology analyzer that evaluates the host response using novel techniques that quantify cellular activation and cell membrane composition ., The aim of this study was to train and validate the IMS to differentiate between arboviral and common bacterial infections in Southeast Asia and compare its performance against C-reactive protein ( CRP ) and procalcitonin ( PCT ) ., 600 adult Indonesian patients with acute febrile illness were enrolled in a prospective cohort study and analyzed using a structured diagnostic protocol ., The IMS was first trained on the first 200 patients and subsequently validated using the complete cohort ., A definite infectious etiology could be determined in 190 of 463 evaluable patients ( 41% ) , including 89 arboviral infections ( 81 dengue and 8 chikungunya ) , 94 bacterial infections ( 26 murine typhus , 16 salmonellosis , 6 leptospirosis and 46 cosmopolitan bacterial infections ) , 3 concomitant arboviral-bacterial infections , and 4 malaria infections ., The IMS detected inflammation in all but two participants ., The sensitivity , specificity , positive predictive value ( PPV ) , and negative predictive value ( NPV ) of the IMS for arboviral infections were 69 . 7% , 97 . 9% , 96 . 9% , and 77 . 3% , respectively , and for bacterial infections 77 . 7% , 93 . 3% , 92 . 4% , and 79 . 8% ., Inflammation remained unclassified in 19 . 1% and 22 . 5% of patients with a proven bacterial or arboviral infection ., When cases of unclassified inflammation were grouped in the bacterial etiology group , the NPV for bacterial infection was 95 . 5% ., IMS performed comparable to CRP and outperformed PCT in this cohort ., The IMS is an automated , easy to use , novel diagnostic tool that allows rapid differentiation between common causes of febrile illness in Southeast Asia . | Distinguishing arboviral infections , such as dengue , from bacterial causes of febrile illness is of great importance for clinical management and antimicrobial stewardship ., In resource-limited countries , costly and expertise-reliant diagnostic assays cannot be performed routinely ., The Infection Manager Software ( IMS ) is a novel diagnostic algorithm equipped on an automated Sysmex hematology analyzer , making use of the principle that different infections evoke different changes in blood cell number and cell phenotype ., In a cohort of adult Indonesian patients presenting to hospital with an arboviral and/or bacterial infection , we first trained and subsequently evaluated the diagnostic performance of the IMS to distinguish common causes of acute febrile illness ., The authors show that the IMS has a reasonable sensitivity for detection of arboviral and bacterial infections and a high specificity ., In comparison with the commonly used biomarkers C-reactive protein ( CRP ) and procalcitonin ( PCT ) , the performance of the IMS was comparable to CRP and better than PCT ., The authors conclude that the IMS is a novel , automated , easy to use diagnostic tool that allows rapid differentiation between common causes of febrile illness in Southeast Asia . | antimicrobials, medicine and health sciences, pathology and laboratory medicine, chikungunya infection, drugs, immunology, tropical diseases, microbiology, parasitic diseases, salmonellosis, bacterial diseases, signs and symptoms, antibiotics, neglected tropical diseases, pharmacology, infectious diseases, inflammation, arboviral infections, immune response, diagnostic medicine, microbial control, biology and life sciences, viral diseases, malaria | null |
journal.pcbi.1002726 | 2,012 | Spatially Pooled Contrast Responses Predict Neural and Perceptual Similarity of Naturalistic Image Categories | Complex natural images are categorized remarkably fast 1 , 2 , sometimes even faster than simple artificial stimuli 3 ., For animal and non-animal scenes , differences in EEG responses are found within 150 ms 4 and a correct saccade is made within 120 ms 5 ., This speed of processing is also found for other scene categories 6 and may require less attentional resources compared to artificial images 7 , 8 ., This suggests that relevant visual information is rapidly and efficiently extracted from early visual responses to natural scenes ., However , the neural computations involved in this process are not known ., Importantly , natural images differ from other image types such as white noise in low-level properties ( e . g . , sparseness ) , leading to the suggestion that the visual system has adapted to these low-level properties 9 ., This idea paved the way for optimal coding models for natural images 10 , 11 and successful predictions of response properties of visual neurons 12 ., Recent work identified statistical properties that differ even within the class of natural images , e . g . between natural scene parts 13 , 14 or natural image categories 15 , showing that image statistics such as power spectra of spatial frequency content or distributions of local image features are informative about scene category ., The fact that it is mathematically possible to distinguish categories based on image statistics , however , does not imply that they are used for categorization in the brain ., Image statistics may not be sufficiently reliable , or their computation may not be suitable for neural implementation 12 , 16 ., We recently showed that statistics derived from the frequency histogram of local contrast – summarized by two parameters of a Weibull fit , Fig . 1A – explain up to 50% of the variance of event-related potentials ( ERPs ) recorded from visual cortex 17 ., These parameters inform about the width and shape of the histogram , respectively , and appear to describe meaningful variability between images ( Fig . 1B ) ., Importantly , we found that these parameters can be reliably approximated by linear summation of the output of localized difference-of-Gaussians filters modeled after X- and Y-type LGN cells , suggesting that this global information may be available to visual cortex directly from its early low-level contrast responses 17 ., Moreover , we found that output of contrast filters with a larger range of receptive field sizes captures additional image information 18 ., This is not surprising since objects in natural scenes appear at many distances and hence spatial scales 19 ., In the present implementation , the model first estimates at which scale relevant contrast information is present , as well as characteristics of the distribution of contrast strengths at those scales ., This model , which approximates early visual population responses based on spatially pooled contrasts , was able to explain almost 80% of ERP variance to natural images 18 ., These previous findings suggest that images with more similar contrast response statistics evoke more similar early visual activity ., Could these responses already contain relevant information about the stimulus for rapid categorization ?, The two parameters appear to index meaningful information such as degree of clutter , depth and figure-ground segmentation 17 , but how the two dimensions in Fig . 1B influence perception has not been examined ., The goal of the current study was thus to explore what type of visual information is contained in the variance of the earliest visual contrast responses that is so well described by these two parameters ., Specifically , we were interested in whether these parameters cannot only predict variance in visual activity , but also ‘variance in perception’ ., In other words , do images with more similar contrast statistics also lead to more similar perceptual representations , and perhaps ultimately , to similar images being considered a single category ?, We aimed to answer this question in a data-driven manner , by investigating, 1 ) which images group by similarity early in visual processing and, 2 ) whether this grouping matches with perceived similarity of those images ., For the first part of this question , we obtained reliable evoked responses to individual images ., The advantage of this approach relative to traditional ERP analysis ( which is based on averaging many trials across individual images within an a priori determined condition ) is that it provides much richer data 20–24 that can be used for model selection ., We used these single-image evoked responses to compute dissimilarities in ‘neural space’ , similar to the pattern analysis approach used in fMRI 25 , 26 ., This allowed us to track , over the course of the ERP , to what extent the representation of an image is ( dis ) similar to all images in the data set ., For the second part of the question , we needed to obtain an image-specific behavioral judgment of perceived visual similarity ., However , simply judging similarity of natural scenes is problematic , because these images obviously contain rich semantic content: there are many features of natural scenes that can be similar or dissimilar , which is likely to lead to different categorization strategies by different subjects ., Also , it is uncertain to what extent specific semantic tags that are provided by the researcher ( e . g . ‘openness’ or ‘naturalness’ , 27 ) , can be uniformly interpreted as a relevant stimulus dimension that has a linear mapping to processing in early vision ., Therefore , to explore the variance explained by contrast response statistics in a bottom-up way , we used stimuli that were simplified model images of natural scenes ( ‘dead leaves’ , Fig . 2A ) , which have similar low-level structure as natural scenes ( e . g . 1/f power spectra ) but are devoid of semantic content ., These images are created by filling a frame with objects - much like fallen leaves can fill a forest floor – and are used in computer vision to study , for example , how the appearance and the distribution of these objects influences the low-level structure of natural scenes 28 ., By manipulating properties of the objects in a controlled manner , we created distinct image categories , and then tested whether differences between these categories in contrast statistics matched with behaviorally perceived similarity by letting human observers perform a same-different categorization task on all combinations of image categories ., Specifically , we used the space formed by the two Weibull parameters to compute geometric distances between images in contrast statistics , and used these distances as quantitative predictors of dissimilarity 29–31 ., We thus tested whether these parameters can predict the extent to which image categories induced dissimilar single-image EEG responses ( experiment, 1 ) and whether they match with perceptual categorization at the behavioral level ( experiment 2 ) ., We predicted that images with very different Weibull statistics would appear less similar , i . e . be less often confused than images from categories with similar statistics ., By using controlled images that we quantified using a model originally derived from contrast responses to natural images , we aim to build a bridge between findings obtained with systematic manipulation of artificial stimuli and those obtained with more data-driven natural scene studies ., For purpose of comparison , and to better understand which statistical information is captured by the Weibull parameters , we also tested two other global contrast statistics ( Fig . 2C ) ., Following 32 we calculated the intercept and slope of the average power spectrum to parameterize spatial frequency information , a commonly used measure of low-level information in scene perception ., In addition , we followed 33 to derive the skewness and kurtosis of the contrast distribution for a range of spatial scales: these higher-order properties of distributions have previously been suggested ( e . g . 34 , 35 to reflect low-level differences between images that are relevant for perceptual processing ., We find that Weibull statistics explain substantial variance in evoked response amplitude to the dead leaves images , predicting clustering-by-category of occipital ERP patterns within 100 ms of visual processing ., In addition , they correlate with human categorization behavior: specific confusions were made between categories with similar Weibull statistics ., By comparison , Fourier power spectra and skewness and kurtosis can be used for accurate classification of image category , but fail to predict neural clustering and behavioral categorization ., These convergent results provide evidence for relevance of pooled contrast response statistics in rapid neural computation of perceptual similarity ., The experiments reported here were approved by the Ethical Committee of the Psychology Department at the University of Amsterdam; all participants gave written informed consent prior to participation and were rewarded with study credits or financial compensation ( 7 euro/hour ) ., Gray-scale dead leaves images ( 512×512 pixels , bit depth 24 ) were generated using Matlab ., Images contained randomly placed disks that were manipulated along 4 dimensions ( opacity , depth , size and distribution ) to create 16 categories ., Disks were either opaque or transparent; intensity at the outer edges of the disk was either constant ( leading to a 2D appearance ) or decaying ( 3D appearance ) , and disk size was determined by drawing randomly from a range of small , medium or large diameters ( exact settings as in 28 ., Twelve categories were created by systematically varying these properties of power-law distributed disks ., Four more categories were created using medium-diameter , exponentially distributed disks that could be 2D or 3D and opaque or transparent ., For each category , 16 images were created using these category-specific settings: the random placement and use of ranges of diameter sizes ensured that each of these 16 images was unique ., This procedure thus resulted in a total of unique 256 images , divided into 16 distinct categories , which were used for experimentation ( Fig . 2A ) ., If we set out all 256 dead leaves images against the three sets of image statistics ( Weibull parameters , Fourier parameters and skewness/kurtosis ) , stimuli cluster by category in all cases , with Fourier parameters leading to the most separable clusters ( Fig . 4A–C ) ., There were considerable correlations between the various parameters ( Fig . 4D; individual correlations plots in Fig . S2 ) ., Skewness and kurtosis correlated highly ( ρ\u200a=\u200a0 . 91 , p<0 . 0001 ) , but other significant correlations are observed as well , for example between Fourier slope and the Weibull beta parameter ( ρ\u200a=\u200a0 . 57 , p<0 . 0001 ) and also between the two Weibull parameters ( ρ\u200a=\u200a0 . 48 , p<0 . 001 ) ., A correlation of similar magnitude was also observed 17 for natural scenes , supporting the notion that the dead leaves stimuli used here have similar low-level structure as natural stimuli ., Interestingly , however , the ‘similarity spaces’ formed by each set of parameters are quite different between the various models ., If Weibull parameters determine the axes of the similarity space ( Fig . 4A ) , highly cluttered images with many strong edges ( e . g . 2D opaque stimuli with small disks ) are located in the upper right corner ( high gamma , high beta ) ; images containing fewer edges ( e . g . with larger disks ) are found more on the left ( low gamma ) ; and most of the transparent stimuli , with weak edges , cluster together in the bottom of the space ( low beta ) ., For Fourier intercept and slope ( Fig . 4B ) , transparent categories are highly separated across the space: however , most images with strong edges end up in a similar part of the space ( low slope , high intercept ) ., Based on either skewness or kurtosis ( Fig . 4C ) , a few categories are distinct , but most tend to cluster together ., These qualitative results suggest that all parameters are informative about clustering of image categories , but that they index different image properties ., Importantly , they give rise to different predictions about which categories should lead to similar evoked responses based on overlapping parameter values ., We tested these predictions using the single-image ERP data ., Whether low-level statistics are indeed actively exploited during scene or object categorization is a topic of considerable debate ., Whereas some studies report that manipulation of low-level properties influences rapid categorization accuracy 54 , 55 as well as early EEG responses 56 , 57 , other studies have shown that not all early visual activity is obliterated by equation of those properties 58–60 and , conversely , that early sensitivity to diagnostic information is revealed in stimuli that do not differ in low-level statistics 20 , 61 ., We find that , at least for our set of simplified models of natural scene images , early differences in ERPs are correlated with low-level contrast statistics that are themselves also directly predictive of perceptual similarity ., It is however likely that the degree to which low-level properties are relevant for processing of natural image categories is highly dependent on stimulus type and context , even within actual natural scene stimuli: for example , low-level information may influence rapid detection of faces to a larger extent than objects 22 and the effects of low-level statistics on animal detection may interact with scene category ( man-made vs . natural ) 62 ., In addition , the present work is very different from these previous reports in that our experiments did not require formation of a high-level representation but only a same-different judgment ., There are also notable differences between our ERP effects and those obtained with standardized object/scene categories: our maximum explained variance was found at around 100 ms , whereas those studies report sensitivity starting at 120 ms and onwards 63–66 ., Maximal sensitivity of evoked activity to faces and objects is found at lateral-occipital and parietal electrodes ( PO , e . g . 58 ) , whereas our correlations are clustered around occipital electrode Oz ., This suggests that the dead leaves images may mostly engage mid-level areas of visual processing , such as those sensitive to textural information , e . g . V2 24 , 67–69 ., Our results implicate that clustering of image similarities at this level of processing can , in principle , already predict perceptual similarity – in turn , these similarities can be derived from Weibull contrast statistics ., Given that for natural scenes , the Weibull statistics explain similar amounts of variance in EEG activity as reported here , we can hypothesize that image similarities as predicted by Weibull statistics are also present in evoked activity to actual natural scenes ., If Weibull statistics indeed approximate meaningful global information in natural images , which image features do they convey ?, By manipulating computational image categories in their perceptual appearance , we were able to get a better understanding of the information contained in the Weibull parameters ., They appear to index the amount of clutter , i . e . are related to occlusion and object size ., These properties may be relevant for natural scene categorization: a forest has a higher degree of clutter ( high gamma ) and lower mean edge strength ( high beta ) compared to a beach scene ., An image containing a few strong edges ( low beta ) that are sparsely distributed ( low gamma ) has high probability of coinciding with a single salient object , for example a single bird against an empty sky , suggesting that these statistics may be relevant for object detection in natural scenes ., Here , behavioral confusions ( and corresponding dissimilarities in ERP signals ) were found between stimuli without coherent edge information ( transparent stimuli with either large or small disks ) , or that were highly cluttered ( opaque stimuli with small disks ) which were exactly the categories that overlapped in Weibull parameter values ., For comparison , we computed Fourier power spectra and higher-order properties of the contrast distribution ( skewness and kurtosis ) , two sets of statistics that each index different sources of information in natural images: spatial frequency content and central moments of the contrast distribution , respectively ., Deviations in the power spectra of natural images inform about variations in contrast across spatial scales: the slope and intercept parameters describe the ‘spectral signature’ of images 32 which is diagnostic of scene category 15 ., Skewness and kurtosis were proposed to be relevant for texture perception 35 , 70 which in turn can be important for feature detection 53 , 71 and the presence of featureless regions of images 34 , 72 ., Our results confirm that both frequency content and central moments of the contrast distribution inform about image properties: both lead to accurate image classification ., However , in the present study they did not predict neural and behavioral categorization patterns , suggesting that these statistics may not be plausible computations involved in visual processing of the dead leaves images ., Even though we used controlled , computationally defined image categories , it is still possible that an image property other that the contrast statistics tested here will provide a better prediction of the ( neural and behavioral ) data , for example one of the manipulations used to create the image categories ( e . g . , opacity ) ., However , neither the observed clustering-by-category of ERPs in the RDM , nor the pattern of categorization errors in behavior mapped clearly onto one of the manipulations used to create the categories ( e . g . , opaque vs . transparent; as is visible in Fig . 7B , there are also differences within opaque and transparent categories , and this complex pattern of clustering is only predicted by Weibull statistics ) ., Why is the Weibull model better than widely used contrast statistics in predicting early neural and perceptual similarity ?, Although higher order moments of distributions can be diagnostic of textural differences , they may in practice be difficult for the visual system to represent 35 ., In addition , it has been suggested that rather than amplitude spectra , phase information derived from the Fourier transform 73 , 74 , or the interaction between these two 75 , 76 contains diagnostic scene information ., The reason that higher-order statistics derived from the phase spectrum may contain perceptually relevant information 77 is that they carry edge information ., In the Weibull model , contrasts , i . e . non-oriented edges , are explicitly computed ( as the response of LGN-type neurons ) and evaluated at multiple spatial scales ., The model may thus be able to capture information contained both in power spectra ( scale statistics ) as well as central moments ( distribution statistics ) ., The Weibull parameters appear to reflect different aspects of low-level information: the beta parameter varies with the range of contrast strengths present in the image , reflecting overall contrast energy , whereas the gamma parameter varies with the degree of correlation between local contrast values , reflecting clutter or spatial coherence ., Obviously , the Weibull fit is still a mathematical construct ., However , the two parameters can also be approximated in a more biologically plausible way: with our previous single-scale model 17 , we demonstrated that simple summation of X- and Y-type LGN output corresponded strikingly well with the fitted Weibull parameters ., Similarly , if the outputs of the multi-scale filter banks used here ( reflecting the entire range of receptive field sizes of the LGN ) are linearly summed , we again obtain values that correlate highly with the Weibull parameters obtained from the contrast histogram at minimal reliable scale ( S . Ghebreab , H . S . Scholte , V . A . F . Lamme , A . W . M Smeulders , under review ) ., This suggests that Weibull estimation can in fact be reduced to pooling of neuronal population responses by summation , which is a biologically realistic operation ., Why would summation of contrast responses of low-level neurons convey the same information as the Weibull parameters ?, This is likely a result of the structure of the world itself: distributions of contrast in natural images tend to range between power-law and Gaussian , which is the family of distributions that the Weibull function can capture 78 ., It appears that this statistic simply provides a good characterization of the dynamic range of the low-level input to the visual cortex when viewing natural images ., Since our brain developed in a natural world , early visual processing may take advantage of this regularity in estimating global properties to arrive at a first impression of scene content ., The present results extend our previous findings 17 , 18 with natural images to other image types ( computational categories ) and to prediction of behavioral categorization ., Interestingly , even though the subjects in experiment 1 ( EEG ) were not engaged in categorization of the dead leaves images , their results generalize to the behavioral categorization patterns that were found in experiment 2 , suggesting that similarity of bottom-up responses measured in EEG - in a different person - can be predictive of the perceived similarity during categorization of these images ., This observation is now restricted to computationally defined categories ., An interesting question for future work is whether in construction of high-level categorical representations of natural stimuli - considered a computationally challenging task - the brain actively exploits the pattern of variability of the population response to low-level information , estimated from early receptive field output ., Contrary to the classical view of the visual hierarchy ( e . g . , 79 ) it has been proposed that a rapid , global percept of the input ( gist ) precedes a slow and detailed analysis of the scene 80–83 ., Natural image statistics provide a pointer to information that could be relevant for such a global percept 84 , 85 ., However , the mechanism by which global information can be rapidly extracted from low-level properties is not directly evident from natural image statistics alone ., As explained above , in our model , the statistics are derived from a biologically realistic substrate ( the response of early visual contrast filters ) ., We suggest that to build a realistic model of natural image categorization , it is essential to understand how statistics derived from very early , simple low-level responses can contribute to gist extraction ., In conclusion , our findings suggest that global information based on low-level contrast can be available very early in visual processing and that this information can be relevant for judgment of perceptual similarity of controlled image categories . | Introduction, Materials and Methods, Results, Discussion | The visual world is complex and continuously changing ., Yet , our brain transforms patterns of light falling on our retina into a coherent percept within a few hundred milliseconds ., Possibly , low-level neural responses already carry substantial information to facilitate rapid characterization of the visual input ., Here , we computationally estimated low-level contrast responses to computer-generated naturalistic images , and tested whether spatial pooling of these responses could predict image similarity at the neural and behavioral level ., Using EEG , we show that statistics derived from pooled responses explain a large amount of variance between single-image evoked potentials ( ERPs ) in individual subjects ., Dissimilarity analysis on multi-electrode ERPs demonstrated that large differences between images in pooled response statistics are predictive of more dissimilar patterns of evoked activity , whereas images with little difference in statistics give rise to highly similar evoked activity patterns ., In a separate behavioral experiment , images with large differences in statistics were judged as different categories , whereas images with little differences were confused ., These findings suggest that statistics derived from low-level contrast responses can be extracted in early visual processing and can be relevant for rapid judgment of visual similarity ., We compared our results with two other , well- known contrast statistics: Fourier power spectra and higher-order properties of contrast distributions ( skewness and kurtosis ) ., Interestingly , whereas these statistics allow for accurate image categorization , they do not predict ERP response patterns or behavioral categorization confusions ., These converging computational , neural and behavioral results suggest that statistics of pooled contrast responses contain information that corresponds with perceived visual similarity in a rapid , low-level categorization task . | Humans excel in rapid and accurate processing of visual scenes ., However , it is unclear which computations allow the visual system to convert light hitting the retina into a coherent representation of visual input in a rapid and efficient way ., Here we used simple , computer-generated image categories with similar low-level structure as natural scenes to test whether a model of early integration of low-level information can predict perceived category similarity ., Specifically , we show that summarized ( spatially pooled ) responses of model neurons covering the entire visual field ( the population response ) to low-level properties of visual input ( contrasts ) can already be informative about differences in early visual evoked activity as well as behavioral confusions of these categories ., These results suggest that low-level population responses can carry relevant information to estimate similarity of controlled images , and put forward the exciting hypothesis that the visual system may exploit these responses to rapidly process real natural scenes ., We propose that the spatial pooling that allows for the extraction of this information may be a plausible first step in extracting scene gist to form a rapid impression of the visual input . | visual system, cognitive neuroscience, behavioral neuroscience, cognition, computational neuroscience, psychophysics, biology, computational biology, sensory systems, neuroscience, sensory perception, neuroimaging, coding mechanisms | null |
