Datasets:

sulovexin

/

mysummary

like 0

journal.pcbi.1000059

The Emergence and Fate of Horizontally Acquired Genes in Escherichia coli

The wide variation in bacterial genome sizes was originally detected in the 1960s by DNA reassociation analyses 1 ., And because bacteria have gene dense chromosomes , the differences in genome sizes implied that there were likely to be vast differences in the gene contents of bacterial species ., With the current availability of hundreds of complete genome sequences , it is now possible to establish exactly which genes are present in , as well as those that are absent from , a genome ., Among sequenced bacterial genomes , gene sets vary over 40-fold , ranging from 182 genes in the gammaproteobacterial symbiont Carsonella ruddii 2 to almost 8000 genes in the soil-dwelling acidobacterium Solibacter usitatus ( jgi . doe . gov ) ., The wide variation in genome sizes and gene contents can also be observed between strains within individual bacterial genera or species ., For example , isolates of Frankia that are more than 97% identical in their rRNA sequences–the conventional cutoff value for a bacterial species–can differ by as many as 3500 genes , which represents nearly half of their 7 . 5 Mb genome 3 ., Even among bacterial strains of similar genome sizes , there can be substantial differences in gene repertoires 4 ., Unlike mammals , in which only about 1% of the genes in a genome are unique to a taxonomic order ( e . g . , mouse vs . human 5 ) , the gene contents of bacterial genomes can change rapidly over relatively short evolutionary distances ., The generation of novel gene repertoires is a consequence of the ongoing processes of gene acquisition and gene loss 6–10 ., Although several mechanisms can generate new genes 6 , 11 , 12 , the novel gene sets observed in closely related bacterial strains result largely from gene transfer from distant sources , as duplications and gene rearrangement only rarely produce entirely unique genes in the short timescales in which bacterial gene sets evolve ., Although homolog searches indicate that many genes arise from lateral transfer from other bacteria , most bacteria also contain genome-specific sets of genes that lack any homologs in the known databases ( termed “ORFans” ) 4 , 13 , 14 ., Counteracting the augmentation of bacterial genomes by gene acquisition , gene loss occurs both through large-scale deletions 15 as well as by smaller changes that erode and inactivate individual genes 7 , 9 , 16 ., As observed for acquired sequences , prokaryotes also contain genome-specific sets of inactivated genes ( i . e . , pseudogenes ) , which can comprise up to 41% of their annotated genes 17 ., Taken together , these lineage-specific gene repertoires indicate the need to monitor bacterial genome dynamics–i . e . , the manner in which genes are gained and lost–over short evolutionary timescales ., To this end , comparisons of closely related strains of Bacillus 18 , Staphylococcus aureus 19 and E . coli 7 , 20 have shown that gene acquisitions are prevalent at the tips of the phylogeny and that recently acquired genes seem to evolve more quickly ., However , few studies have examined the fate of these genes within a bacterial lineage or have asked how many or which classes of genes , once acquired , are maintained , disrupted or removed from a genome ., We address these questions by assessing the differences in gene repertoires among 13 sequenced strains of E . coli/Shigella clade ., These strains are closely related , yet display substantial differences in genome size and gene content 21 , 22 , allowing us to pinpoint the introduction and persistence of genes in the lineages leading to these genomes ., We reconstructed the phylogeny of 13 sequenced strains of E . coli and Shigella species based on the concatenated sequences of 169 conserved , single-copy genes ., The relationships and branching orders are well-resolved , well-supported , and congruent with previous studies 23 ., The overall branching order of the resulting phylogeny is very similar to those based on other characters or for more limited sets of sequenced strains 24–26: the uropathogenic E . coli ( UPEC ) form a monophyletic cluster at the base of the tree , and Shigella strains are polyphyletic , with a major lineage derived from the clade containing E . coli K-12 27 ., Based on this tree , we delineated 12 monophyletic clades of varied phylogenetic depths ( of which we designate the corresponding ancestral branches as S1 to S4 , SC , SCE , C1 , E1 , E2 , U1 , U2 and the ancestral branch , Figure 1 ) , which were used to trace the evolutionary history of all genes in these 13 genomes ., By identifying the homologs of genes from the 13 E . coli and Shigella strains in each of 367 microbial genomes , and by mapping the gene distributions in a phylogenetic context , we could infer the ancestry ( vertical or horizontally acquired ) and dynamics ( incidence of acquisition or loss ) of genes among strains ., Acquired genes were classified into two categories: ORFans , which are genes that have no homologs outside of the analyzed E . coli and Shigella strains , and HOPs , which are genes that have homologs outside of the analyzed E . coli and Shigella genomes but are not ancestral to all taxa containing the gene or whose phylogenetic distributions can not be most parsimoniously reconstructed solely through gene loss events ., From the 13 sequenced strains within the E . coli/Shigella clade , we identified a total of 1443 ORFan gene families and 652 HOP gene families ( a family is a group of homologs ) ., Gene family sizes ranged from one gene , for ORFans or HOPs present in a single genome ( representing 11% and 32% of the total number of families , respectively ) to 13 genes , for ORFans or HOPs with homologs present in all 13 genomes ( representing 13% and <1% of the total number of families , respectively ) ., We inferred the branch on which ORFans and HOPs originated by reconstructing the most parsimonious series of events that would give rise to their present-day distributions ., By this approach , all HOPs could be assigned to a particular clade , but only 1177 of the 1443 ORFan families were assigned unequivocally , and together these constitute the set considered in subsequent analyses ., Only 8 ORFans ( <1% ) could be classified to a particular COG category , whereas 151 ( 23% ) of the HOP families could be assigned to a COG other than ‘poorly characterized’: these included Metabolism ( 10% ) , Cellular Processes and Signaling ( 8% , mostly in the category Cell Wall/Membrane/Envelope Biogenesis ) and Information Storage and Processing ( 5% ) ( Supplementary Table S1 ) ., The numbers of acquired ORFans and HOPs vary substantially across strains and lineages ( Figure 1 ) , with the largest difference occurring in the gene set acquired by the ancestor to all tested strains in which ORFans are approximately four times more common than HOPs ., This is in contrast to genes confined to a single E . coli or Shigella genome , where we identify ∼40% more HOPs than ORFans ., This difference is not affected by the fact that 20% of ORFans could not be placed onto a specific branch , because singleton ORFans are among the easiest genes to assign ., Taken together , these distributions suggest that HOP genes originate more frequently , but ORFans are more likely to persist ., Overall , ORFans constitute between 9% and 14% of the protein coding genes per genome , and HOPs account for at most 5% of the protein coding genes per genome ., Cumulatively , ORFans outnumber HOPs; however , HOPs represent a larger proportion of the acquired DNA in all strains as they are , on average , longer than ORFans ( 853 bp vs . 308 bp respectively ) ( Table 1 ) ., There is an association between genome size and the amount of ORFan and HOP-derived DNA ( r2\u200a=\u200a0 . 75 and r2\u200a=\u200a0 . 72 , respectively ) per genome; however , it is not simply a matter that the strains with the largest genomes have acquired the most DNA ., For example , Shigella dysenteriae and E . coli EDL933 have gained identical amounts of DNA from ORFans and HOPs despite an 800 kb difference in their genome sizes ., ORFans are more A+T-rich than HOPs ( 44% vs . 47% G+C , respectively ) , and such differences in base composition are evident along most lineages ( Supplementary Figure S1 ) ., When examining a single lineage at increasing phylogenetic depths , there is no clear trend towards increased G+C contents , G+C content of the third codon position or increased gene lengths of ORFans or HOPs with duration in the E . coli genome , although this has been observed previously for acquired genes assessed over substantially longer evolutionary timescales 20 ., This indicates that the elapsed time since the divergence of the 13 tested strains from their common ancestor has been insufficient to adjust acquired genes to the nucleotide composition of their host genome ., Recently acquired ORFans and HOPs occur more often in multigene clusters than do those assigned to older branches ., For example , in E . coli CFT073 , which contains the largest numbers of both ORFans and HOPs , about half of the ORFans confined to this strain are adjacent to another ORFan of the same age ., Going back to the next branch that subsumes this strain ( U2 ) , only a third of the ORFans reside next to another ORFan; and among those ORFans originating in the ancestor to the E . coli/Shigella clade , only 14% are situated next to another ORFan ., The average cluster sizes of ORFans along these three branches are 1 . 59 , 1 . 34 , and 1 . 09 genes , indicating that ORFan genes are gained in clusters that subsequently shrink through fragmentation and gene loss ., For the same lineages , a similar trend is observed for HOPs , although it is not as pronounced ( with 1 . 34 , 1 . 34 and 1 . 21 genes per cluster for singleton , U2-specific and ancestral HOPs , respectively ) ., This decrease in gene cluster size is not due to the preferential insertion of new genes near older acquired genes , as we analyzed the cluster sizes of sets of ORFans and HOPs per introgression event ( i . e . , those originating on the same internal branch ) ., A few of the clustered ORFans were located near genes of known phage functions , but a recent exhaustive study into viral ORFans has suggested that phages may play a lesser role in transferring ORFans to prokaryotes than previously thought 28 ., Since the split from their common ancestor , the 13 E . coli and Shigella species have accumulated between 180 and 350 kb of foreign DNA per strain ( Table 1 ) ., Aside from these additions , each of these strains has also lost between 30 and 190 kb of DNA that has been acquired and maintained in other strains ., The two EHEC strains ( E . coli EDL and Sakai ) show the highest net gain of DNA , whereas the Shigella strains , E . coli K-12 and W3110 show the lowest ., To compare the rates at which lineages vary in rates of DNA gain and loss , we calculated the amounts of DNA acquired and lost in relation to the branch lengths in the tree relating the 13 tested genomes ., The rates on individual branches indicate that closely related strains can differ by over two orders of magnitude in the rates at which newly acquired DNA is gained and retained ( E . coli Sakai vs . S . dysenteriae ) but less than 20-fold in the rates at which such DNA is lost ( E . coli EDL933 vs . S . boydii ) ( Supplementary Table S2 ) ., It should be noted that branch lengths can also vary for other reasons ( such as variation in substitution rates and differing rates of recombination ) , but these are most likely compensated due to the extensive gene set employed here ., Gene acquisition rates for both ORFans and HOPs are higher on the internal branches leading to the EHEC and UPEC strains , and in contrast , rates of loss for acquired DNA are highest on all branches descending from the SC ancestor leading to the Shigella species ., Taken together , strains that gain the lowest amounts of DNA , lose the highest amounts of acquired DNA with the result that their genomes have lower numbers of unique genes ., There has been a continual gain and loss of ORFans and HOPs during the evolution and diversification of E . coli ( Supplementary Figures S2 and S3 ) , and based on the distribution of ORFans and HOPs in the 13 tested genomes , HOPs have a higher rate of origination , but ORFans are more likely to be retained ., Since many , possibly most , genes are transient and not present in any contemporary genome , it is not possible to monitor the full complement of genes that are gained and lost in these lineages by comparing their present-day gene repertoires ., However , the patterns of retention of genes assigned to evolutionary lineages of different ages offer a glimpse into the fate of acquired sequences ., Among genes that originated in the ancestor to all 13 strains examined , 78% ( 61% of the ORFans and 95% of the HOPs ) were lost in one or more of the descendant lineages , whereas 96% ( 95% of the ORFans and 97% of the HOPs ) of the genes acquired on the next older branch , SCE , were lost ., Overall , genes acquired on the ancestral branch have higher retention rates than those genes acquired on more recent branches ., Combining the numbers of ORFans and HOPs , Shigella spp ., ( including S . dysenteriae ) show significantly lower retention rates compared to E . coli strains ( 65% vs . 88%; p<0 . 01 ) , which is not surprising since Shigella species have the highest rates of loss of recently acquired genes ., The lower retention rates in Shigella spp ., result from both significantly more gene inactivations ( 11% vs . 5% in E . coli , p<0 . 01 ) and gene losses ( 24% vs . 8% lost in E . coli , p<0 . 01 ) , and though disruptions occur to a similar extent in both sets of acquired genes , HOPs are more frequently lost than retained as inactivated genes ( Table 2 , Supplementary Figure S3 ) ., Also , ORFans are inactivated predominantly by truncations , whereas HOPs are more often disrupted by insertion sequences ( Supplementary Table S3 ) ., Although pseudogenes have been shown to be largely genome-specific 7 , 9 , 16 , it was expected that some would be retained in multiple lineages over the short evolutionary time-span examined in this study ., However , more than half of the inactivated ORFans and HOPs exist only in a single genome , whereas their functional homologs are usually present in several genomes ( data not shown ) ., Similarly , over half of the losses of acquired genes are also genome specific ( i . e . , losses of the only member of a gene family ) , confirming the high turnover rate observed for inactivated DNA ., Gene gain and loss are ongoing processes in microbial genomes , resulting in a diversity in genome sizes , even among closely related strains within a bacterial species 3 , 29 ., By comparing the genome contents of sequenced representatives of the E . coli/Shigella clade , and by mapping the phylogenetic distribution of every gene present in these genomes , we find that the rates of change in novel genes can differ over 200-fold between strains and lineages ., Moreover , genes of different phylogenetic origins arise and persist at very different rates ., For example , ORFan genes , i . e . , those with no homologs outside of the group of bacteria examined , emerge less frequently than do genes originating by acquisition from other bacteria ( termed “HOPs” ) , but are , on average , about eight times more likely to be maintained ., Of the genes acquired on the ancestral branch , nearly 39% of the ORFans , but only 5% of HOPs , are present in all 13 genomes indicating that they now provide functions integral to all strains ., The difference in the persistence of ORFans and HOPs is surprising because those genes acquired from other bacteria ( i . e . , HOPs ) typically encode functional proteins in the donor and could be immediately useful to the recipient , whereas the ORFans , whose origins are less certain , have probably never served a function in a cellular genome prior to their acquisition ., The disparity in the types of properties conferred by these two classes of genes is supported by their assignment to known functional categories: whereas nearly a quarter of the HOPs could be designated a COG category , less than 1% of ORFans could ., Although ORFans are often poorly annotated and resist functional characterization by comparative approaches ( partially due to their characteristically short length and atypical composition ) , several lines of evidence indicate that they encode functional proteins 20 , 30 , including structural in vitro analyses on E . coli ORFans ( unpublished data ) ., Therefore , the retention of ORFans may reside in the fact that they confer truly novel ( but as yet unknown ) functions , as opposed to traits that are apt to be redundant to the recipient organism ., Alternatively , as ORFans are generally thought to be derived from selfish mobile elements ( but see 28 ) , some might be perpetuated by encoding selfish functions themselves ., The distributions of ORFans and HOPs show that sequences that do not provide a useful function are eliminated and that bacterial genomes are not repositories of non-functional genes ., This parallels the situation observed for pseudogenes , which , due to their rapid removal , are largely strain- or genome-specific 7 ., Because the most-recently acquired genes are the least likely to supply an immediately useful function , we might expect that the newest genes in a genome are the most rapidly removed 18 ., Indeed , comparing the two oldest branches indicates that while 33% of the genes gained on the ancestral branch are lost in each extant genome , 42% of the genes gained on a younger branch ( SCE ) are lost ., From the present dataset , it is difficult to assess how this trend continues because relatively few genes are introduced on each branch ( only 9 and 13 genes on SC and S4 respectively ) , and in younger clades , there are successively fewer genomes from which the gene can be eliminated ., However , the low numbers of genes mapped to these internal branches probably reflects the fact that relatively few acquired genes are being maintained ., The density of sequenced genomes has allowed the use of phylogenetic methods to assess the dynamics of gene contents within several bacterial species , and has shown that rates of DNA gain and loss are often strain or lineage specific ., Based on the same genomes analyzed in the present study , Hershberg and co-workers 26 found that the rates of gene loss in Shigella species were consistently higher than in related strains of E . coli , presumably due to reduced selection brought about by their small effective population sizes ., Our data agree with these findings , and additionally , show that Shigella species also have lower rates of gene acquisition and lower rates of retaining acquired genes ., Taken together , the inactivation and subsequent deletion of resident genes coupled with decreased levels of gene acquisition and subsequent persistence accounts for the reduced size of Shigella genomes ., Applying a similar approach , Vernikos and co-workers 31 analyzed the genes acquired by the strains of Salmonella enterica for which genome sequences are available ., In Salmonella , most of the acquired genes have low GC-contents and are still “ameliorating” , i . e . , adjusting their base composition towards that of the host genome 31 , 32 , similar to results observed for acquired sequences in the Gammaproteobacteria as a whole 20 ., That amelioration has been observed in studies on Salmonella and the Gammaproteobacteria , but not in E . coli , is due to the fact that the sequenced strains of E . coli span a much shorter timescale and have not yet accumulated sufficient numbers of mutations to noticeably alter the average base composition of genes ., In addition to assessing genome dynamics by following the presence and absence of acquired genes , we also traced the formation of pseudogenes to more closely monitor the mechanisms by which genes are inactivated and eliminated from these genomes ., Pseudogenes in our analyses have restricted distributions , and nearly half of the inactivated ORFans and HOPs occur in only a single genome ., In that the formation of pseudogenes is an ongoing process , their very restricted distributions denote that inactivated genes are eliminated rapidly from the genome and imply that newly acquired genes that are not immediately functional are also subject to rapid removal ., Although such assessments of gene contents are based only on those genes now present in contemporary genomes , the recognition of pseudogenes can provide additional insights into the evolution and dynamics of genomes ., The inclusion of pseudogenes in the present analysis provides some indication that high numbers of genes are gained and lost without leaving traces of their introgression 7 , 33 ., In conclusion , comparative genomics of multiple closely related strains provides high-resolution assessments and quantifications of gene fluxes in an evolutionary context 31 , 34 , and allows specific estimations of the processes of gene inactivation and deletion ., Within the sequenced strains of E . coli and Shigella spp ., , we detected large differences among closely related lineages in the rates of gene acquisition and loss , but also differences in gene retention rates due to the source of acquired genes ., The higher retention rate observed in the functionally obscure ORFan genes suggests that there are unknown adaptive benefits to these small acquired genes ., To trace the history of each gene in the sequenced E . coli and Shigella genomes , it is first necessary to resolve the phylogenetic relationships among these 13 strains ., We based this phylogeny on the core set of single-copy genes identified by Lerat et al . 35 as showing virtually no evidence of lateral gene transfer within the Gammaproteobacteria ., The seven sequenced E . coli genomes ( E . coli K-12 36 , E . coli W3110 , E . coli Sakai 37 , E . coli EDL933 38 , E . coli CFT073 21 , E . coli UTI89 39 and E . coli 536 ) and six sequenced Shigella genomes ( S . flexneri 301 40 , S . flexneri 2457 , S . flexneri 8401 41 , S . dysenteriae 22 , S . boydii 22 and S . sonnei 22 ) were searched via BLASTP 42 for orthologs of these core genes , applying an E-value<1−10 and a match length >75% ., Of the 203 genes identified by Lerat et al . 35 , 169 single copy genes met these criteria of orthology and were used for phylogenetic reconstruction ., Concatenated sequences of these 169 genes from all 13 E . coli and Shigella genomes were aligned using MAFFT 43 and the alignment was edited to remove gaps using Gblocks 44 ., A maximum likelihood tree ( using DNAML module of PHYLIP; http://evolution . genetics . washington . edu/phylip . html ) was generated using the concatenated orthologous sequences of Salmonella enterica as the outgroup ., The genome sequences of the 367 prokaryotes ( 339 bacteria and 28 archaea ) available at the time of this study were retrieved from GenBank ( ftp . ncbi . nih . gov/genbank/genomes/Bacteria/; August 2006 ) , and an in-house database was created by extracting protein sequences from all but the 13 E . coli and Shigella genomes ., Newly acquired genes can be of two types: ORFans , genes with no detectable homolog in the databases , and HOPs ( heterogeneous occurrence in prokaryotes 20 ) , genes with homologs in distantly related species ., ORFans in each of the 13 E . coli and Shigella genomes were identified as described previously in Daubin and Ochman 20 ., In brief , all protein sequences from these genomes were compared with the database using BLASTP , applying an E-value cutoff of 0 . 01 to uncover distant homologs ., Those genes without a match at this relaxed cutoff were considered to be potential ORFans ., To eliminate possible artifacts due to annotation errors , we queried gammaproteobacterial genomes with all putative ORFans using TBLASTN and excluded those with matches having E-value cutoff<10−5 and alignment lengths >50% ., The distribution of all remaining ORFans among the 13 strains of E . coli and Shigella were obtained by comparing the ORFans from each E . coli and Shigella genome with the remaining 12 genomes using TBLASTN with an E-value cutoff of 10−5 ., Based on their distribution among strains , ORFans were assigned to clades of the E . coli phylogeny ., The orthology of ORFans present in more than one strain was confirmed by genome context ., In contrast to ORFans , HOPs have homologs in other prokaryotic genomes ., To qualify as a HOP , a protein must be restricted to an E . coli clade , absent from closely related genomes , and have a homolog in a more distantly related prokaryotic genome ., We performed BLAST analyses to identify genes that displayed such sporadic distributions ., For example , the 9 HOPs restricted to clade S1 ( Figure 1 ) were present in S . flexneri 301 and S . flexneri 2457 , lacked homologs in the other E . coli and Shigella genomes , but had homologs in some distantly related genomes ., We mapped the branch on which a gene was acquired by reconstructing the parsimonious scenario that explains the present-day gene distribution 18 , such that the path that invokes the lowest number of events was viewed as the most evolutionarily plausible ., In these reconstructions , gene gains and losses were viewed as individual and equally likely events ., The genes acquired on each branch are listed in Supplementary Table S4 ., Classification of the identified ORFans and HOPs to Clusters of Orthologous Groups ( COGs ) 45 was performed using in-house scripts ., Pseudogenes were identified by using Ψ-Φ as described previously 7 , 9 , 16 ., In this procedure , the annotated proteins from each genome were queried against the complete nucleotide sequence of every other strain with E-value cutoffs of 10−15 and sequence identities >75% ., The Ψ-Φ program suite uses the TBLASTN output to return lists of predicted disrupted genes , which are manually curated ., To identify gene-inactivating mutations , the predicted pseudogenes were aligned against their orthologs using CLUSTALW 46 ., Gene-inactivating mutations were grouped into five classes: frameshifts ( insertions or deletions of 1 or 2 nucleotides in length ) , deletions ( >2 nucleotides in length ) , insertions ( >2 nucleotides in length ) , truncations ( large deletions at either or both ends of a coding sequence ) , nonsense mutations , or a combination of different classes .