journal.pgen.1004868 | 2,014 | Genome-Wide Analysis of DNA Methylation Dynamics during Early Human Development | In mammals , DNA methylation is essential for normal development and plays critical roles in repression of transposable elements , maintaining genome stability , genomic imprinting and X-chromosome inactivation ., DNA methylation patterns are relatively stable in somatic cells but genome-wide reprogramming of DNA methylation occurs in primordial germ cells and preimplantation embryos 1–3 ., During mouse preimplantation development , the maternal genome is passively demethylated in a replication-dependent manner while some oocyte-specific methylated regions maintain maternal allele-specific methylation at the blastocyst stage 4 , 5 ., In contrast , the paternal genome is actively and rapidly demethylated through the oxidation of 5-methylcytosine ( 5mC ) to 5-hydroxymethylcytosine ( 5hmC ) by ten-eleven translocation-3 6 ., In spite of the global demethylation , imprinted differentially methylated regions ( DMRs ) and some transposable elements ( e . g . intracisternal A-particles ( IAPs ) ) are specifically protected from demethylation 1 ., During human preimplantation development , the paternal genome is reported to be actively demethylated as in the mouse 7 , 8 , but the regulatory mechanism and the genome-wide DNA methylation patterns in early embryos are not well understood ., Recently , two studies employed reduced representation bisulfite sequencing ( RRBS ) of human gametes and early embryos to characterize the human methylome very early in development 7 , 9 ., According to these studies , the paternal genome is rapidly and globally demethylated after fertilization whereas demethylation of the maternal genome is more limited and some oocyte-specific methylated regions maintain monoallelic methylation during preimplantation development , similar to the mouse genome ., RRBS is known to cover 5–10% of genomic CpGs , favoring those contained within CpG islands ( CGIs ) and promoter regions ., To obtain an unbiased and more complete representation of the methylome during early human development , we performed whole genome bisulfite sequencing ( WGBS ) of human gametes and blastocysts that covered>70% of genomic CpGs ., We found human-specific regulation of DNA methylation in various regions including oocyte-methylated CGIs , gene bodies and tandem repeat-containing regions ., We performed WGBS of human oocytes , sperm , blastocysts and neonatal blood cells ., For ethical reasons , we used only surplus germinal vesicle ( GV ) or metaphase I ( MI ) oocytes and blastocysts obtained from female patients undergoing in vitro fertilization ( IVF ) treatment ., Sperm and blood cells were collected from healthy donors ( see Materials and Methods for details ) ., WGBS libraries were constructed using the amplification-free post-bisulfite adaptor tagging ( PBAT ) method 10 for all samples except the oocytes , which required PCR-amplification ( PCR cycles =\u200a10 ) to increase the read depth ( Table 1 ) ., For each cell type , 87–96% of genomic CpGs were covered by at least one read , which was comparable to the reported methylome maps of mouse gametes 5 , 11 , 12 and human sperm 13 ., We also compared two oocyte PBAT libraries prepared with and without PCR-amplification ( Oocyte ( +PCR ) and Oocyte ( −PCR ) ) ( S1A Figure , S1B Figure , and Table 1 ) ., The methylation levels of individual CpGs were highly correlated ( r\u200a=\u200a0 . 83 ) between these two libraries ., Furthermore , the average methylation levels were very similar: Oocyte ( +PCR ) at 53 . 1% versus Oocyte ( −PCR ) at 54 . 8% ., These data demonstrate that our PCR-amplification protocol did not lead to significant bias in our data sets ., Non-CpG methylation was observed in human oocytes , especially at CpA sites ( mean =\u200a5 . 6% ) , with a positive correlation between CpG and non-CpG methylation ( S1C Figure , S1D Figure ) ., Non-CpG methylation was not a significant feature of sperm or blastocysts ( <1% ) ., In the following analyses , only CpGs covered with ≥3 reads were considered for oocytes and those covered with ≥5 reads were considered for the other samples ., We confirmed that three imprinted DMRs and two pluripotency genes frequently observed to be abnormal in poor quality oocytes or embryos 14 , 15 were normally methylated in our WGBS data ( S1E Figure ) ., We also compared our WGBS data with recently reported RRBS data of human oocytes , blastocysts and inner cell mass ( ICM ) and WGBS data of ICM 7 , 9 ., Our data substantially increased the coverage of genomic CpGs compared with the reported data ( S1F Figure , S1G Figure ) ., The methylation levels of CGIs showed high correlations ( r\u200a=\u200a0 . 96 ) between our WGBS data and the reported RRBS data ( oocyte: S1H Figure , blastocyst: S1I Figure ) , validating the WGBS data ., Similar to findings for the mouse , human oocytes showed an intermediate methylation level of CpGs and blastocysts were globally hypomethylated ( Fig . 1A ) ., To further characterize global DNA methylation changes , we used a system of sliding windows of 20 CpGs with a step size change of 10 CpGs ., Windows were classified as increasing ( or decreasing ) if the methylation levels increased ( or decreased ) by>20% and the changes were statistically significant ( Benjamini-Hochberg ( BH ) corrected P<0 . 05 ) ., We found that 57% and 83% of windows showed decreased methylation levels in blastocysts compared with oocytes and sperm , respectively ( Fig . 1B ) ., In contrast , >90% of windows showed increased methylation in ES or blood cells compared with blastocysts ( Fig . 1B ) ., To explore the differences in demethylation dynamics between parental genomes , we focused on windows hypermethylated in one gamete and hypomethylated in the other ., In this study , we defined regions that were ≥80% methylated as hypermethylated and those that were ≤20% methylated as hypomethylated ., Windows hypermethylated in sperm and hypomethylated in oocytes ( sperm-specific methylated windows ) were abundant in intergenic regions ., In contrast , oocyte-specific ones showed a relatively uniform distribution ( S2A Figure ) ., In blastocysts , oocyte-specific methylated windows showed intermediate methylation levels ( median =\u200a35 . 1% ) , in contrast to the nearly complete demethylation of sperm-specific ones ( Fig . 1C ) ., Almost all windows hypomethylated in both gametes remained hypomethylated and very few windows ( 0 . 04% ) were hypermethylated in blastocysts , suggesting that genome-wide de novo methylation occurred after implantation ( Fig . 1C and S2B Figure ) ., Consistently , the methylation patterns of oocytes and blastocysts were very similar to each other ( Figs . 1E , F ) , suggesting that the global methylation pattern of the maternal genome , but not the paternal genome , was inherited by blastocysts ., Next , we examined specific genomic features: CGIs , promoters and transposable elements ., CGIs and promoters hypermethylated in sperm remained methylated in ES and blood cells ., On the other hand , oocyte-specific methylated CGIs showed variable methylation levels and oocyte-specific methylated promoters were preferentially demethylated in ES and blood cells ( S3A Figure and S3B Figure ) ., In addition , the promoter methylation patterns of sperm , but not of oocytes , showed high correlations with those of ES and blood cells ( r>0 . 8 , Fig . 1D ) ., These data highlighted the unique promoter methylation profile of oocytes ., Short interspersed nuclear elements ( SINEs ) , long interspersed nuclear elements ( LINEs ) , long terminal repeats ( LTRs ) and DNA repeats were essentially highly methylated in ES and blood cells , whereas 20–30% and 3–8% of repeat copies were hypomethylated in oocytes and sperm , respectively ( S2B Figure ) ., These transposable elements were demethylated similarly to other genomic regions in blastocysts ( S3 Figure ) ., Germline DMRs ( gDMRs ) frequently serve as imprinting control regions 16 and we were interested in how many gDMRs exist in the human genome ., Among the 67 known imprinted DMRs 17 , 46 DMRs were classified as gDMRs according to the following definition: DMRs hypermethylated in one gamete and hypomethylated in the other ( Fig . 2A , B and S1 Table ) ., Of these , 15 reportedly placenta-specific DMRs were lost in blood cells ( Fig . 2A , C ) ., The other 31 gDMRs showed intermediate methylation levels in blood cells , but about one-third of these gDMRs were not maintained in ES cells ( H9 ES cells: Fig . 2A , H1 and HUES6 ES cells: S4A Figure ) , indicating the instability of gDMRs in human ES cells ., Importantly , oocyte-specific methylated autosomal CGIs showed methylation levels very similar ( median =\u200a37 . 5% ) to gDMRs ( median =\u200a39 . 2% ) in human blastocysts ( Fig . 2D ) ., We confirmed monoallelic methylation of four autosomal CGIs in human blastocysts by using conventional bisulfite sequencing ( Fig . 2E and S4B Figure ) ., We also analyzed two X-linked CGIs hypermethylated in oocytes and found that these CGIs showed high methylation levels in male blastocysts ( the X chromosome of male blastocysts is derived from oocytes ) and monoallelic methylation in female blastocysts ( Fig . 2F ) ., Consistently , X-linked CGIs with oocyte-specific methylation showed higher methylation levels than autosomal ones in blastocysts ( the WGBS data were derived from a pool of blastocysts ) ( Fig . 2D ) ., A similar tendency was also observed in the sliding window-based analyses ( S2C Figure ) ., These data suggested that a substantial number of oocyte-specific methylated CGIs may maintain maternal allele-specific methylation in human blastocysts ., In contrast , most oocyte-specific methylated CGIs were significantly demethylated compared with gDMRs in mouse blastocysts ( Fig . 2D ) ., In mouse oocytes , gene-body methylation levels are reported to positively correlate with the transcription levels 5 ., In human oocytes , a positive correlation between gene-body methylation and transcription levels was also observed ., Interestingly , there was an expression-level boundary at around log2 ( RPKM ) =\u200a−5 ( RPKM: reads per kilobase per million ) ( Fig . 3A ) ., Genes with log2 ( RPKM ) >−5 and <−5 may be transcriptionally active and inactive genes , respectively ( Fig . 3B ) ., We analyzed previously reported mouse methylome and transcriptome data and found that a bimodal distribution of gene body methylation was also observed while there was a boundary at around log2 ( RPKM ) =\u200a0 ( Fig . 3A ) ., It is unclear whether the difference between the human and mouse expression-level boundaries reflects experimental or functional differences ., We found that 971 genes showed differential gene body methylation between human and mouse oocytes ( Fig . 3C and S2 Table ) ., Gene ontology ( GO ) analysis revealed an abundance of genes encoding cell adhesion molecules with human-specific gene body hypermethylation ( Fig . 3D ) , which could have important roles during human oogenesis ., In mouse oocytes , Dnmt3l and Zfp57 are highly expressed and essential for DNA methylation regulation 18 , 19 whereas human DNMT3L is undetectable in oocytes 20 ., Here we found that the gene body regions of DNMT3L and ZFP57 were hypomethylated in human oocytes and neither gene was expressed ( Figs . 3E , F ) , implying that DNMT3L and ZFP57 might not be essential for regulation of DNA methylation in human oocytes ., As described above , global methylation changes of SINEs , LINEs , LTRs and DNA repeats were very similar to other genomic regions in early human embryos ( S3 Figure ) ., We further analyzed mean methylation levels of CpGs in various classes of these transposable elements ( Fig . 4A , see also S3 Table for details ) ., These repeat classes showed similar methylation changes: ∼60% methylated in oocytes , ∼80% methylated in sperm , ES and blood cells and ∼30% methylated in blastocysts ., These data suggested that SINEs , LINEs , LTRs and DNA repeats were essentially not resistant to genome-wide demethylation after fertilization ., Mouse IAPs are known to be protected from demethylation during preimplantation development 5 , 21 ., To identify transposable elements specifically protected from demethylation during human preimplantation development , we screened repeat copies overlapping windows showing>70% methylation in blastocysts ( 0 . 3% of all windows ) ( S4 Table ) ., We found that SINE-VNTR-Alu ( SVA ) subfamilies , especially SVA_A , frequently overlapped the>70% methylated windows ( Fig . 4B ) ., SVA_A also showed the highest methylation level in blastocysts ( 59 . 2% ) whereas the other repeat sequences were <50% methylated ( Fig . 4A and S3 Table ) ., SVA is a hominid-specific repeat family that remains active in the human genome 22 ., Similar to mouse LTRs 5 , methylation levels of CpGs within SVAs are positively correlated with CpG density in human oocytes and blastocysts ( Fig . 4C and S5 Figure ) ., LTR12 subfamilies , which are LTRs of HERV9 , also tended to overlap the>70% methylated windows ( Fig . 4B ) ., Interestingly , both SVA and LTR12 subfamilies contain CpG-rich variable number tandem repeats ( VNTRs ) 22 , 23 ., We also noticed that whereas the MER34C2 consensus sequence does not contain VNTRs , MER34C2 copies overlapping the>70% methylated windows were all tandemly repeated in a single genomic locus ( Fig . 4D ) ., VNTRs were also found in the two paternal gDMRs ( Fig . 4E ) ., VNTRs were not a common feature of the maternal gDMRs , but a significantly higher proportion of the maternal gDMRs did contain VNTRs as compared with all CGIs ( gDMRs: 11/44 , CGIs: 1763/27718 , chi-square P\u200a=\u200a4 . 1×10−7 ) ., Therefore , we focused on CGIs hypermethylated in both gametes and found that CGIs containing VNTRs were preferentially protected from demethylation in blastocysts ( Fig . 4F ) ., A comparison between VNTRs of>70% and <50% methylated CGIs in blastocysts revealed that VNTRs with more repeats tended to be protected from demethylation , whereas no sequence motif was found ( Fig . 4G ) ., These data suggested that VNTRs might underlie silencing of specific transposable elements and the protection of paternal gDMRs ., We also found that alpha satellite ( ALR ) , which is a tandemly repeated DNA family found in centromeric and pericentromeric regions 24 , was hypermethylated in human oocytes ( 80 . 6% ) ( Fig . 4H ) ., Interestingly , DNMT3B was highly expressed in human oocytes ( Fig . 3E ) , and DNMT3B is reported to interact with centromere protein CENP-C and contribute to DNA methylation of ALR 25 ., Thus , it is possible that DNMT3B is involved in DNA methylation of ALR in human oocytes ., This work reports the genome-wide DNA methylation patterns of human gametes and blastocysts at single-base resolution ., Our WGBS data of oocytes and blastocysts substantially increase the coverage of genomic CpGs adding to the reported RRBS data of oocytes and blastocysts and WGBS data of ICM 7 , 9 ., We confirmed that the paternal genome was globally demethylated as previously reported ., However , the oocyte-specific methylated regions maintained intermediate methylation levels in human blastocysts ( median =\u200a35 . 1% ) ., Consistently , the methylation patterns of oocytes and blastocysts were very similar to each other , suggesting that the global methylation pattern of the maternal genome was inherited by blastocysts ., Furthermore , oocyte-specific methylated CGIs showed methylation levels very similar ( median =\u200a37 . 5% ) to gDMRs ( median =\u200a39 . 2% ) ., These data appear not to support replication-dependent global demethylation of the maternal genome during human early development , because oocyte-specific methylated regions should show ≤25% methylation after one replication-dependent global demethylation event ., In mouse blastocysts , most oocyte-specific methylated CGIs were significantly demethylated compared with gDMRs , which may reflect the passive demethylation of the maternal genome 1 , 2 ., These data strongly suggest that the maternal genome is demethylated to a much lesser extent in human blastocysts than in mouse blastocysts ., We classified known imprinted DMRs 17 and discovered that there were at least 46 gDMRs in the human genome including 15 specific to the placenta ., Our data suggested that a substantial number of oocyte-specific methylated CGIs may also maintain mono-allelic methylation in human blastocysts whereas they were essentially lost through hypermethylation or hypomethyaltion in blood cells ., It is suggested that a significant portion of gene transcripts show mono-allelic expression in human 8-cell embryos and morulae 26 , and the oocyte-specific methylated CGIs could regulate mono-allelic expression of some genes in human preimplantation embryos ., In the mouse genome , ∼25 well defined gDMRs have been identified and only the Gpr1 DMR is reported to be placenta-specific 27 , 28 ., The demethylation resistance of oocyte-specific methylated CGIs during early human development may , in part , explain the increased number of placenta-specific gDMRs in the human genome ., Interestingly , we found that ZFP57 was not expressed in human oocytes ., Because replication-dependent global demethylation of the maternal genome is not likely to occur during human preimplantation development , we speculate that the protection of gDMRs by ZFP57 may be dispensable in human oocytes ., These data contribute to our understanding of the regulatory mechanism of human-specific genomic imprinting ., Both human and mouse oocytes showed bimodal gene body methylation patterns associated with transcription ., While it is unclear whether transcription is the only determinant , transcription may be an important determinant of the oocyte methylomes ., In mammals , DNMT3A and DNMT3B are de novo DNA methyltransferases whereas DNMT3L acts in a recruiting role ., In mouse oocytes , Dnmt3a and Dnmt3l are essential for de novo DNA methylation , whereas Dnmt3b is poorly expressed and essentially dispensable 11 , 29 ., In contrast , in human oocytes DNMT3B showed ∼10-fold higher expression than DNMT3A , and DNMT3L was not expressed , suggesting that DNMT3B may be the critical de novo DNA methyltransferase during human oocyte growth ., Interestingly , centromeric satellite repeats were highly methylated in human oocytes ., These regions are known to be hypomethylated in mouse oocytes 30 ., Human DNMT3B is reported to interact with centromere protein CENP-C and contribute to DNA methylation of centromeric satellite repeats 25 ., Similarly , centromeric satellite repeats are demethylated in Dnmt3b mutant mice 31 ., Therefore , the differential expression pattern of DNMT3B could explain this human-specific hypermethylation of centromeric satellite repeats in oocytes ., It is suggested that evolutionarily young SINEs and LINEs are demethylated to a milder extent than older ones during human preimplantation development 7 ., We found that SVAs and some LTRs containing CpG-rich VNTRs were much more preferentially protected from demethylation than SINEs and LINEs in human blastocysts ., Paternal gDMRs also contained VNTRs and many VNTR-containing CGIs remained highly methylated in human blastocysts ., Therefore , VNTRs might underlie the protection of paternal gDMRs and specific transposable elements from demethylation ., The maintenance of DNA methylation of SVAs may be especially important because SVAs are currently active in the human genome and are involved in various human diseases 22 , 32 ., While the underlying mechanism of the protection of VNTR-containing regions is currently unknown , it is noteworthy that VNTRs are related to RNA-directed DNA methylation in plants 33 ., Many transposable elements including SVAs are expressed in human early embryos 7 , 9 and it is interesting to speculate that RNA might be involved in the demethylation resistance of VNTR-containing regions ., Overall , this work highlights both conserved and species-specific regulation of DNA methylation during early mammalian development ., Our WGBS data of human gametes and blastocysts not only provide information to support our understanding of normal human developmental processes but also will be useful in interpreting studies on assisted reproductive technologies ( ARTs ) ., ARTs in humans are associated with an increased risk of imprinting disorders 34 , 35 , and our data will aid in the safety evaluation of ARTs and the preimplantation epigenetic diagnosis of human embryos ., Human oocytes , sperm , blastocysts and umbilical cord blood cells were obtained with signed informed consent of the donors or the couples , and the approval of the Ethics Committee of Tohoku University School of Medicine ( Research license 2013-1-57 ) , associated hospitals , the Japan Society of Obstetrics and Gynecology and the Ministry of Education , Culture , Sports , Science and Technology ( Japan ) ., Altogether , 202 surplus oocytes and 80 surplus blastocysts were obtained from female patients ( ages 26–43 ) undergoing IVF treatment ., The patients were healthy women with no habitual drug use and no particular past or familial disease history ., We collected morphologically normal GV and MI oocytes from preovulatory follicles by intravaginal ultrasound-guided follicular aspiration after controlled ovarian hyperstimulation ., To remove cumulus cells and the zona pellucida , oocytes were treated with hyaluronidase solution ( JX Nippon Oil & Energy Corporation , Tokyo , Japan ) and Tyrodes solution-Acidified ( JX Nippon Oil & Energy Corporation ) according to the manufacturers instructions ., Blastocysts were obtained by culturing early cleavage-stage embryos in Global Medium ( LifeGlobal , Guilford , CT ) overlaid with mineral oil ., We used morphologically normal expanding or expanded blastocysts ., The number of ICM cells is similar to , or a little lower than , that of trophectoderm ( TE ) cells in blastocysts at this stage 36 ., Because ICM and TE cells show similar methylation levels 7 , 9 and the available embryos in this study were limited , we performed WGBS using whole blastocysts ., Ejaculated sperm samples with normal volume , counting and rates of mortality were collected ., Only motile sperm cells isolated by the swim-up method 37 were used ., Oocytes and blastocysts were incubated in a lysis solution ( 0 . 1% SDS , 1 mg/ml proteinase K , 50 ng/µl carrier RNA ( QIAGEN , Valencia , CA ) ) for 60 min at 37°C and then 15 min at 98°C ., Genomic DNA was purified with phenol/chloroform extraction and ethanol precipitation ., Sperm genomic DNA was prepared as described 38 ., Genomic DNA of cord blood cells was purified with phenol/chloroform extraction and ethanol precipitation ., Isolated genomic DNA was spiked with 5% ( for oocytes and blastocysts ) or 0 . 5% ( for sperm and cord blood cells ) unmethylated lambda DNA ( Promega , Madison , WI ) ., Bisulfite treatment was performed using the MethylCode Bisulfite Conversion Kit ( Invitrogen , Carlsbad , CA ) ., PBAT libraries were prepared as previously described 10 ., Briefly , the first-strand DNA was synthesized with the Klenow fragment ( 3′-5′ exo- ) ( NEB , Beverly , MA ) using BioPEA2N4 ( 5′-biotin-ACA CTC TTT CCC TAC ACG ACG CTC TTC CGA TCT NNN N-3′ ) ., The biotinylated first-strand DNA was captured using Dynabeads M-280 Streptavidin ( Invitrogen ) ., The second-strand DNA was synthesized with the Klenow fragment ( 3′-5′ exo- ) using PE-reverse-N4 ( 5′-CAA GCA GAA GAC GGC ATA CGA GAT NNN N-3′ ) ., After removing the first-strand DNA , the second strand was double stranded with Phusion Hot Start II High-Fidelity DNA Polymerase ( Finnzymes , Woburn , MA ) using Primer-3 ( 5′-AAT GAT ACG GCG ACC ACC GAG ATC TAC ACT CTT TCC CTA CAC GAC GCT CTT CCG ATC T-3′ ) ., For an oocyte PBAT library , PCR-amplification was performed with KAPA HiFi HotStart Uracil+ ReadyMix ( 2× ) ( Kapa Biosystems , Woburn , MA ) using primers , ( 5′-CAA GCA GAA GAC GGC ATA CGA GAT-3′ ) and ( 5′- AAT GAT ACG GCG ACC ACC GAG ATC T-3′ ) ., The following program was used for the PCR-amplification: 10 cycles of 98°C for 15 sec , 65°C for 30 sec and 72°C for 30 sec ., Concentrations of the PBAT libraries were measured by quantitative PCR ( qPCR ) using the Kapa Library Quantification Kit ( Kapa Biosystems ) ., PBAT libraries were sequenced on the HiSeq 2000 or HiSeq 2500 platform ( Illumina , CA , USA ) with 100-bp single-end reads using the TruSeq SR Cluster Kit v3-cBot-HS and the TruSeq SBS Kit v3-HS ( Illumina ) ., Sequenced reads were processed using the Illumina standard base-calling pipeline ( v1 . 8 . 2 ) and the first 4 bases were trimmed to remove random primer sequences ., The resulting reads were aligned to the reference genome ( UCSC hg19 ) using Bismark 39 ( v . 0 . 9 . 0 ) with default parameters ., For the oocyte library prepared with PCR-amplification , identical reads were treated as a single read to remove PCR duplicates ., The methylation level of each cytosine was calculated using the Bismark methylation extractor ., For CpG sites , reads from both strands were combined to calculate the methylation levels ., Except for S1 Figure , methylation levels of CpGs covered with ≥3 reads were analyzed for oocytes and those of CpGs covered with ≥5 reads were analyzed for the other samples ., Bisulfite conversion rates were estimated using reads that uniquely aligned to the lambda phage genome and were>99% for all samples ., In this study , mC and hmC were indistinguishable because bisulfite sequencing cannot differentiate hmC from mC ., We also included available RRBS data of human oocytes , ICM and blastocysts 7 , 9 and WGBS data of human ICM 7 , ES cells ( H1 , H9 and HUES6 ) and mouse oocytes 11 ., Processed methylation data were downloaded from NCBI GEO ( http://www . ncbi . nlm . nih . gov/geo ) for ICM ( Accession number: GSE49828 and GSE51239 ) , blastocysts ( Accession number: GSE51239 ) , H1 ( Accession number: GSM429321 ) , H9 ( Accession number: GSM706059 ) and HUES6 ES cells ( Accession number: GSM1173778 ) ., The RRBS data from biological replicates were combined ., For mouse oocytes 11 , the raw reads were mapped to the reference genome ( UCSC mm9 ) and analyzed as described above ( only CpGs covered with ≥5 reads were used ) ., Annotations of Refseq genes , CGIs and repeat sequences were downloaded from the UCSC Genome Browser ., Refseq genes shorter than 300 bp ( encoding microRNAs or small nucleolar RNAs in most cases ) were excluded from our analyses ., Promoters were defined as regions 1 kb upstream and downstream from transcription start sites of Refseq transcripts ., For calculation of the mean methylation levels , we analyzed only CGIs and promoters containing ≥10 CpGs with sufficient coverage for calculation of the methylation levels ., Similarly , we considered only repeat copies containing ≥5 CpGs for calculation of the mean methylation levels of repeat copies ., The gene bodies were defined as transcribed regions of Refseq transcripts except for promoters ., When several Refseq transcripts were assigned to a Refseq gene , the transcribed regions were merged into a single gene body ., Regions and names of the 67 imprinted DMRs were defined as previously reported 17 ., The CpG density was defined for each CpG site as the density of CpGs within 100 bp upstream and downstream regions ( the number of CpGs was divided by 200 ) ., Gene ontology analyses were performed using the Database for Annotation , Visualization and Integrated Discovery ( DAVID ) 40 ., The list of human and mouse homologs including HomoloGene IDs was downloaded from Mouse Genome Informatics ( MGI , http://www . informatics . jax . org/ ) ., VNTRs were identified using Tandem Repeats Finder 41 ( alignment parameters =\u200a2 , 5 , 7; minimum alignment score =\u200a150; maximum period size =\u200a500 ) ., Sequence motifs among VNTRs were searched using DREME 42 ( the consensus patterns of VNTRs of>70% and <50% methylated CGIs in Fig . 4G were used as positive and negative sequences , respectively ) ., Transcriptome data of human and mouse oocytes were previously reported 5 , 43 ., The raw reads from biological replicates were combined and analyzed using Avadis NGS software with default parameters ( version 1 . 5 , Strand Scientific Intelligence ) ., Methylation levels of CpGs were visualized using Integrative Genomics Viewer ( IGV ) software ( http://www . broadinstitute . org/igv/ ) ., Heatmaps and scatter plots were generated using the heatmap . 2 function of the gplots package and the heatscatter function of the LSD package in R ( http://www . R-project . org/ ) , respectively ., Violin plots were generated using the vioplot package ( http://neoscientists . org/~plex/ ) ., We used a sliding window of 20 CpGs with a step size of 10 CpGs ( the mean length was ∼2 kb ) for consideration of the successful identification of imprinted DMRs using sliding windows of 10 CpGs 44 and 25 CpGs 17 ., We considered only windows containing ≥10 CpGs with sufficient coverage for calculation of the methylation levels ( 84% of windows were covered in all samples shown in Fig . 1 ) ., Windows were classified as increasing ( or decreasing ) if the methylation levels increased ( or decreased ) by>20% and the changes were statistically significant according to Students t-test with BH correction ( P<0 . 05 ) ., DNA samples were treated with sodium bisulfite using an EZ DNA Methylation Kit ( Zymo Research , Orange , CA ) and PCR-amplified using TaKaRa EpiTaq™ HS ( Takara Bio , Shiga , Japan ) ., The PCR products were cloned into the pGEM-T Easy vector ( Promega ) and individual clones were sequenced ., The following primers were used: chr5: 662 , 283–663 , 402: ( 5′-GGG GTT AAG ATG GGA GTT ATG A-3′ ) and ( 5′-TAA ACA ACC CAA TCC CCA CA-3′ ) , chr12: 20 , 704 , 525-20 , 706 , 004: ( 5′-GGG AGG AGG AGG AGT AGT AGG A-3′ ) and ( 5′-CCC ACT AAA AAC AAA ATC AAT ACC-3′ ) , chr15: 89 , 952 , 271-89 , 953 , 061: ( 5′-GAT TTT TGT TAA TGA TTG GGT AGG A-3′ ) and ( 5′-CCC CAC AAT ATC TAC CCT CAT A-3′ ) , chr21: 32 , 716 , 044-32 , 716 , 485: ( 5′-AGA AGT TAA GGG GGA AAG ATG A-3′ ) and ( 5′-TTC ACA AAT TAC ACC CAC TAC CTC-3′′ ) , chrX: 3 , 732 , 573-3 , 734 , 579: ( 5′-TTA ATG GGG TAA AGG GGT TAG A-3′ ) and ( 5′-ACC AAA TAA ACC CCA CCC AAA C-3′ ) , chrX: 153 , 694 , 352-153 , 694 , 774: ( 5′-GTG GGG TTT AAG GAA GGA GGT A-3′ ) and ( 5′-CAA TCA CCC ACA CAC AAC TCC-3′ ) ., The sex of blastocysts was determined by PCR amplification of the male-specific SRY locus using bisulfite-converted DNA with the following primers: Forward: ( 5′ -TGA AAT TAA ATA TAA GAA AGT GAG GGT TG- 3′ ) and Reverse: ( 5′ -CCA CAC ACT CAA AAA TAA AAC ACC A- 3′ ) ., All sequencing data are deposited in the Japanese Genotype-phenotype Archive under the accession number JGAS00000000006 . | Introduction, Results, Discussion, Materials and Methods | DNA methylation is globally reprogrammed during mammalian preimplantation development , which is critical for normal development ., Recent reduced representation bisulfite sequencing ( RRBS ) studies suggest that the methylome dynamics are essentially conserved between human and mouse early embryos ., RRBS is known to cover 5–10% of all genomic CpGs , favoring those contained within CpG-rich regions ., To obtain an unbiased and more complete representation of the methylome during early human development , we performed whole genome bisulfite sequencing of human gametes and blastocysts that covered>70% of all genomic CpGs ., We found that the maternal genome was demethylated to a much lesser extent in human blastocysts than in mouse blastocysts , which could contribute to an increased number of imprinted differentially methylated regions in the human genome ., Global demethylation of the paternal genome was confirmed , but SINE-VNTR-Alu elements and some other tandem repeat-containing regions were found to be specifically protected from this global demethylation ., Furthermore , centromeric satellite repeats were hypermethylated in human oocytes but not in mouse oocytes , which might be explained by differential expression of de novo DNA methyltransferases ., These data highlight both conserved and species-specific regulation of DNA methylation during early mammalian development ., Our work provides further information critical for understanding the epigenetic processes underlying differentiation and pluripotency during early human development . | DNA methylation reprogramming after fertilization is critical for normal mammalian development ., Early embryos are sensitive to environmental stresses and a number of reports have pointed out the increased risk of DNA methylation errors associated with assisted reproduction technologies ., Therefore , it is very important to understand normal DNA methylation patterns during early human development ., Recent reduced representation bisulfite sequencing studies reported partial methylomes of human gametes and early embryos ., To provide a more comprehensive view of DNA methylation dynamics during early human development , we report on whole genome bisulfite sequencing of human gametes and blastocysts ., We show that the paternal genome is globally demethylated in blastocysts whereas the maternal genome is demethylated to a much lesser extent ., We also reveal unique regulation of imprinted differentially methylated regions , gene bodies and repeat sequences during early human development ., Our high-resolution methylome maps are essential to understand epigenetic reprogramming by human oocytes and will aid in the preimplantation epigenetic diagnosis of human embryos . | sequencing techniques, biochemistry, developmental biology, embryology, high throughput sequencing, molecular biology, genomic imprinting, embryo development, genetics, biology and life sciences, dna, epigenetics, dna modification, molecular biology techniques, dna methylation | null |
journal.pcbi.1006025 | 2,018 | Gap junction plasticity as a mechanism to regulate network-wide oscillations | Oscillatory patterns of neuronal activity are reported in many brains regions with frequencies ranging from less than one Hertz to hundreds of Hertz ., These oscillations are often associated with cognitive phenomena such as sleep or attention ., Local field potential measurements in the neocortex and thalamus show the prevalence of delta oscillations ( 0 . 5-4Hz ) and spindle oscillations ( 7-15Hz ) during sleep 1 ., Theta oscillations ( 4-10Hz ) are also reported in hippocampus and other brain regions 2 ., Gamma oscillations ( 30-100Hz ) observed in the cortex are thought to be involved in attention 3–6 , perception 7 , 8 and coordinated motor output 9 , 10 ., Thus , at the minimum , oscillations are present during the normal functioning of neural circuits ., However , oscillations are also associated with pathological circuit dynamics , such as hyper-synchronous activity during epileptic seizures 11 ., Altered gamma-frequency synchronizations may also be involved in cognitive abnormalities such as autism 12 or schizophrenia 13 ., Thus , given both the functional and pathological effects of oscillations , a homeostatic mechanism is necessary to regulate oscillatory behavior ., Several mechanisms can lead to the emergence of oscillations ., They can arise in homogeneous population of excitatory neurons , where the positive feedback loop of excitation is only limited by the refractoriness of the neurons 14 ., Alternatively , oscillations can also arise in a coupled network of excitatory and inhibitory neurons , where the excitatory and inhibitory neurons burst in opposing phase ., 15–19 ., Finally , gap junctions between inhibitory neurons promote synchronous oscillatory patterns 20–24 ., The inhibitory network oscillations primarily involve fast-spiking interneurons ., These neurons represent a large proportion of GABAergic interneurons 25 ., They are the main cells targeted by thalamocortical synapses transmitting sensory information to the cortex 26 ., They are coupled via chemical synapses and gap junctions ., Gap junctions are mostly found between neurons of the same class 26–28 but they can also connect different subtypes , such as fast-spiking and regular spiking cells 26 , 29 , 30 ., Moreover , there is evidence of the critical role of fast-spiking parvalbulmin ( FS ) interneurons in the emergence of cortical gamma activity in the cortex of rodents in response to sensory stimuli 31–34 ., Two main properties of FS interneurons have been found critical in the existence of gamma oscillations ., Firstly , FS interneurons selectively amplify gamma frequencies through subthreshold resonance 33 ., Secondly , gap junctions between inhibitory interneurons 27 have been shown to enhance synchrony 24 , 26 , 35–41 ., A computational model with both properties , inhibitory neurons with subthreshold resonance , connected by gap junctions , has been shown to support gamma oscillations 24 , 42–46 ., Recently , gap junction plasticity has been experimentally demonstrated 47–51 ., For example , the gap junctions between rod cells in the retina can vary their conductance during day and night cycles 52 ., Moreover , they can experience bidirectional long-term plasticity in an activity-dependent manner 49 , 53 , 54 ., High frequency stimulation of a coupled pair of thalamic reticular nucleus ( TRN ) neurons induces gap junction long-term depression ( gLTD ) 55 ., This occurs only when the TRN neurons burst ., There is no data yet on the long-term potentiation of cortical gap junctions ., However , 56 show that the pathways leading to gLTD are calcium-dependent which suggest that gap junction long-term potentiation ( gLTP ) could also be the result of an activity-dependent mechanism ., Other passive mechanisms , such as gap junction connexin turnover could compensate for long-term depression as well 57–62 ., Given the existence of gap junction plasticity and the omnipresence of oscillations in cortex , we wondered whether gap junction plasticity can regulate network-wide gamma oscillations in cortex ., To that end , we developed a computational model of a network of excitatory and FS inhibitory neurons ., As demonstrated analytically by 24 , we observed two different network behaviors depending on the gap junction strength ., For weak gap junction strength , the network exhibits an asynchronous regime , whereas for strong gap junctions , the network synchronizes into coherent gamma oscillations with bursting activity ., We then modelled the gap junction plasticity observed by 55 showing that bursting activity leads to gLTD ., The plastic network sets itself at the transition between the asynchronous regime , where sparse spiking dominates , and the synchronous regime , where network oscillations dominate and burst firing prevails ., Thus , our model shows that gap junction plasticity maintains the balance between the asynchronous and synchronous network states ., This is robust to different possible gLTP rules ., We then show that the network allows for transient oscillations driven by external drive ., This demonstrates that transient , plasticity regulated oscillations can efficiently transfer information to downstream networks ., Finally we show that gap junction plasticity mediates cross-network synchronization and allows for robust information transfer trough frequency modulation ., Critically , gap junction plasticity allows for the recovery of oscillation mediated information transfer in the event of partial gap junction loss ., To study the effect of gap junction plasticity , we developed a network of coupled inhibitory and excitatory neurons in the fluctuation-driven state ( Fig 1A ) ., The Izhikevich model was used for the inhibitory neuron population to fit the fast-spiking inhibitory neuron firing pattern 63 ., Excitatory neurons are modelled by leaky integrate-and-fire models ., As in 24 , the excitatory neurons act as low pass-filters for their inputs while the FS neurons have a sub-threshold resonance in the gamma range 42–46 ., To demonstrate this , we injected an oscillatory current of small amplitude in a single cell and recorded the amplitude response for different oscillatory frequencies ., Excitatory neurons better respond to low frequency inputs , while FS neurons respond maximally for gamma inputs ( Fig 1B ) ., This is in line with the experimental evidence of Cardin et al . showing that FS-specific light stimulation amplifies gamma-frequencies 33 ., All neurons have chemical synapses but only inhibitory neurons are also coupled via gap junctions ( Fig 1A ) ., The gap junctions are modelled such that a voltage hyperpolarization ( depolarization ) in one neuron induces a voltage hyperpolarization ( depolarization ) in the connected neuron ., The current contribution of gap junction coupling is proportional to the difference of voltages between the coupled neurons , multiplied by the gap junction strength γ ( Fig 1C ) ., Moreover , when one neuron spikes , it emits a spikelet in the coupled neuron ., We model this by a positive inhibitory to inhibitory electrical coupling , which we add on top of the negative inhibitory to inhibitory chemical coupling ( see Materials and methods ) ., In order to understand the effects of gap junction plasticity , we initially considered the network without plasticity ., We first explored the network behavior for different values of the mean gap junction strengths γ and mean external drive to the inhibitory neurons νI ., As demonstrated by 24 , our network exhibits two regimes ( Fig 1D ) : an asynchronous irregular ( AI ) regime and a synchronous regular regime ( SR ) ., The AI regime occurs for networks with weak external drive and weak gap junctions ., In this regime the network is in the fluctuation driven regime so that the neurons spike due to variations in their input ., The SR regime occurs for strong external drive and strong gap junctions ., This regime leads to the emergence of gamma oscillations ., Mathematically , the network undergoes a Hopf bifurcation 24 , 39 ., The oscillations arise as the network directly inherits the resonance properties of the individual neurons ., This is mediated through the gap junction coupling which effectively allows positive coupling through their spikelets ., Moreover , the gap junctions reduce sub-threshold voltage differences between neurons which promotes synchrony ., The excitatory neurons are not necessary for the oscillations but they amplify the dynamics ( see 24 for mathematical derivations ) ., When placed in the SR regime , the network oscillates in the gamma-range at a frequency near the single neuron resonance frequency ( Fig 1E and 1F ) ., In addition , we observe that the spiking activity is characteristic to the network regime , with bursting activity in the synchronous regime and spikes in the asynchronous regime ( Fig 1G–1I ) ., To summarize , increased gap junction coupling and input drive into the network promotes gamma oscillations ., To explain the relationship between network activity and gap junction plasticity , we first model the simplest case of plasticity between a pair of electrically coupled neurons ., We then apply the plasticity rule to a population of neurons and investigate the effects on the network dynamics ., To determine how gap junction plasticity can alter network dynamics , we developed a model of the plasticity based on experimental observations ., 55 have shown that bursts in one or both neurons in an electrically coupled pair lead to long-term depression ( gLTD ) ., Therefore , we modeled gLTD as a decrease in the gap junction strength that is proportional to the amount of bursting ., The constant of proportionality , αgLTD serves as the learning rate ., To infer αgLTD , we reproduced the bursting protocol in Haas et al . , where a neuron bursting for a few milliseconds , 600 times for 5 minutes , leads to 13% decrease ( Fig 2A ) ., Activity-dependent gap junction long-term potentiation ( gLTP ) has not been reported experimentally yet in the mammalian brain ., There is evidence for activity dependent short-term potentiation in vertebrates 53 , 64 ., However , without potentiation , all gap junctions would likely become zero with time ., To address this concern , we hypothesize that gap junctions can undergo gLTP and we modeled it such that single spikes induce gLTP by a constant amount given by the potentiation learning rate αgLTP ( Fig 2B , first half ) ., Furthermore , we considered activity-independent gLTP rules in the supplementary materials ( S1 Fig ) ., Our plasticity model therefore potentiates gap junctions under spiking activity and depresses under bursting activity ., Therefore , we wondered how gap junction plasticity can alter network dynamics ., We previously quantified the amount of spiking versus bursting in our network for different levels of fixed gap junction strength and mean drive ., For low levels of both , the network is spiking whereas for high levels of both the network is bursting ., The spiking to bursting transition ( Fig 1G ) corresponds to the bifurcation ( Fig 1D ) from asynchronous irregular to synchronous oscillations at gamma frequency ., When inhibitory neurons are oscillating , they fire a burst of spikes at the peak of the oscillations ( Fig 1I , γ = 5 ) ., Therefore , when gap junctions are plastic , the network steady state can be found on the side of the bifurcation that balances the amount of potentiation due to spiking activity with the amount of depression due to bursting activity ., The depression learning rate is inferred from Haas et al . , while the potentiation learning rate is left as a free parameter ., We found that a strong relationship exists between gap junction plasticity and network synchrony ., When the network is in the AI regime , characterized by low prevalence of bursting activity , gap junction potentiation dominates ., However , for a strong mean coupling strength , the emergence of oscillations is associated by high bursting activity which leads to depression of the gap junctions ., Therefore gap junction plasticity in our network maintains a tight balance between asynchronous and synchronous activity ., Depending on the value of αgLTP , the position of the plasticity fixed point lies either in the asynchronous regime ( low αgLTP , Fig 2C ) or in the synchronous regime ( high αgLTP ) ., For high values of αgLTP , potentiation is fast while for low values , the potentiation is slow ., We wondered how gap junction plasticity would interact with time-varying inputs ., For the following experiment we consider slow gLTP ., First , we let the network reach its steady state with a low level of drive ( Fig 2E , beginning ) ., As previously observed , the mean gap junction strength reaches a value which sets the network near the AI/SR transition ., Then , we proceeded by injecting an additional constant current to the network ., This new current baseline induces network level oscillations ( Fig 2E , transition ) ., However , over time the mean gap junction strength decays due to the gap junction plasticity mechanism ., This gap junction depression is followed by a loss of synchrony and the network reaches its new steady state ( Fig 2E , end ) , again near the border of asynchronous and synchronous regimes ., We measured the response of read-out neurons which receive projections from the excitatory and inhibitory neurons in our network ( Fig 2D ) ., At the onset of the current step , the network undergoes transient oscillations ., When the gap junctions are plastic , the downstream neurons increase their spiking activity only for a few hundred milliseconds during the transient oscillations and then became almost quiescent again ( Fig 2F , second panel ) ., This contrasts with the simulation of a static network where the downstream keep a high firing rate ( Fig 2F , third panel ) ., These results suggest that synchronous activity is a powerful signal to provoke spiking in downstream neurons ., But oscillations and high firing rates of downstream neurons are also metabolically costly 65 ., With transient oscillations however , the downstream neurons only sparsely fire when the stimulus changes but not when it is predictable ., Thus , the regulation of oscillations mediated by gap junction plasticity allows for sparse but salient information transfer ., We now sought to study the functional implications of fast gLTP ., As stated before , this synchronizes the network into gamma oscillations ., Synchronization between networks is considered to be one possible mechanism of information transfer 66–69 ., We wondered whether gap junction coupling could mediate cross-network synchronization , and how gap junction plasticity would regulate this synchronization ., To test this hypothesis , we considered two subnetworks having different oscillation frequencies and coupled by gap junctions ( Fig 3A ) ., A fast network oscillates at a gamma frequency and therefore is called the gamma-network ., Then , a slow-network oscillates at a slower frequency as the membrane time constant of its inhibitory neurons is chosen to have a larger value ., Indeed , previous analyses show that the network frequency in our model is inherited from the single neuron resonance frequency of inhibitory neurons 24 , 70 ., As a result , increasing the membrane time constant of the inhibitory neurons results in a decrease of the network oscillation frequency ( Fig 3B–3D ) ., Cross-network gap junctions reduce the frequency and phase difference between the gamma- and slow-network ( Fig 3E and 3F ) and larger differences of subnetwork resonant frequencies require a larger number of cross-network gap junctions for the networks to oscillate in harmony ( Fig 3E and 3G ) ., Their common frequency lies between the resonant frequencies of the decoupled networks ., Importantly , cross-network synchronization requires the subnetworks to be in phase ., If the gamma- and slow-network do not share enough gap junctions , there is little mutual information and no correlation in their population activities ( Fig 3H and 3I ) , despite having a common oscillation frequency in some cases ( Δfres = 0; number of shared GJs = 0 on Fig 3I ) ., However , for small differences in the subnetworks resonant frequency Δfres , increasing the number of shared gap junctions induces the oscillations to lock together ., The networks oscillate in phase ( Fig 3F , end of first row ) as reflected in their mutual information ( Fig 3H , dark blue area ) and their correlation ( Fig 3I , dark red area ) ., In summary , two networks in the SR regime with different resonance frequencies and/or out-of-phase can synchronize if they are coupled by gap junctions ., Furthermore , a large number of shared gap junctions is required for large differences of resonant frequency ., As gap junctions can synchronize two oscillating populations of neurons , we wondered whether the same synchronization would occur with one population in the AI regime ., First , we initialized the gamma-network in the AI regime while the slow-network was initialized in the SR regime ( Fig 4A ) ., After coupling the gamma- and slow-network together , we found that , while the oscillation frequency of the gamma- and slow-network matched ( Fig 4B ) , the two networks could not synchronize ., The networks were always out-of-phase with very weak correlation between the population activities ( Fig 4C and 4D ) ., The results were similar if the gamma- and the slow-network were initialized in the reverse synchronous and asynchronous parameter regimes , respectively ( not shown ) ., Cross-network synchronization is not robust when one network is not oscillatory ., Given these constraints on cross-network synchronization , we wondered if gap junction plasticity could remedy the situation and allow for robust cross-network synchronization ., To test this hypothesis , we repeated the simulation protocols with the gamma- and slow-network initialized in the asynchronous and synchronous regimes ( respectively ) and with plastic gap junctions ., Here we considered the case where the gLTP rates were slow ., As shown previously , gap junction plasticity regulates oscillations such that the network in the asynchronous irregular regime transitions to the oscillatory regime ( Fig 4E ) ., The oscillation frequencies of these two networks match ( Fig 4F ) ., Strikingly , even with a large resonant frequency difference , the gamma- and slow-network now synchronize through a small number of shared gap junctions ( Fig 4G and 4H ) ., This indicates that gap junction plasticity allows for cross-network synchronization that is robust to the underlying neuronal parameters for small numbers of shared gap junctions ., We hypothesized that cross-network synchronization mediated by plasticity allows information transfer ., To investigate this , we considered a similar network architecture as previously studied , with two networks , an input-network and an output-network ., The input-network receives an input projected by random weights to its neurons ., The output-network is connected to the input-network with a small number of gap junctions and inhibitory chemical synapses ., First , to demonstrate the information transfer capability of the network , we consider static gap junctions with oscillatory inputs to the input-network ., The stimulus information is transmitted to the output-network via the frequency modulation of the synchronized oscillations and not by spike transmission nor amplitude modulation ( Fig 5A–5D ) ., When sharing gap junctions , the input- and output-network synchronize together ( Fig 5A ) and their spiking activity is locked ( Fig 5B ) ., As the amplitude of the input signal increases , the spiking activity increases in the input-network but not in the output-network ( Fig 5C ) ., For a network in the SR , there is a positive correlation between the signal amplitude and the network oscillation frequency ( Figs 1E and 5D ) ., This frequency modulation is transferred from the input- to the output-network ., Thus , the input amplitude can be estimated from the oscillation frequency of the output-network , despite the absence of chemical synapses between the input-network and the output-network ( Fig 5E ) ., However , this synchrony code is only possible for signals below a certain frequency ( Fig 5F and 5G ) ., Indeed , the instantaneous oscillation frequency is estimated by measuring the period between consecutive peaks of the population activity ., For example , oscillations at 50 Hz have a period of 20 ms . Variations happening within those 20 ms are compressed to a single period value and thus are not transferred via frequency modulation ., Mechanisms for estimating the input value from the oscillation frequency of the output-network are discussed further in the methods section ., Finally , we tested if this synchrony code was valid for non-oscillatory signals ( Fig 5H ) ., We found that non-oscillatory , slowly varying random signals could also be robustly transmitted from the input- to the output-network with gap junction coupling ( Fig 5I ) ., As gap junction plasticity can regulate oscillations , we tested whether the plasticity can make this synchrony code robust to parameter variations or potential gap junction loss ., First , as previously shown , gap junction plasticity enhances the ability of networks to synchronize ., If initialized in the AI regime and with static gap junctions , there is no information transfer via frequency modulation ( Fig 5J , left panel ) ., However , with plasticity and fast gLTP , the oscillations are regulated and the network synchrony is recovered which results in successful information transfer ( Fig 5J left panel ) ., A critical amount of oscillation power and a critical number of shared gap junctions are required for information transfer , after which increasing each of them does not yield significant improvement ( Fig 5J ) ., Furthermore , we studied whether gap junction plasticity could restore information transfer if gap junctions were deleted ., While there is loss in the quality of the transfer when static gap junctions are removed , plastic gap junctions maintain the quality of the transfer by increasing the strength of the remaining gap junctions ., This mechanism compensates for the missing gap junctions ( Fig 5J and 5K ) ., To summarize , gap junction plasticity expands the necessary conditions for information transfer ., It regulates oscillations , and by promoting phase-locking of oscillations , it contributes to the propagation of information to downstream networks ., Finally , if some gap junctions are failing , due to protein turnover perhaps , the remaining ones can increase their strength through plasticity ., This helps to maintain accurate information transfer ., Despite being less common than chemical synapses , gap junctions are ubiquitous in the central nervous system ., Example includes the inferior olivary nucleus 71–73 , the thalamic reticular nucleus 74 , 75 , the hippocampus 36 , 76 , the retina 52 , 77 , the olfactory bulb 78 , the locus coeruleus 79 , or also the neocortex 80 , 81 ., Moreover , they drastically alter the firing activity of their connecting neurons 82 , 83 , as well as the network dynamics 20–24 ., Furthermore , gap junctions between inhibitory interneurons are reported in many cortical regions where global oscillations of neural activity are observed 21 , 27 , 84 , 85 ., These inhibitory neurons exhibit sub-threshold resonance that amplifies a specific frequency range 33 ., Therefore , gap junction induced synchrony and inhibitory neurons frequency preference are a possible substrate for global oscillations in these cortical regions ., Our work is consistent with recent results showing that together gap junction strength and sub-threshold resonance of inhibitory neuron promote oscillations of neuronal activity 24 , 70 ., There has been a recent interest in modelling gap junction plasticity ., Snipas et al . 