Introduction, Results, Discussion, Materials and Methods

Bacterial species , and even strains within species , can vary greatly in their gene contents and metabolic capabilities ., We examine the evolution of this diversity by assessing the distribution and ancestry of each gene in 13 sequenced isolates of Escherichia coli and Shigella ., We focus on the emergence and demise of two specific classes of genes , ORFans ( genes with no homologs in present databases ) and HOPs ( genes with distant homologs ) , since these genes , in contrast to most conserved ancestral sequences , are known to be a major source of the novel features in each strain ., We find that the rates of gain and loss of these genes vary greatly among strains as well as through time , and that ORFans and HOPs show very different behavior with respect to their emergence and demise ., Although HOPs , which mostly represent gene acquisitions from other bacteria , originate more frequently , ORFans are much more likely to persist ., This difference suggests that many adaptive traits are conferred by completely novel genes that do not originate in other bacterial genomes ., With respect to the demise of these acquired genes , we find that strains of Shigella lose genes , both by disruption events and by complete removal , at accelerated rates .

Changes in genetic repertoires can alter the adaptive strategy of an organism , especially in bacteria , in which genes are continually gained and lost ., Mapping the gains and losses of genes in the densely sequenced clade of Escherichia coli and Shigella shows that these genomes harbour two types of acquired genes: HOPs , which are those acquired genes with homologs in distantly related bacteria; and ORFans , which are genes without any known homologs ., Surprisingly , the two classes of acquired genes display very different patterns of gain and loss ., HOPs are acquired more frequently , though they rarely persist in the recipient genomes ., In contrast , ORFans are much more likely to be maintained over evolutionary timescales , suggesting that despite their unknown origins , they will more often confer novel and beneficial traits to the recipient genome .

computational biology/comparative sequence analysis, evolutionary biology/bioinformatics, evolutionary biology/evolutionary and comparative genetics

journal.ppat.1002884

Ago HITS-CLIP Expands Understanding of Kaposis Sarcoma-associated Herpesvirus miRNA Function in Primary Effusion Lymphomas

Kaposis sarcoma-associated herpesvirus ( KSHV ) or Human Herpesvirus type 8 ( HHV-8 ) is associated with Kaposis sarcoma ( KS ) and two lymphoproliferative disorders: primary effusion lymphomas ( PEL ) and a subset of multicentricCastlemans disease ( MCD ) 1–3 ., In KS tumors and PEL viral gene expression is highly restricted to the latency-associated region which encodes four proteins and the viral microRNAs ( miRNA ) ., MiRNAs are 21 to 23 nucleotide ( nt ) long , non-coding RNAs that preferentially bind to 3′UTRs of mRNAs to prevent translation and/or induce degradation ( for review see 4 ) ., The first viral miRNAs were identified in 2004 in Epstein-Barr virus ( EBV ) -infected Burkitts lymphoma cells 5 and subsequently more than 140 miRNAs have been identified in all herpes viruses studied thus far with the exception of Varicella Zoster virus ( for review see 6 , 7 ) ., The 12 KSHV miRNA genes 8–11 can each give rise to two different mature products 12 , miR and miR* ., MiR-K12-10 is moreover edited 13 bringing the total number of mature miRNAs to 25 ., KSHV miRNAs are expressed during the latent phase of infection and expression has been detected in tissues and biopsies of classical and AIDS-associated KS as well as in PEL and MCD 14–16 ., Since aberrant expression of miRNAs is associated with many human diseases including cancer 17 , it was hypothesized early on that KSHV-encoded miRNAs may contribute to pathogenesis and/or tumorigenesis by de-regulating host cellular gene expression ., Until recently , only a small number of target genes have been identified mainly by combining bioinformatics predictions with gene expression profiling and 3′UTR luciferase reporter assays in cells that either ectopically express viral miRNAs or in tumor cell lines in which viral miRNAs are inhibited by antagomir approaches 18–21 ., Although limited in number , the initially reported targets immediately suggested that KSHVmiRNAs contribute to the regulation of pathogenesis-relevant processes such as angiogenesis , apoptosis , cell cycle control , endothelial cell differentiation , and immune surveillance ( for review see 6 , 7 ) ., Moreover , one KSHV miRNA , miR-K12-11 , shares the same seed sequence as human miR-155 , one of the first “oncomirs” discovered 22 , 23 ., MiR-K12-11 was shown to mimic miR-155 function to induce a splenic B cell expansion in a NOD/SCID mouse model 24 ., Investigating the combinatorial nature by which viral miRNAs expressed within a background of tissue-specific host miRNAs interact with their cognate transcriptomes requires genome-wide ribonomics-based approaches ., Recently , high-throughput sequencing of RNA isolated by crosslinking immunoprecipitation ( HITS-CLIP ) and Photoactivatable-Ribonucleoside-Enhanced Crosslinking and Immunoprecipitation ( PAR-CLIP ) techniques have been developed that are based on the enrichment of Ago-miRNA-mRNA complexes from cells after UV cross-linking 25 , 26 ., While HITS-CLIP uses 254 nm UV light to cross-link RNA protein complexes , in PAR-CLIP cells are first treated with nucleoside analogs such as 4-thiouridine ( 4-SU ) that are incorporated into nascent mRNAs , which are then cross-linked at 365 nm ., After cross-linking , RNase treatment , and immunoprecipitation , small RNAs representing both miRNAs and their bound targets are extracted and converted into small RNA libraries that are analyzed by high-throughput sequencing ., Very recently , Gottwein et al . reported a list of more than 2000 putative KSHV miRNA targets that were identified by PAR-CLIP in BC-1 and BC-3 cells 27 ., Here we report on a detailed HITS-CLIP analysis of two commonly studied PEL cell lines , BCBL-1 and BC-3 , which are both KSHV-positive but represent different B cell developmental stages 28 , 29 ., We identified 1170 and 950 genes , respectively , that were enriched for clusters of sequence tags containing KSHV miRNA seed sequence matches within 3′ UTRs and exons ., Comparative analysis between both PEL cell lines revealed dramatic differences in Ago-associated miRNA repertoires and in the number and nature of miRNA targets , further supporting the idea that miRNA regulation can be highly cell type-and developmental stage-specific ., In addition , comparison of our HITS-CLIP data with the PAR-CLIP data set reported by Gottwein et al . revealed about 42% overlap , which suggests that neither method enriches miRNA targets in a saturating manner ., In summary , we have identified KSHV miRNA targets highly enriched for the gene ontology terms apoptosis , glycolysis , and lymphocyte activation , which will provide an important resource to further delineate the role of KSHV-encoded miRNAs for viral biology and pathogenesis ., To identify genes that are targeted by KSHV and human miRNAs in latently KSHV-infected cells Ago HITS-CLIP was performed in the KSHV-positive and EBV-negative PEL cell lines BCBL-1 and BC-3 . BCBL-1 cells are post germinal center B cells characterized by rearranged immunoglobulin loci 30 ., BC-3 cells are pre-B cells , which have not undergone antigen-dependent B cell maturation 31 ., As a result , both cell lines harbor significantly different transcriptomes 28 , 29 , 32 ., HITS-CLIP was performed according to Chi et al . with minor changes of the immunoprecipitation ( IP ) and library construction protocols ( for details see Materials and Methods and Text S1 ) ., IP of cross-linked and RNase-treated Ago-miRNA-mRNA-complexes from 1–2×108 cells yielded two complexes migrating approximately at 110 kDa and 130 kDa ( Figure 1A , B ) ., While the smaller complex contained only short 20–25 nt long RNAs ( presumably miRNAs ) , the 130 kDa complex contained two different RNA species: short RNAs ( miRNAs ) and 50–70 nt long RNAs ( presumably target mRNAs ) ( Figure 1C , D ) ., Both short and long RNA species derived from the 130 kDa complex were extracted and processed separately for library construction and deep sequencing ( in the following referred to as miRNA libraries and mRNA libraries , respectively ) ., To account for biological variance as observed in published HITS-CLIP data sets 25 , 33–35 we performed three biological replicates for each cell line ( BR1-3 ) ., As additional quality control one BCBL-1mRNA library ( BR1 ) was sequenced in two technical replicates ( TR1 , 2 ) ., High throughput sequencing of six mRNA and five miRNA ( 2 from BCBL-1 , 3 from BC-3 ) libraries was performed as 40 nts single strand runs and yielded more than 250 million sequence tags ( 16–31 million per run ) ., To validate known and identify potential new host and viral miRNAs , sequence tags from miRNA libraries were aligned to miRBase ( http:/mirbase . org/ , release 17 ) using BLAST , and in addition analyzed using the miRDeep software package 36 ., Nearly 90% of the miRNA library reads originated from human and KSHV miRNAs and comparison across BRs revealed a very high correlation of R2>0 . 92 ( Figure 1E and Figure S1A , and S1B ) ., The comparison between BCBL-1 and BC-3 was lower ( R2\u200a=\u200a0 . 65; Figure S1C ) , indicating significant differences in Ago-associated miRNA profiles betweenBCBL-1 and BC-3 as described in detail below ., Sequencing reads from all mRNA libraries were uploaded to the CLIPZ database , an open source software package specifically developed for the analysis of HITS-CLIP and PAR-CLIP data 37 , and annotated to the human genome ( hg19 ) ., The correlation for technical replicates was R2\u200a=\u200a0 . 88 ( Figure S1D ) ., Observed correlations across biological replicates ( R2\u200a=\u200a0 . 53–0 . 72; Figure S1E , F ) were comparable to previously reported HITS-CLIP data sets 25 , 34 ., mRNA libraries were analyzed for clusters of overlapping reads using the CLIPZ sequence cluster tool ., Of all the clusters aligning to mRNAs about two thirds were located in exons and one third in introns ., Read distribution within exons largely reflected the current understanding of miRNA targeting , as the majority aligned to 3′UTRs and CDS ( Figure 1F ) , and about 4% to 5′UTRs ( 7–8% after adjusting for possible UTR length bias; see Text S1 ) ., Read distribution within exons is also in agreement with recently published Ago HITS-CLIP and PAR-CLIP data sets 25–27 , 34 , 38 , 39 ., With respect to intron/exon distribution no significant differences were observed between mRNA libraries from BCBL-1 and BC-3 cells ( Figure 1F and Figure S2 ) ., A miRNA was counted as present if it was sequenced with at least one read in each BR and the average count over all BRs was at least, 10 . In BCBL-1 , all 25 KSHV miRNAs were recovered , inBC-3 cells all except for miR-K12-9 and -9* , which are highly polymorphic and not expressed 12 , 14 ., However , we note that 7 KSHV miRNAs in both cell lines were detectable at very low read numbers ( below 200 reads/million total reads; Figure 2A ) ., We also detected 370 and 306 human miRNAs in BCBL-1 and BC-3 , respectively ., As observed previously by Chi et al . 25 the 30 most abundantly expressed miRNAs represent 94% of all miRNA reads ( the top 20 contribute 90% ) , suggesting that only a small number of miRNAs act as major players in miRNA-mediated regulation of gene expression ., A comparison of the miRNA libraries showed remarkable differences in the miRNA composition between the two PEL cell lines ., While in BCBL-1 82% of the miRNA reads originate from human miRNAs , in BC-373% are KSHV-derived ( Figure 1G ) ., A more detailed analysis revealed that in BCBL-19 KSHV miRNAs rank within the top 30 , with the most frequent one , miR-K12-4-3p , at position4 ., The three human lymphocyte-specific miRNAs hsa-miR 30a , 30d , and 142-3p occupy more than 50% of all RISCs in BCBL-1 cells ( Figure 2B ) ., In contrast , inBC-3 the top 5 miRNAs ( miR-K12-3 , -1 , -4-3p , -10a , and 10b ) , as well as 15 of the top 30 miRNAs originate from KSHV ( Figure 2C ) , contributing 74 . 5% of the top 30 and 71% of all miRNAs associated with Ago ., At the same time , the read counts of miR-30a , -30d and miR-142-3p are dramatically decreased from 50% of all miRNA reads in BCBL-1 to 12% in BC-3 ., Also , individual viral miRNAs were associated with Ago at highly different frequencies in BCBL-1 and BC-3 cells ., For example miR-K12-3 , the most prevalent miRNA in BC-3 , was 10-fold less abundant in BCBL-1 ( Figure 2A ) ., These results indicate that especially in BC-3 cells the KSHV miRNAs out-compete human miRNAs by displacing them from RISC complexes ., In addition , we note that the pattern of human miRNA abundance differs between PEL cell lines , with some of the most abundant miRNAs in one cell line being found at much lower levels in the other ., These abundantly expressed human miRNAs in BCBL-1 included miR-146 , a major regulator of the inflammatory response 40 , which was detected in BC-3 at almost 30-fold lower read numbers ( data not shown ) ., Vice versa , miR-155 , whose aberrant expression is associated with multiple malignancies 41 , 42 is not expressed in BCBL-1 but within the top 30 in BC-3 ., The differential Ago-association of KSHV and host miRNAs between these PEL cell lines suggests that marked differences may also exist in their respective miRNA targetomes ., For identification of putative miRNA targets , mRNA-derived clusters of overlapping reads were built on the human genome ( hg19 ) within each BR followed by a search for overlapping clusters across BRs ( superclusters ) ., Super clusters were called at two different stringency criteria: clusters present in two of three BRs ( stringency 2of3 ) or in all three BRs ( 3of3 ) ., Super clusters matching these criteria were then scanned for the presence of 7-mer seed matches ( nt 2 to 8 ) of KSHV and the top 30 human miRNAs ., Seed sequences for KSHV miRNAs that were recovered at very low frequencies ( Figure 2A ) were initially included in the search but not considered for the final target lists ( for BCBL-1: miR-K12-6p , 11* , 2* , 5* , 10a* , and 1*; for BC-3 additionally miR-K12-9 , 9*; for exclusion criteria see Text S1 ) ., Seed match-containing clusters were further filtered for alignment to annotated transcripts and sufficient coverage ( for details see Text S1 ) ., Clusters that passed all filtering steps showed a tight width distribution of 41–150 nts ( ∼64% in BCBL-1 and 84% in BC-3 ) , with more than 90% of all clusters being between 41 and 300 nts wide ( Figure 3A and Table S1 ) ., The 41–300 nts wide , seed match-containing clusters , their associated genes , and targeting miRNAs identified at stringencies 2of3 and 3of3 were compiled into putative miRNA target lists for each cell line ., We observed that clusters wider than 300 nts often consisted of overlapping peaks of different sizes , which didnt allow the identification of biologically meaningful seed pairing sites without individual visual inspection of each cluster ., These clusters were therefore not included in the main target lists , but are listed as potential additional targets in separate tables ., Prior to further analysis , we also asked whether target enrichment was correlated to transcript abundance , or biased by 3′UTR length and/or sequence composition ., As expected , Ago HITS-CLIP recovered highly abundant transcripts with higher frequency ( Table S2 , Figure S3A and Text S1 ) ., With respect to 3′UTR length we found a weak bias towards longer 3′UTRs , however , short 3′UTRs ( <300 bases ) were enriched 5-fold higher than expected if enrichment would only be due to 3′UTR length instead of target specificity ( Figure S3B ) ., Finally , targets were enriched for genes with low GC content ( Figure S3C ) , which may reflect less secondary structure and therefore better RISC accessibility ., We note , however , that the overall variation in GC content across transcripts is moderate , with most transcripts being in the range of 35–55% GC ., BCBL-1 data ( 2of3 ) yielded 1516 clusters ( 41–300 nts wide ) corresponding to 1170 transcripts , which contained one or more of the 18 included KSHV miRNA seed matches ( Figure 3B ) ., Using the highest stringency by calling clusters across all three BRs ( 3of3 ) yielded 648 clusters representing 552 transcripts ., Stringency 2of3 inBC-3 yielded 1135 clusters ( 950 transcripts ) targeted by 16 KSHV miRNAs , which was reduced to 470 clusters and 413 transcripts at the highest stringency ( 3of3 ) ., Comparing putative KSHV miRNA targets of both cell lines revealed that 50% or 468 of the transcripts targeted in BC-3 cells ( 2of3 ) were also targeted in BCBL-1 ( Figure 3B ) ., Complete target lists can be found in Tables S3 and S4 ., Remarkably , despite the much larger number of KSHV miRNAs in BC-3 cells , the overall number of KSHV miRNA targets in the two cell lines is not very different and even lower in BC-3 ., Only the percentage of transcripts targeted exclusively by KSHV miRNAs is higher in BC-3 than in BCBL-1 ( 47 vs . 33% of all KSHV miRNA target transcripts , respectively; Figure 3C ) ., Conversely , in congruence with the much higher levels of Ago-associated cellular miRNAs in BCBL-1 , the overall number of transcripts containing human miRNA seed matches ( with or without additional KSHV miRNA seed matches ) as well as the number of transcripts exclusively targeted by host miRNAs was much higher in BCBL-1 than in BC-3 ( Figure 3C ) ., In addition , the overall number of seed match-containing clusters and targets in BC-3 cells is smaller , which may be a result of the reduced miRNA complexity ., These data show that the miRNA targetome in BC-3 cells is dominated by viral miRNAs ., A small proportion of the mRNA library reads , ranging from 0 . 15% to 1 . 05% ( average 0 . 43% ) , originated from KSHV transcripts ., Similar to miRNA expression levels and target numbers , these reads differed between both cell lines ( Figure 4A ) ., Overall , in BCBL-1more viral transcripts were enriched than in BC-3 , which could be a result of the larger number of cellular miRNAs associated with Ago in BCBL-1 , as described above ., A prominent peak was present in both cell lines in the 3′UTR of K2 , the viral interleukin6 ( vIL-6; Figure 4A , B ) , which is expressed in a subset of tumor cells at low levels during latency 43 ., In addition , strongpeaks originated from the K12/Kaposin and ORF71/vFLIP 3′UTRs , and across the vFLIP/vCyclin ( ORF72 ) transcripts ( more prominent in BC-3 than in BCBL-1 ) , as well as minor peaks at the miRNA cluster region and K5 ., Moreover , BCBL-1 showed additional peaks within K4 , T1 . 1/PAN , RTA/ORF50 , ORF58 , and ORF59 ( Figure 4A ) ., Within ORF50 , some peaks were located within the open reading frame as well as downstream; we detected small clusters of reads over one of the potential miR-K12-5 binding sites 44 and over the miR-K12-9* target site 45 in the putative 3′UTR of RTA ., Reads originating from the miRNA cluster likely represent the 1 to 2% miRNA reads recovered from the mRNA target libraries as well as pre-miRNA sequences 39 ., We further validated the prominent peak within the 3′UTR of vIL-6 , which contained a miR-K12-10a seed match by luciferase reporter assay as described below ., While overall the enrichment of viral 7mer2-8 seed match-containing clusters was low across the viral genome , some ORFs and/or putative 3′UTRs contained clusters with host miRNA seed matches ., Tracks showing enriched read clusters on the KSHV genome for all viral and the top 30 human miRNAs are provided in the supporting information ( Dataset S1 , S2 , S3 , S4 ) ., As a first validation of the target data set , we analyzed the read distribution over seed matches of31 experimentally confirmed KSHV miRNA targets reported by multiple groups 18 , 19 , 21–24 , 44 , 46–52 ., From these , 16 were enriched by Ago HITS-CLIP and all but two showed enriched read clusters harboring the experimentally confirmed seed match ( Table S5A ) ., Some transcripts contained additional clusters with seed matches for other viral miRNAs ., Figure S4 shows the read distribution of eight previously characterized targets visualized as wiggle plots in the UCSC genome browser ., The target interactions of miR-K12-11 , an ortholog of the oncomir miR-155 22 , 23 , with BTB and CNC homology 1 , basic leucine zipper transcription factor 1 ( BACH1 ) , Src-like-adaptor ( SLA ) , FBJ murine osteosarcoma viral oncogene homolog ( FOS ) , and CCAAT/enhancer binding protein beta ( C/EBPβ ) have been confirmed by 3′UTR mutagenesis 22–24 ., The 3′UTR of BACH1 contains four , FOS and SLA each contain two , and C/EBPβ one seed match for miR-K12-11 ., Three of the BACH1 sites previously demonstrated to be important for miR-K12-11 targeting were indeed enriched by HITS-CLIP; for the other three transcripts all miR-K12-11 seed match sites were occupied by clusters , although sometimes only in one biological replicate ., Further comparison of recovered miR-K12-11 targets with a list of 151 putative miR-155 targets reported by multiple groups 22 , 23 , 53–61 revealed 30 commonly targeted transcripts ( Table S5B ) ., Dolken et al . reported on114putative KSHV miRNA targets that were enriched using immunoprecipitation in the absence of cross-linking ( RIP-CHIP ) 46 ., Of these , 33 overlap with our data set including NHP2 non-histone chromosome protein 2-like 1 ( NHP2L1 ) and leucine rich repeat containing 8 family , member D ( LRRC8D ) , which both recovered high frequency clusters for the validated miR-K12-3 target sites ( Figure S4 and Tables S5A , S6A ) ., Also , Thrombospondin1 ( THBS1 ) was previously shown to be targeted by multiple KSHV miRNAs 19 ., Correspondingly , the HITS-CLIP data revealed seed match-containing clusters for miR-K12-1 , -3 , -3* , -6-3p , and, -11 . We note that all of the previously reported target sites for THBS1 and LRRC8D consist of a 6-mer seed match and are therefore not included in the overall target lists ( Table S3 and S4 ) , but could be confirmed by manual investigation of the seed match sites ( Figure S4 ) ., Figure 5 shows the read distribution for eight newly identified targets: Annexin A2 ( ANXA2 ) , CCAAT/enhancer binding protein alpha ( C/EBPα ) , major histocompatibility complex , class I , C ( HLA-C ) , protein tyrosine phosphatase , non-receptor type 11 ( PTPN11 ) , stress-induced-phosphoprotein 1 ( STIP1 ) , tumor protein p53 inducible nuclear protein 1 ( TP53INP1 ) , tumor protein D52 ( TPD52 ) , and tyrosine 3-monooxygenase/tryptophan 5-monooxygenase activation protein , epsilon polypeptide ( YWHAE ) , and their corresponding KSHV miRNAs ., Both targeting and miRNA-specificity for these transcripts were further validated by 3′UTR luciferase assays ( see below ) ., This initial target validation demonstrates that our experimental HITS-CLIP conditions in combination with stringent filtering of clusters across biological replicates yielded a reliable working list of putative targets for KSHV miRNAs in BCBL-1 and BC-3 cells ., Very recently , Gottwein et al . reported more than 2000 putative KSHV miRNA targets that were identified using PAR-CLIP in BC-1 and BC-3 cells 27 ., We found that 830 or 42% of the putative targets identified by PAR-CLIP in BC-3 were also enriched by Ago HITS-CLIP in at least one biological replicate from BCBL-1 and/orBC-3 ( Table S6B ) ., Potential new KSHV miRNA targets were first aligned to their corresponding miRNA using RNA hybrid ( Figure S5 ) ., We then cloned eight 3′UTRs and four enriched seed match-containing CDS downstream of a luciferase reporter cassette and performed miRNA sensor assays in HEK293 cells ., For each transcript , we tested the predominantly identified miRNA or miRNA combinations and for some additionally the miRNA cluster , which contains 10 of the 12 miRNA genes as previously reported 19 , 23 ., As positive controls , we used the known miR-K12-11 targets BACH1 and C/EBPβ 23 , 24 ., All eight 3′UTRs responded to miRNA expression with a dose-dependent decrease of luciferase expression by at least 20% ( Figure 6A ) ., These included ANXA2 , C/EBPα , HLA-C , high mobility group AT-hook 1 ( HMGA1 ) , interferon regulatory factor 2 binding protein 2 ( IRF2BP2 ) , TP53INP1 , TPD52 , and YWHAE ., We moreover introduced three point mutations in the miR-K12-11 seed match sites in the 3′UTRs of ANXA2 and YWHAE ( Figure S5 ) ., This resulted in a de-repression of both luciferase reporter constructs , thus further confirming the functionality of these target sites ( Figure 6B ) ., Finally , we showed by Western blot analysis a decrease of the TP53INP1 and YWHAE protein levels in the presence of miR-K12-11 ( Figure 7 ) ., Of the four transcripts enriched for CDS seed matches , PTPN11 and STIP1 responded to miRNA expression while HLA-E , and complement component 1 , q subcomponent binding protein ( C1QBP ) did not ., This is in congruence with the literature reporting that miRNA target sites located within exons are less often functionally active 62–65 ., Overall , 10 out of 12 putative targets were functionally confirmed ., In addition , we tested the 3′UTR of vIL-6 , which revealed a strong cluster that contained a 6-mer miR-K12-10a seed match ., The vIL-6 reporter was inhibited in a dose-dependent manner up to40% in the presence of miR-K12-10 ( Figure 6a ) ., This effect was abolished by the introduction of two different miR-K12-10a seed match mutations ( Figure 6B , S5 ) ., These data functionally confirm the first KSHV latency-associated gene to be modulated by a viral miRNA ., Other putative viral targets including RTA , vFLIP , vCyclin and Kaposin are currently under investigation ., All HITS-CLIP-derived KSHV miRNA targets found at analysis stringency 3of3 were subjected to Gene Ontology ( GO ) analysis using DAVID 66 , 67 ., GO analysis was performed against two different backgrounds:, ( i ) the published BCBL-1 and BC-3 transcriptomes 32 and, ( ii ) all human transcripts ., Pathway enrichment was analyzed for five target subsets: all targets enriched in each cell line , cell line-specific targets , and overlapping targets between cell lines ., A partial representation of enriched GO terms is shown in Table 1 , the detailed GO analysis in Table S7 ., Genes involved in highly regulated processes often have long 3′UTRs and thus potentially contain more miRNA target sites ., We therefore tested if GO terms were identified due to a bias for 3′UTR length rather than a functional enrichment ., However , GO terms for three highly regulated processes ( apoptosis , cell cycle , and glycolysis ) showed only very moderate association with intermediate 3′UTR length and no association with long 3′UTRs ( Figure S3D ) ., In both cell lines Ago HITS-CLIP significantly enriched for KSHV miRNA targets involved in different pathways regulating apoptosis ., Among the more than 40 genes were the tumor necrosis factor receptor superfamily member 10b ( TNFRSF10B , miR-K12-1 , -3 ) and the TP53 apoptosis effector PERP ( miR-K12-3; p53 pathway ) , FEM1B ( miR-K12-4-3p; Fas/TNFR1 signaling ) , and Transforming Growth Factor beta Receptors ( TGFBR ) 1 ( miR-K12-2 ) and 3 ( miR-K12-4-3p ) ( TGFBR pathway ) ., The latter two proteins together with growth factor receptor-bound protein 2 ( GRB2 , miR-K12-4-3p ) also signal in the pro-apoptotic P70S6K pathway ., Moreover , we identified the tumor suppressor phosphatase and tensin homolog ( PTEN , miR-K12-4-3p , -7 ) , a negative regulator of the anti-apoptotic Akt/PKB ., Finally , we recovered several known apoptosis targets: cyclin-dependent kinase inhibitor 1A and 1B ( CDKN1A ( miR-K12-11 ) , 1B ( miR-K12-K12 ) 22 , 27 , which are also involved in cell cycle control , Caspase 3 ( CASP3 ) , which was recently reported as miR-K12-1 , -3 , and -4-3p target 52 , and BCL2-associated transcription factor 1 ( BCLAF1 , miR-K12-2 ) , which appears to have dual roles in PEL cells ., While it was originally characterized as pro-apoptotic factor 68 , 69 , Ziegelbauer et al . found that in KSHV-infected cells BCLAF1 impairs apoptosis and also regulates lytic viral replication by sensitizing latent cells to reactivation stimuli 18 ., The most enriched GO term in both cell lines was glycolysis ( 11 genes , p<4×106 ) ., Recently , it was shown that KSHV infection of endothelial cells induces the Warburg effect during latency 70 , which is observed in many human tumors and results in increased aerobic glycolysis and decreased oxidative phosphorylation 71 ., Interestingly , initial experiments showed that latent KSHV infection of SLK cells leads to increased oxygen consumption ( data not shown ) ., However , testing 293 cells engineered to express the KSHV miRNA cluster containing 10 of the 12 miRNAs 19 failed to show a similar effect ., Additional studies in SLK and primary endothelial cells are currently ongoing ., In BCBL-1 , Ago HITS-CLIP enriched for two inhibitors of the NFκB pathway , lectingalactoside-binding soluble 1 ( LGALS1 , miR-K12-10b ) 72 and Interleukin 10 ( IL-10 , miR-K12-12* ) 73 ., This suggests that , in addition to positively regulating NFκB via the latency-associated vFLIP 74 , KSHV further reinforces this crucial pathway for PEL cell survival by miRNA regulation ., IL-10was also part of the GO term lymphocyte activation , which was highly enriched in BCBL-1-specific targets ., We recovered 13 genes of this pathway including growth and/or differentiation factors such as inosine 5′-monophosphate dehydrogenase 1 ( IMPDH1 , miR-K12-7 ) , early growth response 1 ( EGR1 , miR-K12-4-3p ) , CD48 ( miR-K12-7* ) , and bone marrow stromal cell antigen 2 ( BST2 , miR-K12-8* ) ., In BC-3 cells , putative KSHV miRNA targets were enriched for factors that inhibit cell proliferation and G1 to S phase transition via multiple pathways ., Among them were four inhibitors of Cyclin-dependent Kinase 2 , including the previously characterized targets CDKN1A ( p21; miR-K12-1 ) 22 , CDKN1B ( p27; miR-K12-11 ) and Protein Phosphatase 2A ( PP2A; miR-K12-1 ) 27 , as well as the WEE1 homolog ( WEE1; miR-K12-1 , -12 ) ., PP2A in addition causes G1 arrest via the BTG protein pathway ., Moreover , p21 and p27 together with WEE1 block the progression of the cell cycle via E2F activation ., We note that p21 and p27 expression is also regulated by vCyclin 75 , 76 ., The downregulation of these transcripts by KSHV miRNAs suggests a release of the cell cycle arrest and increased proliferation ., It was previously reported that KSHV miR-K12-7 targets MHC class I polypeptide-related sequence B ( MICB ) 47 , which is also targeted by HCMV and EBV miRNAs 47 , 77 ., HITS-CLIP did not enrich for MICB , which might be expressed at levels too low to detect in PEL cells ., However , HITS-CLIP enriched for KSHV miRNA targets involved in antigen presentation in the context of cellular immunity , i . e . the Major Histocompatibility ( MHC ) class I alpha-chain genes ( HLA-C , -E , -F , and –G ) , and also genes involved in the process of loading and transporting MHC , i . e . calreticulin ( CALR , miR-K12-4-3p , -10b ) and adaptor-related protein complex 3 , delta 1 ( AP3D1 , miR-K12-3 ) ., We note that KSHV additionally encodes two E3 ubiquitin ligases , ORF K3 ( MIR1 ) and OFR K5 ( MIR2 ) that potently downregulate MHC I on the surface of infected cells 78 , 79 , suggesting another concerted action of KSHV miRNAs and proteins ., Several members of the ubiquitin conjugating ( TMEM189-UBE2V1 , miR-K12-1;UBE2V2 , miR-K12-11; UBE2L3 , miR-K12-1; and UBE2D3 , miR-K12-2 , -4-3p ) and ubiquitin ligase ( UBE3C , miR-K12-3 ) families were enriched in both cell lines , while negative regulators of kinases ( e . g . CDKN1A , CDKN1B , and PP2A ) were BC-3-specific ( Table 1 , Table S7 ) ., Potential signaling pathways modulated by these kinases and the question whether ubiquitin-dependent protein turnover is modulated by KSHV miRNAs needs to be experimentally addressed ., The best characterized KSHV miRNA targets so far are mostly involved in regulating immune evasion ( MICB ) , pro-apoptotic pathways ( BCLAF1 ) , and cell cycle control ( BACH1 , FOS , THBS1 , CDKN1A , and C/EBPβ ) ; for review see 7 , 81 ., The Ago HITS-CLIP-derived targetome shows strong enrichment for genes involved in these pathways , thus significantly expanding what to this point was solely based on single target gene studies ., In addition , GO analysis suggests new host cell pathways to be targeted , such as glycolysis , lymphocyte activation and the ubiquitin/proteasome pathway , opening up additional interesting themes for functional studies ., Finally , one clearly emerging concept from this HITS-CLIP data set is that multiple key pathways and processes such as the NFκB pathway , MHC class I-mediated immune surveillance , and cell cycle control can be co-regulated by both virally encoded proteins and miRNAs ., MiRNA library analysis revealed strong differences in Ago-associated miRNAs in BCBL-1 and BC-3 cells , with KSHV miRNAs comprising 18% of all miRNA reads in BCBL-1 , and an astonishing 73% in BC-3 , and numbers of single KSHV miRNAs being up to 10-fold higher in BC-3 ., Similar results for the overall KSHV versus human miRNA count in both cell lines were obtained by the recent PAR-CLIP study 27 ) ., Interestingly , several studies have analyzed KSHV miRNA expression in additional PEL cell lines and found differences not only with respect to overall expression levels but moreover also differences in the relative abundance of specific viral miRNAs 15 , 82 ., The fact that such expression differences likely affect targeting further supports the notion that miRNA targetomes are strictly context dependent ., Surprisingly , despite the much higher levels of KSHV miRNAs in BC-3 cells compared to BCBL-1 , we identified similar KSHV miRNA target numbers in both cell lines , which were even 15–20% lower in BC-3 ., Only the number of transcripts exclusively targeted by KSHV miRNAs was slightly higher in BC-3 ( Figure 3C ) ., In contrast , we found that the number of genes targeted by human miRNAs ( either exclusively or with additional KSHV sites ) , was almost 2-foldhigher in BCBL-1 than in BC-3 ., Thu

Introduction, Results, Discussion, Materials and Methods

KSHV is the etiological agent of Kaposis sarcoma ( KS ) , primary effusion lymphoma ( PEL ) , and a subset of multicentricCastlemans disease ( MCD ) ., The fact that KSHV-encoded miRNAs are readily detectable in all KSHV-associated tumors suggests a potential role in viral pathogenesis and tumorigenesis ., MiRNA-mediated regulation of gene expression is a complex network with each miRNA having many potential targets , and to date only few KSHV miRNA targets have been experimentally determined ., A detailed understanding of KSHV miRNA functions requires high-through putribonomics to globally analyze putative miRNA targets in a cell type-specific manner ., We performed Ago HITS-CLIP to identify viral and cellular miRNAs and their cognate targets in two latently KSHV-infected PEL cell lines ., Ago HITS-CLIP recovered 1170 and 950 cellular KSHVmiRNA targets from BCBL-1 and BC-3 , respectively ., Importantly , enriched clusters contained KSHV miRNA seed matches in the 3′UTRs of numerous well characterized targets , among them THBS1 , BACH1 , and C/EBPβ ., KSHV miRNA targets were strongly enriched for genes involved in multiple pathways central for KSHV biology , such as apoptosis , cell cycle regulation , lymphocyte proliferation , and immune evasion , thus further supporting a role in KSHV pathogenesis and potentially tumorigenesis ., A limited number of viral transcripts were also enriched by HITS-CLIP including vIL-6 expressed only in a subset of PEL cells during latency ., Interestingly , Ago HITS-CLIP revealed extremely high levels of Ago-associated KSHV miRNAs especially in BC-3 cells where more than 70% of all miRNAs are of viral origin ., This suggests that in addition to seed match-specific targeting of cellular genes , KSHV miRNAs may also function by hijacking RISCs , thereby contributing to a global de-repression of cellular gene expression due to the loss of regulation by human miRNAs ., In summary , we provide an extensive list of cellular and viral miRNA targets representing an important resource to decipher KSHV miRNA function .

Kaposis sarcoma-associated herpesvirus is the etiological agent of KS and two lymphoproliferative diseases: multicentricCastlemans disease and primary effusion lymphomas ( PEL ) ., KSHV tumors are the most prevalent AIDS malignancies and within Sub-Saharan Africa KS is the most common cancer in males , both in the presence and absence of HIV infection ., KSHV encodes 12 miRNA genes whose function is largely unknown ., Viral miRNAs are incorporated into RISCs , which regulate gene expression mostly by binding to 3′UTRs of mRNAs to inhibit their translation and/or induce degradation ., The small subset of viral miRNA targets identified to date suggests that these small posttranscriptional regulators target important cellular pathways involved in pathogenesis and tumorgenesis ., Using Ago HITS-CLIP , a technique which combines UV cross-linking , immunoprecipitation of Ago-miRNA-mRNA complexes , and high throughput sequencing , we performed a detailed analysis of the KSHV miRNA targetome in two commonly studied PEL cell lines , BCBL-1 and BC-3 and identified 1170 and 950 putative miRNA targets , respectively ., This data set provides a valuable resource to decipher how KSHV miRNAs contribute to viral biology and pathogenesis .