86 developed of model of gap junction coupling that would exhibit short-term plasticity ., By combining a 36-state model of gap junction channel gating with Hodgkin-Huxley equations 87 , they show that gap junction channel gating , induced by bursting activity , could lead to short term depression ., In future work , it would be interesting to combine this model of gap junction short-term plasticity with our model ., Chakravartula et al . 88 introduced a new type of adaptive diffusive coupling in a network of Hindmarsh-Rose neurons 89 , 90 ., They assumed that connections between pairs of neurons would follow a Hebb’s law 91 , where neurons with simultaneous activity would strengthen their connection , while others with dissimilar activity would weaken their coupling ., They observe the emergence of locally synchronized groups of neurons , whose synchronization could be transient or permanent ., Their results are consistent with ours showing synchronization of subnetworks coupled with gap junctions ., Recently , Haas et al . 55 reported the first experimental evidence of activity-dependent gLTD of gap junctions of interneurons in the thalamic reticular nucleus , even though the mechanism remains to be investigated 62 ., Also Sevetson et al . 56 found that calcium-regulated mechanisms support gap junction gLTD in the thalamic reticular nucleus ., The mechanisms are similar to those observed for the plasticity of chemical synapses ., We designed a rule for activity-dependent gLTD consistent with those results ., We assumed that a cortical fast-spiking interneuron would exhibit the same plasticity properties as a thalamic reticular neuron because gap junctions are mostly made from the connexin Cx36 throughout the central nervous system 74 , 92 ., To our knowledge , there is no study yet on activity-dependent gLTP of gap junctions ., However recent studies suggest that gLTD and gLTP share a common pathway 48 , 56 ., Therefore , we propose a rule for activity dependent gLTP , assuming that low frequency spiking activity leads to gap junction potentiation ., However , our results do not depend on the exact formulation of gLTP ., As we have shown , an activity-independent rule yields similar behavior ( supplementary material , S1 Fig ) ., Moreover , we did not observe significant changes by modelling asymmetrical gap junctions ( supplementary material , S2 and S3 Figs ) ., Our model demonstrates that the regulation of oscillations is mediated by gap junction plasticity ., Fast potentiation leads to bursting activity while slow potentiation leads to asynchronous irregular activity ., Our first hypothesis assumed that the potentiation is slow and the network is in the AI regime ., Thus , at the steady-state , gamma power is weak or non-existent ., Evidence from Tallon-Baudry et al . and Ray et al . 93 , 94 is consistent with our results ., When no stimulus is provided or task required , electroencephalogram recordings show that power in the gamma-band is weak ., After the onset of a sensory stimulus , gamma oscillations can be detected in cortical areas ., This has been reported for example with visual stimuli triggering gamma oscillations in the mouse visual cortex 95 ., In our model , the neurons oscillate transiently when receiving a constant external stimulation ., This mechanism operates by crossing the bifurcation boundary between the AI and SR regime ., However , over time the mean gap junction strength decays due to the additional bursting activity ., The gap junction depression leads to a loss of synchrony and the network returns to the AI regime ., Therefore we predict a loss in gamma power for sustained stimulus ., A similar mechanism may be involved in the reduction of gamma oscillation induced by slow smooth movements 96 , 97 ., We wondered what could be the functional role of this transient oscillatory regime ., Projecting the excitatory activity of our network model to downstream neurons revealed that they fire sparsely , for a short duration after stimulus onset , and are quiescent otherwise ., Thus , gap junction plasticity could efficiently encode the change in incoming stimuli ., This could allow for energy conservation as oscillations are energetically expensive 65 ., Moreover , 98 show that cortical circuits near the onset of oscillations could promote flexible information routing by transient synchrony ., The role of gamma oscillations is highly debated 94 ., They could play no role and simply be a marker of the excitation-inhibition interaction ., However others studies suggest they could be involved in information transfer ., It is thought that retinal oscillations carry information to the visual cortex 99 ., Moreover they could serve as inter-area communication by promoting coherence in neural assemblies which would align their windows of excitation ., This would allow for effective spike transmission 68 , 94 , 100 ., Furthermore , Roberts et al . 101 observed high gamma coherence between layers 1 and 2 of macaque’s visual cortex by dynamic frequency matching ., Here , we demonstrate one potential mechanism for information transmission through gamma oscillations ., Our networks make use of gamma frequency modulation to transmit information in a robust manner , similar to the principle used for FM radio broadcasting ., The amplitude of the input signal modulates the oscillation frequency , which increases almost linearly with the amplitude ., Our model demonstrates that gap junction plasticity robustly mediates network oscillations and cross-network synchronization ., If some gap junctions are removed , the remaining gap junctions become stronger and compensate for the missing ones ., Thus , gap junction plasticity insures the phase-locking of the coupled network and it allows for information routing ., In particular , there is evidence suggesting that gap junctions could promote long-distance signaling by implementing frequency modulation of calcium waves in astrocytes 102 ., Moreover , correlation was found during gamma activity between amplitude and frequency modulation of local field potential of CA3 pyramidal neurons of anesthetized rats 103 ., In addition , our network models could also represent the subnetworks of the TRN , with each connected to a separate excitatory neuron of thalamus 104 ., However , TRN inhibitory neurons exhibit longer bursts than those of cortical fast-spiking neurons , due to long lasting T-current ( about 50ms ) and further work is necessary to make predictions on this brain region behaviour 105 ., Failure to regulate oscillations , could be the origin of several cognitive pathologies ., Disruption of brain synchrony in the inferior olive is thought to contribute to autism due to the loss of coherence in brain rhythms 106 ., Excess of high frequency network wide oscillations in the cortex have been observed to also correlate with autism in young boys 12 ., The inferior olive differs for its density of gap junction being the highest in the adult brain 71 , 72 ., It may be involved in the generation of tremors in Parkinson’s disease , however the severity of induced tremors in Cx36 knockout mice remained the same as in wild-type mice 107 , 108 ., This could be due to gap junctions made from other connexins ( such as Cx43 ) taking over for the knocked-out ones ., Recent studies highlight the critical role of gap junctions and their plasticity in efficient cognitive processing 109 ., As experimental and computational techniques improve , new efforts can further unveil their properties and expand our understanding of cortical functions ., Our computational model shows that gap junction activity-dependent plasticity may play an important role in network-wide synchrony regulation ., We model Fast Spiking ( FS ) interneurons with Izhikevich type neuron models 63 ., This model offers the advantage to reproduce different firing patterns as well as a low computational cost 112 ., The voltage v follows | Introduction, Results, Discussion, Methods | Cortical oscillations are thought to be involved in many cognitive functions and processes ., Several mechanisms have been proposed to regulate oscillations ., One prominent but understudied mechanism is gap junction coupling ., Gap junctions are ubiquitous in cortex between GABAergic interneurons ., Moreover , recent experiments indicate their strength can be modified in an activity-dependent manner , similar to chemical synapses ., We hypothesized that activity-dependent gap junction plasticity acts as a mechanism to regulate oscillations in the cortex ., We developed a computational model of gap junction plasticity in a recurrent cortical network based on recent experimental findings ., We showed that gap junction plasticity can serve as a homeostatic mechanism for oscillations by maintaining a tight balance between two network states: asynchronous irregular activity and synchronized oscillations ., This homeostatic mechanism allows for robust communication between neuronal assemblies through two different mechanisms: transient oscillations and frequency modulation ., This implies a direct functional role for gap junction plasticity in information transmission in cortex . | Oscillations of neural activity emerge when many neurons repeatedly activate together and are observed in many brain regions , particularly during sleep and attention ., Their functional role is still debated , but could be associated with normal cognitive processes such as memory formation or with pathologies such as schizophrenia and autism ., Powerful oscillations are also a hallmark of epileptic seizures ., Therefore , we wondered what mechanism could regulate oscillations ., A type of neuronal coupling , called gap junctions , has been shown to promote synchronization between inhibitory neurons ., Computational models show that when gap junctions are strong , neurons synchronize together ., Moreover recent investigations show that the gap junction coupling strength is not static but plastic and dependent on the firing properties of the neurons ., Thus , we developed a model of gap junction plasticity in a network of inhibitory and excitatory neurons ., We show that gap junction plasticity can maintain the right amount of oscillations to prevent pathologies from emerging ., Finally , we show that gap junction plasticity serves an additional functional role and allows for efficient and robust information transfer . | cell physiology, resonance frequency, medicine and health sciences, action potentials, neural networks, nervous system, membrane potential, junctional complexes, electrophysiology, neuroscience, gap junctions, neuronal plasticity, computer and information sciences, animal cells, resonance, physics, cellular neuroscience, cell biology, anatomy, synapses, physiology, neurons, biology and life sciences, cellular types, physical sciences, neurophysiology | null |
journal.pntd.0007528 | 2,019 | Potential effects of heat waves on the population dynamics of the dengue mosquito Aedes albopictus | Originated from Southeast Asia , Asian tiger mosquito ( Aedes albopictus ) is the most prevalent vector in all continents except the Antarctica 1 , 2 , 3 , 4 , 5 ., The pathogens it transmits pose a severe threat to human health by global epidemics , including dengue and Zika arboviruses ., For instance , the dengue incidence has increased six-fold from 1990 to 2013 , with cases more than doubled every decade 6 ., This historical evidence suggests a crucial need to develop effective disease control and intervention strategies in order to minimize the risk of epidemic spread and infection 7 , 8 ., Meanwhile , the Zika virus , being detrimental to children born with microcephaly and neurological disorders , has spread from Brazil to twenty-six other countries or territories in the Americas within one year 1 ., Despite the increasing infections and rapid spread of these arboviruses , no effective antiviral treatment exists ., Thus , controlling the development of mosquito vectors becomes a viable option for curbing the disease transmission , especially in regions with limited public health resources 9 ., The life cycle and transmission of most infectious agents are inextricably linked to climate 10 ., Ae ., albopictus is a small-bodied ectotherm; its population abundance and dynamics are firmly regulated by meteorological factors 11 and are sensitive to climate change 5 , 12 ., Temperature influences many aspects of Ae ., albopictus’ life cycle in a non-linear fashion 13 , 14 , 15 ., Lukewarm temperature fosters the development of mosquito at the stages of egg incubation 13 , larval pupation 14 , and pupal eclosion 15; and it shortens the extrinsic incubation period , eventually expediting the transmission cycle and adult production 15 ., However , when temperature exceeds a certain threshold , the effects on the mosquito development become contrastingly different and even detrimental 15 ., It was tested that the duration of the gonotrophic cycle or the oviposition extended and the number of laid eggs decreased when the temperature rose above 35 . 0°C 16 ., Findings from forecasting models also proved that the mosquito population tended to decrease in certain tropical regions under extremely hot weather 17 ., The mechanism leading to the population dynamics of Ae ., albopictus has yet to be elucidated 14 , 15 , 16 , 18 ., An obstacle to the identification is the uncertainty of climatic conditions , such as the onset , peak , and duration of extreme weather events , which are globally heterogeneous and regionally specific ., Furthermore , seeking the theoretic pathway to the mosquito development is becoming more challenging , since the global climate manifests a higher degree of oscillation 19 , 20 , 21 ., A coupled global climate model predicts that heat waves , as common extreme weather events , will become more frequent and longer-lasting in the second half of the 21st century 22 ., Despite few instances exploring the statistical links between heat waves and the mosquito ecology , the climate-driven mechanism has been poorly understood 23 ., Specifically , little is known about how heat wave characterisitics ( e . g . , the onset day of a heat wave , the duration of a heat wave ) affect the development ., This existing knowledge gap obfuscates developing effective strategies to prevent and control mosquito-borne epidemics ., Many studies have employed controlled experiments to identify the response of Ae ., albopictus to extremely high temperature 14 , 15 , 16 , 18 ., These studies , however , cannot capture the full range of parameters in the mosquito’s life stages , since the development process is relatively slow , complicated , and unrepeatable ., Statistical methods ( e . g . , multivariate regression models ) are able to establish the long-term association between environmental factors and population growth , but they are invariably focused on the aquatic stages ( e . g . , larvae ) and are thus unable to characterize the growth parameters in the aerial stages ( i . e . , adults ) and explain the intricacy of the transition between stages 24 ., Most importantly , the few recorded heat wave events at data collection sites pose a considerable challenge to the model validation ., To overcome the data issue , computer-based simulations of the weather processes offer an alternative solution 25; however , very few existing studies are focused on the impact of extreme weather 23 ., In addition , most simulation studies rely on statistical models while overlooking the intrinsic process of the development within the mosquito’s life-history stages ., The mechanistic population model , which establishes the multi-stage development of the mosquito by a series of differential equations , has become popular in the entomological research of mosquitoes 26 , 27 , 28 ., Recently , Jia et al . 29 proposed a mechanistic population model that accounts for the diapause behavior , referring to the inactive state in which the mosquito is unable to hatch and ceases from the development in order to survive extreme environmental conditions ( e . g . , high temperature , extreme desiccation ) ., This model , termed the mechanistic population model of Ae ., albopictus with diapause ( MPAD ) , has been further explored in this paper to identify the mechanistic associations between heat waves and the population abundance of Ae ., albopictus ., A 35-year historical heat wave dataset was employed to extract key climatic elements ., Finally , a rich set of mathematical simulations were conducted to thoroughly investigate the important mechanisms responsible for the population dynamics of Ae ., albopictus caused by heat waves ., Our study area is in Guangzhou ( 113 . 23°E , 23 . 17°N ) ( Fig 1 ) —the largest city of Southwest China with over 12 . 7 million population and a population density of 1 , 708 residents per km2 30 ., This mega-city has a distinct subtropical climate with an average annual temperature of 21 . 9°C and an annual rainfall ranging from 1 , 370 to 2 , 353 mm ., The humid and warm climate is favorable for Ae ., albopictus to survive and grow ., In 2014 , an unprecedented outbreak of dengue fever occurred in Guangzhou , causing 37 , 305 cases of infections 30 ., This outbreak was attributed to the combined effects of the urban heat island and climate change , including more frequent and intense heat wave events 31 ., The theoretical foundation of the study is the climate-driven and process-based MPAD model 29 , 32 ., The MPAD model formulates the continuous development of Ae ., albopictus in a seven-stage process using a bottom-up approach , as shown in Eqs ( 1 ) through ( 7 ) ., These seven stages include eggs ( E , including non-diapause E0 and diapause Edia , Eq ( 1 ) ) , larvae ( L , Eq ( 2 ) ) , pupae ( P , Eq ( 3 ) ) , emerging adults ( Aem , Eq ( 4 ) ) , blood-fed adults ( Ab , Eq ( 5 ) ) , gestating adults ( Ag , Eq ( 6 ) ) , and ovipositing adults ( Ao , Eq ( 7 ) ) ., In each equation ( representing one development stage ) , the variation of daily population abundance ( marked in the prime notation ) is determined by ( 1 ) the accumulated population from the last stage , ( 2 ) the mortality at the current stage , and ( 3 ) the population developing into the next stage ., The life-history traits are driven by both climate-dependent parameters and climate-independent parameters ., The climate-dependent parameters include daily mean temperature , daily accumulated precipitation , and daily photoperiod ., These variables , derived from the experimental results 12 , 16 , 33 , 34 , are given in S1 Table and S2 Table ., One highlight of the MPAD model is the consideration of diapause ., Diapause-related parameters , as indicated by the subscript dia in Eqs ( 1 ) through ( 7 ) , are defined to indicate whether the mosquito eggs are dormant or whether adults suspend the hatching activity under extreme conditions 35 , 36 ., The performance of the MPAD model was evaluated in our previous work by comparing against field Ae ., abopictus container index ( CI ) in two Chinese cities: Guangzhou and Shanghai 29 ., The coefficient of determination ( r2 ) was 0 . 84 in Guangzhou and 0 . 90 in Shanghai , which showed a significant improvement over previous mechanistic population models ., The better performance was attributed to the inclusion of diapause-related parameters and the modification of temperature-driven parameters ., These adjustments are of critical importance in regions characterized by considerable seasonality ( e . g . , temperate zones ) , where the intra-annual dynamics of mosquito population only emerges with one peak ., The heat wave ( HW ) is an extended period of continuously hot weather , typically followed by a high level of humidity 22 ., However , since local acclimatization and adaptation influence the impact of extreme heat , there is no globally accepted measure of heat waves 37 ., A widely used strategy is to define heat wave locally using both intensity and duration indicators 38 , 39 ., Here , we first adopted one heat wave definition given by the China National Standard: a heat wave refers to an extreme weather event where the daily maximum temperature is greater than or equal to 35 . 0°C for at least three consecutive days ( HW Definition I ) 40 ., To extract the historical heat wave events , we acquired all available daily temperature and precipitation measurement data in Guangzhou from the China Meteorological Data Sharing Service System , and generated a 35-year climate dataset spanning from 1980 to 2014 41 ., We also derived the photoperiod data from the National Oceanic and Atmospheric Administration 42 ., Using the temperature dataset , we identified all heat waves in the study area ., Heat wave events operate at both fast and slow rates with various degrees of severity ., These processes can be characterized by the onset day ( OHW , the first day in day of year DOY when a heat wave occurs ) , the duration ( DHW , the period of consecutive heat wave days ) , and the average daily mean temperature ( TaveHW ) ., Their descriptive statistics are shown in Table 1 . The frequency distributions ( fit by trend curves ) of their characteristics are summarized in Fig 2 . The only year without an occurrence is 1996 , after which an increased frequency can be identified ( Fig 2A ) ., The occurrences have a strong seasonality , where the most frequent DOYs range from mid-July to mid-August ( Fig 2B ) ., More than two-thirds of events ( n = 86 ) have lasted three to four days ( Fig 2C ) ., In addition , the peak of TaveHW ranges from 29 . 7–30 . 5°C ( Fig 2D ) and the peak of the maximum of the daily maximum temperature ( TmaxHW ) is around 35 . 5–36 . 8°C ( Fig 2E ) ., As the heat wave is a complex extreme weather event , the estimates of the recurrence probabilities of heat waves are used as the proxy for the temporality of their occurrences 43 , which are calculated from the probability distributions ( Pdf ) of OHW , DHW , and TaveHW , as given by Eqs ( 8 ) through ( 10 ) ., In order to evaluate the effects of heat waves on population abundance , one assumption is to make the non-heat wave conditions constant across the period of observation without inter-annual variability ., Thus , we calculated the averaged annual daily mean temperature ( T ) over 35 years using Eq ( 11 ) , as shown in Fig 3A ., We also derived the time series of two other key climatic variables required by the MPAD model: the daily accumulated precipitation ( P ) and the photoperiod ( PP ) ( Eq 11 ) ., The temperature series T was then replaced by the temperature of a heat wave event ( THW , the red line in Fig 3B ) during the heat wave DOYs ( Eq 12 ) ., This new synthetic temperature series was labeled as T’ ( Fig 3B ) ., A total of 127 such temperature series T’ were generated ., In addition , we also found that using a single year is insufficient to identify the climate-driven mechanism , as the result is largely dependent on the conditions in Year 1 27 , 28 , 29 ., Thus , we extrapolated the 3-year temperature curve by placing T’ in Year 2 ( Fig 3C ) ., We then designed experiments to test the effect of the 3-year temperature curve on the population abundance in Year 2 and Year 3 . i = 1…35 ( year ) j = 1…365 ( day ) X = T , P , PP, T′={THWOHW≤DOYs≤OHW+DHWTotherDOYs, ( 12 ), The stage of blood-fed adults is of critical importance in the disease ecology ., During this period , the mosquito becomes an active transmission vector of disease pathogens 44 ., For this reason , we used the daily population abundance of the blood-fed adults to examine the heat wave effects ., Here we compared the population dynamics between two groups of blood-fed adults: the control group ( A ) under the non-heat wave scenario ( T ) and the test group ( AHW ) under the heat wave scenario ( T’ ) ., After deriving the daily population abundance of A and AHW by the MPAD model , we calculated the relative difference in population ( R ( j ) ) , as shown in Eq ( 13 ) ., We then derived the duration of consecutive days ( RD ) when this relative difference exceeds 10% as a proxy for the heat wave effect , as shown in Eq ( 14 ) ., R ( j ) =|AHW ( j ) −A ( j ) |A ( j ) , j=1…365, ( 13 ), RD=|tE−tB|, ( 14 ), where tB denotes the first day when R ( j ) exceeds 10% and tE denotes the last day when R ( j ) exceeds 10% ., Then we tested the effect of each heat wave characteristic ., Specifically , the proposed indicator RD is treated as a function of three heat wave variables ( OHW , DHW , and TaveHW ) , as shown in Eq ( 15 ) ., To test the contribution of each climatic factor , we designed three groups of sensitivity analysis ., In each group , only one factor was treated as a test variable while the two other factors were held constant as controlled variables , as shown in Table 2 . For example , in the first group ( {RD}~OHW ) , the value of OHW was randomly drawn from its probability distribution ( Eq ( 8 ) ) for 1 , 000 times , while the two other factors DHW and TaveHW were selected as the combinations of their first , second , and third quartiles ( the values were drawn from Table 1 ) ., This group of simulation generated a total of 9 , 000 runs ., The given heat wave definition generated a total of 127 synthetic heat wave temperature series T’ ., These series of T’ served as the input into the MPAD model , further generating 127 daily blood-fed adult population abundance curve AHW as the outcome ., Comparatively , the population abundance A under the non-heat wave scenario T was also derived ., The overlay of simulated heat wave population curves AHW is shown in Fig 4 , which reveals that the historical heat waves only occurred briefly from early summer into early autumn ( DOYs∈144 , 271 , DHW∈3 , 18 ) ., Their effects on the population abundance were also limited to the time period when heat waves stroke and would not carry over to winter or the next year ., In addition , the heat wave occurrences mostly suppressed the mosquito development rather than promoted it , as demonstrated by comparing A and one selected AHW ( Fig 4 inset ) ., We further examined how the population dynamics responded to the variation of individual heat wave characteristics , including OHW ( Fig 5A and 5B ) , DHW ( Fig 5C ) , and TaveHW ( Fig 5D ) ., For each test variable , we held the other two variables constant and included three specific cases for discussion ., In addition , we generated the population abundance under the non-heat wave scenario T ( black curve in Fig 5 ) and derived its peak at DOY 192 ., Fig 5A shows the examples of three heat waves with different onset days ( i . e . , OHW = 169 , 179 , and 189 ) ., These scenarios , with an onset day earlier than DOY 192 , generated population curves similar to that under the non-heat wave scenario ., The early onset of heat wave slightly advances the emergence of the population peak but has no cascading effect on the late stage development ( DOY > 225 ) ., However , when heat waves occur after DOY 192 ( i . e . , OHW = 205 , 214 , and 244 ) , the population curves largely shift , where a greater level of variation is observed ( Fig 5B ) ., Fig 5C shows three heat waves with different durations ( i . e . , DHW = 4 , 7 , and 18 ) , which demonstrates that the longer the event lasts , the greater extent it suppresses the population growth ., Lastly , Fig 5D shows three scenarios under different temperature conditions ( i . e . , TaveHW = 29 . 