biochemistry, rna, nucleic acids, viral immune evasion, virology, viral persistence and latency, viruses and cancer, biology, rna stability, microbiology

journal.pcbi.1007106

Modeling the daily rhythm of human pain processing in the dorsal horn

The processing of pain engages a wide variety of neural circuits across the nervous system including those in the spinal cord , brainstem , thalamus , and cortex ., More specifically , it is thought that the dorsal horn ( DH ) , an area of the spinal cord , serves as the initial processing center for incoming nociceptive , or painful signals , with the midbrain and cortex providing top-down modulation to that circuitry 1 ., As a result , there is a tradition of modeling pain processing by focusing exclusively on spinal cord circuitry ., This circuitry receives information about stimulation of peripheral tissues from several types of primary afferent nerve fibers ., These afferents have their cell bodies in the dorsal root ganglia ( DRG ) , a cluster of nerve cell bodies located exterior to the spinal cord , and their axons ( or fibers ) target the DH 2 ., Responses to innocuous stimulation are carried by rapidly conducting Aβ-fibers 3 , whereas nociceptors ( i . e . , nerve fibers that detect painful stimuli ) are only activated when a stimulus exceeds a specific threshold ., There are two major classes of nociceptive fibers: fast conducting Aδ-fibers that mediate localized , fast pain and small-diameter C-fibers that mediate diffused , slow pain ., Among the neuronal populations in the DH , the projection neurons ( PNs ) receive input from all fibers and constitute the majority of the output from the dorsal horn circuit up to the brain ., In this article , we introduce a biophysically-based , mathematical model of the nociception-processing neural circuit in the DH , which expands on our earlier work 4 ., We are particularly interested in using the model to investigate mechanisms for daily ( i . e . , diurnal ) modulation of pain sensitivity ., In many clinical conditions , pain sensitivity follows a daily cycle 5 , that is , it exhibits a trough in the late afternoon and a peak sometime after midnight for humans 6 , but it is currently unclear how much of that rhythmicity is derived from daily fluctuation in the underlying causes of the pain versus rhythmicity in the neural processing of pain ., Within the experimental pain literature , rhythmic influences on pain sensation occur regardless of whether pain responses are measured subjectively or objectively 7–10 , suggesting that the rhythmic modulation of pain responses occurs at the level of basic nociceptive processing ., This rhythmic modulation of pain sensitivity also increases with pain intensity 9 , 11 , 12 ., Furthermore , rhythmic influences on pain sensitivity are detectable in experiments involving a variety of different kinds of nociceptive stimuli , including cold , heat , electric current , pressure , and ischemia ( see Tables 1–2 in 6 ) ., Interestingly , experimental studies have also shown daily rhythmicity in tactile discrimination in nearly opposite phase to pain sensitivity , namely highest tactile sensitivity occurring in the late afternoon and lowest in the morning 13 ., There are several hypotheses for the source of the daily rhythm in pain sensitivity , including central nervous system , spinal , and peripheral mechanisms 5 , 14–18 ., Recent studies show that cells in the DRG rhythmically express the primary genes responsible for generating an intrinsic 24-hour , or circadian , rhythmicity of other physiological processes , including Bmal1 , Clock , Per1 and Per2 15 , 16 ., In addition , the rhythm in behavioral nociception followed the gene expression rhythm 15 and disruption of their expression affected behavioral pain responses 16 ., These findings motivate our use of a spinal cord model to test questions regarding daily influences on pain processing ., As such , the model assumes that the daily modulation occurs at the level of primary afferent input to the spinal cord circuitry ., Additionally , we specifically model the portion of experienced pain that arises from nociceptive input to the spinal cord and ignore any potential sources of top-down modulation ., As concerns the connections between neuron populations in the DH , there are several proposed circuitries for the processing of touch , nociception , and itch ( see , e . g . , 19 , 20 ) ., In this work , we take an approach similar to previous models of spinal cord nociception processing ( e . g . , 21 ) and employ the network architecture in the DH proposed by the gate control theory of pain 22 ., In doing so ( and when we introduce daily modulation ) , we note that the aim of our work here is focused on the processing of painful , noxious stimulation , not mechanical , non-noxious stimulation , which we acknowledge may have a different circuitry ( for a review of circuitries for mechanical pain and itch , see 23 ) ., The gate control theory of pain 22 , 24 posits that the neural circuitry in the DH exhibits a gating mechanism that is modulated by activity in the Aβ- and C-fibers 25 ., Specifically , nociceptive C-fiber-facilitated activity in the DH circuit is inhibited by Aβ-fiber activity ., When the amount of painful stimuli ( i . e . , activity in the C-fibers ) outweighs the inhibition from the Aβ-fibers , the “gate opens” and activates the PNs ( and the experience of pain ) ., Although the gate control theory of pain 22 is a simplification and not a complete representation of the physiological underpinnings of pain processing 25 , it has been a productive starting point for several mathematical and computational models of pain 21 , 26–28 ., For our model of the DH circuit , we implement a neuronal population firing-rate model formalism 29 , 30 to describe the population activity of projection , inhibitory , and excitatory neurons in the DH ., Our choice of this commonly-used model formalism is based on the large number of afferent fibers and neurons in the DH , and the assumption that the majority of information flow in the DH circuit is through firing rates of neural populations rather than in specific spike timing within the populations 30 , 31 ., An advantage of this formalism is its biophysical basis and relative simplicity , thus making our model an accessible theoretical framework for experimental and clinical investigations of diverse physiological processes modulating pain processing in humans ., The rest of the paper is organized as follows ., In the Methods section , we formulate the equation system of the neural circuit for pain processing in the DH , describing the time evolution of the average firing rates of the excitatory and inhibitory interneuron , and PN populations in response to input on the afferent nerve fibers ., The model includes NMDA-mediated synaptic input from the C-fibers to the PNs that depends on postsynaptic activity ., We also describe the use of a Poisson process to simulate neural spikes on the afferent fibers that represent the input from the DRG to the DH and the interactions between the afferent fibers incorporated in our model , respectively ., In the Results section , we present validation studies for our model including reproduction of the wind-up phenomenon ., With our principal aim to investigate the daily rhythmicity of pain sensation , we apply the model to predict the daily modulation of the pain inhibition phenomena ., As a novel application of the model , we investigate effects of experimentally-observed dysregulation of inhibition within the DH circuit under neuropathic pain conditions ( i . e . , a chronic condition with persistent pain experience associated , e . g . , with peripheral nerve damage ) on the daily modulation of pain sensitivity ., We find that dysregulation of Aβ-fiber dependent presynaptic inhibition of C-fiber signaling can account for it ., Finally , we discuss limitations and future modifications , as well as importance and application , of our model in the Discussion ., We construct a model describing the spinal processing of nociceptive stimuli in humans by considering the average firing rate of three populations of neurons in the DH: the PNs ( P ) , inhibitory ( I ) interneurons , and excitatory ( E ) interneurons , in response to the average firing rate of the Aβ- , Aδ- , and C-afferent fibers ( see Fig 1 ) ., In this work , we expand on a model developed in 4 , which follows the modeling approach similar to 26 with the exception that our model predictions are in terms of average firing rates of neuron populations 29 instead of average membrane potentials ., In contrast to our previous model in 4 , the new elements of the model introduced in this work consist of including, i ) Poisson processes to generate spiking activity on the input nerve fibers ,, ii ) NMDA receptor-mediated synaptic interactions and, iii ) an additional inhibitory interneuron population I2 , and, iv ) removal of the connection to the midbrain ., These four modifications allow us to, i ) represent a biologically realistic fiber input to which the model is robust ,, ii ) reproduce experimentally observed frequency effects during wind-up ,, iii ) expand the model parameter range that replicates patterns seen in experiments on neuropathy , and vi ) focus solely on modeling spinal-cord processing of pain , respectively ., As concerns the general structure of connections between the neuronal populations in the circuit ( dashed rectangle in Fig 1 ) , we follow previous models of pain and use the circuitry presented , e . g . , in 21 ., Briefly , PNs receive direct synaptic input from the three afferent fiber types , Aβ-fibers excite inhibitory interneurons and C-fibers excite excitatory interneurons ., Both interneuron populations synapse onto the PNs and the inhibitory interneurons inhibit the excitatory interneurons ., We also include Aβ-dependent presynaptic inhibition of C-fiber activity mediated through an additional inhibitory interneuron population ( I2 ) that is modeled indirectly see Eq ( 5 ) ., We assume that the input to our model circuit is a stimulation of the afferent fibers that has been pre-processed in the DRG ., Based on fiber input and the connections between the neuron populations in the DH , our model computes the activity of the PNs , ( P in Fig 1 ) , whose output directly corresponds to the amount of pain experienced 32 ., We note that there are many nuances in the perception of pain , including those originating in the cortex; however , we model the portion of pain that stems from nociceptive input to the spinal cord since it has been shown that pain perception correlates strongly with the firing rate of the PNs in the spinal cord 32 , 33 ., According to the formalism of firing-rate models , e . g . , 29 , we assume that the rate of change of the average firing rate in spikes per second ( Hz ) of the projection , inhibitory , and excitatory neuron populations , fP , fI , and fE , respectively , is determined by a nonlinear response function ( see Fig 2A ) ., These response functions determine the average firing rate response of a neuron population to a combination of external inputs ( i . e . , stimulations of the afferent fibers pre-processed in the DRG ) and the firing rates of the presynaptic neuron populations ( see Fig 1 ) ., In the absence of input from other neuron populations and afferent fibers , the average firing rate of the neuron population decays exponentially ., These assumptions yield the following set of equations for the average firing rate of each population:, d f P d t = P ∞ ( g A β P f A β ( t ) + g A δ P f A δ ( t ) + ( g CP + g NMDA ) f C ( t ) + g EP f E - g IP f I ) - f P τ P , d f E d t = E ∞ ( g CE f C ( t ) - g IE f I ) - f E τ E , d f I d t = I ∞ g A β I f A β ( t ) - f I τ I , ( 1 ), where t is time in seconds , τP = 0 . 001 s , τE = 0 . 01 s , and τI = 0 . 02 s are the intrinsic time scales of the projection , excitatory , and inhibitory neuron populations , respectively ., Weights gij denote the strength of the external input or connections from presynaptic neuron populations i ( i = Aβ , Aδ , C , P , E , I ) to neuron population j ( j = P , E , I ) ., We indicate inhibitory synaptic input with a negative sign and excitatory synaptic input with a positive sign ., We define the functions ( of time t ) for the external inputs , fAβ ( t ) , fAδ ( t ) , and fC ( t ) in the next section ., The model includes N-methyl-D-aspartate ( NMDA ) type synapses from the C-fibers to the P population in the following way: to represent postsynaptic voltage-dependent removal of the magnesium ( Mg+ ) block on NMDA receptors , we assume that the synaptic weight , gNMDA , depends on the average firing rate of the P population ( fP ) , and thus , we consider gNMDA as a variable that changes as a function of time ( see Fig 2B ) , similar to 34:, d g NMDA d t = M ∞ ( f P ) - g NMDA τ NMDA , ( 2 ), where τNMDA = 1 s is the intrinsic time scale of the synaptic weight , gNMDA ., We assume a sigmoidal shape for the monotonically increasing firing rate response functions of the neuronal populations P∞ , E∞ , I∞ , and the synaptic weight response function M∞ , and use hyperbolic tangent functions to represent them as follows:, P ∞ ( x ) = max P 1 2 1 + tanh ( 1 α P ( x - β P ) ) , E ∞ ( x ) = max E 1 2 1 + tanh ( 1 α E ( x - β E ) ) , I ∞ ( x ) = max I 1 2 1 + tanh ( 1 α I ( x - β I ) ) , M ∞ ( x ) = max M 1 2 1 + tanh ( 1 α M ( x - β M ) ) , ( 3 ), where maxP , maxE , maxI , and maxM are the maximum firing rates of the projection , excitatory , and inhibitory populations , and the maximum synaptic strength of the NMDA-mediated input , respectively ., In Eq ( 3 ) , the shape of the response functions is determined by the input x at which the average firing rate of the projection , excitatory , and inhibitory neuron population reaches half of its maximum value , x = βP , x = βE , and x = βI , respectively ( see Fig 2A ) ., The slope of the transition from non-firing to firing in the projection , excitatory , and inhibitory neuron population is given by 1/αP , 1/αE , and 1/αI , respectively ., The activation of the NMDA synapse , M∞ ( fP ) , is modeled as an increasing function of the firing rate of the projection neurons , representing the resulting increase in synaptic strength as postsynaptic membrane potentials depolarize and the magnesium block of the NMDA receptors is released 34 ., We choose parameter values for the response functions in such a way that the input-output curve of the projection , excitatory , and inhibitory neuron populations agrees qualitatively with experimental observations ., Hence , we assume the inhibitory interneuron population has a nonzero resting firing rate , as has been reported in 1 , 2 , and a higher maximum firing rate than that of the projection and excitatory interneuron populations , as has been assumed in a biophysically detailed model of the DH circuit 21 ., In our model assumptions for the response functions , we mimic the model predictions of 21 that agree with data from experimental observations in 35 , 36 ., As concerns the NMDA activation , we assume a similar sigmoidal shape but with a very slow rise time modeling the slow removal of magnesium ions from blocking the NMDA receptors with increase in cell activity ., As the magnesium blockage is removed , the NMDA channels are clear to be activated and further depolarize the cell , resulting in an increase in the firing rate of the PNs ., The function M∞ ( fP ) models this activation of the NMDA channels resulting from the removal of the magnesium ions ( see Fig 2B ) ., All values of the parameters discussed above that we use in the numerical simulations of our model are listed in S1 Table ., To model input from the DRG , we simulate 1000 afferent fibers of three types that project to the DH ., We note that our choice of 1000 fibers is based on the number of afferent fibers experimentally observed in one nerve bundle that projects to a skeletal muscle in the rat 37 , which is on the order of 1000 37 , 38 ., These three afferent fiber types differ not only in diameter sizes but also in the level of myelination ., As a result , impulses are transmitted at different speeds in the three fiber types ., The majority ( 82% ) of these fibers are slow C-fibers ( with an average conduction velocity of 0 . 5-2 m/s ) , 9% are Aδ-fibers ( with an average conduction velocity of 5-30 m/s ) , and 9% are Aβ-fibers ( with an average conduction velocity of 30-70 m/s ) 3 , 38 ., We assume that the times of initiation of activity in each of these fibers in response to nociceptive stimulation are roughly equivalent , resulting in the distribution of arrival times to the DH that has been experimentally observed , e . g . , in Fig 1 of 39 ., We aim to model nerve fiber activity from a brief nociceptive stimulus at the periphery ( see Fig 1a in 39 ) ., To do this , we use a Poisson process to simulate spike trains in the afferent fibers at a given firing rate ., The activity of the afferent fibers in response to a brief nociceptive stimulus at t = 0 . 5 s can be seen in the raster plots in Fig 3A , where each small bar represents one spike/action potential and each row represents the activity in one afferent fiber over the course of 1 second ., We consider the activity in 90 Aβ- , 90 Aδ- and 820 C-fibers with baseline frequency of 1 Hz and stimulus response frequency of 40 , 20 , and 20 Hz , respectively ., Each fiber has an increased firing rate for a set amount of time ( 10 ms for both Aβ- and Aδ-fibers and 210 ms for C-fibers ) chosen to replicate the response in the PNs as measured experimentally in 39 ., We choose these increased firing rates for the afferent fibers to simulate a response to a nociceptive stimulus ( see 33 and 40 for spiking dynamics of afferent fibers in response to varying levels of nociceptive stimuli ) and a low background drive to simulate spontaneous activity of the fibers 41 ., To compute the average firing rate in each of the three fiber groups , we compute an instantaneous firing rate by counting the number of spikes in a one-millisecond window of time , and then use a moving average with a time window of 10 ms to create a smooth firing rate function ., As a result , our simulated input to the spinal cord on fibers with different conductance speeds reproduces the observed pattern 39 of fast , brief Aβ- and Aδ-fiber activity ( i . e . , first pain ) followed by delayed , longer lasting C-fiber activity ( i . e . , second pain ) ., When simulating our model , we use these smoothed average firing rates ( see Fig 3B ) representing the response in the three fiber groups to a brief nociceptive stimulus as input to the DH circuit model ., Pain sensitivity follows a daily cycle in many clinical conditions 5 ., There is strong evidence supporting rhythmicity in response to acute nociceptive stimuli 8 , 11–13 , 42 ., In experiments where a rhythm in pain sensitivity was detected , its pattern is remarkably consistent , with pain sensitivity peaking during the hours when there is no daylight ( and when humans are typically asleep ) , that is , from midnight to 5 AM 5 ., In previous work , we analyzed experimental data reporting on the daily rhythm in human pain sensitivity from four studies investigating: 1 ) the threshold for forearm pain in response to heat ( n = 39 , 40 ) , 2 ) the threshold for tooth pain in response to cold ( n = 79 , 13 ) , 3 ) the threshold for tooth pain in response to electrical stimulation ( n = 56 , 13 ) , and 4 ) the threshold for nociceptive pain in response to electrical current ( n = 5 , 8 ) ., The data points from these studies are shown in Fig 4A and details on the derivation of these data points can be found in 6 ., We note here that we aligned the data to the subject’s typical or scheduled wake time ( i . e . , 0 hours after wake ) and thus , clearly , this data represents a daily rhythm in pain sensitivity that includes sleep-wake-cycle effects that cannot be uncoupled from an endogenous circadian rhythm ., The result is that , in this work , we discuss pain sensitivity as a function of hours since morning wake time to align our results with these data sources ., The data strongly suggest a sinusoidal profile , and thus we fit a sinusoidal function to the data using Matlab’s 43 curve fitting scheme ( cftool ) ( see solid curve in Fig 4A ) ., We hypothesize that this best-fit sinusoid ( R2 = 0 . 73 and root mean-square error of 4 . 69 ) represents a prototypical daily rhythm in pain sensitivity for humans , with a sharp peak in pain sensitivity occurring close to midnight ( following 18 hours of waking ) , and that then decreases during the night to reach a minimum in pain sensitivity in the afternoon ( following 9 hours of wake , or approximately 4pm ) ., Experimental work also suggests a daily rhythm in the sensitivity of touch ( see Figs 1 and 2 in 13 ) with the highest sensitivity for tactile discrimination occurring in the late afternoon and the lowest sensitivity in the late morning 13 ., Since cells in the DRG ( that contains the cell bodies of the afferent fibers ) rhythmically express clock genes responsible for generating rhythmicity of other physiological processes 15 , we assume in our model that daily modulation occurs at the level of primary afferent input to the spinal cord ., Furthermore , these experimental observations motivate us to introduce rhythmicity in the model input from Aβ-fibers that exhibits nearly a 12-hr shift from the rhythm of the C-fiber-model inputs ., We note here that although we consider rhythmicity in the Aβ-fibers 13 , our modeling work focuses on describing processing of nociceptive stimuli ., Thus , our model does not simulate processing of strictly mechanical stimuli which may use different circuitry from that of nociceptive stimuli ., We use the sinusoidal curve obtained from fitting the experimental data in Fig 4A , with the slight modification of making the period exactly 24 hours , to modulate the Aβ- and C-fiber activity as a function of the time of day in hours since typical morning wake time ., We implement daily rhythmicity in the firing rates of the Aβ- and C-fibers by varying their stimulation response frequencies , R A β ( t ^ ) and R C ( t ^ ) , respectively , with approximately opposite phases ., The average firing rates of the fibers ( 40 Hz for Aβ- fibers and 21 Hz for C-fibers ) were estimated from experiments of receptor activity in the human hand 40 ., This yields equations for the firing rates of the fibers over the day as follows:, R A β ( t ^ ) = 6 sin ( π 12 t ^ ) + 40 , R C ( t ^ ) = 1 2 sin ( π 12 t ^ + 2 . 8 ) + 21 , ( 4 ), where t ^ denotes time , in hours since morning wake time ( see blue and green curves in Fig 4B ) ., The amplitudes of the daily modulation of response frequencies ( ±6 Hz for Aβ-fibers and ±0 . 5 Hz for C-fibers ) were chosen to fit the model’s simulated pain signal , namely the firing rate of the projection neuron population , to the experimental measurements of pain sensitivity , as described below ., To model the effects of the Aβ-dependent presynaptic inhibition of C-fiber activity mediated through an additional inhibitory interneuron population ( I2 ) , we assume that the I2 population is only activated by high , stimulus-induced activity of the Aβ-fibers and that its activity tracks the daily modulation of R A β ( t ^ ) but at a lower firing rate ., As a result , presynaptic inhibition lowers the stimulus response frequency of C-fiber activity , R C ( t ^ ) , as follows:, R C eff ( t ^ ) = R C ( t ^ ) - g A β C ( R A β ( t ^ ) - 30 ) , ( 5 ), where gAβC scales the effects of the presumed I2 activity ( see black dashed curve in Fig 4B ) , and the -30 mimics the lower I2 firing rate ., This presumed level of I2 activity maintains effective C-fiber activity on the same scale as the original C-fiber activity , see blue solid and black dashed lines in Fig 4B ., We note that while the daily modulation of the stimulus response frequencies governing spikes on the afferent fibers is on the order of hours , our model output changes on the order of fractions of seconds ( e . g . , τP = 0 . 001, s ) ., Because of such a difference in time scales , there is only a small change in the stimulus frequencies R A β ( t ^ ) and R C ( t ^ ) during the response to a brief nociceptive stimulus ., Hence , we consider specific time points at a constant t ^ in a 24-hour period ( see Fig 4B ) when generating the ( daily modulated ) response of afferent fibers to stimulation ., We compare the 24-hour rhythm in pain sensitivity computed by our model with the sinusoidal curve representing the human daily pain sensitivity fitted to experimental data in Fig 4A ., Introducing the rhythmicity of fiber responses described above , we simulate our model equations at 7 time points over the 24-hour day , recording our model output ( firing rate of the projection ( P ) neuron population ) for each time point ., To compare with the experimental curve , we compute variation as a percent of the mean by calculating the mean of the average response firing rates of the P population to stimuli given over the whole day , and comparing the firing rate at each time point during the day to that mean firing rate ., Fig 5A shows the model pain sensitivity as a percent of the average over the day ( blue curve ) as compared to the experimental pain sensitivity ( black dashed curve ) ., Notice that the average firing rate of the P population , as shown in Fig 5B , is above 25 Hz which can be considered as a threshold for pain ( see 44 ) ., Furthermore , for the daily rhythmicity of pain sensitivity , the model output represented in terms of percent of mean ( firing rate of the P population ) closely follows experimental results ( see Fig 5A ) ., To simulate response to a brief painful stimulus at the periphery , we construct average firing rate functions for activity of the Aβ- , Aδ- and C-fibers based on the time of day as input to the DH , and calculate the resulting behavior of the PNs as described by the equations in ( 1 ) ., Fig 6 displays the average firing rate of the P population in response to nociceptive stimuli at two time points during the 24-hr day ., Our model reproduces the average firing-rate pattern of the populations of neurons in the DH when the three afferent fibers differ in their conductance speeds , as noted by three distinct activations of the PNs in Fig 6 ., We follow 44 , and interpret the painful response as the firing rate of the PNs crossing a threshold of 25 Hz ., The average firing rate of the P population is qualitatively similar to that seen experimentally ( e . g . , see Fig 1a in 39 ) and agrees with the daily variation in pain as reported in 13 ( lower sensitivity in the afternoon and higher sensitivity at night ) ., Note here that we are only considering nociceptive stimulation of the afferent fibers as mechanical stimulation may follow a different circuit within the DH or more complicated activation of the different afferent fibers ., To quantify the amount of pain experienced from the stimulation of the afferent fibers , we take the average firing rate of the PNs over the period of time when the C-fibers’ response has reached the DH ( see blue rectangle in Fig 6 ) ., Note that the amount of time that the C-fiber response is activated is constant across the day and we consider the average firing rate above 25 Hz as painful 44 ., The parameters for this model were chosen to give painful responses ( i . e . , firing rate of the P population above 25 Hz ) , but also to allow the neuron populations to reach their maximal firing rates during times of day with highest pain sensitivity ., We note that the input from the spinal cord is only one component to the overall experience of pain ., The P population reaching a maximum represents the maximum possible nociceptive response from this portion of the spinal cord ., Thus , ( and as concerns all of the simulations of our model ) a maximal firing rate of the P population does not necessarily correspond to the maximal pain experience ., Additionally , the chosen parameter set allows our model to sufficiently capture experimentally-observed phenomena such as wind-up and pain inhibition , but we recognize that this is not the only set of parameters that would yield these results ., For a complete description of the parameter value choices , see S1 Table ., In addition to the example model output in Fig 6 , we further validate our DH circuit model by showing that it reproduces wind-up —that is , increased ( and frequency-dependent ) excitability of the neurons in the spinal cord due to repetitive stimulation of afferent C-fibers 45 ., Wind-up serves as an important tool for studying the role of the spinal cord in nociception and has often been used as an example phenomenon to validate single neuron models of the DH ( see 21 , 27 , 28 , for example ) ., However , both the physiological meaning and the generation of wind-up remain unclear ( see 46 for a review ) ., There are several possible molecular mechanisms proposed for the generation of wind-up 46 ., Earlier work on single neuron models suggests that wind-up is generated by a combination of long-lasting responses to NMDA-receptor-mediated synaptic currents and membrane calcium currents providing for cumulative depolarization of the PNs 27 ., Indeed , calcium conductances and NMDA receptors of the projection/deep dorsal horn neurons are included in all previous models of the DH circuit 21 , 27 , 28 ., In addition , the study done in 28 emphasizes the effect ( direct or via influencing the dependence of the deep dorsal horn neurons on their intrinsic calcium currents ) NMDA and inhibitory synaptic conductances have on the extent of wind-up in the deep DH neurons 28 ., As noted in the Methods section , we incorporate NMDA synapses into our model for the DH circuit by taking into account that the dynamics of the synaptic weight of the connection from the C-fibers to the PNs , gNMDA , depends on the average firing rate of the P neuron population see Eq ( 2 ) ., We assume that the dynamics of gNMDA are much slower than those of the neuron populations ( τNMDA = 1 s while , e . g . , τP = 0 . 001 s ) ., As a result , in response to a repeated stimulus ( i . e . , when the model input as shown in Fig 3 is presented to the DH circuit at a frequency of 2 Hz ) , the average firing rate of the P population during the C-response increases ( see top panel in Fig 7A ) and the synaptic weight gNMDA exhibits slowly increasing dynamics in response to the increased activity in the P population ( see bottom panel in Fig 7A ) ., For a repeated stimulus at 2 Hz , the latency , which we consider as the time from the start of the stimulus ( t = 0 . 5 s ) to the time when the average firing rate of P exceeds 25 Hz ( i . e . , considered as painful ) , decreases with the stimulus index ( i . e . , index 1 denotes the first stimulus in the repeated sequence ) , see Fig 7B , as seen in experiments 47 ., However , the increase in the average firing rate of the P population depends on the frequency of the repeated stimulation , with optimal effects seen experimentally at stimulation frequencies between 1-3 Hz 46 ., Our model captures the phenomenon of wind-up , as well as the frequency dependency ., For example , when the model input is repeated at a frequency of 2 Hz , the mean of the average firing rate of the P population during the C-response ( see blue box on bottom of Fig 6 ) increases from about 25 Hz during the first stimulus to about 50 Hz during the fifth stimulus similar to previous modeling results 21 , while in the case of a stimulus repeated at 0 . 5 Hz , the mean P firing rate during the C-response does not change as a function of the stimulus index ( Fig 8A , yellow curve vs blue curve ) ., We note that we simulate frequencies up to 3 . 22 Hz as this is the highest frequency we can model without an overlap in the P neuron responses ( see Fig 7A , top ) ., We include it here to show the general trend of wind-up in response to an increase in frequency ., It’s clear to see that as the frequency increases , and the responses are allowed to interact , the result would be a yet faster rise in the firing rate to its maximum due to the additive na

Introduction, Methods, Results, Discussion

Experimental studies show that human pain sensitivity varies across the 24-hour day , with the lowest sensitivity usually occurring during the afternoon ., Patients suffering from neuropathic pain , or nerve damage , experience an inversion in the daily modulation of pain sensitivity , with the highest sensitivity usually occurring during the early afternoon ., Processing of painful stimulation occurs in the dorsal horn ( DH ) , an area of the spinal cord that receives input from peripheral tissues via several types of primary afferent nerve fibers ., The DH circuit is composed of different populations of neurons , including excitatory and inhibitory interneurons , and projection neurons , which constitute the majority of the output from the DH to the brain ., In this work , we develop a mathematical model of the dorsal horn neural circuit to investigate mechanisms for the daily modulation of pain sensitivity ., The model describes average firing rates of excitatory and inhibitory interneuron populations and projection neurons , whose activity is directly correlated with experienced pain ., Response in afferent fibers to peripheral stimulation is simulated by a Poisson process generating nerve fiber spike trains at variable firing rates ., Model parameters for fiber response to stimulation and the excitability properties of neuronal populations are constrained by experimental results found in the literature , leading to qualitative agreement between modeled responses to pain and experimental observations ., We validate our model by reproducing the wind-up of pain response to repeated stimulation ., We apply the model to investigate daily modulatory effects on pain inhibition , in which response to painful stimuli is reduced by subsequent non-painful stimuli ., Finally , we use the model to propose a mechanism for the observed inversion of the daily rhythmicity of pain sensation under neuropathic pain conditions ., Underlying mechanisms for the shift in rhythmicity have not been identified experimentally , but our model results predict that experimentally-observed dysregulation of inhibition within the DH neural circuit may be responsible ., The model provides an accessible , biophysical framework that will be valuable for experimental and clinical investigations of diverse physiological processes modulating pain processing in humans .

Human pain sensitivity follows a daily ( ∼24 hour ) rhythm ., In particular , humans experience the highest sensitivity to pain in the middle of night and lowest in the afternoon ., Patients suffering from neuropathy , a disease resulting from nerve damage leading to an increase in pain sensitivity , experience an approximately 12-hour shift in their rhythmicity such that the highest sensitivity occurs in the afternoon ., Neuropathy is a difficult condition to treat since it is often unfeasible to locate the damaged nerve and it is also unclear how this damage causes a shift in rhythmicity and an increase in pain ., Understanding the mechanism underlying the shift in rhythmicity may lead to improvements in the knowledge of the transmission of pain from the damaged nerve to the pain-processing center in the spinal cord , and thus better treatment protocols ., We have built a population-based model to describe this transmission with a particular focus on daily rhythms ., We show that our model reproduces experimentally-observed rhythmicity of both normal pain responses , as well as neuropathic pain ., Our model predicts that a potential mechanism underlying the shift in rhythmicity for neuropathic pain is a change in the interaction of the nerve fibers from inhibition to excitation .

somatosensory system, medicine and health sciences, pathology and laboratory medicine, nervous system, neuroscience, signs and symptoms, nerve fibers, interneurons, neuropathic pain, sensory physiology, spinal cord, animal cells, neural pathways, pain, cellular neuroscience, pain sensation, diagnostic medicine, neuroanatomy, cell biology, anatomy, physiology, neurons, biology and life sciences, sensory systems, cellular types, afferent neurons