1 , 29 . 6 , and 30 . 7 ) , where the resulting effects on the population abundance are not significant ., One noticeable pattern in all of these scenarios is that the population grows when the heat wave strikes but plummets after a short period ., Several factors may contribute to this phenomenon ., Environmentally , long-lasting heat waves can dry up shallow bodies of water and subsequently deprive mosquitos of breeding grounds ., Physiologically , heat waves can also cause most mosquito species to spawn at once and then dry in unison when weather becomes extreme ., Several genes of heat shock protein—known to overcome high temperature stress—tend to show downregulation in larvae when subject to thermal stress at 39°C 45 ., Besides the visual assessment , we further quantified the effects via statistical regressions based on the experimental design in Table 2 ., Specifically , we used RD—consecutive days when the relative difference in the population abundance exceeds 10%—as a population index representing the heat wave effects ., Fig 6 shows the associations between RD and OHW , DHW , TaveHW ., In Fig 6A–6C , the relationship between RD and OHW generally follows a quadratic form ( average r2 is around 0 . 90 ) with the trough appearing in late July ( DOY 203–204 ) ., We noticed that in Fig 6C , in addition to the quadratic curve , two peaks emerge in early June ( DOY 160 ) and late September ( DOY 265 ) when both DHW and TaveHW are at their third quantiles ( i . e , blue curve in Fig 6C ) ., In Fig 6D–6F , a significant linear correlation is observed between RD and DHW only with a large OHW ( i . e . , late heat wave onset , blue lines in Fig 6D–6F ) ., In Fig 6G–6I , RD and TaveHW have a piecewise association , which is relatively flat before TaveHW = 30 . 5 and follows a linear pattern afterwards ., The full list of the mathematical relationships are included in S3 Table ., The definition of a heat wave event is regionally specific 39 ., Since there is a lack of consensus about the heat wave definition , we would like to examine if our results are robust when a different definition applies ., To test the sensitivity of the MPAD model , we adopted two other heat wave definitions that have been previously employed in Guangzhou 46 , 47 ., The second definition is less restrict: a heat wave is defined as ≥ 2 consecutive days with the daily mean temperature at or above the 95th percentile of the year ( HW Definition II ) 46 ., The last definition is a stricter criterion: a heat wave is defined as ≥ 7 consecutive heat days with the daily mean temperature at or above the 95th percentile ( HW Definition III ) 47 ., Based on the same experimental design , we extracted the historical heat waves according to each new definition ., Their descriptive statistics are shown in S4 Table ., Then , we simulated the mosquito population AHW under each new heat wave definition and tested the relationship between RD and the three heat wave variables OHW , DHW , and TaveHW following the simulation design in Table 2 ., The results are shown in S1 Fig ( for HW Definition II ) and S2 Fig ( for HW Definition III ) ., A total of 489 heat waves were extracted by using HW Definition II ., It can be observed from the results that the correlation patterns are in consistent with HW Definition I . However , when HW Definition III was employed , only 12 heat waves were extracted ., With the few identified events , we were unable to establish a significant correlation pattern ., It is thus demonstrated that our simulation results are robust , when a sufficient number of observations can be generated using a new definition ., There is much epidemiological evidence demonstrating how climate variations and trends affect human health outcomes 55 , 56 , 57 ., Despite the many explorations on the disease pathogens , the complicated interplay between heat waves and Ae ., albopictus remains unclear ., This paper explores the variability of Ae ., albopictus responding to heat waves events using a 35-year historical climate dataset via mathematical modeling and a simulation design ., Our simulation results reveal that the unusual onset of a heat wave and a relatively high temperature over an extended period are the two primary factors inhibiting the population development ., As the frequency and severity of heat waves are likely to increase in the future 22 , this study provides insights into assessing the potential effects on the mosquito introduced by the global climate ., Understanding this climate-driven mechanism is crucial to developing effective strategies to prevent and control dengue fever , Zika , as well as other far-reaching mosquito-borne epidemics . | Introduction, Materials and methods, Results, Discussion | Extreme weather events affect the development and survival of disease pathogens and vectors ., Our aim was to investigate the potential effects of heat waves on the population dynamics of Asian tiger mosquito ( Aedes albopictus ) , which is a major vector of dengue and Zika viruses ., We modeled the population abundance of blood-fed mosquito adults based on a mechanistic population model of Ae ., albopictus with the consideration of diapause ., Using simulated heat wave events derived from a 35-year historical dataset , we assessed how the mosquito population responded to different heat wave characteristics , including the onset day , duration , and the average temperature ., Two important observations are made: ( 1 ) a heat wave event facilitates the population growth in the early development phase but tends to have an overall inhibitive effect; and ( 2 ) two primary factors affecting the development are the unusual onset time of a heat wave and a relatively high temperature over an extended period ., We also performed a sensitivity analysis using different heat wave definitions , justifying the robustness of the findings ., The study suggests that particular attention should be paid to future heat wave events with an abnormal onset time or a lasting high temperature in order to develop effective strategies to prevent and control Ae ., albopictus-borne diseases . | Understanding the population dynamics of Asian Tiger mosquito ( Ae . albopictus ) –the most prevalent vector of global epidemics including West Nile virus , dengue fever , Zika–could shed lights on improving the understanding of vector transmission as well as developing effective disease control strategies ., It is widely acknowledged that the life cycle of Ae ., albopictus is firmly regulated by meteorological factors in a non-linear way and is sensitive to climate change ., Our study extends the understanding about how extreme heat events manipulate the mosquito population abundance ., We adopted an existing mechanistic population model of Ae ., albopictus , combined with a rich set of simulated heat wave events derived from a 35-year historical dataset , to quantify the mosquito’s responses to different heat wave characteristics ., We found that an abnormal onset time and a lasting high temperature play the most important role in affecting the mosquito population dynamics ., We also performed a sensitive analysis by changing the definition of the heat wave , justifying the rigor of the conclusion ., This research provides implications for developing public health intervention strategies: to control dengue fever , Zika , as well as other far-reaching mosquito-borne epidemics , priority should be given to heat wave events with an abnormal onset time or a lasting high temperature . | ocean waves, invertebrates, medicine and health sciences, atmospheric science, population dynamics, experimental design, animals, research design, developmental biology, physiological processes, diapause, population biology, infectious disease control, insect vectors, research and analysis methods, climate change, infectious diseases, oceanography, marine and aquatic sciences, disease vectors, insects, arthropoda, mosquitoes, eukaryota, climatology, earth sciences, physiology, biology and life sciences, species interactions, organisms | null |
journal.pbio.2005956 | 2,018 | Integrative network-centric approach reveals signaling pathways associated with plant resistance and susceptibility to Pseudomonas syringae | Plant immunity is generated by the activation and coordination of several protein kinase-based signal transduction pathways into cellular defense responses 1 , 2 ., Kinases modify the activity status of other proteins through specific biochemical modifications ( substrate phosphorylation ) or by recruiting proteins in signaling complexes ., Signaling pathways transmit pathogen signals from the cell periphery to intracellular compartments and trigger changes in gene expression , hormone-based signaling , and defense compound production 3 ., To survive in plant tissues and ensure spread to other plants , pathogens must overcome plant defenses and redirect their energetic and nutrient resources ., The constant tug-of-war between plants and pathogens has generated a complex immune system in plants , and equally multifaceted assault and endurance mechanisms in pathogens ., Plant pathogens such as the gram-negative flagellated bacterium Pseudomonas syringae can colonize a broad range of plants , an ability at least partly determined by an extensive and versatile effector repertoire 4 , 5 ., P . syringae subverts the basal immunity in part by attacking components of signaling pathways activated by pathogen-associated molecular patterns ( PAMPs ) or secreted effectors ., PAMP-triggered immunity ( PTI ) is induced by PAMP perception by pattern recognition receptors ( PRRs ) , some of which are receptor-like kinases ( RLKs ) ., Upon PAMP recognition , PRRs activate membrane-associated receptor-like cytosolic kinases ( RLCKs ) , cytosolic mitogen-activated protein ( MAP ) kinase ( MAPK ) cascades , and other cytosolic kinases , including Ca2+-dependent kinases 6 ., Effector-triggered immunity ( ETI ) , the second layer of immunity , is activated by direct or indirect recognition of effectors , followed by activation of signaling pathways and induction of defense responses and programmed cell death ( PCD ) ., However , most intracellular effectors are not recognized by the plant and instead are thought to impair the plant’s ability to sustain an efficient immune response , a condition described as effector-triggered susceptibility ( ETS ) 7 ., Work with Arabidopsis , tomato ( Solanum lycopersicum ) , and Nicotiana benthamiana has identified specific P . syringae effectors that inactivate plant kinases 8 , 9 ., For example , the AvrPto effector binds membrane-associated kinases , including the PRRs FALGELLIN-SENSING2 ( FLS2 ) , EF-TU RECEPTOR ( EFR ) , and the BRI1-ASSOCIATED RECEPTOR KINASE1 ( BAK1 ) co-receptor to disrupt PTI and promote bacterial virulence 10 ., The AvrPto effector also induces host resistance in some tomato genotypes by interacting with the Pto kinase , which activates Pseudomonas resistance and fenthion sensitivity ( Prf ) resistance protein , resulting in ETI 11 ., Another effector called HopAI1 represses PTI through its interactions with the cytosolic MAPKs , MPK3 and MPK6 12 ., Recent work suggests that pathogen effectors may interact with not only a few targets but with multiple host targets , indicating that the breadth of effector–plant interactions are only beginning to be understood ., Proteome-scale interactomics 13 , 14 revealed an impressive number of putative effector-interacting proteins alongside fundamental properties of plant–pathogen interaction networks , such as effector convergence on network hubs ., Furthermore , using a different methodology ( i . e . , global transcriptional profiling of Arabidopsis defense-related mutants coupled with modeling ) 15 identified regulatory relationships between immune-related subnetworks and highly interconnected network components ., However , in these studies , the physical layout of the underlying plant cellular networks targeted by pathogens remained out of the reach of the analytic and experimental methodologies utilized ., To better understand how protein kinases contribute to basal immunity , we first sought kinase targets that interact with multiple effectors , a characteristic of defense-associated host proteins 13 , 14 ., Five P . syringae effectors ( AvrPto , HopA1 , HopAI1 , HopAF1 , and HopM1 ) were selected based on two main criteria: high prevalence among the P . syringae isolates 5 and a known ability or potential to suppress defense responses 10 , 16–19 ., HopA1 disrupts the formation of a protein complex involved in activating basal immunity and ETI 20 ., HopAF1 interacts with the methylthioadenosine nucleosidase proteins MTN1 and MTN2 to disrupt ethylene ( ET ) production 18 ., HopM1 binds HopM1 interactor 7 ( MIN7 ) , disrupting vesicle trafficking and reducing callose deposition 19 ., These five effectors suppress different parts of the cellular immune response in plants , suggesting that they may interact with distinct host proteins ., Here , we identify targets of bacterial effectors in plant cells , perform an in-depth functional analysis of a set of multi-effector–interacting kinases to superimpose effector-specific pathways over the plant–effector interaction space , and characterize the properties of the plant defense network ., We developed a multipronged approach consisting of identification of in vivo pairwise interactions between 279 tomato kinases and five effectors from the model tomato pathogen , Pseudomonas syringae pv ., tomato ( Pst ) ( HopA1 , HopAI1 , HopAF1 , AvrPto , and HopM1 ) ., Next , we characterized the role of 35 multi-effector–interacting kinases in PTI , ETS , ETI , and PCD ., We created new methodologies for data integration and generated signaling networks to facilitate visualization of the protein kinase networks involved in defense ( Fig 1 ) ., This network-centric approach allowed us to compare signaling networks associated with different levels of plant immunity and led to identification of novel defense-associated kinases ., The approach and the results obtained are described in the sections that follow and in the Supporting information ( S1 Materials and Methods ) ., To better understand how diverse effectors may be targeting plant protein kinases , we tested interactions between 279 tomato kinases 21 and five effectors ., A total of 1 , 170 pairwise kinase–effector ( K-E ) interactions were tested in tomato protoplasts using the split luciferase complementation assay ( SLCA ) , where luciferase activity indicates reconstitution of the N-terminal and C-terminal domains of the enzyme fused to interacting bait or prey proteins ., The split luciferase complementation ( SLC ) data analysis is described in the Supporting information ( S1 Materials and Methods ) ., Interactomics primary data are available at https://figshare . com/s/35c4aab65174c67a496e ) ; the MATLAB code for the SLC data analysis is provided in S3 Data ., Several controls were included in each SLC experiment to ensure reproducibility of the method across replicates , including a positive control ( protoplasts with full-length luciferase ) , a negative control ( untransformed protoplasts ) , and a reference interaction set between the AvrPto effector and the Pto kinase , which have been shown to interact in planta 22 ., In addition , an AvrPtoI96A ( AvrPto with an Ile to Ala mutation ) and Pto kinase pair were included as a control for interaction strength because the I96A mutation inhibits effector function and interaction with the Pto kinase 10 ., The SLCA screen was reproducible with low variability of luminescence signals across technical replicates ( S1A Fig ) ., High correlation was observed for the signals for full-length luciferase and controls ( reference set of positive and negative interactions ) among plates ( S1B Fig ) ; the K-E signals and the control sets did not show correlation , indicating a lack of a measuring bias in the protocol ( S1C Fig ) ., The signals from the AvrPto–Pto and AvrPtoI96A–Pto interactions were highly correlated with an average 3-fold reduction in signal for the AvrPtoI96A–Pto interaction ., A multiple regression model of Pto–AvrPtoI96A versus Pto–AvrPto and Luciferase signals has R2 = 0 . 887 ( adjusted R2 = 0 . 886 ) and regression coefficients of 7 . 17 × 10−3 ( Luciferase ) and 3 . 06 × 10−1 ( Pto–AvrPto ) ( S1D Fig ) ., Moreover , the luminescence signal of the K-E pairs showed a wide dynamic range , indicating that there are no physical limitations in measuring the luminescence produced in the SLCA ( S1E Fig ) ., Out of the 279 kinases tested for interactions with AvrPto , HopA1 , HopAI1 , or HopAF1 , 133 ( 48% ) interacted with at least one effector and were named Kinase Effector Interactors ( KEIs ) ., No significant interactions were identified for HopM1 out of the 30 tested kinases , suggesting that this endomembrane-specific effector 19 may not associate with kinases ., The K-E interaction network contains 321 significant interactions of 133 kinases with four effectors and includes previously confirmed K-E interactions ( Fig 2A; S1 Table ) ., Among the 133 kinases , approximately 70% are multi-effector interactors , out of which 38 interact with all four effectors and 24 are shared by HopA1 , HopAI1 , and HopAF1 ( Fig 2B ) ., To estimate the relative affinity of K-E interactions , a metric called the “interaction strength coefficient” ( normalized signal fold change ) was used to quantify the difference in reconstituted luciferase activity between each tested pair and the reference interactions ., On average , the interactions of KEIs with HopA1 or HopAI1 were twice as strong when compared with AvrPto , possibly due to the better reconstitution of the luciferase , higher affinity , or low dissociation of K-E complexes ( Fig 2C ) ., Notably , two known interaction pairs ( HopAI1–MPK6 and HopAI1–MPK4 ) were the strongest among all control interactions tested for these MAPKs ( inset of Fig 2A ) ., The distribution of the fold change interaction values off all K-E interactions tested is shown in S2A Fig . An analysis of the candidate KEIs along the spectrum of kinase classes 23 revealed that the effectors interacted mostly with leucine-rich repeats ( LRR ) -type RLKs and RLCKs from Class 1 ( 42% ) , kinases from Class 2/Raf-like ( 59% ) , and Class 4/MAPKs and calcium-responsive kinases ( 47% ) ( Fig 2D; S2B Fig ) ., A group of KEIs was selected for functional characterization based on their ability to putatively interact with multiple effectors ., The 35 focus KEIs ( S2 Table ) were silenced in N . benthamiana , a relative to tomato and host for Pseudomonas syringae DC3000 strains lacking the HopQ1-1 avirulence gene , due to its amenability for transformation and high efficiency of gene silencing 24 , 25 ., Virus-induced gene silencing ( VIGS ) constructs containing a fragment of an Escherichia coli gene ( EC1 ) served as a negative control ., After confirmation that KEI expression was silenced , the plants were inoculated with an effectorless Pst strain ( D29E ) 26 and four single-effector strains expressing AvrPto , HopA1 , HopAF1 , or HopAI1 in the D29E background ( S3 Table; S3A Fig; S1 Data ) ., In the EC1 control , the presence of some effectors ( AvrPto , HopAI1 , or HopAF1 , but not HopA1 ) in D29E led to a moderate but significant increase in Pst growth compared to D29E ( Fig 3A ) , indicating that these effectors can contribute to Pst virulence in isolation from the broader repertoire ., Among the 35 KEIs , seven KEIs influenced D29E growth compared with the EC1 control , indicating a role in basal immunity ( Fig 3B ) ., The majority of these KEIs—including RLKs ( KEI188/LYK4 , KEI72/SOBIR1 , KEI156 , and KEI161/RKL1 ) , RLCKs ( KEI149/PTI1-like ) , and the Ca2+-regulated KEI255/CIPK25—promoted bacterial growth when silenced , while silencing of one kinase ( KEI339 ) inhibited D29E growth ., In comparison , silencing of 17 KEIs caused a significant change in the growth of single-effector strains compared with the EC1 control ( Fig 3C–3F; S3B and S3C Fig ) ., Silencing of SOBIR1 , a key component of PTI 27 , 28 , affected the growth of D29E and the HopA1- and AvrPto-carrying strains ., Moreover , silencing of KEI342/SlBAK1 ( one of the tomato BAK1 homologs that may facilitate basal immunity 29 ) or of KEI327/SlMPK1 ( a kinase with high similarity to AtMPK6 and a possible role in PTI 30 ) interfered with the growth of single-effector strains exclusively ., Interestingly , the majority of KEIs required for D29E response were RLKs ., On the other hand , cytosolic kinases were preponderant in plant response to D29E + HopAI1 , + HopAF1 or + AvrPto , showing a 4- , 2 . 8- , and 2 . 3-fold increase , respectively , relative to the RLKs ( Fig 3G ) ., Among the KEIs with significant contributions to bacterial growth , 52% participated in plant response to D29E + HopA1 and + HopAF1 and 36% to D29E + AvrPto and HopAI1 , while only 12% were necessary for defense against D29E ( S3D Fig ) ., KEIs had significant positive or negative effects on the growth of Pst strains , indicating that KEIs promote either immunity or ETS , but not both ( S3E Fig ) ., Overall , when mapping the sign of variation ( Fig 3H ) , most KEIs classified as RLK/RLCKs promoted basal immunity , while KEIs promoting ETS mainly included kinases from the other cytosolic and MAPK-like ( 60% and 30% , respectively ) ., Some protein kinases have been shown mediate cellular response to multiple types of stresses 31 ., To determine if KEIs are similarly involved in multiple response pathways , we tested the focus KEIs in ETI and MAPK-mediated PCD responses ., The HopQ1-1 effector is an avirulence factor in N . benthamiana , in which it is recognized by an unknown R protein 32 ., To test the ETI in KEI-silenced plants , we quantified the size of the necrotic lesion 32 triggered by the inoculation with a D29E + HopQ1-1 strain as a proxy for quantification of PCD ( S4A Fig; S1 Data ) ., Eleven out of the 35 tomato KEIs tested were required for PCD , including RLKs ( KEI37/LYC10 , KEI161/RKL1 , and AtFLS2 ) , RLCKs ( KEI7/PBL8 ) , MAPKK kinases MAP3Ks ( KEI20/SlCTR1 ) , SnRKs ( KEI250/CIPK6 ) , GSK3/Shaggy-like ( KEI272/SK13 ) , and KEI339 ( Fig 4A; S4B Fig ) ., Silencing of KEI72/SOBIR1 , a known positive cell death modulator 33 , KEI7/PBL8 , and KEI20/SlCTR1 impaired PCD the most ., KEI160/NtIRK , previously associated with antiviral defense and regulation of the R-gene–mediated PCD in N . benthamiana 34 , facilitated PCD ., The results indicate that many KEIs mediate ETI-associated PCD by exerting exclusively positive regulatory roles ., To test the role of KEIs in the development of MAPK-dependent PCD , KEI-silenced plants were infiltrated with constitutively active MKK7 or MKK9 ., Both MAPK kinases ( MAP2Ks ) are known to participate in multiple immune-related processes 35 , 36 , and their prolonged expression induces activation of MPK3 and MPK6 and PCD 1 ., Lesion size was significantly altered in 23 of the tested KEI-silenced lines following MAP2K expression ., MKK7-triggered , MKK9-triggered PCD was modified in several lines , seven of which were required for both MKK7- and MKK9-mediated pathways ( Fig 4B and 4C ) ., The RLKs and RLCKs functioned as both positive and negative PCD regulators , compared with other classes ( Fig 4D ) ., Notably , silencing of the PCD negative regulator KEI342/BAK1 37 inhibited both MKK7- and MKK9-PCD ., To determine how these phenotypes may be related , we performed correlation analyses between bacterial growth and lesion size measurements across the KEI lines ( Fig 4E ) , as described in S1 Materials and Methods ., A significant correlation ( R > 0 . 6 ) was observed across bacterial growth assays , but no correlation was found between bacterial growth and PCD treatments , suggesting these responses utilize distinct signaling pathways ., To obtain a global view of the link between the KEIs’ structural class and their contribution to immune phenotypes , we plotted a phylogenetic tree and visualized significant contributions to immunity based on our assays ( Fig 4F and S2 Data ) ., Interestingly , the phylogenetic tree highlighted the clear difference in structural class between the PTI- versus ETS-promoting kinases ( RLKs/RLCKs versus MAPKs , CIPK/SnRKs , ribosomal protein S6 kinase ( S6K ) , AGC kinases , and glycogen synthase kinase3 GSK3-like , respectively ) ., The involvement of KEIs in multiple stress responses prompted the development of functional signaling networks to understand how defense networks are modified during different immune responses ., To construct the networks , we calculated the co-occurrence frequency of the focus KEIs in various functional assays to evaluate the degree of KEIs phenotype overlap , indicative of functional association among KEIs ., Using a set of logical rules and prior information ( Fig 5A ) , a KEI signaling network was generated with the nodes ( KEIs ) ordered hierarchically within the canonical structure of a signaling pathway: RLK → RLCK → RAFs/MAPKs → cytosolic kinases ( Materials and methods; S1 Materials and Methods ) ., Indirect evidence positioned RAFs upstream of MAPK cascades and at a similar hierarchical level with MAP2Ks 2 , 31 , 38 , 39 ., Directed edges weighted by co-occurrence values link the nodes ., To generate the signaling network , KEIs with similar phenotypes were grouped in modules , in which the position of nodes within the same hierarchical level or kinase structural class was based on their regulatory strength rank and sign of regulation , while the redundant edges between successive network levels were removed ., Using these criteria , we collapsed the weighted composite graphs into a minimal network providing an overview of the KEI pathways critical to immune-related plant phenotypes ( Fig 5B ) ., Next , we generated networks representing the response of the minimal network under our eight experimental conditions , called stimuli-specific networks ( SSNs ) , to reveal how the signaling network responded ( Fig 5C ) ., A comparison of the infection-response networks demonstrated that most of the D29E network is maintained across single-effector networks , with the exception of avirulence-inducing HopQ1-1 ., Addition of these single effectors affected the network topology , such that a larger number of RLK and RLCKs played a role ., The networks doubled in diameter ( the average length of shortest paths between all pairs of nodes ) and had longer distance ( shortest path index ) between any two nodes , relative to the D29E network , indicative of activation of a more diverse and complex defense network ( Fig 6A , S5A Fig ) ., For example , KEI104-BSK7 promoted immunity against strains containing HopA1 , HopAF1 , and AvrPto , but not the D29E strain ., Interestingly , addition of virulence-promoting effectors ( HopAI1 , HopAF1 , and AvrPto ) activated ETS pathways , including cytosolic kinases KEI339 , KEI323/S6K2 , and KEI318/SRK2C ., In the PCD networks , ( HopQ1-1 , MKK7 , and MKK9 ) , the signaling pathways were markedly distinct ., While the MKK7 network is primarily comprised of KEIs that repress cell death , the MKK9 and HopQ1-1 networks were comprised of KEIs that promote cell death , densely populated with several RLK modules feeding into many cytosolic KEIs , and with most of the components functioning as negative regulators of the PCD; few nodes were shared ., In the MKK7 and MKK9 networks , the diameter and path length indices were similar to HopQ1-1 ( Fig 6A; S5A Fig ) ., Signaling flow through SSNs converged onto a set of KEIs associated with the global control of transcription and translation , ion and nutrient homeostasis , and extracellular acidification ., Examples include KEI250/CIPK6 40–42 , KEI33/CIPK11 43 , KEI255/CIPK25 44 , KEI323/S6K2 45 , and KEI311/KIN10 46 ., To measure relative importance of the nodes in the SSNs we used maximum clique size algorithm ( MCC ) 47 , which finds clusters of the largest size in a given network; sub-graphs of essential nodes were derived based on their MCC rank for the PTI , ETS ( Fig 6B ) and ETI , PCD ( Fig 6C ) ., Cytosolic kinases from the RLCK and MAPK-like families were preponderant essential nodes in both MCC-ranked graphs ., Another parameter measuring centrality in networks is the betweenness centrality ( BC ) index , also regarded as a measure of the control potential of a node within a network 48 ., Among all SSNs , the average BC indices were higher for MKKs and HopQ1-1 networks , indicating the importance of individual nodes on signaling outcome ( S5B Fig ) ., In contrast , the signaling networks associated with PTI and ETS ( D29E , HopA1 , HopAI1 , HopAF1 , and AvrPto ) were smaller and had fewer high-control nodes ( low-centrality nodes ) , implying decreased efficiency in signal transmission ., To determine if the SSN networks could be used to predict the performance of genes in the defense response , we tested 18 KEIs for their role in basal immunity in N . benthamiana ., In this assay , immunity is first induced by inoculation with a non-pathogen ( P . fluorescens ) , followed by inoculation with the ETI-inducing P . syringae 49 ., In the region where the inoculation areas overlap , little visible cell death develops , likely because of induced defense responses , limiting bacterial proliferation and secretion of the HopQ1-1 avirulence protein 50 ., Most KEI-silenced lines had significantly increased cell death in the area infiltrated with both strains , indicating an impaired immune response as compared with the EC1 control and known immunity-promoting kinases ( BAK1 and FLS2 ) ( Fig 6D; S1 Data ) ., Six of the eight highly MCC-ranked KEIs , including KEI149/PTI1-like , KEI104/BSK7 , KEI86/PBL5 , and KEI7/PBL8 , had statistically significant phenotypes; others , including KEI156 , 151/BIR2 , 160/IRK , 323/S6K2 , 304/LeMKK3 , and the PTI-promoting 72/SOBIR1 , also exhibited significant differences in the cell death intensity compared with controls ( Fig 6E ) ., KEI91 LRR RLK , which had no significant phenotypes in the ETS , ETI , or PCD phenotyping , showed control-level cell death ., The known and curated protein–protein interactions ( PPIs ) for the Arabidopsis homologs of these KEIs were extracted from public databases and used to generate a network ( Fig 6F ) , as described in the Supporting information ( S1 Materials and Methods—KEI signaling network analysis ) ., The network has a PPI enrichment p-value of 3 . 