journal.pbio.0060135

The Evolution of the DLK1-DIO3 Imprinted Domain in Mammals

Genomic imprinting is a process that causes genes to be expressed according to their parental origin and is evident in plants and mammals ., Many imprinted genes are located in clusters regulated by a single imprinting control element , whose function across the whole imprinted domain depends on DNA methylation acquired differentially in the male and the female germlines 1 ., It is not known how or why mammalian imprinting evolved; however , its emergence is associated with the evolution of a placenta 2 , 3 , and the correct dosage of imprinted genes is important in prenatal growth , postnatal metabolism 4 , and neurodevelopment 5 ., Where tested , the majority of imprinted genes are expressed and imprinted , sometimes specifically , in the placenta 6 , suggesting that even distantly related placental mammals such as metatherians ( marsupials ) will have imprinting , while oviparous mammals , the prototherians ( monotremes ) , will not ., Assessment of the imprinting status of a few individual mammalian imprinted genes is consistent with these data ., The orthologues of four genes imprinted in mouse and human are clearly imprinted in marsupials 7–10 , and no evidence of imprinting has been found in monotremes , although only three genes have been tested to date 8 , 11 , 12 ., The Dlk1-Dio3 imprinted domain in eutherian mammals contains the protein-coding genes Delta-like homologue 1 ( Dlk1 ) , Retrotransposon-like gene 1 ( Rtl1/Mart1 ) , and the type 3 deiodinase ( Dio3 ) expressed from the paternally inherited chromosome , and multiple long and short non–protein coding RNAs including microRNAs ( miRNAs ) and C/D small nucleolar RNA ( snoRNA ) genes expressed solely from the maternally inherited chromosome ( Figure 1A ) ., Seven imprinted miRNAs are located within anti-Rtl1 , and over forty are located further downstream including within the miRNA-containing gene Mirg ( Figure 1A ) ., All of the genes in the domain are developmentally regulated and expressed in a range of embryonic and extraembryonic cells types with postnatal expression being found predominantly in the brain 13–15 ., In mouse , imprinting is regulated by an intergenic differentially methylated region ( IG-DMR ) , located 75 kb downstream of Dlk1 , that becomes methylated during spermatogenesis but remains unmethylated in the maternal germline 16 , 17 ., When a targeted deletion of the IG-DMR is inherited maternally , an epigenetic switch occurs causing the maternally inherited chromosome to behave like the paternally inherited chromosome; no effect is seen when the deletion is paternally inherited ., The IG-DMR is also differentially methylated in human 17 , and recently identified patients with deletions and epimutations in the DLK1-DIO3 region indicate that this element likely acts as the imprinting control element in human 18 ., Tight linkage and strong conservation of Dlk1 and Dio3 is maintained in all vertebrates ., The two genes are located 10 . 5 kb apart in Takifugu rubripes , approximately 370 kb apart in chicken , and 830 kb in human and mouse ( Figure 1B ) ., To determine the sequence and organization of the region in marsupial and monotreme mammals , we cloned and sequenced the region between DLK1 and DIO3 in the platypus , Ornithorhynchus anatinus , and the tammar wallaby , Macropus eugenii ., Bacterial artificial chromosome ( BAC ) clones containing the orthologous DLK1 and DIO3 genes were identified 19 ., Thirteen overlapping wallaby BACs and seven overlapping platypus BACs were isolated from genomic libraries , then initially characterized using a parallel landmark content mapping and fingerprinting strategy 20 , and sequenced ( Figure S1 and Table S1 ) ., This genomic sequence represents complete coverage of the domain in both species and was generated independently of the whole-genome sequencing projects for these organisms ., The wallaby sequence is 1 , 510 . 8 kb and slightly smaller than that of the South American marsupial Monodelphis domestica ( 1 , 637 . 8 kb plus 26 gaps ) ., The marsupial region is therefore approximately twice as long as its eutherian orthologue ( Figure 1B ) ., The region in platypus is 594 . 8 kb , which is 28% smaller than in mouse ., For both wallaby and platypus , DLK1 and DIO3 genes were identified , cDNAs characterized , and the genes subjected to imprinting analysis ( Figure 2 and Figures S2 and S3 ) ., For wallaby , fetal tissues , yolk sac placenta , and pouch young samples were dissected ., Platypus fetal material is unavailable , so the analysis was conducted on primary adult skin fibroblasts cultured from two male and one female platypus; therefore the analysis in that species is limited ., Several single nucleotide polymorphisms ( SNP ) ( Figure 2A ) were identified for DLK1 from wallaby tissues that included one sample ( 2386 ) from a homozygous mother allowing allele-specific activity to be determined ., Both maternally and paternally inherited alleles of DLK1 were expressed in all wallaby fetal , extraembryonic , and pouch young tissues analysed ( Figure 2A and Figure S2C ) ., Similar SNP and restriction fragment length polymorphism analysis showed biallelic expression of DLK1 in platypus ( Figure 2B ) ., Two polymorphisms were identified in wallaby DIO3 , a G/A SNP at nucleotide 94 in the coding region of the gene and a CTT insertion/deletion ( indel ) at nucleotide 1 , 187 in the 3′ untranslated region ( UTR ) ( Figure 2A and Figure S3 ) ., Both polymorphisms were present in nine animals , suggesting co-segregation of the variant alleles ., Direct sequencing of cDNAs amplified across both polymorphisms indicated there was preferential expression from the G/-CTT allele ( Figure 2D ) ., Quantitative real-time , reverse-transcriptase PCR ( RT-PCR ) proved that wallaby DIO3 was expressed from both parental chromosomes ., However , an allelic bias towards the -CTT allele was observed in all samples tested regardless of parental origin ( Figure 2E ) ., Expression analysis of two polymorphisms in platypus DIO3 confirmed biallelic expression in this species ( Figure 2C ) ., Comparative sequence analysis of the Dlk1-Dio3 genomic landscape between eutherian and other noneutherian mammals can identify the dynamic changes that are associated with and have the potential to contribute to imprinting ., Figure 3A and Table 1 show the relative GC and repeat sequence content of the region in seven genomes; three eutherian species ( human , mouse , dog ) , two marsupials ( opossum and tammar wallaby ) , one monotreme ( platypus ) , and one bird ( chicken ) ., The eutherian GC content , %CpG and number of CpG islands was significantly higher than the genome average ( p < 0 . 01 using Chi-squared test ) in contrast to marsupial and monotreme mammals , and chicken , that all lack imprinting at this domain ., Repeat content was analysed using the most recent previously unreleased platypus repeat database ( kindly provided by R Hubley , Repeatmasker ) ., Eutherian LINE content is consistent with the genome-wide average; however , there is a paucity of LINEs in the region between Dlk1 and Mirg ( miRNA-containing gene ) in the eutherians ( Table 1 ) ., The majority of repeats identified in the DLK-DIO3 domain in the marsupials are LINE1 repeats ., This is consistent with the high number of LINEs identified in the opossum genome and suggests that expansion in the DLK1-DIO3 region , as in the marsupial genome as a whole , is due to LINE1 insertion ., The opossum region has a slightly larger proportion of SINEs than expected from the genome average ., The SINE content is also greater in the tammar wallaby , although the whole-genome sequence for this species is not currently available for comparison ., The relative repeat content in platypus is greater than eutherians despite the region being smaller in this species ( Figure 3A ) ., The majority of repeats in the platypus DLK1-DIO3 region are SINEs and the more ancient LINE2s ., Interestingly; there is a notable absence of long terminal repeat ( LTR ) elements at this locus in platypus ( Figure 3A ) ., The chicken region is devoid of any SINE elements which is consistent with the whole genome analysis of this species ., Hence platypus and marsupials have greater SINE content in the domain than do the eutherian mammals with imprinting ., This is consistent with the SINE depletion previously reported when comparing imprinted with nonimprinted domains in mouse and human 21 , 22 ., Together , these findings indicate that selection against SINE repeats is an evolutionary feature of imprinted domains ( see Discussion ) ., Detailed comparative sequence analysis was conducted between the Dlk1-Dio3 domain in the seven vertebrates ., Using a threshold of 55% nucleotide sequence identity over 80 bp , which recognizes the Dlk1 exons in all seven sequences , 141 evolutionary conserved regions ( ECRs ) were identified across sub-groups representing eutherians , marsupials , platypus , and chicken ( Table S2 ) ., Of the 141 ECRs found , 22 . 7% ( 31 ) were common to all seven vertebrates , 15 . 6% ( 22 ) were common to all mammals , and another 16 were found in all therian mammals ., Six were found only in platypus and chicken ., Figure 3B illustrates the number of ECRs arranged according to the sub-classes of vertebrates in which they are identified ., In mammals , 27 . 7% were identified in at least one eutherian , one marsupial and platypus , whereas 24 . 8% were found in at least one species representing each therian infraclass ., Although the greatest number of ECRs is found within the mammalian species , the more ancestral ECRs ( the 31 found in all species studied ) are on average larger , having a mean length of 494 bp compared with the mean length of all ECRs at 340 bp and suggesting greater functional constraint ., We used the 31 ECRs found in all vertebrates to align the Dlk1-Dio3 domain and subdivide it into 30 inter-ECR zones for further comparative analysis ( Figure 4A ) ., Exons of Dlk1 and Dio3 are represented by vertebrate ECRs 1–3 and 30–31 , respectively ., The intergenic distribution of the ECRs is not uniform throughout the domain with two-thirds being located in the 3′ half of the domain ., One of the ECRs , approximately 3 kb upstream of DIO3 , contains a highly conserved putative CTCF binding site in all therian species ., The amount of sequence in each of the 30 inter-ECR zones relative to the overall size of the domain was quantified for each vertebrate ( Figure 4B ) ., This provides a measure of the overall expansions/contractions between species ., The regional changes between marsupial , monotreme , and eutherian mammals across the domain are not uniform ., The most striking differences between the mammals lie in zone 3 ( between vECR3 and vECR4 ) , zone 6 , and zone 7 ., Zone 3 which is located between the last exon of Dlk1 and the conserved intron 5 region of Gtl2 , is expanded in eutherians ( Figure 4B ) ., This expansion does not appear to be caused by LINEs , because LINE1 and LINE2 repeats are equivalently represented in eutherians and marsupials ( Figure S4 ) ., As shown , this zone contains a higher proportion of SINE elements than reported for the whole genome and compared with the entire domain ., However , the increased SINE content does not explain the expansion of zone 3 , which is due to the acquisition of unique sequence , including the imprinting control region ( the IG-DMR ) and presumably other eutherian specific regulators ., In contrast , eutherian zone 6 , located between Gtl2 and Rtl1 , is smaller than in marsupials , platypus and chicken , implying either that it contracted or that it is resistant to expansion ., This latter explanation is favoured , because in marsupials expansion is predominantly due to LINE1s , and in platypus to LINE2 repeats and SINEs ., This exclusion suggests an important previously unrecognized eutherian specific function for that zone ( Figure 4 ) ., The eutherian specific expansion of zone 7 , as for zone 3 , is not associated with the insertion of repetitive sequences , compared with marsupials ., Rather , zone 7 represents the region located between Rtl1 and Mirg , which , in eutherians , contains approximately 50 miRNA genes and three clusters of C/D snoRNA genes , all expressed from the maternally inherited chromosome 23 ., With one exception ( see below ) , our analysis failed to find homologous sequences in marsupials , platypus , or chicken ., Instead , the zone contains LINE1 repeats in marsupials and as before , LINE2s and SINEs in platypus ( Figure S4 ) ., Therefore eutherians acquired transcribed non–protein coding RNAs in a zone that appears resistant to expansion by LINEs and SINEs ., Interestingly , the acquisition of snoRNA genes in the imprinted Prader Willi-Angelman syndrome locus also corresponds to the acquisition of imprinting 12 ., In the mouse , all the imprinted non–protein coding transcripts in the domain require the imprinting control element and sequences 5′ to Gtl2 for their activity on the maternally inherited chromosome ., They are all expressed in the same orientation , and data suggest that they are at least in part associated with a single long transcription unit 17 , 24 ., ECRs specifically associated with Gtl2 were identified by phylogenetic footprinting ( Figure 5A ) ., Two approaches were undertaken to determine whether GTL2 and other non–protein coding transcripts were present within the domain; expression analysis of DLK1-DIO3 intergenic ECRs and the amplification from cDNA of randomly selected sequences from the wallaby region ( Figure S5A and Table S3 ) ., Five mammalian ECRs were found in the vicinity of Gtl2 , of which three were common to all vertebrates; one corresponds to exon 5 of NM_144513 ( ECR19 ) , and the remainder appear to be intronic ., One of the intronic ECRs ( ECR18 ) was previously identified in intron 8 of Y13832 22 ., An additional ECR ( ECR14 ) located close to exon 1 was identified and found to be inverted in eutherians ( Figure 5A ) ., This and the three vertebrate ECRs were expressed at very low levels in wallaby tissues , with no transcriptional activity from the other two ., RT-PCR analysis of 29 additional , randomly selected sequences in wallaby located between Gtl2 and Mirg identified weak transcriptional activity from five sequences , including one mammalian Mirg-specific ECR ( Figure S5 ) ., Quantitative RT-PCR comparing the relative expression of ECR19 and one of the random sequences ( Ran3 ) with DIO3 expression in the same samples confirms expression from the GTL2-like locus in marsupials is between 1 . 1 × 10−4 and 4 . 2 × 10−4 lower in fetal head and pouch young brain ( Figure 5 ) ., Polymorphisms located in ECR19 and the MIRG-ECR were used to demonsrate that this low level of transcription is biallelic ( Figure 5B and Figure S4 ) ., It was of particular interest to determine whether the protein-coding , retrotransposon-like gene Rtl1 ( also known as Peg11/Mart1 ) was present in non-eutherian mammals ., Rtl1 is a member of the Ty3-Gypsy family of LTR retrotransposons with closest similarity to the Sushi-ichi class 25 ., In mouse and human , it has lost its LTRs , encodes a protein essential for normal placental development and fetal growth and viability ( M . Ito , A . Ferguson-Smith , unpublished data , and 26 ) , and is expressed from the paternally inherited chromosome ., Its levels are regulated by miRNAs processed from an antisense transcript on the maternally inherited chromosome that are 100% complementary to the Rtl1 mRNA ( Figure 1A ) 17 , 27 , 28 ., Another member of this family , Peg10 located on mouse Chromosome 6 , was recently shown to be imprinted in wallaby fetus and placenta ( but is absent in the platypus ) , and its repression on the maternally inherited chromosome is associated with differential methylation in the body of the gene 9 ., We could not demonstrate RTL1 sequences in the platypus or chicken domain ., However , we did find sequences related to Rtl1 in the appropriate position in marsupials but , interestingly , it is extensively degraded with very few regions of homology remaining ( Figure 5D ) ., No expression of the most highly conserved region was found in fetal and pouch young tissues ( Figure 6A ) ., This suggests that Rtl1 retrotransposed into the locus prior to the divergence of marsupial and eutherian mammals and , in the absence of functional selection , it degraded in marsupials but acquired a growth regulatory function in eutherians coincident with the evolution of imprinting ., A number of miRNAs that are antisense to Rtl1 are transcribed from the maternal chromosome in eutherians ., Using the miRNA prediction programme miR-abela 29 , no miRNAs were found to be conserved between all vertebrates , and none were conserved between eutherians and marsupials ., A single predicted miRNA was conserved between the marsupials ( 74% identity ) ( Figure S6B and S6C ) ., Interestingly , this was located in the vicinity of the eutherian miR127 , which is transcribed antisense to Rtl1 and along with seven others , contributes to the stability of the Rtl1 mRNA through an RNAi-dependent mechanism 27; a function that would not be evident in marsupials that lack this gene ., The sequence of the predicted processed miRNA from marsupial miR127 though common to both marsupials is less similar to eutherians and RT-PCR analysis failed to amplify the primary transcript or predicted hairpin from wallaby fetal head or pouch young brain cDNAs ( Figure S6D ) ., These data suggest that this is not a functional miRNA , and sequence similarity is due to miR127 being located within RTL1 ., A small number of conserved CpG islands and CpG-rich regions were found to be shared between eutherians , marsupials , and platypus and their methylation status was determined ., They included the promoters of Dlk1 and Dio3 and the differentially methylated region in the last exon of eutherian Dlk1 , known as the Dlk-DMR 16 , 30 ., Each region was analysed by methylation-sensitive Southern blots with genomic DNA from platypus and wallaby and from wallaby sperm ., Results are shown in Figure 6 ., The ECR at intron 5 in Gtl2 ( Y13832 ) is CpG-rich , and this too was analysed ., As in eutherians , the DLK1 and DIO3 CpG-island promoters are completely unmethylated on both parental chromosomes ., The Gtl2 ECR is partially methylated on both parental chromosomes in mouse , and has the same pattern in platypus and wallaby ., In mouse , the Dlk-DMR is hypermethylated on the paternally inherited chromosome and in sperm , and hypomethylated on the maternally inherited chromosome 16 , 30 ., Platypus and wallaby genomic DNA showed hypermethylation of the locus similar to that seen on the paternal chromosome in the mouse ., Wallaby sperm was also hypermethylated ., This suggests that the methylation state of the mouse paternal chromosome resembles the methylation state of the mammalian domain prior to the emergence of imprinting and implies that hypomethylation of the maternal chromosome evolved with imprinting ., In eutherians , Dlk1 and Dio3 are developmentally important genes that are expressed in numerous embryonic and extraembryonic tissues ., Here we have shown that DLK1 and DIO3 are both biallelically expressed in marsupial fetus , placenta , and neonatal pouch young ., DLK1 was recently shown to be expressed biallelically in adult brain , liver , and kidney in the South American marsupial , Monodelphis domestica; however , analysis of imprinting in embryonic and extraembryonic tissues was not conducted in that study 31 ., We also demonstrate biallelic expression of both genes in platypus ., Because fetal material is not available , biallelic expression of these genes during platypus development can only be inferred ., Together , our results indicate that imprinting of the whole DLK1-DIO3 domain evolved after the divergence of metatherian and eutherian mammals ., Comparative sequence analysis of the DLK1-DIO3 region in seven different amniote vertebrates ( representing Eutheria , Metatheria , Prototheria , and Aves ) demonstrates that the overall genomic landscape in this region is GC-rich in eutherians but not in the other species studied ., It has previously been postulated that GC-rich isochores in eutherians were once located on GC-rich microchromosomes in the ancestral amniote 32 ., The elevated GC content in eutherians but not in the noneutherian species suggests that the increase occurred in eutherians rather than existing as an ancient isochore ., A number of results suggest that the DLK1-DIO3 is a recombination hot spot and under purifying selection in eutherian species where it is imprinted ., First , elevated GC content correlates with increased levels of recombination 32 ., Second , the introns of DLK1 are shorter in the eutherians than in the noneutherian species ( Figure S2B ) , and decreased intron length is associated with high recombination rates 33 ., Third , the reduced SINE content in the eutherian indicates the region is under purifying selection , especially because SINEs are usually associated with GC-rich regions ., Interestingly , the region between vertebrate ECR1 and ECR8 , which encompasses Dlk1 , Gtl2 , Rtl1 , snoRNAs , and miRNAs , is particularly devoid of LINEs , indicating that this region is under even greater constraint ( Table 1 and Figure 4A ) ., Finally , the eutherian DLK1-DIO3 regions are also all located close to the telomeres , whereas in noneutherian species , they are located mid-chromosome 19 ., A correlation of elevated recombination levels at sub-telomeric regions has previously been reported 34–36; however , it is possible that this sub-telomeric position is the result of increased breakage in GC-rich regions 37 ., Imprinted domains have previously been shown to be associated with elevated GC content 38–41 , short introns 42 , and reduced SINE content when compared to nonimprinted regions in eutherians 21 , 43 ., Our finding that this comparison can be extended to the same domain between mammals that imprint and those that do not strongly suggests that imprinted domains are under purifying selection perhaps to constrain domain size such that cis-acting elements can function correctly ., None of the ECRs maps to the position of the eutherian imprinting control element ., Whether any of the ECRs plays a functional role in the regulation of the domain is currently under investigation ., Those specific to subgroups such as oviparous vertebrates , or the sixteen ECRs specific to therian mammals , might relate to the regulation of specific functions such as the development of extraembryonic structures in therians ., Expression analysis has provided evidence that Gtl2 and other noncoding transcripts existed throughout amniote evolution , suggesting that Gtl2 did not arise from an eutherian-specific retrotransposition event that triggered imprinting at the domain as has been previously suggested 31 ., Our results show that weak regional non–protein coding transcriptional activity can occur in some places across the domain in noneutherian mammals and suggest that the process repressing the protein-coding genes on the maternal chromosome in eutherians ( driven by the imprinting control region upstream from Gtl2 ) facilitated stronger expression from these non–protein coding transcripts ., The appearance of functional miRNAs and C/D snoRNAs within the locus may therefore have been a consequence of the acquisition of imprinting with the strongly expressed Gtl2 gene , providing an ideal host transcript ., It is not known whether the duplications that gave rise to the miRNA clusters occurred before or after evolution of imprinting at the locus ., Interestingly , a role for these miRNAs in the trans-regulation of neural and placental processes has been inferred 44 ., A functional role for these transcripts in the regulation of the neighbouring imprinted protein-coding genes also cannot be ruled out ., Furthermore , the emergence of a regulatory relationship between RTL1 and its reciprocally imprinted miRNA-containing antisense transcript is also intriguing ., In contrast to the more distal miRNA clusters to which they are not related , these seven anti-RTL1 miRNAs are not likely to have arisen through duplication/divergence events ., Rather , these may have evolved as a host defence mechanism associated with the retrotransposon properties of RTL1 , and evolved with it to modulate its expression 27 as it acquired an endogenous function ., During the course of evolution , the genomic landscape of the Dlk1-Dio3 region has undergone a number of changes ( Figure 7 ) ., Most significantly , the region has become imprinted ., This analysis has proven that imprinting in this domain emerged after the divergence of marsupials and eutherian mammals ., This provides evidence that mammalian imprinting evolved at different loci at different times in response to selective pressures acting on different domains , suggesting an adaptive process ., Prior to the divergence of metatherians from eutherians , the Sushi-ichi retrotransposon Rtl1 , inserted between DLK1 and DIO3 , gained no function and was degraded in marsupials ., In marsupials , the region expanded 2-fold through the insertion of LINE repeats ., As the eutherian lineage evolved through selective regional changes , Rtl1 evolved into a new gene acquiring a vital function in growth and development ., This gain of function may indeed have driven imprinting at the domain , conferred through the acquisition of the imprinting control element ., Gtl2 and associated transcripts became up-regulated on the maternal chromosome in eutherians , and miRNAs and C/D snoRNAs specifically evolved in the region ., Once imprinted , gene expression was fixed in the region it underwent purifying selection , correlating with an increase in GC content , reduction in Dlk1 intron size , and selection against SINE and LINE insertions ., Comparison of these results with similar detailed analyses on domains acquiring imprinting prior to the divergence of marsupials and eutherians will provide further insight into the relationships between dynamic changes in genomic landscape and the evolution of imprinting ., RNA was extracted using the GenElute mammalian total RNA miniprep kit ( Sigma ) following the manufacturers protocol ., cDNA was synthesized using Superscript III RNase H− Reverse Transcriptase ( Invitrogen ) following the manufacturers instructions ., The RT-PCRs were primed using random hexamer primers or the following gene-specific primers; platypus DLK1 5′-GAACGTTTATTTTACAAAAGATAGCTG-3′ , wallaby DIO3 5′-CGGGCACTCACAGAGTTACA-3′ , and platypus DIO3 5′-GACTCCGTCTCCGAGAACAT-3′ , and 5′-TGAACATCTTACAAAAACCAACAAA-3′ ., cDNA was amplified using Hot Start KOD polymerase ( Novagen ) , PCR conditions are as described in 19 ., For particularly GC-rich regions ( e . g . , platypus DIO3 ) 1× Polymate ( Bioline ) was also added to the PCR reaction ., Primer sequences and annealing temperatures can be found in Table S3 ) ., PCR fragments were gel purified using Qiaquick Gel Extraction Kit ( Qiagen ) and sequencing was performed ., For ECR and random sequence expression analysis , cDNA was generated as above using random hexamers and PCR amplification performed using either Hot Start KOD polymerase ( Novagen ) or Taq polymerase ( Bioline ) , using conditions described in 19 ., The primer sequences and annealing temperatures can be found in Table S4 ., Custom TaqMan assays were produced using the Assays-by-Design facility at Applied Biosystems ., 1 μl of cDNA was amplified in a 12 . 5-μl reaction 1× TaqMan Universal PCR Master Mix ( Applied Biosystems ) and 1× specific assay as per the manufacturers instructions ., CT ( threshold cycle ) values for both the VIC and FAM probes were recorded and the difference between them ( ΔCT ) was calculated ., Samples were analysed in triplicate ., Genomic DNA from homozygous individuals , was used as controls to ensure no cross hybridisation occurred between the two probes ., The ΔCT of cDNAs was compared with a standard curve of ΔCT values from two homozygous gDNAs mixed at different ratios ( 49:1 , 9:1 , 4:1 , 7:3 , 3:2 , 1:1 , 2:3 , 3:7 , 1:4 , 9:1 , and 1:49 ) , and the percentage expression from each allele was extrapolated ., This method was adapted from 45 ., The primers used were as follows: MeDIO3UTR-F , 5′CTTCCCTCCTCCCCAAATTCC-3′; MeDIO3UTR-R , 5′-TGCAGTCAACAAAGTGGAGGAA-3′; + allele probe , 5′- ( VIC ) -TTCTCTCCTTGGTTTTT- ( MGB ) -3′; and – allele probe , 5′- ( FAM ) -TTTTTTCTCTCGGTTTTT- ( MGB ) -3′ ., Assays were performed using the SensiMix NoRef kit ( Quantace ) ., The amplification of each primer pair was determined using a serial dilution of cDNA ( 1 , 1/5 , 1/25 , 1/125 , 1/625 ) ., Reactions were performed in triplicate , and the average CT value of each dilution was used to generate a standard curve ., The slope of the curve when plotted to log10 was used to determine the efficiency of amplification ( E ) for each primer set using the following equation: E = 10 ( −1/slope ) and the relative fold expression calculation ( 2– ΔCT ) was corrected for the amplification efficiency 46 ., The relative gene expression was calculated using the following equation: Expression Ratio = EΔCT ( sample ) /EΔCT ( reference ) ., Sample reactions were performed in triplicate ., Three fetal head samples were used ( from between d22 and d25 RPY ) and three PY brain samples ( from between D17 and D20 post partum ) ., The primers used were as follows: MeDIO3QF , 5′- CCGAGGGCTACAAGATCTCA-3′; MeDIO3QR , 5′- CACGTTTGTTTGGGGTTCTT-3′; MeECR19F , 5′-GCGGCTTCACAAATTTATTTTC-3′; MeECR19R , 5′-CAACTCTGCACAGATGGATGA-3′; MeRan3F , 5′-CAGCTGGATCCAATTTGACA-3′; and MeRan3R , 5′-TTGGACCATGATCCTGGAAT-3′ ., Genomic DNA was extracted using standard protocols 47 ., 10 μg of restriction enzyme–digested DNA was separated on 0 . 5× TBE 1% agarose gels and transferred to Hybond-N+ ( GE Healthcare ) nylon membranes ., Filters were pre-hybridised in ULTRAhyb solution ( Ambion ) at 42 °C for a minimum of 2 h ., Probes were labelled with α-32P-dCTP using the Megaprime DNA labelling system ( GE healthcare ) and added to the hybridization buffer ., Filters were incubated at 42 °C overnight , washed to the stringency of 2XSSC/0 . 1%SDS , and exposed to a phosphoimager screen ., Probes used for the analysis can be found in Table S3 ., With the exception of novel platypus and tammar wallaby sequence generated here , sequences for computational analysis were downloaded from the UCSC genome browser 48 ., These sequences are: human sequence , build March 2006 , chr14:100210000–101150000; mouse sequence , build February 2006 , chr12:109850000–110780000; dog sequence , build May 2005 , chr8:71946000–72800000; opossum sequence , build January 2006 , chr1:315760000–317570000 ( reverse complement ) ; and chicken sequence , build May 2006 , chr5: 51365000–51832000 ., The repeat content for the Dlk1-Dio3 region in each species was determined using RepeatMasker version: open-3 . 1 . 8 49 ., The specific repeat library was used for each species and the default parameters ., For the platypus sequence , a command line version of the software was used with a pre-release of the latest library for this species ( kindly provided by R . Hubley , RepeatMasker , Institute for Systems Biology , USA ) ., CpG islands were predicted by CpGplot 50 using the default parameters and a window of 200 ., Two different programs were used to predict ECRs: zPicture 50 and mVista 51 , 52 ., The LAGAN algorithm 53 was used in the mVista alignment ., The translated anchoring option ( where one or more of the alignments steps are performed on translated sequence ) was used , because it can improve the alignment of distant homologues ., The default setting of >70% identity and >100 bp in length was used between marsupials and platypus ., Greater than 55% identity and >80 bp was used for the alignments of eutherians to other species and chicken to other species ., For both programs , human , mouse , dog , and chicken sequences were repeat masked using th

Introduction, Results, Discussion, Materials and Methods

A comprehensive , domain-wide comparative analysis of genomic imprinting between mammals that imprint and those that do not can provide valuable information about how and why imprinting evolved ., The imprinting status , DNA methylation , and genomic landscape of the Dlk1-Dio3 cluster were determined in eutherian , metatherian , and prototherian mammals including tammar wallaby and platypus ., Imprinting across the whole domain evolved after the divergence of eutherian from marsupial mammals and in eutherians is under strong purifying selection ., The marsupial locus at 1 . 6 megabases , is double that of eutherians due to the accumulation of LINE repeats ., Comparative sequence analysis of the domain in seven vertebrates determined evolutionary conserved regions common to particular sub-groups and to all vertebrates ., The emergence of Dlk1-Dio3 imprinting in eutherians has occurred on the maternally inherited chromosome and is associated with region-specific resistance to expansion by repetitive elements and the local introduction of noncoding transcripts including microRNAs and C/D small nucleolar RNAs ., A recent mammal-specific retrotransposition event led to the formation of a completely new gene only in the eutherian domain , which may have driven imprinting at the cluster .

Mammals have two copies of each gene in their somatic cells , and most of these gene pairs are regulated and expressed simultaneously ., A fraction of mammalian genes , however , is subject to imprinting—a chemical modification that marks a gene according to its parental origin , so that one parents copy is expressed while the other parents copy is silenced ., How and why this process evolved is the subject of much speculation ., Here we have shown that all the genes in one genomic region , Dlk1-Dio3 , which are imprinted in placental mammals such as mouse and human , are not imprinted in marsupial ( wallaby ) or monotreme ( platypus ) mammals ., This is in contrast to a small number of other imprinted genes that are imprinted in marsupials and other therian mammals and indicates that imprinting arose at each genomic domain at different stages of mammalian evolution ., We have compared the sequence of the Dlk1-Dio3 region between seven vertebrate species and identified sequences that are differentially represented in mammals that imprint compared to those that do not ., Our data indicate that once imprinted gene regulation is acquired in a domain , it becomes evolutionarily constrained to remain unchanged .

genetics and genomics

A comparative analysis of genomic imprinting between mammals that imprint and those that dont has provided insights into how and why imprinting evolved.

journal.pgen.1000698

Quantifying Adaptive Evolution in the Drosophila Immune System

Hosts face an ever-changing array of parasites to which they must adapt , and parasites are widely believed to be one of the most important and universal selection pressures in natural populations ., Consistent with this view , immune genes in several taxa are known to evolve faster than other genes , and sometimes significantly faster than the neutral rate – a signature of adaptive evolution 1 , 2 , 3 ., Indeed , many studies of one or a few immune genes have identified the action of positive selection in Drosophila , including Relish 4 , the Scavenger Receptors 5 RNAi genes 6 , TEPs 7 , Persephone 8 and others 2 ., More recently , complete genome sequencing of multiple Drosophila species found that immune-related genes have high rates of amino-acid substitution , and are more likely to show evidence of adaptive evolution than other genes 1 , 9 ., Here we go beyond the yes/no detection of selection , to quantify the additional adaptation that occurs in proteins of the immune system over and above that which occurs in the rest of the genome ., The rate at which natural selection fixes new mutations can be estimated by comparing the amount of polymorphism within populations to divergence between species at synonymous and nonsynonymous sites 10 , 11 , 12 , 13 , 14 ., Approaches of this kind have been used to estimate the genome-wide rate of adaptive evolution , and found that it is often surprisingly high 10 , 13 , 15 , 16 , 17 ., However , the nature of the selection pressures underlying this evolution remains unknown ., One approach to answering this question is to compare estimated rates of adaptive evolution between proteins with different functions ., Moreover , focussing on genes where we have a strong expectation of elevated positive selection also has a further benefit; there is an ongoing debate about the extent to which the high genomic estimates represent artefacts of processes such as population demography 15 , 18 , 19 , and testing the a priori hypothesis that immunity genes will have increased adaptive rates can address this issue ., To assess the role of pathogens and other parasites as a cause of molecular evolution we have resequenced population samples of most of the best-characterised immunity genes in the Drosophila melanogaster genome , together with position-matched ‘control’ genes with no known immune function ., This provides a quantitative estimate of the impact of parasite-mediated selection on the rate of adaptive evolution , and suggests that immunity genes have double the genome-average rate ( Figure 1 ) ., We found that this was not caused by a generally elevated rate in immunity genes ., Instead , most immunity genes show similar rates of adaptive evolution to the rest of the genome , with only a small subset evolving under very intense selection ( Figure 2 ) ., These genes tend to be concentrated in a few pathways , which we argue are likely to be hotspots of host-parasite coevolution ( Figure 3 ) ., Interestingly , these pathways are known to be suppressed by pathogens , and this suggests that active parasite-suppression of the immune system is an important cause of this adaptive evolution ., Furthermore , when independent lineages are compared , similar genes show accelerated rates of adaptation ( Figure 4 ) ., This suggests that despite their dynamic nature , host-parasite interactions may create similar selective pressures in related species , leading to replicable signatures at the molecular level ., The proportion of amino acid substitutions that were fixed by natural selection ( denoted α ) can be estimated using extensions of the McDonald-Kreitman test 16 , which compares non-synonymous and synonymous changes , and contrasts within-species polymorphism to fixed differences between species ., We have extended existing maximum likelihood approaches 15 , 23 , 24 to estimate separate α values for immunity and non-immunity genes , and for different classes of immunity genes ( see Materials and Methods ) ., We found that the proportion of substitutions attributable to positive selection in immune genes is approximately 50% greater than the genome average ., Based on the divergence between D . simulans and D . melanogaster and polymorphism in Kenyan populations of both species , we estimated that 65% of amino acid substitutions in immunity genes have been fixed by selection ( 95% bounds bootstrapping across genes within categories: 55–72% , Figure 1A ) ., This is significantly higher than our estimate for non-immunity genes , which is very close to previous genome-wide estimates ( reviewed in 10 ) ( α\u200a=\u200a41%; 95% bounds are 31–50%; difference from immunity genes: p\u200a=\u200a0 . 004 , inferred by bootstrapping ) ., The effect remained highly significant when data from all populations were combined , though absolute estimates of α were slightly lower ( immune: α\u200a=\u200a58%; non-immune: α\u200a=\u200a33%; p\u200a=\u200a0 . 004; Figure S10 ) ., Since the exclusion of rare variants led to slightly higher estimates of α ( Figure S16 ) , this effect is probably caused by the enlarged sample size containing a higher proportion of ( low-frequency ) mildly-deleterious non-synonymous variants , which can cause α to be underestimated 23 ., Estimates of α in the Greek ( Athens ) populations had greater variance and failed to detect a significant difference between immunity and non-immunity genes ( Figure S10B ) , as might be expected because the relatively low genetic diversity of this population means we have little statistical power to accurately infer α 14 ., The proportion of amino acid substitutions fixed by selection ( α ) will clearly be affected by the number of substitutions not fixed by selection , i . e . , the number of effectively neutral substitutions fixed through genetic drift ., Therefore , it is possible that the higher α of immunity genes does not reflect any increase in the absolute number of adaptive substitutions per non-synonymous site ( denoted a 16 ) ., This possibility has been little explored , because a , unlike α , is difficult to estimate as a multi-gene average , and because single-gene estimates of either statistic tend to be imprecise ., Here we use an approach that allows us to obtain relatively stable estimates of a for individual genes ( see Materials and Methods ) , which can then be averaged across immune and non-immune genes ., Using Kenyan populations of D . melanogaster and D . simulans , we estimated that since their common ancestor , selection has fixed an average of 10 . 6×10−3 adaptive substitutions per non-synonymous site in immunity genes , but only 5 . 7×10−3 in other genes ( difference between immunity and control genes: p\u200a=\u200a0 . 02; Figure 1B ) ., This difference in the absolute number of adaptive substitutions corresponds to 50% increase in the proportion ( α ) described above , and suggests that natural selection is fixing adaptive substitutions in immunity genes at nearly double the genome average rate ., The high rate of adaptive evolution that we found in immunity genes could be driven either by a general elevation in the strength of selection across all immunity genes , or by a few key genes experiencing intense selection pressures ., To investigate this , we examined the distribution of a across genes ., Although mean a is higher for immunity genes than other genes ( Figure 1B ) , the modal class is the same , i . e . , centred on zero in both cases ( Figure 2A versus Figure 2B ) , and the difference in mean is driven by a subset of immune genes with unusually high a ( Figure 2C; this results in a significantly higher variance for immunity genes ) ., The wider distribution of a across immunity genes suggests that most of these genes experience similar selection pressures to the rest of the genome , while a small subset are under substantially stronger selection ., This is consistent with the analyses of D . simulans genome sequences that found little evidence that immunity genes as a group are outliers in terms of recurrent adaptive evolution 17 ., Thus it appears that host-parasite arms races may involve a relatively small subset of the immune system ., This analysis could be confounded if our estimates were less accurate for immune genes than control genes , but this is unlikely for two reasons ., First , the immunity genes tend to be longer than control genes , which will reduce the variance of a estimates and make our analysis conservative ( Figure 2C ) ., Second , the pattern remains significant and quantitatively almost identical if the analysis is restricted to genes with more than 500 non-synonymous sites ( Figure S17 , S18 ) ., Clues as to the nature of the selection pressures acting on immune genes can be gained from looking at which functional classes of immune gene are experiencing the strongest selection 1 , 2 ., To examine how selection pressures differ between immune genes with different functions , we classified the genes in two different ways ., First , we classified genes according to the branch of the immune system in which they function: the humoral , cellular , melanisation and antiviral RNAi responses ., We found little variation between the first three categories ( α\u200a=\u200a51% , 62% and 63%; per-site a\u200a=\u200a0 . 009 , 0 . 010 and 0 . 012 , respectively ) , and individually no category was significantly different from non-immunity genes ( Figure 1A and Figure 1B ) ., However , RNAi genes were an exception to this , showing approximately twice the proportion of adaptive substitutions as compared to non-immune genes ( α\u200a=\u200a88% vs . 41%; p<0 . 001 ) , and seven times the number of adaptive substitutions per site ( a\u200a=\u200a0 . 042 vs . 0 . 0057; p<0 . 001; Figure 1 ) ., This is consistent with previous results , which found that some RNAi genes evolve rapidly under positive selection 6 , 25 ., Second , we classified immune genes ( excluding those involved in RNAi ) according to their mode of action: pathogen recognition , signalling cascade , and antimicrobial peptides ( AMPs ) ., This categorisation gave a superior fit to the data according to model selection techniques ( see Materials and Methods , and Table S2 ) and was also a significantly better fit than randomly assigning genes to categories of the same size ( randomization test: p<10−3 ) ., Using this alternative categorisation , no group was significantly higher than non-immune genes , although signalling molecules did have a marginally higher α but not a ( estimated α\u200a=\u200a57% vs . 41%; p\u200a=\u200a0 . 085 ) ., Consistent with previous results 26 , 27 , AMPs showed no evidence of adaptive evolution ( were not detectably different from α\u200a=\u200a0; Figure 1A ) , undergo significantly less adaptive evolution than RNAi , signalling and cellular recognition genes ( p<0 . 014 in each case ) , and undergo marginally less adaptive evolution than non-immune genes ( estimated α\u200a=\u200a−13% vs . 41%; p\u200a=\u200a0 . 082 ) ., Alternative analyses using other populations and outgroups resulted in a qualitatively identical pattern ( Figures S10 , S11 , S12 , S13 , S14 , S15 ) , except that the use of D . yakuba as an outgroup resulted in the signalling molecules having a significantly higher α than the controls ( p<0 . 031; Figure S14A and S14B ) ., Because the high rate of adaptive evolution in immune system genes is caused mainly by a subset of genes under very strong selection ( Figure 1 and Figure 2 ) , we investigated how these genes are distributed across the immune system ( Figure 3 ) ., The two main signalling pathways in the immune system are the Toll and IMD pathways , and of these the IMD pathway has a much higher rate of adaptive evolution than the Toll pathway ( IMD: mean estimated a\u200a=\u200a0 . 023; Toll: mean a\u200a=\u200a0 . 009; difference between Toll and IMD p\u200a=\u200a0 . 039 by bootstrapping within classes ) ., Within the Toll pathway , the extracellular molecules are under stronger selection than the cytoplasmic ones ( extracellular: mean a\u200a=\u200a0 . 015 , cytoplasmic: mean a\u200a=\u200a0 . 005 , p\u200a=\u200a0 . 033 ) ., The antiviral RNAi genes again show strong adaptive evolution 6 ( mean estimated a\u200a=\u200a0 . 032 ) ., Elsewhere , TEP I and PGRP-LD are also under exceptionally strong selection 1 , 7 ., It has been suggested that the phagocytosis receptor Dscam , which can produce up to 18 , 000 differently spliced isoforms , may allow Drosophila to mount specific immune responses 28 , 29 ., However , despite having over 22 kbp of coding sequence from Dscam , we were unable to find any evidence of adaptive evolution in this gene , indicating that this gene is not subject to arms-race selection ., If the immune system adapts to parasites in similar ways in related species , then we would expect to see the same genes experiencing positive selection in different lineages 30 ., Alternatively , each species could respond differently , resulting in different genes being positively selected in different lineages 30 ., To address this question , we estimated the rate of adaptive evolution separately for each of the lineages leading to D . simulans and D . melanogaster from the common ancestor of the two species ., The pattern of α ( and a ) across different pathways and functional categories of genes was very similar between the two lineages ( Figures S12 , S13 ) , suggesting that the broad distribution of selection pressures between immune functions is the same ., For example , in both lineages antiviral RNAi genes have the highest rates of adaptive evolution and antimicrobial peptides have the lowest rates ., Estimates of a along these individual lineages are associated with high levels of noise due to the short length of the branches; furthermore , the measurement error will be negatively correlated across the two lineages ., Despite these sources of error , however , the data show a significant positive correlation in immunity gene a estimates between the two lineages ( Figure 4 ) , and this suggests that individual genes , and not just categories of gene , are under similar selection pressures in both lineages ., This correlation was not significantly different to that that found in the non-immunity genes , indicating that there is no greater tendency for parasites to cause lineage specific selection than other selective agents ( Figure 4 ) ., The analyses presented above can identify selection that has occurred over millions of years , but recent selective sweeps can also be detected though reductions in genetic diversity ., In both D . melanogaster and D . simulans there was no significant difference in the diversity of synonymous sites ( πs ) between immunity and non-immunity genes ( Kenyan D . melanogaster: πs\u200a=\u200a1 . 60% vs . 1 . 55%; Kenyan D . simulans: 2 . 46% vs . 2 . 62%; Figure S19 , Figure S20 , Table S3 ) ., Furthermore , if the immune genes are split into functional categories , only the diversity of the antiviral RNAi genes is significantly lower than the control genes ( D . melanogaster πs\u200a=\u200a0 . 80% , p<0 . 001; D . simulans πs\u200a=\u200a1 . 01% , p<0 . 001 . Figure S19 , Figure S20 , Table S3 ) ., This is consistent with RNAi genes having the highest rates of adaptive substitution in the immune system , and suggests a high proportion of them may have recently experienced selective sweeps in both species ., Furthermore , none of the immune genes had unusually high levels of polymorphism , suggesting host-parasite coevolution in Drosophila has not resulted in the ancient polymorphisms like those seen in vertebrate MHC genes and some plant resistance genes 31 , 32 ., It is known that flies are infected by different parasites in different populations , and this could lead to local adaptation where different alleles of a gene are favoured in different populations 33 , 34 , 35 , 36 , 37 ., However , we could not detect any differences between immune genes and the controls in the amount of population structure in either D . melanogaster or D . simulans ( Figure S21 ) providing no evidence to suggest that local adaptation of immune genes is common ., However , it should be noted that our statistical power to detect genetic structure may be extremely low , and the effects of local adaptation on patterns of nucleotide variation may be small 38 ., We also compared the amino acid diversity ( πa ) of the immunity and control genes , as this may reflect differences in selective constraint or the effects of balancing selection ., In all eight populations πa was slightly higher in the immune genes , and in three populations the difference was significant ( Figure S22 , Figure S23 , Table S3 ) ., Compared to the control genes , immune signalling molecules tend to have lower amino acid diversity , while antimicrobial peptides and recognition molecules in the cellular immune system have significantly higher amino acid diversity ( Figure S22 , S23 ) ., These differences correspond to the estimated number of substitutions occurring by genetic drift ( Figure S24 ) , but not to differences in πs , implying that they are caused by differences in selective constraint , rather than long-term balancing selection maintaining amino acid polymorphisms ., We have found that the rate of adaptive substitution in immunity genes is nearly double the genome average ., This is the first quantitative estimate of the rate at which natural selection drives protein evolution in genes of the immune system relative to the genome as a whole , and confirms that adaptation to parasites is an important force driving evolution ., There are several reasons why parasites may be a powerful selection pressure ., Firstly , parasites can cause high rates of mortality and morbidity , and therefore have a large impact on the fitness of their hosts ., Secondly , the direction of parasite-mediated selection continually changes , due to coevolutionary arms races between hosts and parasites 39 , and ecological factors altering the composition of the parasite community ., Finally , parasites generally have shorter generation times , and ( in the case of viruses ) elevated mutation rates , potentially giving them an edge in the ‘arms-race’ ., This means that hosts may often be maladapted to their current set of parasites , and therefore under strong selection to evolve resistance ., We have also found that the high rate of adaptive substitution of immunity genes is driven by a small subset of immune genes under strong selection , while the majority of immunity genes have similar rates of adaptive evolution to the rest of the genome ., This suggests that rapid ‘arms-race’ coevolution may only involve a small subset of molecules in the immune system ., Since there is a tendency for these strongly-selected genes to cluster by pathway or protein-family , these clusters may reflect hotspots for coevolutionary interaction with parasites ., By examining the function of these groups of strongly-selected genes , we can gain clues regarding the underlying molecular processes that drive this coevolution ., It is striking that almost all of these genes fall within the IMD signalling pathway and the antiviral RNAi pathway ( Figure 3 ) ., It is known that both signalling pathways and RNAi are targeted by parasite molecules that suppress the immune response , and it has been suggested that this suppression may cause much of the adaptive evolution seen in immunity molecules 1 , 2 , 4 , 25 , 40 ., The Toll pathway tends to have lower rates of adaptive evolution ., It is unclear why this is , although it may reflect the pathogens with which it interacts , or constraint from its other functions in development 41 ., In contrast to the signalling pathways , the PGRPs and GNBPs that act as receptors for the Toll and IMD pathways are not positively selected , possibly reflecting their role in binding to highly conserved pathogen molecules 7 ., Unlike many other organisms ( especially vertebrates 42 ) , AMPs in Drosophila show less adaptive evolution than most genes ., This contrasts with the high rate of AMP gain and loss in the Drosophila phylogeny 1 , and suggests that whatever process favours the duplication of AMPs does not result in strong selection on their protein sequence ., Our results also imply that AMPs may be weakly constrained , with genetic drift fixing amino acid substitutions at a relatively high rate ., This may be a consequence of gene duplication , as duplicated genes often have elevated rates of amino acid substitution 43 ., It is interesting to note that components of the antiviral RNAi pathway also mediate defence against transposable elements 44 , 45 , 46 , and these ‘genomic parasites’ may be an important selective force on these genes 25 ., Indeed , several RNAi genes with no reported anti-viral function 25 , 47 , 48 , and other genes involved in chromatin function 17 , show evidence of rapid adaptive evolution in Drosophila ., At the phenotypic level , many organisms show evidence of convergent evolution , with different species evolving similar adaptations in response to similar selection pressures ., However , it is unclear whether convergence is also common in molecular evolution , or whether molecular evolution is idiosyncratic , with each species following a unique evolutionary pathway 30 ., One way to address this question is to test whether the same genes are evolving adaptively in different species 30 ., At a broad level , we found that similar functional classes of immunity genes tend to have elevated rates of adaptive evolution in both the D . melanogaster lineage and the D . simulans lineage ., At a finer scale , the rate of adaptive evolution of individual genes is correlated in the two lineages ( despite the very high levels of noise associated with these single-lineage estimates ) ., Because this correlation was not significantly different in immunity genes and our control genes , this suggests the fluctuating selection pressures associated with host-parasite coevolution do not result in unusually high rates of lineage-specific selection ., Together these results suggest that the immune system of these two closely related species experience similar selection pressures , and adapt to those selection pressures in similar ways ., Previous studies on immunity genes have applied various tests of adaptive evolution , and found that a higher than average fraction of immunity genes test ‘positive’ ( e . g . , 1 , 2 ) ., However , the statistical power of these tests will depend on factors such as selective constraint and gene length , and these could differ between immunity and non-immunity genes , even if their rates of adaptive substitution were identical ., Furthermore , such confounding factors will be even more important if adaptive substitution is frequent across the genome , meaning that a large proportion of all genes evolve under some degree of positive selection 10 ., Therefore a particular strength of the current approach , which can compare the estimated rates of adaptive evolution across different groups of genes , is that it provides quantitative estimates of the effect size rather than simply counting the number of ‘significant’ tests ., Estimates of the rate of adaptive substitution based on the McDonald-Kreitman test have been subject to some recent criticism as they can be influenced by factors such as population demography 18 , 19 ., However , it seems unlikely the differences observed here are artefacts ., First , we compared loci where we have a strong a priori expectation of adaptive substitution to position-matched control loci ., Second , we found no significant differences in the rate at which genetic drift causes non-adaptive evolution at these loci , such as could mislead the tests ( Figure S24 ) ., Finally , false signatures of adaptive substitution can occur in populations that have experienced bottlenecks or recent expansions , and yet the signal we observed was much stronger in the ancestral Kenyan populations ( Figure S10A ) , and weakest in the more derived populations ( Figure S10B ) , while quantitative estimates of a differed surprisingly little between datasets ., As new sequencing technologies result in ever larger datasets , this approach promises to be a powerful way to identify the selection pressures driving molecular evolution ., Our data not only confirm that parasites are an important driving force in molecular evolution 1 , 2 , they quantify the magnitude of this effect , and show that the rate of adaptive protein evolution in immunity genes is nearly twice the genome average ., This elevated rate in the immune system is due to a subset of genes evolving under intense positive selection , and many of these genes are strongly selected in both D . melanogaster and D . simulans , suggesting that our results may reveal general principles of immune system evolution ., In particular , some of the most strongly selected genes may be targeted by parasite suppressors the immune response , and this may be a key battlefield in coevolution ., These data add to the growing evidence that much adaptive protein sequence evolution is driven by co-evolutionary conflicts within or between genomes 49 , 50 ., Flies were sampled from six populations of D . melanogaster and two populations of D . simulans , covering both their original range in Africa and more recent global expansion ., In each population we extracted genomic DNA from four female flies that were either collected from the wild or were the progeny of crosses between pairs of isofemale lines ( i . e . we sampled eight chromosomes from each population ) ., Targeted genes were amplified by PCR in ∼5 kbp products , and the PCR products from each population were then mixed together , purified on a gel , and sequenced using the Solexa-Illumina sequencing platform to high coverage ( mean >130-fold; Figure S1 ) ., The 36 bp sequencing reads were aligned to the D . melanogaster or D . simulans genome using MAQ 51 allowing for up to 2 mismatches per read , which resulted in 5–16 million mapped reads in each population ., The sites were then assigned to coding or non-coding sequence using the genome annotation , and coding sites were classified as synonymous or non-synonymous ., Positions with less than 20-fold coverage were excluded , as were genes represented by less than 100 bp; however , our results were not strongly affected by the exclusion of sites with less than 50-fold or 100-fold coverage ( Figure S25 ) ., Full details of the Solexa-Illumina sequencing , together with a detailed comparison with traditional Sanger sequencing , are given in Text S1 ., A full listing of loci , their positions and polymorphism counts are given in Table S1 ., To estimate the rate of adaptive substitution , we used a multi-locus , maximum likelihood extension of the McDonald-Kreitman test ., This method is based on Welch 2006 ( ref . 15 , see also 23 , 24 ) , but contains several new features and models ., Software that implements the new methods is available on request from the authors , or from http://tree . bio . ed . ac . uk/software/ ., We compared non-synonymous and synonymous divergence between D . melanogaster and D . simulans with polymorphism from both species ., For each locus , the six observations ( dN , dS , and pN and pS for each species ) , were assumed to have the following expected values:where lS and lN are the number of synonymous and non-synonymous sites , λ\u200a=\u200aμt is the expected neutral divergence between the species , θi\u200a=\u200a4Neμ is the expected neutral polymorphism for species i , ni is the number of alleles sampled for species i ( taken here to be 8 per sampled population ) , and f is the fraction of non-synonymous mutants that are effectively neutral 15 ., The parameters of greatest interest here , α or a , quantify the multiplicative or additive deviation of the observed dN from its expectation under neutrality and purifying selection ., Positive estimates of either α or a are consistent with adaptive protein evolution , while negative values result either from sampling error , or from the presence of mildly deleterious mutations ( which violate the assumptions of the test , contributing to pN but rarely reaching fixation 16 , 52 ) ., This violation can be mitigated by excluding low frequency synonymous and non-synonymous polymorphisms , as this is expected to remove the great majority of mildly deleterious mutations while leaving the neutral pN/pS ratio unaltered 52 , 53 ., To explore this phenomenon , we repeated our analyses excluding all putative polymorphisms with an estimated minor-allele frequency below a range of threshold frequencies ( Figure S3 ) ., Our results were qualitatively unaltered , and so in the main text we report only results with all sampled polymorphisms included in the counts ., To estimate the model parameters it was assumed that observed quantities were Poisson distributed around their expected values 15 , 23 , 24 ., This distribution is derived under the assumption that substitutions and polymorphisms occur as independent events , but this assumption can be violated , e . g . , by linked selection causing the clustering of substitution events in time ., We used three approaches to reduce the impact of such violations ., First , for some parameter types ( selective constraint f and/or adaptive substitution a ) , we assigned separate parameters to each locus , making the extent of stochastic variation irrelevant to the parameter estimates obtained ., Second , we obtained confidence intervals by bootstrapping across loci , rather than using the curvature of the likelihood surface ., Third , we used model-selection criteria that allow for un-modeled over-dispersion ( such as that arising from the clustering of events in time ) ., To avoid over-parameterization associated with assigning large numbers of locus-specific parameters , we assumed that λ ( the neutral mutation rate multiplied by divergence time ) took a single value across all loci ., To model neutral polymorphism , we exploited the correlation between θ at a locus , and its local recombination rate 54 , by fitting the model θ\u200a=\u200amr+b , where r is the local D . melanogaster recombination rate 55 ., Maximum likelihood estimates of m and b were then obtained for each of the two species ., This model has the advantage of providing appropriate estimates of θ for loci where the synonymous polymorphism is not at equilibrium , such as after a recent selective sweep ., Model selection techniques ( see below ) also showed that it was significantly preferred to models in which θ did not vary between loci , and in which each locus had a separate parameter ., Importantly , however , estimates of a were very similar under all three parameterizations ( Figure S26 ) ., Given our chosen model , a data set of k loci was used to fit k+5 nuisance parameters , plus the a or α values of interest ., To choose between different parameterizations of the likelihood model ( see Table S2 ) we used the Akaike Information Criterion , corrected for finite sample size and over-dispersion in the count data 56 ., This criterion is given by QAICc\u200a=\u200a−2lnL/c+2K+K ( K+1 ) / ( n-K-1 ) where lnL is the maximized likelihood for the model , K is the number of parameters it contains , and n is the number of data points ( taken to be 6 times the number of loci ) ., The factor c is the correction for overdispersion , and was estimated by c\u200a= ( 2lnLfull-2lnLsat ) /nfull , where “full” denotes the largest model in the set of models being compared , and “sat” denotes the saturated model , in which the expected values of all data points were set to their observed values ., The conditional likelihood of each model was obtained by converting the QAICc values into Akaike weights 56 ., To compare estimates of adaptive substitution along two independent lineages , we used a variant of the method above , including polymorphism from a single species , and polarizing substitutions on to the D . melanogaster or D . simulan