1 × 10−8 , indicating that most of these kinases are biologically connected among themselves ., To further test our SSNs predictions , we overlapped the information from our eight orthogonal phenotyping assays over the PPI network ., The nodes connected by 70% of the edges ( 17 out of 25 ) co-occurred in various SSNs , indicating they may also be functional partners ., Notably , 65% of edges connecting the interacting and functionally related KEIs co-occurred in more than two SSNs ., For example , the interacting pairs KEI327/MPK6 and KEI323/S6K2 were part of the HopA1 , HopAI1 , and AvrPto SSNs ., Overall , these results indicate the predictive potential of the SSNs for mapping plant defense networks and their response to perturbations ., Plant immunity is generated as a result of numerous coordinated cellular processes ., The study of inducible plant immunity requires approaches that reveal the organization and dynamics of the overall system and generate predictions on how molecular-level interventions can modify plant phenotypes ., Building and characterizing biological networks , as a system-level approach to study plants , is starting to prove its effectiveness in predicting the function of cellular components and identifying biochemical and functional relationships among them 15 , 51–54 ., Here , we describe a network-driven integrative analysis of the plant immune system , which includes in vivo plant–pathogen interactomics and a comprehensive study of kinase targets and identification of signal-specific networks ., Some of the findings revealed by our approach included ( 1 ) that some effectors may bind several tomato kinases and that a proportion of kinases can interact with multiple effectors , ( 2 ) defense-associated kinase networks contain both shared and specific nodes involved in basal immunity , ETS , ETI , and PCD , ( 3 ) effector-triggered kinase networks are larger and more complex compared with a basal-defense network; however , they have fewer nodes with high centrality than unperturbed networks , and ( 4 ) previously uncharacterized kinases are essential for promoting bacterial resistance in N . benthamiana ., A comprehensive characterization of the kinases identified in this study can provide insights into the underlying molecular mechanisms of defense and on the sensitivity and response to perturbations of plant defense networks , and will help identify targets for genome editing in crops ., Our K-E screen predicts that interactions between plant proteins and pathogen effectors occur with a relatively low specificity when compared , for example , with receptor–ligand interactions ., These observations confirm previous assumptions regarding effector promiscuity in target selection 13 , 55 , 56 and are supported by work demonstrating the functional interchangeability of P . syringae effectors 4 ., By associating with multiple elements of a pathway , an effector may increase its chances to interfere successfully with the plant immune response ., Furthermore , it may be evolutionarily beneficial for effectors to maintain the ability to interact with diverse partners to ensure functionality in new plant hosts with divergent immune signaling pathways 57–61 ., On the other hand , HopM1 did not interact with any of the tested tomato kinases , suggesting a degree of target selectivity for some effectors ., Indeed , target selectivity is further indicated by the fact that not all members of a kinase family interacted with the same effector ., While this may be due to our experimental system , because effectors are rarely present individually and high expression of both kinase and effectors likely increased the chances for false positives , the effectors interacted with a mostly shared set of defense-associated kinases , suggesting the functional relevance of these interactions ., Thus , while these interactions will have to be confirmed by additional methods , our findings indicate the effectiveness of using effector interactions as a starting point for genetic characterization ., Interestingly , several effectors interacted with both positive and negative modulators of immunity , demonstrating that interaction alone is not sufficient to predict the role of a host target in defense ( HopA1 and PK3 or BAK1 ) ., During pathogenesis , effectors have additive or synergistic effects on promoting virulence in plants , and the impact of individual effectors on immunity is typically minor or nonsignificant 62 ., The overall contribution of individual effectors is likely dependent on both the relative importance of individual host targets within the defense network and on the status of the network itself , as different sectors are inactivated by other effectors ., In addition , essential kinases such as BAK1 often play dual roles , depending on the status of other regulatory kinases in the cell ., For example , AtBAK1 is essential for activation of PTI , but overaccumulation of AtBAK1 or loss of its negative regulator AtBIR1 can also activate immunity 63 ., In biological networks , elimination of highly connected nodes ( hubs ) increases the diameter of the network 64 and has a deleterious effect on the characteristic path length and network integrity 65 compared with the removal of low-connectivity nodes ., In this study , effectors appeared to neutralize hubs and nodes with high control potential in the network , thus having a detrimental effect on the structural integrity of the plant immune network ., In addition , networks expanded in the presence of effectors , which may indicate plant deployment of new signaling sectors during purturbation ., Interestingly , several susceptibility-linked KEIs were identified across defense networks ., These KEIs may act as negative immune regulators or could be recruited to subvert plant pathways for the benefit of the pathogen 66 , 67 ., Our results postulate that the composition and topology of plant signaling networks are determined by the plant’s ability to identify damage from effectors and activate compensatory pathways ., Conversely , effector strategies to increase pathogen virulence consist in blocking/inactivating the sensor layers ( RLK/RLCK modules ) and recruiting kinases in the lower layers of the network for increasing pathogen fitness ., Comparison of the MKK7 and MKK9 networks suggests an antagonistic relationship between the pathways activated by these MAP2Ks , whereby activation of one may cause inhibition of the other ., MKK7 is a positive regulatory component of the immune response and systemic acquired resistance , operating via salicylic acid ( SA ) synthesis 68 , while MKK9 positively regulates ET signaling through increasing ETHYLENE-INSENSITIVE3 ( EIN3 ) receptor stability 69 ., The complex functional relationship between SA and ET , comprising both synergistic 70 , 71 and antagonistic 72 interactions , provides additional strength to this model ., Together , our network-centered approach has revealed the effect of individual effectors on signaling network topology and has facilitated the identification of novel immune kinases ., However , several questions remain , including how effectors work together to modify the host immune network and if this information can be used to accurately predict the outcome of plant–pathogen interactions ., A combination of systems biology approaches and genome editing has the potential to help address these questions and further the development of resistant plants for agricultural production ., The coding region of the effector genes without the stop codon was cloned into the pENTR/SD/D-TOPO ., The sequence for HopAI1 was amplified from P . syringae pv ., tomato T1 using primers 5′-caccatgctcagtttaaagctgaacacccag and 5′-gcgagtccagggcggtggcatcag ., All other effectors were obtained from P . syringae pv ., tomato DC3000 ., Hrp promoter-driven effectors fused at the C terminus with the HA tag were generated in the destination vector pCPP5372 73 using Gateway cloning ., pCPP5372 carrying different effectors was mobilized into DC3000D29E , a derivative of DC3000D28E lacking HopAD1 , by triparental mating using the helper plasmid pRK2013; Trans-conjugants were selected on KB medium with appropriate antibiotics ., DC3000D28E::ShcM HopM1 has been described previously 74 ., Bacteria were maintained on King’s B medium at 37 °C ., Cloning of the tomato KEIs and the SLC method were described previously 21 ., To create clones for VIGS of orthologous kinases in N . benthamiana , tomato gene sequences were analyzed using bioinformatics tools available at solgenomics . net; the VIGS tool and the optimal gene fragment with the fewest off-targets were used to design primers ., Gene fragments were amplified from N . benthamiana cDNA , cloned into the TOPO pER8 Donor vector using the manufacturer’s protocol , subcloned into the TRV2 expression vector , and transformed into Agrobacterium GV2260 for expression in planta 75 ., Each interaction was tested in 4 to 16 independent assays , and the reconstituted luminescence was recorded at six time points ., The decision to test over the minimum of four times was taken for the pairs showing significant levels of interaction when compared with the reference sets , while up to 16 assays ( four biological replicates ) were used for K-E pairs showing variability or low interaction levels ., The interactions were corrected for multiple testing with a false discovery rate ( FDR ) of of 0 . 05 ., The analysis of SLCAs is described in S1 Materials and Methods ., The KEI-silenced lines were produced by syringe-infiltrating leaves of 2-week-old N . benthamiana plants with the TRV2-KEI Agrobacterium clones along with TRV1-containing Agrobacterium at a 1:1 ratio as described 75 ., The EC1 and FLS2 constructs 49 served as controls and were included in each round of KEI line t | Introduction, Results, Discussion, Materials and methods | Plant protein kinases form redundant signaling pathways to perceive microbial pathogens and activate immunity ., Bacterial pathogens repress cellular immune responses by secreting effectors , some of which bind and inhibit multiple host kinases ., To understand how broadly bacterial effectors may bind protein kinases and the function of these kinase interactors , we first tested kinase–effector ( K-E ) interactions using the Pseudomonas syringae pv ., tomato–tomato pathosystem ., We tested interactions between five individual effectors ( HopAI1 , AvrPto , HopA1 , HopM1 , and HopAF1 ) and 279 tomato kinases in tomato cells ., Over half of the tested kinases interacted with at least one effector , and 48% of these kinases interacted with more than three effectors , suggesting a role in the defense ., Next , we characterized the role of select multi-effector–interacting kinases and revealed their roles in basal resistance , effector-triggered immunity ( ETI ) , or programmed cell death ( PCD ) ., The immune function of several of these kinases was only detectable in the presence of effectors , suggesting that these kinases are critical when particular cell functions are perturbed or that their role is typically masked ., To visualize the kinase networks underlying the cellular responses , we derived signal-specific networks ., A comparison of the networks revealed a limited overlap between ETI and basal immunity networks ., In addition , the basal immune network complexity increased when exposed to some of the effectors ., The networks were used to successfully predict the role of a new set of kinases in basal immunity ., Our work indicates the complexity of the larger kinase-based defense network and demonstrates how virulence- and avirulence-associated bacterial effectors alter sectors of the defense network . | Some bacterial pathogens secrete virulence factors called effectors , which influence host tissues during infection ., The impact of such bacterial effectors on the transmission of immune signals in plants remains poorly understood ., In this study , we developed an integrative network approach to discover interactions between bacterial effectors and a class of host signal-mediating enzymes called protein kinases ., We also characterized the functions of the targets of these kinases in order to understand how bacterial effectors might disrupt the flow of information in signaling pathways within plant cells ., We show that plants activate larger signaling networks when inoculated with pathogens that produce effectors ., We also find that plant signaling networks are specific to individual effectors and that the networks include kinases with both positive and negative effects on plant resistance to pathogens ., We propose that the topology of immune signaling networks is determined by the plant’s ability to activate compensatory pathways in response to the effectors’ network-disruptive actions ., Conversely , pathogens may increase their virulence both by disrupting host signaling at the membrane-located end of the signaling network and by recruiting cytosolic kinases ., This work provides a framework for the study of plant–pathogen communication and could be used to prioritize targets for improving resistance in crops . | cell death, medicine and health sciences, pathology and laboratory medicine, pathogens, cell processes, immunology, signaling networks, microbiology, cell-mediated immunity, network analysis, plants, flowering plants, bacteria, bacterial pathogens, protein kinase signaling cascade, pseudomonas, computer and information sciences, pseudomonas syringae, nicotiana, medical microbiology, microbial pathogens, tomatoes, immune response, fruits, signal transduction, eukaryota, cell biology, immunity, biology and life sciences, cell signaling, organisms, signaling cascades | null |
journal.pcbi.1002968 | 2,013 | RFECS: A Random-Forest Based Algorithm for Enhancer Identification from Chromatin State | Enhancers are distal regulatory elements with key roles in the regulation of gene expression ., In higher eukaryotes , a diverse repertoire of transcription factors bind to enhancers to orchestrate critical cellular events including differentiation 1 , 2 , maintenance of cell-identity 3 , 4 and response to stimuli 5–7 ., While enhancers have long been recognized for their regulatory importance , the fact that they lack common sequence features and often reside far away from their target genes has made them difficult to identify ., Computational techniques relying on transcription factor motif clustering or comparative analyses have had some success in identifying enhancers , but these predictions are neither comprehensive nor tissue-specific 8–13 ., Recently , several high-throughput experimental approaches have been developed to identify enhancers in an unbiased , genome-wide manner ., The first is mapping the binding sites of specific transcription factors by ChIP-seq 14 ., Because this approach requires the knowledge of a subset of transcription factors ( TFs ) that are not only expressed but also occupy all active enhancer regions in the cell-type of interest , identification of all enhancers using this approach is not a trivial task ., The second approach involves mapping the binding sites of transcriptional co-activators such as p300 and CBP 4 , 5 , 15 , which are recruited by sequence-specific transcription factors to a large number of enhancers 6 , 16 , 17 ., Since not all enhancers are marked by a given set of co-activators 18 , 19 , and ChIP-grade antibodies against these proteins may not always be available , systematic identification of enhancers by mapping the locations of co-activators is not generally feasible ., A third approach relies on identifying open chromatin with techniques such as DNase I hypersensitivity mapping 20 ., However , since open chromatin regions can correspond to not only enhancers , but also silencers/repressors , insulators , promoters 21 , 22 or other functionally unknown sequences occupied by nuclear proteins , this approach lacks specificity in enhancer identification ., Finally , a fourth approach interrogates covalent modifications of histones 5 , 23–26 as it was observed that certain histone modifications form a consistent signature of enhancers ., It is on this approach that the present work is focused ., Previously , we and others observed that distinct chromatin modification patterns were associated with transcriptional enhancers 5 , 22 , 27 ., Specifically , active promoters are marked by trimethylation of Lys4 of histone H3 ( H3K4me3 ) , whereas enhancers are marked by monomethylation , but not trimethylation , of H3K4 ( H3K4me1 ) ., This chromatin signature has been used to develop a profile-based method for enhancer discovery 5 ., Both unsupervised 25 , 28 and supervised learning approaches have also been employed to exploit chromatin modification-based differences to identify enhancers ., The supervised machine learning techniques include HMM 8 , 23 , neural networks 24 and genetic algorithm-optimized SVM 26 based approaches , and have proved to be improvements over the profile-based method ., While these methods have led to identification of a great number of enhancers in the human and mouse genomes 3 , 25 , 29 , the current computational techniques have thus far been limited by the small number of the training set samples and limited number of chromatin modifications examined ., Thus , it is possible that these approaches may not fully capture the entire range of chromatin modification patterns at enhancer elements ., With the discovery of ever more histone modifications , it is likely that additional chromatin modifications may distinguish enhancers from other functional elements in the genome ., This additional data should in principle allow us to answer the key question: what is the optimal set of modifications required for enhancer prediction ?, Some researchers have tried to tackle this issue by using algorithms such as simulated annealing 23 or genetic-algorithm optimization 26 ., We sought to develop a method in which the selection of the optimal set is automatically built into the training-process and is easily adapted to a large number of features ., As part of the NIH Epigenome Roadmap project , we have generated genome-wide profiles for 24 chromatin modifications and DNase-I hypersensitivity sites in 2 distinct cell types- human embryonic stem cell ( H1 ) and a primary lung fibroblast cell line ( IMR90 ) 30 ., Additionally , we have experimentally determined a large number of promoter-distal p300 binding sites in each cell type , providing a rich training set for development of accurate and robust enhancer prediction algorithms ., We now describe a random-forest 31 based method for integrative analysis of diverse histone modifications to predict enhancers ., We show that this new algorithm outperforms the existing methods and leads to the automatic discovery of an optimal set of chromatin modifications for enhancer predictions ., Random forests have recently become a popular machine learning technique in biology 32 due to their ability to run efficiently on large datasets without over-fitting , and their inherently non-parametric structure ., Since random forests use a single variable at a time , they can give an automatic measure of feature importance 33 ., Hence , we developed an algorithm based on this random forest technique for the purpose of enhancer prediction ., Conventional random forests utilize a single scalar value associated with each feature at each node of the tree ., In order to train a random-forest for enhancer prediction we wanted to use histone modification profiles at p300 binding sites ., Because the spatial organization of histone modifications along a linear chromosome can be as informative as their actual levels , they are better represented as vectors of binned reads ., Inspired by recent modifications to the random-forest approach such as discriminant random forests 34 or oblique random forests 35 that utilize a linear classifier at each node , we developed a new vector-based random forest algorithm RFECS or Random Forest for Enhancer Identification using Chromatin States ( see Methods ) ., Genome-wide distal p300 binding sites were found using ChIP-seq in H1 and IMR90 cell-lines ., We selected p300 binding sites overlapping DNase-I hypersensitive sites and distal to annotated TSS as active p300 binding sites representative of enhancers ., We found 5899 such p300 binding sites in H1 and 25109 such sites in IMR90 ( Table S1 , S2 ) , and observed several distinct and diverse chromatin states using an unsupervised clustering technique , ChromaSig ( fig . 1A , B ) ., All clusters showed enrichment of H3K4me1 and depletion of H3K4me3 as previously observed 5 ., However , different clusters were characterized by varying levels of histone acetylation , H3K4me2 or H3K27me3 ., Clusters with presence or absence of H3K36me3 may represent genic and intergenic enhancers respectively ., In order to ensure we represented all these different chromatin states at active p300 binding sites , we selected a relatively large number of these sites ( >5000 ) for training as compared to previous methods ., To train the forest , active and distal p300-binding sites ( BS ) were selected as representative of the enhancer class ., As non-enhancer classes , we considered annotated transcription start sites ( TSS ) that overlap DNase-I , and random 100 bp bins that are distal to known p300 or TSS ( see Methods ) ., The confidence of each enhancer prediction is given by the percentage of trees that predict this site to be an enhancer ., In general , a genomic region is predicted as an enhancer if it has a background cutoff greater than 0 . 5 ( >50% trees vote in its favor ) ., At higher cutoffs , confidence of prediction is higher , but fewer enhancers are predicted ., We used Receiver Operating Characteristic ( ROC ) curves to determine optimal parameters for our classification algorithm 36 ., In the case of enhancer predictions , we can only obtain an approximate measure of specificity since we can never be certain that the randomly selected elements of the non-p300 class are all true negatives ., Hence , in addition to the ROC curves generated using 5-fold cross-validation , we also verified parameter selection by comparing the percentage of predicted enhancers at each cutoff that overlap markers of active enhancers ( validation rate ) or TSS ( misclassification rate ) ., The markers of active enhancers include distal DNase-I hypersensitivity sites ( HS ) , p300 binding sites ( excluding those used in training ) , occupancy by CBP or sequence-specific transcription factors known to act at embryonic stem cell enhancers such as NANOG , OCT4 and SOX2 ., In the case of Random forests , the main parameter to be determined is the number of trees ., Since the non-enhancer class is assumed to be several times enriched compared to the enhancer class in the genome , we select a greater number of non-p300 training sites as compared to p300 sites and this proportion is also adjusted using the above-described methods ., Previous algorithms 23 as well as empirical observations showed a width of −1 kb to +1 kb around the p300 binding site as optimal but we further verified this selection by cross-validation in the H1 cell-type ( fig . S1A ) ., The difference in cross-validation curves using a width of 0 . 5 kb or 1 kb is not obvious on the cross-validation curve while a width of 1 . 5 kb clearly shows a sharp drop in the area under the ROC curve ( fig . S1A ) ., When we further made enhancer predictions using all three widths ( fig . S1B , C ) , it can be seen that a width of 1 kb on either side shows best validation and misclassification rates as compared to 0 . 5 or 1 . 5 kb widths ., To determine the optimal number of trees for the random-forest , we examined the area under the ROC curve in H1 and IMR90 and found both to be stable beyond 45 trees ( fig . 2A , B ) ., In order to verify this further , we made enhancer predictions using various number of trees such as 45 , 65 and 85 and compared the validation and misclassification rates ( fig . S2A–D ) ., While H1 appeared to show no change at all ( fig . S2A , , C ) IMR90 showed a slight improvement from 45 to 65 trees ( fig . S2B , D ) ., In the end , we selected 65 trees for training the random forest as it appeared to be optimal for both cases ., The training-set ratio of p300 to non-p300 was set at 1∶7 since the ROC curve did not appear to change much beyond this ratio ., ( fig . S2E , F ), In order to estimate the accuracy of the enhancer prediction by RFECS , we applied this algorithm to chromatin profiles of 24 marks obtained in H1 and IMR90 ., We then calculated the validation rate as the percentage of predicted enhancers overlapping with DNase-I hypersensitivity sites and binding sites of p300 and a few sequence specific transcription factors known to function in each cell type ( true positive markers ) ., We also computed the misclassification rate as the percentage of predicted enhancers overlapping with known promoters ., These overlaps were computed using a window of −2 . 5 to +2 . 5 kb ., Incase , both a true positive marker as well as promoter lay within this window , the criteria used to decide if the enhancer was “validated” or “misclassified” is discussed in detail in the Methods section ., In H1 cells , we obtained a total of 55382 predicted enhancers at the lowest voting cutoff of 0 . 5 ., Over 80% of these predicted enhancers overlap with distal DNase-I hypersensitive sites and the binding sites of p300 , NANOG , OCT4 and SOX2 ., Upon randomly generating enhancer predictions in the H1 genome 100 times , we found the average validation rate to be 18 . 43% and the actual validation rate of 80% to be highly significant with a one-sided t-test p-value of 10∧-256 ., Additionally , we found that 5% of them overlap with UCSC TSS , indicating a low misclassification rate of 5% ( fig . 2C , E , in red ) ., A similar high level of validation rate and low misclassification rate were observed when RFECS was applied to IMR90 cells , where 83581 enhancers were predicted with a validation rate of 85% ( average random validation rate\u200a=\u200a16 . 13% , pvalue\u200a=\u200a2×10∧-279 ) , and misclassification rate of 4% ( fig . 2D , F ) ., Thus , RFECS appears to accurately predict putative enhancer sequences based on chromatin modification state of the genome ., We next tried to assess the linear resolution of RFECS predictions ., We calculated the distance between the predicted enhancers and locations of enhancer markers such as DNase-I hypersensitive sites , or p300 binding sites in each cell type , and found that the majority of predicted enhancers are within 200 bp of these sites ( fig . S3A , B ) ., In H1 , nearly 62% of enhancers lie within 200 bp of an enhancer marker site ( fig . S3A ) , while in IMR90 this value is around 70% ( fig . S3B ) ., Thus , the majority of enhancer predictions also show a high distance resolution in terms of proximity to the validation marker ., We also confirmed that our enhancer predictions showed an activation of gene expression in the proximal TSS ., In order to do this , we compared RNA-seq datasets ( Wei Xie et al . , manuscript under revision ) in H1 and IMR90 using edgeR 37 to identify H1-specific and IMR90-specific TSS ., Then we identified enhancer predictions specific to either H1 or IMR90 using a filter distance of 2 . 5 kb ., When we look at the average distribution of H1-specific enhancers they are clearly enriched in the vicinity of H1-specific TSS as compared to either non-specific TSS or IMR90-specific TSS ( fig . S3C ) and this enrichment is found to significant at distances up to at least 500 kb using a Wilcoxon test ( p-value<10∧-6 ) ., Similarly , in the case of IMR90-specific enhancers , we observe them to be more enriched in the proximity of IMR90-specific TSS as compared to H1-specific TSS ( fig . S3D , p-value<10∧-23 ) ., As further evidence that RFECS accurately predicts enhancers , chromatin modifications at the predicted enhancers showed presence of all chromatin states observed in the training sets comprised of a subset of distal p300 binding sites ( fig . 1 ) ., In H1 , clusters 1 , 2 and 8 of enhancer predictions ( fig . S4 ) are similar to clusters 1–3 of the p300 binding sites ( fig . 1A ) , clusters 3–4 appear to correspond to cluster 5 of p300 BS , while clusters 5–6 look like cluster 4 of p300 BS ., In IMR90 , similar trends could be observed when comparing chromatin states at enhancer predictions ( fig . S5 ) to those of p300 binding sites ( fig . 1B ) ., Further , it can be observed that clusters 3–6 of the enhancer predictions in H1 ( fig . S4 ) that have weaker acetylation and/or enrichment of H3K27me3 also tend to have lower voting percentage of trees ., In summary , we showed that RFECS accurately predicted enhancers in the two cell lines H1 and IMR90 using a set of 24 chromatin modifications ., These enhancers showed high validation rates , low misclassification rates and sharp linear resolution ., To make enhancer predictions , our approach requires a construction of a random forest trained on promoter-distal p300 binding sites ., It is time-consuming and expensive to create a new training set for enhancer prediction in each new cell type , so it is desirable to use a random forest developed in one cell type to predict enhancers in another ., To evaluate the feasibility of such approach , we first trained a random-forest using chromatin modification profiles obtained in H1 , and then applied it to the IMR90 cells ., Compared to RFECS predictions using IMR90 chromatin profiles as training set , RFECS predictions using H1 training dataset reduces the validation rate by ∼5–8% and increases the misclassification rate by ∼2% ( fig . 2C , E black vs red ) ., Similarly , we also developed a random forest using the IMR90 data as the training set and then applied it to H1 ., This led to an average reduction of 2–3% in validation rate ( fig . 2D , black vs red ) ., Therefore , RFECS trained using one cell type may be applied to a different cell type , albeit with slightly lower accuracy ., We sought to examine if this moderate decrease in performance was largely due to cell-type specific differences or was within the limits of technical or biological variability between replicates ., To this end , we trained a random forest on one replicate of a cell-type , and made predictions on the other replicate of the same cell type ., RFECS trained on IMR90 and then applied to the replicate 1 of the H1 profiles ( blue dot vs asterisk ) actually showed a higher validation rate and lower misclassification rate than RFECS trained using replicate 2 of H1 ( fig . 2C , E ) , while similar performance was observed with enhancer predictions on replicate 2 of H1 independent of whether the random-forest was trained on H1 replicate 1 or IMR90 ( green dot vs asterisk ) ., Similar trends were observed when comparing predictions made on individual replicates of IMR90 using either H1-training or training on the other replicate ( fig . 2D , F ) ., In conclusion , predicting enhancers using the random forest built from a different cell type exhibits a modest decrease in performance compared to a same-cell training set ., However , this decrease in performance is comparable to the decrease that can arise due to variability between two replicates of the same cell-type ., With the increasing number of histone modifications being discovered and mapped , determination of the relative importance of each mark in defining genomic elements is important ., An out-of-bag measure of variable importance is a natural by-product of random forest classification scheme 33 wherein the relative importance of each feature is assessed as the increase in classification error upon permutation of feature values across classes ., In both H1 and IMR90 , the variable importance was assessed for random forests trained on 5 cross-sections of data for each of the 2 sets of replicates individually as well as the set of averaged replicates ., Upon ranking histone modifications by variable importance , it is apparent that H3K4me1 and H3K4me3 are the top 2 most robust modifications across replicates and cross-sectional samples in both cell types , followed by H3K4me2 ( fig . 3A , B ) ., This indicates that these 3 modifications maybe the most informative in the prediction of enhancers in any unknown cell type as well ., Beyond the top 3 modifications , there is variability among the cell types ., In IMR90 , the other modifications appear to contribute almost equally , while in H1 there is a much clearer difference in variable importance ., These differences are supported by correlation analyses in H1 and IMR90 ( fig . 3C , D ) ., In H1 , several modifications are highly correlated , which could explain the larger differences in variable importance , as only a few variables maybe needed to form a non-redundant set ., In IMR90 , the correlations are lower and hence each of the modifications may contribute non-redundant information and thus contribute equally to the variable importance ., Modifications that cluster together in both H1 and IMR90 ( shown in the same non-black colors , fig . 3C , D ) suggest cell-type independent redundancy ., Having established the relative importance of each histone modification in predicting enhancers , we next examined the accuracy of predictions using different sets of modifications ., Validation rates obtained by using the minimal set of H3K4me1-3 is within 2% of that for all 24 modifications in H1 ( fig . 4A ) ., Furthermore , this minimal set performs considerably better than the more conventionally selected set of H3K4me1 and H3K4me3 3 , 5 and at times , H3K27ac 38 , 39 ( fig . 4A , B , in black and blue ) ., The set of H3K4me1-2-3 is more comparable to H3K4me1-H3K4me3-H3K27ac in IMR90 but does have a slightly lower misclassification rate ( fig . 4D ) ., In both cases the use of the minimal set of 3 modifications shows a much closer resemblance in performance to all 24 modifications than to the set of 2 marks H3K4me1 and H3K4me3 ( fig . 4A–D ) ., It can also be observed that in conjunction with H3K4me1 and H3K4me3 , using H3K4me2 picks up a larger proportion of enhancers with weaker acetylation enrichment as compared to H3K27ac ( fig . S4 , S5 ) , supporting our prediction of the minimal set ., We also made enhancer predictions using all possible combinations of 3 modifications in chromosome 1 for replicate 1 and replicate 2 of H1 ., The average validation rate for a fixed range of enhancers was compared across replicates and it can be seen the set corresponding to H3K4me1 , H3K4me2 and H3K4me3 ( marked in * ) , is the highest performing combination common to both replicates ( fig . 4E ) ., We also found the performance of the combination of H3K27ac with H3K4me1 and H3K4me3 appears to be comparable in this case ( 3 , fig . 4E ) , validating the use of H3K27ac as a feature for enhancer prediction when H3K4me2 is not available ., Some of the worst performing combinations include H3K9me3 and H4K20me1 ( 4 and 5 , fig . 4E ) , which also show up as variables with least importance in fig . 3A ., In many currently existing datasets , H3K27ac is the more commonly sequenced histone modification as compared to H3K4me2 due to its perception as a marker of active enhancers ., While using H3K4me2 may improve enhancer prediction in some cell-types , use of H3K27ac in addition to H3K4me1 and H3K4me3 marks does show considerable improvement over using just the top 2 marks H3K4me1 and H3K4me3 ( fig . 4A–D ) ., Hence , for many of the currently existing datasets , we could use H3K4me1 , H3K4me3 and H3K27ac as the features in our random-forest with satisfactory performance ., Overall , these comparisons indicate the suitability of selecting H3K4me1 , H3K4me2 and H3K4me3 as three minimal chromatin marks for purposes of enhancer prediction ., Additional chromatin modifications required for improving upon enhancer predictions may depend on cell-type specific characteristics , as indicated by the differences in variable importance between H1 and IMR90 ( fig . 3A , B ) ., We next asked if our enhancer prediction algorithm performed better than several other current techniques for enhancer prediction – CSIANN , ChromaGenSVM and Chromia 23 , 24 , 26 , 39 ., In previous studies , CSIANN and ChromaGenSVM were applied on the histone modification dataset in CD4 T-cells 24 , 26 , 39 ., In order to make a comparison of performance of our method with previous approaches , we applied RFECS to the CD4+ T cell dataset as well and determined parameters using cross-validation ( fig . S6 ) ., Using H3K4me1 , H3K4me3 , and H3K27ac , CSIANN made 21832 predictions 39 and ChromaGenSVM method made 23574 predictions 26 ., We made enhancer predictions using H3K4me1 , H3K4me3 and H3K27ac with RFECS as well as Chromia 23 ., Cutoffs were selected that yielded a similar number of enhancer predictions for both Chromia ( 21895 ) and RFECS ( 22947 ) ( fig . 5A ) , so as to make a fair comparison across methods ., To compare these different sets of enhancer predictions , we computed validation rates by comparing them to TSS-distal DNase-I hypersensitive sites , p300 binding sites , and CBP binding sites and misclassification rates by comparing to known UCSC TSS using a window of −2 . 5 kb to +2 . 5 kb as described in the methods ., ( fig . 5A ) ., The validation rate of RFECS predictions is around 70% , which is considerably higher than the other three methods ( 57% ChromaGenSVM , 51% CSIANN , 60% Chromia ) ., Further , the misclassification rates of RFECS is less than 7% , much lower than the 27% , 35% and 15% rates of ChromaGenSVM , CSIANN and Chromia , respectively ., These results suggested that overall procedure for RFECS , including selection of training set as well as training and prediction using the vector-random-forest , performs better than currently available techniques for enhancer prediction ., In the above comparison , we selected our enhancer-representative training set as p300 peaks called using MACS 40 that were distal to known UCSC TSS and overlapped DNase-I locations while CSIANN and ChromaGenSVM used a training-set of p300 peaks called using SICER previously 41 ., We also wanted to compare the performance of the different algorithms on our own datasets using the same training-set to evaluate the performance of the random-forest based part of the algorithm ., To achieve this , we ran the various enhancer prediction methods on H3K4me1 , H3K4me2 and H3K4me3 datasets of H1 , with help from the author of ChromaGenSVM 26 ( fig . 5B ) ., We tried to make the pre-processing stages of the various algorithms as consistent as possible by merging several replicates of each histone modification files and input files into single bed files and randomly selecting a smaller subset of p300 peaks for training , since these were the requirements of the other algorithms such as CSIANN and ChromaGenSVM ., Incase of CSIANN , the selection of background was hard-coded in the software but in all other cases we used our own background training set as well ., In fig . 5B , it can be observed that RFECS shows a maximum validation rate of around 82 . 8% as compared to 66 . 8% , 57 . 7% and 63 . 3% for ChromaGenSVM , CSIANN and chromia respectively ., Further , RFECS showed the lowest misclassification rate of 4 . 9% as compared to 8 . 3% , 36 , 7% and 10 . 1% rates for the above-mentioned cases ., Hence , the improvement in performance due to RFECS cannot be solely attributed to method of selecting the training-set ., In summary , RFECS shows considerably improved performance over existing enhancer-prediction algorithms in two very different datasets and hence can be considered an advance in the field ., Comparing enhancer predictions across diverse cell-types can contribute to understanding differences in regulatory mechanisms between cell-types ., The ENCODE dataset is an example of a collection of high-throughput datasets such as histone modifications and transcription factor binding data that are available for multiple cell-types 42 ., Having a set of high-confidence enhancer predictions in these cell-types would be a valuable resource ., We trained our random forest on the p300 ENCODE data in H1 and made enhancer predictions in 12 ENCODE cell-types using the three marks H3K4me1 , H3K4me3 and H3K27ac since these were available for all the cell-types ., Validation rates were assessed based on overlap with existing DNAse-I hypersensitivity data while misclassification rates were calculated based on overlap with UCSC TSS ., It can be seen that the majority of cell-types show high validation rates between 80 and 95% , while the misclassification rates lie within acceptable levels of 2–7% ( fig . 6A , B ) ., In order to compare enhancers across cell-types , it is preferable to have enhancer predictions with the same level of confidence ., To determine the appropriate cutoff for multiple number of cell-types , we calculate a False Discovery rate by randomly permuting 100 bp bins across the genome and computing the ratio of enhancers predicted in permuted data/enhancers predicted in real data for various cutoffs of voting percentages ., In fig . 6C , it can be seen that different cell-types show a different relationship with FDR ., For example , at an FDR of 5% , the voting percentage for GM12878 ( solid dark blue ) is 0 . 74 , for Nhek ( dashed cyan ) 0 . 64 and for Hsmm ( solid yellow ) it is 0 . 56 ., Using an FDR of 5% , we obtained a consistent set of high-confidence enhancer predictions in the 12 ENCODE cell-types ., In fig . 6D , the numbers of enhancer predictions in each cell type is shown above the bar ., The validation rates ( in red ) are above 90% for all cell-types except H1 , Hepg2 and GM12878 ., In H1 and Hepg2 , the numbers of DNase-I hypersensitivity sites are relatively less , i . e . ∼150 to 177K as compared to ∼230 to 380K in the other cell-lines ., This may explain the somewhat lower validation rate in these two cell-types ., GM12878 appears to be an outlier and we suspect that enhancer predictions may potentially be improved in this cell line by using a different training set ., In summary , we obtained a high-confidence set of enhancer predictions in multiple ENCODE cell-lines with the same level of confidence ., This will enable more rigorous comparisons of regulatory characteristics of these cell-types in the future ., We describe here a novel machine-learning algorithm to accurately predict enhancers in a genome-wide manner based on chromatin modifications ., We trained this algorithm using novel p300 training sets in H1 and IMR90 and 24 chromatin modifications in each cell-type ., We showed that models trained on one cell-type could be effectively applied on another cell-type ., Random forests enable detection of the most informative features required for a classification task ., In the case of enhancer prediction , we identified a set of 3 histone modifications that appeared to be the most informative and robust across cell-types and replicates ., Such an approach can once again be applied when the number of genome-wide modification maps is expanded in various different cell types and the most informative set of modifications can be further refined ., We show that RFECS outperforms other machine-learning based prediction tools in CD4+ T cells , and can be applied in the future to multiple cell types ., We successfully applied our enhancer prediction tool to 12 cell-lines in the publicly available ENCODE database and obtained a set of enhancers with a consistently high level of confidence across the cell-types ., In the future , we could potentially adapt the RFECS method to detect other regulatory genomic elements that can be observed to have a distinct chromatin signature and find the minimal set of chromatin marks for this purpose ., The ability to detect diverse patterns of features within the training set indicates that the RFECS approach could be used to train on a composite training set comprised of different transcription factors ., Combining information from different enhancer-binding proteins may improve prediction of regulatory elements ., Random forests are non-parametric and have been shown to integrate a large number of diverse features ., This could suggest the addition of other discrete and continuous data types such as sequence or motif based information or DNA methylation to the prediction of genomic elements ., The H1 and IMR90 datasets used in this study were generated as part of the NIH Roadmap Epigenome Project and have been released to the public prior to publication ( http://www . genboree . org/epigenomeatlas/multiGridViewerPublic . rhtml ) ., Briefly , 24 chromatin modifications in human embryonic stem cell ( H1 ) and primary lung fibroblast cells ( IMR90 ) were generated by the Ren lab and deposited under the NCBI Geo accession number GSE16256 ., Additionally , two replicates of H3K9me3 datasets deposited under Geo accession numbers GSM818057 and GSM42829 were used ., Genome-wide binding data for p300 in H1 and IMR90 , and transcription factors NANOG , SOX2 and OCT4 in H1 were generated in the Ren lab using ChIP-seq and deposited under accession numbers GSE37858 , GSE18292 and GSE17917 respectively ., Any data mapped to hg18 was converted to hg19 using liftover tools 43 ., The DNase-I hypersensitivity datasets for H1 and IMR90 were produced by the Stammatoyanopoulos group at UW 44 ., IMR90 DNase-I raw data may be accessed using GSM468792 and narrow peak calls are attached as supplemental information ., Narrow DNase-I peaks in H1 were downloaded from UCSC ENCODE page ( http://hgdownload . cse . ucsc . edu/goldenPath/hg19/encodeDCC/wgEncodeUwDnase/ ) For CD4 , previously generated datasets for p300 41 , CBP 41 and DNase-I 21 data as well as histone modifications 45 , 46 we | Introduction, Results, Discussion, Methods | Transcriptional enhancers play critical roles in regulation of gene expression , but their identification in the eukaryotic genome has been challenging ., Recently , it was shown that enhancers in the mammalian genome are associated with characteristic histone modification patterns , which have been increasingly exploited for enhancer identification ., However , only a limited number of cell types or chromatin marks have previously been investigated for this purpose , leaving the question unanswered whether there exists an optimal set of histone modifications for enhancer prediction in different cell types ., Here , we address this issue by exploring genome-wide profiles of 24 histone modifications in two distinct human cell types , embryonic stem cells and lung fibroblasts ., We developed a Random-Forest based algorithm , RFECS ( Random Forest based Enhancer identification from Chromatin States ) to integrate histone modification profiles for identification of enhancers , and used it to identify enhancers in a number of cell-types ., We show that RFECS not only leads to more accurate and precise prediction of enhancers than previous methods , but also helps identify the most informative and robust set of three chromatin marks for enhancer prediction . | Enhancers are regions in the genome that can activate the expression of a gene irrespective of their location with respect to the gene ., Identifying these elements is critical in understanding regulatory differences between different cell-types ., Since enhancers lack characteristic sequence features and can be far away from the gene they regulate , their identification is not trivial ., Experimentally determining the genome-wide binding sites of transcriptional co-activator p300 is one way of finding enhancers but it can only identify a subset of enhancers ., A few years ago , it was observed that the binding sites of p300 are marked by distinctive , post-translational histone modifications ., Several groups have exploited this discovery to predict genome-wide enhancers based on their similarity to the histone modification profiles of p300 binding sites ., We here report a novel algorithm for this purpose and show that it has much greater accuracy than existing methods ., Another unique feature of our algorithm is the ability to automatically deduce the most informative set of histone modifications required for enhancer prediction ., We expect that this method will become increasingly useful with the expanding number of known histone modifications and rapid accumulation of epigenomic datasets for various cell types and species . | computer science, biology | null |
journal.pcbi.1005452 | 2,017 | Heterogeneous firing responses predict diverse couplings to presynaptic activity in mice layer V pyramidal neurons | The specific activation of subpopulations within neocortical networks appears to be the core mechanism for the cortical representation of sensory features ., The details of how such specific activations happen are therefore key questions in systems neuroscience ., As a primary source for specific activation , the neocortex is characterized by some degree of specific circuitry: neurons differ in their afferent connectivity ., A classic example can be found in the primary visual cortex , layer IV simple cells specifically sample their input from ON and OFF cells in the thalamic nucleus 1 , 2 ., Additionally , neocortical neurons also vary in their electrophysiological properties: for example , heterogeneous levels in the action potential threshold are routinely measured in vivo 3–5 ., Thus , an emerging refinement is that the sensitivity of a neuron to a given feature do not only results from its stimulus specificity ( e . g . orientation selectivity as a result of a specific afferent circuitry ) , but from the combination of its stimulus specificity and its biophysical specificity ., Two somato-sensory cortex studies illustrates this point precisely ., In Crochet et al . 3 , during active touch , the spiking probability of a neuron ( its sensitivity to whisker touch ) follows from the combination of the reached level of synaptically-driven membrane potential deflection ( its stimulus-specificity resulting from afferent circuitry , as quantified by post-synaptic reversal potentials ) and its threshold for action potential triggering ( its biophysical specificity ) ., A similar result was found in the study of Yang et al . 5 for texture recognition , where the combination of those two quantities was shown to predict choice-related spiking ., Those results therefore suggest that heterogeneity in the biophysical properties of neocortical neurons might have an impact on their functional role during sensory processing ., In the present work , we further investigate the interaction between stimulus specificity and biophysical specificity in the light of the variability in the biophysical features reported in our previous study 6 , namely that single neurons in juvenile mice cortex not only vary in their excitability ( linked to the action potential threshold ) but also in their sensitivity to the properties of the membrane potential fluctuations ., Our previous communication introduced those new dimensions in the biophysical specificity , we aim here at understanding their functional impact ., To this purpose , we implemented various stimuli onto layer V pyramidal cells ( we varied the properties of presynaptic activity in the fluctuation-driven regime ) , and we investigated whether individual neurons would differentially respond to those inputs as a result of their various firing rate responses 6 ( their various biophysical specificities ) ., Our study investigates the properties of single cell computation in the regime of low synchrony population dynamics 7 , 8 ( the analogous , at the network level , of the fluctuation-driven regime at the cellular level ) and aims at describing effects mediated by slow population dynamics ( T ≥20-50ms ) ., In this context , the cellular input-output function of a neocortical neuron corresponds to the function that maps the presynaptic variables to the spiking probability of the neuron , which is often called the transfer function of the neuron ., The framework of our approach to the transfer function is illustrated in Fig 1A ., Our cellular model has five presynaptic variables: four of them are presynaptic firing rates ( stationary release probabilities at the synapses ) as those constitute the primary input variables in this rate-based paradigm ., To investigate their differential contribution , the proximal and distal parts of the dendritic trees have been treated separately and each of them has two presynaptic rates corresponding to the excitatory and inhibitory input ( hence four rate variables: νep , νip , νed and νid ) ., The main motivation for this separation is to distinguish between two types of projections onto neocortical pyramidal neurons: synaptic inputs from the local network are thought to be more proximal while the distal apical tuft receives input from more distant cortical areas and thalamic locations 9 ., Additionally , a global synchrony variable has been introduced for presynaptic events ., This reproduces the effect of multi-innervation of a cell by its presynaptic afferent and , more importantly , the effect of pairwise correlations associated to neocortical dynamics 10 ., The synchrony degree in the presynaptic activity has been suggested to vary with stimulus statistics in the primary visual cortex 11 , 12 what motivates its introduction as a separate variable in our model ., As illustrated on Fig 1A , we introduce a two-step procedure to determine the input-output function of a single cell in the fluctuation-driven regime ., The idea behind this two-step approach relies on the fact that the action potentials are initiated at the axon initial segment 13 ( i . e . electrotonically close to the soma ) so that the fluctuations at the soma could determine the firing probability uniquely ( see Discussion for the approach’s limitations ) ., We thus split the relation from presynaptic quantities ( the input ) to the spiking probability ( the output ) into the following successive steps ., The first step consists in evaluating how dendritic integration will shape the membrane potential fluctuations at the soma for given values of the presynaptic input variables ., The fluctuations are quantified by their mean μV , their amplitude σV ( given by the standard deviation of the fluctuations ) and the fluctuations speed τV ( given by the autocorrelation time of the fluctuations ) ., This first step will be performed analytically as the passive properties and simplified morphology of the dendritic model enables a mathematical treatment of this question ., The second step consists in determining how the somatic fluctuations ( μV , σV , τV ) are translated into action potential output ., This last step is computed thanks to a fitted function directly constrained by experiments ., Indeed , by analyzing the spiking response of pyramidal neurons recorded in vitro as a function of these somatic fluctuations ( μV , σV , τV ) , we previously found that a parametric function could reliably describe the relation between the fluctuations properties and the firing rate output in individual neurons 6 ., Thus , because the experiments were performed as a function of these somatic variables ( μV , σV , τV ) , the same experimental data obtained previously can be used in the present framework ., Within this framework , one individual cell ( indexed by k ) is therefore described by two functions:, 1 ) a dendritic integration function ( μV , σV , τV ) =FkD ( νep , νip , νed , νid , s ) that accounts for synaptic integration up to the somatic Vm fluctuations and, 2 ) a parametric function νout=Fkν ( μV , σV , τV ) that translates the somatic Vm fluctuations into a spiking output ., The final input-output function νout=Fk ( νep , νip , νed , νid , s ) of cell k is thus the result of the composition: Fk=Fkν∘FkD ( see Fig 1A ) ., From our previous study , we benefit from a set of firing response functions {Fkν}k∈1 , n of n = 30 cells ., To complete the framework , we now need to associate a dendritic morphology for each of this cell to be able to calculate the associated set of dendritic integration functions {FkD}k∈1 , n ., This is the focus of the next two sections ., We first present our dendritic morphology model and then derive an association rule from input resistance to dendritic morphology based on in vitro recordings of somatic input impedance ., The morphology of our theoretical model is a lumped impedance somatic compartment in parallel with a dendritic arborescence of symmetric branching following Ralls 3/2 branching rule ( see Fig 1B and Methods ) ., This morphology is of course a very reductive description of pyramidal cells: it does not discriminate between the distinct apical trunk and the very dense basal arborescence ., Also , branching in pyramidal cell morphologies have been shown to deviate from Ralls 3/2 branching rule ., Nonetheless this simplified model contains the important ingredient for our study: the fact that the transfer impedance to the soma of a synaptic input will strongly depend on its location on the dendritic tree ., Indeed , as observed experimentally 14 , distal events will be more low-pass filtered than proximal events in this model ., We spread synapses onto this morphology according to physiological densities 15 and describe synaptic events as transient permeability changes of ion-selective channels ( see Methods ) ., We arbitrarily separate the dendritic tree into two domains: a proximal and a distal domain ( delimited by their distance to the soma , see Fig 1B ) ., The distal part was taken as the last eighth of the dendritic tree to reproduce the large electronic distance to the soma characterizing distal synapses 14 ., Following experimental evidences 14 , we set a higher synaptic efficacy for distal synapses ., The synaptic parameters take physiological values 16 and can be found on Table 1 ., The passive parameters as well as the individual morphologies are estimated in the next section ., We will use the firing response functions of the cells of our previous study 6 ( the “firing response dataset” ) ., For each of this cell , we therefore need an estimate of the parameters of the dendritic model ( described above , i . e . passive properties and morphology parameters ) ., The purpose of this section is thus to derive an association rule from the input resistance ( a quantity that we have for all cells of the “firing response dataset” , see Fig 2C ) to the parameters of the dendritic model ., We based this estimate on the comparison between the dendritic model behavior and the properties of the somatic input impedance in layer V pyramidal neurons ( n = 13 cells , measured with intracellular recordings in vitro ) ., The key property on which this characterization relies is the fact that the input impedance at the soma cannot be accounted for only by the isopotential somatic compartment ( i . e . a RC circuit ) ., The input impedance shows the contribution of the dendritic tree in parallel to the soma 17 ., Indeed , both the modulus and the phase of the input impedance show deviations from the RC circuit impedance ( see the comparison in Fig 2B ) : see for example the exponent of the power law scaling of the modulus ( -1 exponent for the single compartment and ∼-0 . 7 for pyramidal cells ) or the decreased phase shift around 100Hz ., As this behavior is the consequence of the electrotonic profile along the dendritic tree , we used it to estimate the parameters of our simplified dendritic model ., We first average all data ( shown on Fig 2A ) to obtain a mean input impedance ( shown on Fig 2B ) representative of a mean cellular behavior ., We then performed a minimization procedure to obtain both the passive properties and the morphology corresponding to this average behavior ( see Methods ) ., The obtained passive properties were compatible with standard values , e . g . the resulting specific capacitance was 1 . 05 μF/cm2 , close to the commonly accepted 1 μF/cm2 value , thus suggesting that the procedure could capture the physiological parameters of pyramidal cells , see Table 1 for the other parameters ., Most importantly , the surface area was physiologically realistic , so that when using synaptic densities , we obtain an accurate number of synapses ( see below ) ., A representation of this mean morphology can be seen on Fig 2D ., Pyramidal cells show a great variability in input impedance , for example their input resistance almost spans one order of magnitude ( both in the present n = 13 cells , see the low frequency modulus values in Fig 2A , as well as in the firing response dataset , see bottom in Fig 2C ) ., We found that varying the size of the morphological model within a given range around the mean morphological model could partially reproduce the observed variability in the input impedance profiles ( see Fig 2B ) ., Size variations corresponds to a linear comodulation of the, 1 ) tree length Lt ,, 2 ) the diameter of the root branch Dt and, 3 ) the length of the somatic compartment LS ( see Fig 2C for the range of their variations ) ., On Fig 2A , the cells have been colored as a function of their input resistance while on Fig 2B , we vary the size of the size of the morphological model ., Large cells ( blue , low input resistance ) tend to have a lower input resistance and phase shift than the small cells ( red , high input resistance ) ., Note that this simplistic account of morphological variations only very partially describes the observed behavior in pyramidal cells ., In particular ,, 1 ) it strongly underestimates the variations of phase shifts at medium and high frequencies ( f>20Hz ) and, 2 ) the relationship between size and impedance modulus at high frequencies ( f>100Hz ) is poorly captured ., Those discrepancies are likely to be due to the details of dendritic arborescence that are not captured by the strong constraints of our dendritic model ( symmetric branching , diameter rules , number of branches , etc… ) ., Despite those discrepancies , size variations in our morphological model constitute a reasonable first approximation to account for cellular variety within the layer V pyramidal cell population ., This characterization , combined with the analytical tractability of the model ( see Methods ) allowed us to construct a map between input resistance at the soma and size of the morphological model ( the passive properties are set as identical , the one fitted on the mean impedance behavior ) ., Thus , for each neuron in our previous firing response dataset , because we have its input resistance at the soma , we can associate a given morphology ., The association rule is shown in Fig 2C ., We now check what is the number of synapses obtained from the combination of our fitted morphologies with the physiological synaptic densities ., We found a number of synapses of 3953 ± 1748 ( mean and standard deviation across the n = 30 cells ) with a ratio of excitatory to inhibitory numbers of synapses of 4 . 5 ± 0 . 1 ., The fact that those numbers fall within the physiological range constitutes a validation of our approach ( the morphology estimate through input impedance profile characterization ) ., We now want to translate the five variables of the model in terms of membrane potential fluctuations properties at the soma ( μV , σV , τV ) , i . e . determining the function FkD for each cell k depending on its dendritic parameters ., This constitutes the first step to obtain the final input-output function of individual cells ( see Fig 1A ) ., Investigating dendritic integration for detailed morphological structures is made difficult by the fact that this has to be done numerically with a relatively high spatial and temporal discretization ., In the fluctuation-driven regime , one also needs to sample over long times ( T ≫ τV ∼ 20ms ) to obtain the statistical properties of the somatic Vm resulting from dendritic integration ., In addition , this study relies on n = 30 different morphologies and we will explore a five dimensional parameter space ( the five variable of our model ) ., Under those conditions , if performed numerically , the computational cost of such a study is clearly prohibitive ., We briefly describe here , why , in our simplified model , an analytical treatment is possible and thus renders this investigation feasible ( see details in the Methods and in S1 Text ) ., The key ingredient is the ability to reduce the dendritic tree to an equivalent cylinder 17 , we only adapted this reduction to the changes in membrane permeability associated to the high conductance state 18 ., Two approximations underlie our estimation:, 1 ) the driving force during an individual synaptic event is fixed to the level resulting from the mean bombardment 19 and, 2 ) the effect at the soma of a synaptic event at a distance x in a branch of generation b , corresponds to the 12b−1 fraction of the post-synaptic response to the stimulation made of synchronous events at distance x in all the 2b−1 branches of the generation b ., Luckily , the combination of those approximation is a favorable situation ., Indeed , hypothesis, 1 ) overestimates the size of post-synaptic events ( because the driving force is not fixed , it diminishes during the PSP time course ) while hypothesis, 2 ) underestimates the size of post-synaptic events ( because of the 2b−1 − 1 synchronous events in neighboring branches , the membrane conductance is higher than in the case of a single event , consequently neighboring events have a shunting effect that artificially decreases the response ) ., In addition , both of those approximation are likely to hold when single events are of low amplitude compared to the amplitude of the massive synaptic bombardment ( see e . g . Kuhn et al . 19 for the validity of the first hypothesis ) ., In Fig 3 , we compare the analytical approximation to the output of numerical simulations performed with the NEURON software 20 ., We varied the five variables of the model around a mean synaptic bombardment configuration ( see next section ) ., Some discrepancies between the approximation and the simulations appeared , in particular one can see a ∼1mV shift in the standard deviation σV of the fluctuation ( meaning that single events are underestimated in the analytical treatment , so that hypothesis 2 is the most problematic one ) ., Because the synchrony controls the amplitude of the fluctuations ( Fig 3B and next section ) , the analytical estimate could therefore be seen as an accurate estimate , modulo a shift in the synchrony ( see Fig 3B , an increase of 0 . 18 in the synchrony corrects for the ∼1mV shift in σV ) ., Importantly , the trend in the variations of the fluctuations as a function of the model variables is globally kept between the analytical estimate and the numerical simulations ., This relatively good agreement therefore shows that our analytical estimate is a valid tool to study dendritic integration in the fluctuation-driven regime ., We now implement various types of presynaptic activity and investigate the properties of the resulting membrane potential fluctuations at the soma ., In addition , we represent the variations of the somatic input conductance ( relative to the leak input conductance ) because , as it is routinely measured in intracellular studies in vivo , this quantity allows a comparison between the model and experimentally observed activity levels ., On Fig 4 , we present those different protocols , on the left ( panel A ) , one can see how the five variables of the model are comodulated for each protocol ( color coded , see bottom legend ) and on the right ( panel B ) , one can see the resulting properties of the membrane potential fluctuations ., We present those results only for the medium-size model , but it was calculated for the morphologies associated to all cells ., The variability introduced by the various morphologies is shown in S2 Fig and we found that the qualitative behavior discussed in this section was preserved in all cells ., We first introduced a baseline of presynaptic input corresponding to a low level of network activity: gtotsomagL∼1 . 7 , compared to ∼3–4 in activated states , reviewed in 18 ., This baseline activity is a mix of proximal and distal activity with a low degree of synchrony ( s = 0 . 05 ) ., Similarly to Kuhn et al . 19 , the inhibitory activity is adjusted to obtain a balance of the Vm fluctuations at -55mV ., The firing values of this baseline level are very low ( νed=νep = 0 . 2Hz for the excitation and νid=νip = 1 . 2Hz for the inhibition ) in accordance with the sparse activity characterizing mammalian neocortical dynamics 3 , 10 ., On top of this non-specific background activity , we will now add a specific stimulation ., We consider four types of presynaptic stimulations: Note , that in addition to the sparse activity constraints or the balance constraints , the criteria for the ranges of the model variables was chosen to have the fluctuations in the same domain ., For example , we investigated a lower activity range for the distal part ( variations of νdist ) than for the proximal part ( variations of νprox ) to avoid an explosion of σV , the range for the synchrony increase followed the same criteria ., For each one of the n = 30 cells of our previous study 6 , we have, 1 ) a morphological model ( see previous sections ) and, 2 ) a firing response function νout=Fkν ( μV , σV , τV ) ., Thanks to the previous analytical approximation , we can translate the five model variables ( νep , νip , νed , νid , s ) into the stationary fluctuations properties ( μV , σV , τV ) that , in turn , the function Fkν translate into a spiking probability ., Thus , we finally get the full input-output function ( within our theoretical framework ) as illustrated on Fig 1A ., We show on Fig 5 the response of four cells to the different types of presynaptic activity described in the previous section ( those four cells were chosen as they were representative of different firing response behaviors , see Figs 5 and 6B in 6 ) ., The input-output relationships show qualitative and quantitative differences , we briefly discuss them here and we perform a more rigorous analysis on the full dataset in the next section ., First , we can see that individual cells have a very different level of response to the baseline level of synaptic activity ( initial response in Fig 5 ) ., Cell 1 has a baseline at ∼ 10−2 Hz while Cell2 or Cell3 have response above 1Hz , i . e . two orders of magnitude above ., Importantly , those cells have different preferences for particular types of stimulations ., Cell 1 responds more to unbalanced activity whereas Cell 2 and Cell 4 respond more to an increase in synchrony and Cell 3 responds preferentially to proximal activity ( within this range ) ., This is what we mean by preferential coupling: individual neurons will respond preferentially to a particular type of synaptic activity ., An even more pronounced discrepancy appears for proximal activity: the response can be either increased ( Cell 1 and Cell, 3 ) or decreased ( Cell 2 and Cell, 4 ) with respect to the baseline level ., Given the relative invariance of the fluctuations properties for each cell ( see previous section and S2 Fig: despite the various morphologies , the same input creates the same fluctuations ) , those differences can only be attributed to the various firing responses of individual cells ( the diversity in the Fkν functions found experimentally 6 ) ., We conclude that heterogeneous firing responses induce diverse coupling to presynaptic activity ., We now make this analysis more quantitative by computing the responses for all n = 30 cells ., We get their response to the baseline level νbsl and their mean response change for each stimulation type ( the mean over the range of scanned presynaptic input ) : δνubl for the unbalanced activity , δνprox for the proximal activity , δνdist for the distal activity and δνsynch for an increased synchrony ., We show the histogram of those values in the left column of Fig 6 ., We now investigate how the response of an individual cell relates to its biophysical specificity ., It is defined here by four quantities 6 ( see also Methods ) : We first analyze the response to baseline activity νbsl ., When log-scaled ( Fig 6A ) , the distribution is approximately normal and spans 2–3 orders of magnitude ., This log-normal distribution of pyramidal cell firing rates during spontaneous activity seems to be a hallmark of mammalian neocortical dynamics ( see e . g . 10 in human neocortex ) ., We investigated what properties of the firing responses could explain this behavior , we therefore looked for correlations between our measures of the firing responses in the fluctuation-driven regime 6 and the baseline responses ., Not surprisingly , we found a very strong linear correlation between the excitability 〈Vthreeff〉D and the baseline response level , the other characteristics do not have an impact ( Pearson correlations , see values in Fig 6A ) ., It should be stressed that presynaptic connections are homogeneous across cells in this model , those results therefore show that the typical log-normal distribution of firing rates could very naturally emerge as a result of the normal distribution observed in pyramidal cells excitabilities 6 , thus suggesting that no specific circuitry might be needed to explain this neocortical property ., Despite the important differences in the fluctuations they create ( see Fig 6B ) , the responses over cells to unbalanced activity , distal activity and an increased synchrony share a very similar behavior ., First , those stimuli produce systematically an increase in firing rate ( n = 30/30 cells ) ., The firing increase again show a strong heterogeneity over cells , covering two orders of magnitude ( see log y-axis on Fig 6B , 6D and 6E ) ., Again , this variability in responses was highly correlated with the excitability ., Surprisingly , the response was not dependent on any other of the characteristics of the firing response ., For example , because synchrony controls the standard deviation σV , the variability observed during an increase in synchrony could have been linked to the to the sensitivity to the standard deviation 〈∂ν∂σV〉D , but this effect was not significant ( see Pearson correlation in Fig 6E ) ., For those three protocols , none of the sensitivities to the fluctuations properties had a strong impact on the individual cellular responses ( c<0 . 4 and p>0 . 01 , Pearson correlations , see Fig 6 ) ., This analysis therefore revealed that , for those type of synaptic activities , those properties of the firing response have negligible impact compared to the very strong effect of the variability in excitabilities ( see Discussion ) ., The response to proximal activity also showed a great variability but with a qualitatively different behavior ( Fig 6C ) ., Notably , firing could be suppressed or increased ., This variability was independent of the excitability of the cells but was correlated with the sensitivity to the speed of the fluctuations 〈∂ν∂τVN〉D ., Indeed , the proximal stimulation implies a strong variations of the fluctuations speed ( i . e . decreasing τV , while keeping moderate variations of σV and , by design , a constant μV ) thus rendering the sensitivity to the fluctuation speed the critical quantity for this stimulation type ., Those results therefore show that the response to proximal activity of an individual cell is controlled by its level of sensitivity to the speed of the fluctuations ( see Discussion ) ., Despite the current weaknesses of our description ( see next section ) , we believe that having an analytical model for dendritic integration in the fluctuation-driven regime is a useful tool for many problems in theoretical neuroscience ., The main advantage of this model is that one can very naturally plug in physiological parameters ( because surface area as well as transfer resistance to soma can take physiological values ) while still allowing an analytical treatment ( though see deviations of the approximations in Fig 3 ) ., In the theoretical analysis of neural network dynamics , the literature is almost exclusively based on the reduction to the single-compartment ( reviewed in 22 ) ., Though being approximate , our framework thus opens the path toward a detailed mathematical analysis of recurrent network dynamics containing neurons with extended dendritic structures ., Additionally , It must be stressed that formulating the cell response with the somatic fluctuations properties ( μV , σV , τV ) as an intermediate variable is very powerful because it allows one to apply the same measurements to various models ., For example , with a single-compartment model it is easy to translate these variables into excitatory and inhibitory activities 19 ., We showed here that it is also possible to obtain relations with synaptic inputs occurring in dendrites in a simplified morphological neuron model ., The latter model uses the same measurements , so no experiments need to be redone ., We could in principle also apply the same approach to more complex models and obtain more realistic transfer functions ., This phenomenological two-step procedure thus offers a flexible complementary approach to the analytical approaches tackling the problem of the spiking behavior in presence of an extended dendritic structure 23–25 ., The proposed theoretical framework for single-cell computation nonetheless suffers from several weaknesses ., First , even if our description captures the various electrotonic distances associated to various synaptic locations ( the crucial ingredient here to discriminate between proximal and distal inputs ) , the morphological model appears as a very poor description of layer V pyramidal cell ., Deviations from the symmetric branching hypothesis and Ralls branching rule will have a significant impact on dendritic integration in the fluctuation-driven regime ., To investigate those effects within the framework proposed in our study , one could benefit from the large body of theoretical work on the derivation of Green’s function for arbitrary branched passive dendritic trees 26–29 ., Another important limitation of our description lies in the absence active mechanisms in dendrites 30 ., It is therefore a question how much those mechanisms could affect the picture provided in our study ., Preliminary numerical analysis performed in presence of NMDA and Ca2+ currents 31 ( see S4 Fig and S5 Fig ) , showed that , provided excitation balances inhibition ( a situation where NMDA channels keep a relatively low level of stationary activation ) and provided synchrony do not reach a too high level ( unlike for s≥0 . 4 , where excitatory events almost systematic lead to NMDA spikes ) , the qualitative behavior of the cellular input-output function remains unaffected ., Thus , even if the lack of dendritic mechanisms underestimates the coupling values reported here ( cells are less excitable in the passive setting , leading to attenuated spiking ) , the absence of qualitative differences suggest that the cell-to-cell variability reported in this study would be poorly affected ., Finally , our description assumes a unique correspondence between average presynaptic quantities , somatic fluctuations and output spiking ., The compartmentalized nature of active dendritic integration is very likely to break this hypothesis: the same average input could lead to very different output responses when targeting different functional subunits ., We conclude that , given the complexity of synaptic integration in neocortical cells , the proposed approach only constitutes a very first approximation of the cellular input-output function ., Quantifying its weaknesses and improving this picture should be the focus of future investigation ., Very naturally , a key quantity to explain the various levels of neuronal responses is the cellular excitability ., Indeed , the response to baseline activity , unbalanced activity , distal activity or an increase in synchrony is strongly correlated with cellular excitability ., More surprisingly , the sensitivity to the speed of the fluctuations also have a crucial impact on the response for one type of synaptic activity: proximally targeting synaptic input ., In our previous communication 6 , theoretical modeling suggested that a high sensitivity to the speed of the fluctuations was enable by a high level of sodium inactivation ( as only fast fluctuations allow to deinactivate sodium channels ) and a high density of sodium channels ( as it corresponds to a sharp spike initiation mechanism that enables to extract fast fluctuating input , revi | Introduction, Results, Discussion, Methods | In this study , we present a theoretical framework combining experimental characterizations and analytical calculus to capture the firing rate input-output properties of single neurons in the fluctuation-driven regime ., Our framework consists of a two-step procedure to treat independently how the dendritic input translates into somatic fluctuation variables , and how the latter determine action potential firing ., We use this framework to investigate the functional impact of the heterogeneity in firing responses found experimentally in young mice layer V pyramidal cells ., We first design and calibrate in vitro a simplified morphological model of layer V pyramidal neurons with a dendritic tree following Ralls branching rule ., Then , we propose an analytical derivation for the membrane potential fluctuations at the soma as a function of the properties of the synaptic input in dendrites ., This mathematical description allows us to easily emulate various forms of synaptic input: either balanced , unbalanced , synchronized , purely proximal or purely distal synaptic activity ., We find that those different forms of dendritic input activity lead to various impact on the somatic membrane potential fluctuations properties , thus raising the possibility that individual neurons will differentially couple to specific forms of activity as a result of their different firing response ., We indeed found such a heterogeneous coupling between synaptic input and firing response for all types of presynaptic activity ., This heterogeneity can be explained by different levels of cellular excitability in the case of the balanced , unbalanced , synchronized and purely distal activity ., A notable exception appears for proximal dendritic inputs: increasing the input level can either promote firing response in some cells , or suppress it in some other cells whatever their individual excitability ., This behavior can be explained by different sensitivities to the speed of the fluctuations , which was previously associated to different levels of sodium channel inactivation and density ., Because local network connectivity rather targets proximal dendrites , our results suggest that this aspect of biophysical heterogeneity might be relevant to neocortical processing by controlling how individual neurons couple to local network activity . | Neocortical processing of sensory input relies on the specific activation of subpopulations within the cortical network ., Though specific circuitry is thought to be the primary mechanism underlying this functional principle , we explore here a putative complementary mechanism: whether diverse biophysical features in single neurons contribute to such differential activation ., In a previous study , we reported that , in young mice visual cortex , individual neurons differ not only in their excitability but also in their sensitivities to the properties of the membrane potential fluctuations ., In the present work , we analyze how this heterogeneity is translated into diverse input-output properties in the context of low synchrony population dynamics ., As expected , we found that individual neurons couple differentially to specific form of presynaptic activity ( emulating afferent stimuli of various properties ) mostly because of their differences in excitability ., However , we also found that the response to proximal dendritic input was controlled by the sensitivity to the speed of the fluctuations ( which can be linked to various levels of density of sodium channels and sodium inactivation ) ., Our study thus proposes a novel quantitative insight into the functional impact of biophysical heterogeneity: because of their various firing responses to fluctuations , individual neurons will differentially couple to local network activity . | medicine and health sciences, action potentials, dendritic structure, nervous system, membrane potential, electrophysiology, neuroscience, ganglion cells, neuronal dendrites, animal cells, biophysics, physics, cellular neuroscience, cell biology, anatomy, synapses, pyramidal cells, physiology, neurons, biology and life sciences, cellular types, physical sciences, neurophysiology | null |
Subsets and Splits