Introduction, Results, Discussion, Materials and Methods

It is estimated that a large proportion of amino acid substitutions in Drosophila have been fixed by natural selection , and as organisms are faced with an ever-changing array of pathogens and parasites to which they must adapt , we have investigated the role of parasite-mediated selection as a likely cause ., To quantify the effect , and to identify which genes and pathways are most likely to be involved in the host–parasite arms race , we have re-sequenced population samples of 136 immunity and 287 position-matched non-immunity genes in two species of Drosophila ., Using these data , and a new extension of the McDonald-Kreitman approach , we estimate that natural selection fixes advantageous amino acid changes in immunity genes at nearly double the rate of other genes ., We find the rate of adaptive evolution in immunity genes is also more variable than other genes , with a small subset of immune genes evolving under intense selection ., These genes , which are likely to represent hotspots of host–parasite coevolution , tend to share similar functions or belong to the same pathways , such as the antiviral RNAi pathway and the IMD signalling pathway ., These patterns appear to be general features of immune system evolution in both species , as rates of adaptive evolution are correlated between the D . melanogaster and D . simulans lineages ., In summary , our data provide quantitative estimates of the elevated rate of adaptive evolution in immune system genes relative to the rest of the genome , and they suggest that adaptation to parasites is an important force driving molecular evolution .

All organisms are attacked by an ever-changing array of pathogens and parasites , and it is widely supposed that the ensuing host–parasite “arms race” must drive extensive adaptive evolution in genes of the immune system ., Here we have taken advantage of new sequencing technologies and analytical approaches to quantify the amount of adaptation that is occurring in immunity genes relative to the rest of the genome ., We sampled two species of fruit fly ( D . melanogaster and D . simulans ) from eight different populations around the world , and sequenced 136 immunity and 287 non-immunity genes from these samples ., Based on the differences in the sequences between the two species , and the genetic diversity within each species , we have estimated that natural selection drives twice as much change in immune-related proteins as in proteins with no immune function ., Interestingly , the rate of adaptation is also more variable among immunity genes than among other genes in the genome , with a small subset of immunity genes evolving under intense natural selection ., We suggest that these genes may represent hotspots of host–parasite coevolution within the genome .

evolutionary biology, genetics and genomics/population genetics, genetics and genomics/genetics of the immune system, evolutionary biology/genomics

journal.pgen.1004966

Genomics of Divergence along a Continuum of Parapatric Population Differentiation

During ecological speciation , divergence along the genome has been observed to be heterogeneous in numerous taxonomic groups e . g . , 1–4 ., Typically , the average genome-wide divergence is low , interspersed with regions of exceptional differentiation ., However , studies describing divergence patterns across the genome have found regions of exceptional differentiation to be either numerous and small 4 or few and large 5 , 6 , the latter sometimes referred to as ‘genomic islands’ ., A variety of explanations have been proposed for the observed heterogeneity in genomic divergence , including stochastic processes such as genetic drift , but also deterministic mechanisms such as locus-specific reduction of gene flow in the vicinity of genes causing reproductive isolation , hitchhiking around selected variants , or variation in recombination and mutation rates 7 ., Generally , genetic drift , population expansion , migration , and other demographic events affect the whole genome , whereas natural selection modified by local environmental differences impact only those regions of the genome that affect the respective phenotypes and fitness ., It is not known whether or not genomic patterns such as the variation of divergence and recombination along the genome tend to follow a predictable evolutionary trajectory as populations proceed along a speciation continuum 7 ., We investigated the early phase of divergence using lake-river stickleback population pairs varying in their degree of genetic differentiation ., If divergence patterns are driven by locus-specific effects of gene flow and divergent selection , the extent of divergence is expected to be more localized than widespread , in line with the “island view” 6 ., These regions might hold “speciation genes” maintaining reproductive isolation between species including genes underlying a fitness reduction in hybrids 8 ., Furthermore , “divergence hitchhiking” , the accumulative effect of selectively advantageous loci , predicts a positive correlation between genomic divergence and island size progression 9 ., An alternative explanation posits that the lack of differentiation across most of the genome is due to shared ancestral polymorphism rather than ongoing gene flow 10 , 11 , whereas regions of high differentiation represent regions influenced by selection at linked sites 12 ., Such a hitchhiking pattern may be caused by both advantageous ( positive selection ) and deleterious alleles ( background selection ) ., Therefore , if adaptation alone ( assuming some degree of geographic separation ) shapes the genomic landscape , population genetic processes unrelated to the extent of overall genomic differentiation govern divergence patterns ., Disentangling such alternative scenarios is a crucial yet challenging step in understanding the genomics of divergence , especially in parapatry where the current and historic extent of migration and gene flow contribute to the overall genomic patterns ., We tested predictions inherent to the different scenarios explaining genomic patterns of divergence using whole-genome sequencing data of replicated population pairs of three-spined sticklebacks varying in their degree of genetic differentiation ., Five population pairs were sampled from connected lakes and rivers from the United States ( Us ) , Canada ( Ca ) , Norway ( No ) , and from two sites in Germany ( G1 and G2; Fig . 1 and S1 Table ) ., As ice sheets covered these regions during the last glaciation , these populations represent recent colonization events ( ~12 000 years ago ) ., Both lake and river populations are derived from marine ancestors that became landlocked during de-glaciation , and in which ecotype differentiation between watersheds has occurred repeatedly ., Some phenotypic traits such as feeding morphology 13 , brain development 7 , and parasite resistance 14 seemingly differentiated in parallel with habitat ( i . e . lake and river ) suggestive of local adaptation ., Furthermore , experimental studies have shown evidence for local adaptation to lake and river habitats mediated by parasites 15 ., Hence , contrasting the differentiation between populations from distinct ecosystems permits us to study the onset of divergence , which might eventually lead to complete reproductive isolation ( i . e . speciation ) ., Here , we scan genomic divergence patterns and evaluate differences and commonalities across a wide geographic sampling of parapatric population pairs to uncover the relative importance and interaction of evolutionary factors like drift , selection , and recombination during adaptive divergence ., One consistent difference between lake and river habitats is that lake fish posses a higher parasites diversity than parapatric river fish ., From previous work on three-spined sticklebacks , lakes and rivers in Northern Germany are known to harbour distinct parasite communities 14 , 16 ., Despite the relatively low sample size for individual locations in this study ( n = 12–17 ) , this ecological difference between lakes and rivers is here confirmed on a broader geographic scale ( Fig . 1 ) ., From each of the ten sampled populations , six stickleback genomes were sequenced using a combination of paired-end and mate-pair libraries on the Illumina HiSeq platform to an average genomic coverage of 26-fold ( S2 Table ) ., Instead of sequencing many individuals with low coverage , a small number of genomes per population was chosen to be sequenced to high coverage ., This approach takes advantage of the greater resolution of single nucleotide polymorphisms ( SNPs ) and copy number variations ( CNVs; evaluated in greater detail in a companion paper 17 ) plus increased genotype accuracy within each individual to decipher the divergence mechanisms acting towards an apparent repeated differentiation between lake and river fish ., Besides evaluating allele frequencies , the high individual sequence coverage permits us to infer haplotypes and examine recombination patterns ., After stringent quality filtering , we accessed 297 , 437 , 667 bp from the 20 autosomes ( 380 , 547 , 835 bp ) ., SNP density varied from 3 to 10 SNPs per kilobase ( kb ) within each population ( S3–S4 Tables ) ., For each of the five parapatric comparisons , pairwise genome-wide averages of divergence ( FST ) ranged from 0 . 10 to 0 . 28 , disclosing a varying degree of differentiation in the ascending order of Us , G2 , No , G1 , and Ca ( Table 1 ) ., The parapatric pairs emerge as repeated independent differentiation events ( neighbor joining tree , Fig . 1A ) except for the German populations , despite belonging to different draining systems ( North Sea versus Baltic Sea ) ., Due to low land levels and historically varying water levels , water bodies and connections across Northern Germany have most likely fluctuated over time ., Thus the two lake and river population pairs in Germany ( G1 and G2 ) might have been originally connected ., Because of this , G1 and G2 share some postglacial history , common ancestral variation , and divergence while currently the two water systems are physically separated ., Specifically , studies on the German system have proposed parasite communities as a promising candidate mediating divergent selection , pointing out their role in local adaptation 15 , 18 ., As a further global perspective of this hypothesis , we find a signal of isolation-by-adaptation ( partial mantel test: r = 0 . 622 , P = 0 . 0007 ) shown by a significant association of genome-wide FST and parasite community ( jaccard distance of parasite sums across individuals , counts were 4th square root transformed ) while correcting for geographic distance ( geodetic distance between GPS coordinates of each sampling location ) ., As we detected isolation-by-adaptation at a spatial scale beyond which gene flow occurs , this signal might be most likely caused by a loose linkage between locally adapted loci and the genome-wide neutral regions 19 ., This result suggests a role of parasites for the local adaptation of freshwater stickleback populations ., Spatial heterogeneity along the genome was analyzed between parapatric populations by applying a genome scan approach , which averaged genetic divergence ( FST ) in 10 kb and 100 kb non-overlapping windows across the 20 autosomes ( Fig . 2 ) ., The shape of the distribution of FST values across the genome qualitatively matches a skewed Poisson distribution , suggestive of divergence with gene flow ( S1 Fig . ) 9 ., The pronounced right tail of the distributions aided the identification of outlier windows , which are significantly different from the genome-wide average ., Outlier windows were detected for each population pair as the top 1% of the empirical distribution in addition to being significantly differentiated compared to a random permutation of markers across windows , applying a false discovery rate ( FDR ) of 0 . 01 ., Using the exact same approach comparing marine and freshwater populations , regions known to be under strong divergent selection such as Eda and Atp1a1 were detected as outliers demonstrating the robustness and reliability of the applied methods ( details see Methods ) ., Across all five parapatric lake-river comparisons , we identified a total of 1 , 530 extreme 10 kb outlier windows , in which 47 are shared between at least two of the five population pairs , a proportion that is slightly more than expected by chance ( 10 , 000 permutations of random sampling gave on average 28 overlaps , one-tailed P = 0 . 0006 ) , but none of the windows were shared across all five population pairs ., Although we found a weak positive correlation of FST along the genome between the five lake and river ecotype pairs ( Fig . 2 and S2 Fig . ) , there is a negative correlation of FST among the 1 , 530 outlier windows ( Pearson correlation ranging from r = -0 . 2531 to -0 . 1064 , all P<10-4 ) ., These results indicate that outlier windows in one population pair are often windows of low FST in the other population pairs ., Hence , outlier windows are not the same across the different population pairs ., Annotations for all genes overlapping common outlier windows can be found in S5 Table ., None of these outlier windows overlapped with those detected in a previous lake and river comparison of different stickleback populations on the Haida Gwaii archipelago 20 ., Thus outliers of exceptional differentiation appear to be locally specific for lake-river ecotypes on a wide geographic scale as well as on a narrow scale 20 , 21 ., This is in contrast to earlier comparisons between marine and freshwater sticklebacks where few loci are repeatedly found under divergent selection on a global scale 22 , 23 ., Our results are in line with the notion that the repeated differentiation between derived freshwater stickleback populations occurs as a response to different ecological pressures specific to their local environment 24 ., This might reflect locally specific parasite communities , aside from the general trend of an increase in parasite diversity in lakes compared to rivers ., However , genomic diversification seem to be an inevitable consequence following the dispersal across habitats , reinforcing the concept that local adaptation is a major contributor to the evolution of species ., In order to further understand processes shaping the heterogeneity of genomic divergence , we evaluated if divergence is widespread or localized along the genome ., Divergence hitchhiking predicts a trend towards an increase in size of divergent regions with overall population differentiation 8 , 19 ., Conversely , if size was largely determined by the strength and duration of selection , the size of divergent regions would be independent of overall population differentiation ., To test these predictions in our dataset , we exploited our comprehensive sequencing resolution to identify precise borders and dimensions of regions of exceptional differentiation ., Amongst the 1 , 530 outlier windows , adjacent outlier windows were combined into 794 continuous outlier “regions” of exceptional differentiation estimated to the nearest 1 kb ( S6 Table ) ., The size of a region of exceptional differentiation was determined utilizing barrier strength ( b , ref 25 ) to contrast local divergence to the genome-wide average ., We found a high degree of size heterogeneity among divergent regions within and across population pairs , with no evidence that the size of these regions increases with higher levels of genome-wide differentiation ( Table 1 , S3 Fig . ) ., This also holds true when recombination rates are taken into account ( see below ) ., Therefore , the genomic pattern of divergence observed across a continuum of population differentiation suggests that selection at linked sites drives the observed pattern rather than the interplay of gene flow and divergent selection , consistent with the perspective of geographically specific local adaptation ., However , additional factors such as soft sweeps resulting from adaptation based on standing genetic variation might also contribute to the observed patterns , further complicating interpretations ., To further explore if the observed divergence patterns are indeed facilitated by selection and not induced by drift alone , we investigated fine-scale linkage patterns and their effects on genomic heterogeneity across a populations ., For each population , we estimated the realized population-scaled recombination rates ( ρ/Θ ) along the genome ., Both a local reduction of gene flow mediated by divergent selection and selection with the hitchhiking of linked neutral sites are predicted to produce a negative correlation between FST and recombination rate 12 , 26 , however this association would be unlikely mediated by drift alone ., In addition , divergence hitchhiking predicts that over time , linkage will extend along the genome and eventually encompass large tracts of the genome 27 ., In our study , realized recombination rates in regions of exceptional differentiation were often significantly reduced compared to genome-wide estimates ( Fig . 3 ) ., We found that genome-wide recombination rates tended to decrease with increasing overall differentiation ( Fig . 3 ) ., However , realized recombination rates in divergent regions are not significantly correlated with genome-wide differentiation , adding to the growing lack of empirical evidence for divergence hitchhiking 28 ., These results suggest that either actual recombination rates coincide regions of the genome , which become divergent , or selection drives local reductions in realized recombination rates ., The coalescent-based population recombination rates ( 4Ner ) estimated in this study are simultaneously affected by the variation in genomic structure within and across populations , which may influence actual recombination rates , as well as by selection ., Hence , selection might have locally reduced realized recombination rates in certain genomic regions or actual recombination has been reduced due to the intrinsic genomic structural variations thereby promoting genomic divergence ., Previous studies evaluating large-scale map-based recombination patterns in sticklebacks have also found a correlation between recombination and divergence , suggesting that genome structure , via its influence on recombination , is important in understanding patterns of genomic differentiation 29 , 30 ., Here , the low correlation in divergence ( FST ) between different population pairs ( Fig . 2 ) suggests that local factors specific to each population pair drive genomic differentiation , and that population specific selection reduces realized recombination , particularly if genomic structure is conserved across populations ., However , it is possible that genome structure is not so strongly conserved across these geographically distant pairs ., Structural variations such as inversions and CNVs have been shown to be abundant within stickleback populations 31 ., A companion paper 17 highlights the prevalence of CNVs among and between the populations studied here , in which CNVs tend to also be population specific ., These findings indicate that genome structure might be more variable than expected , and therefore might hold potential for promoting genomic differentiation in a population specific manner ., We cannot here distinguish between selection-induced influences on realized recombination rates , and actual variation in recombination rates due to differences in genome structure and resultant effects on patterns of genomic differentiation ., Further understanding of genome structure’s influence on recombination rates , and its variability within and across populations , will be crucial for disentangling the combined influences of selection and recombination on patterns of genomic variation ., Relative divergence ( FST ) in regions with low levels of recombination might be misleadingly interpreted as conclusive evidence for a local reduction of gene flow ., For this reason , measurements of absolute divergence such as Dxy have been suggested as a complement to more reliably identified regions of locally reduced gene flow 10 , 12 , 32 ., However , absolute divergence measurements are unreliable statistics for nascent populations and in non-equilibrium situations during population differentiation ., Hence , we aim to disentangle different mechanisms shaping regions of exceptional differentiation by assessing selective sweep signatures in one or both populations of each parapatric pair ., Utilizing the base pair resolution of our whole genome sequence data , we evaluated allele frequency spectra to differentiate between molecular signatures of selection among individual regions of exceptional differentiation ., In divergent regions differentiated due to a local restriction of gene flow mediated by selection , the spectrum is not expected to be affected locally and should reveal a signature of neutral evolution 12 ., The opposite is true for regions resulting from selection with hitchhiking at linked sites , which causes a characteristic skew of the spectrum ., An excess of rare alleles is expected in a population experiencing a selective sweep 33 , or in both populations in the case of background selection 34 ., Distortions in the allele frequency spectrum were calculated for each population as Tajima’s D ( TD ) across the genome in 100 kb windows and in each region of exceptional differentiation ., Genome-wide averages of TD varied from 0 . 0385 to 0 . 5936 suggesting predominantly neutral evolution across the genome with no indication for an excess of low frequency polymorphism in any of the populations ., TD values within regions of exceptional differentiation were shifted towards negative values except for the Alaskan river ( Us_R , Fig . 4A ) ., These negative shifts of TD are consistent with selection as a major mechanism responsible for localized divergent regions along the genome ., In order to quantify the relative contribution of different mechanisms shaping the genomics of speciation , we partitioned individual regions of exceptional differentiation into four mutually exclusive categories with different molecular signatures of evolution based on contrasting local TD values to the genome-wide average ( Table 1 and Fig . 4B–F ) ., The minority of divergent regions is consistent with background selection ( 12% , TD reduced in both populations , Fig . 4B ) , whereas adaptation seems to shape most of the divergent regions ( 48% ) , consistent with the influential role of selection ., Divergent regions with signals of positive selection ( TD reduced in one of the two populations ) should harbor those genes responsible for local adaptation ., Genes in divergent regions with a signature of positive selection in lakes ( Fig . 4C ) were overrepresented with functions involved in structural molecule activity ( 18 out of 260 annotated genes , P = 0 . 0018 ) , while genes in divergent regions with signals of positive selection in rivers ( Fig . 4D ) were overrepresented with functions involved in G-protein coupled receptor activity ( 15 out of 105 , P = 0 . 0038 ) , antiporter activity ( 6 out of 36 , P = 0 . 0280 ) , and drug transmembrane transporter activity ( 4 out of 8 , P = 0 . 0367 ) , suggesting functions in environmental response ., Divergent regions with neutral TD patterns ( TD in both populations similar to genome-wide average , Fig . 4E ) potentially harbor genes restricting gene flow ., Despite the prominent occurrence of neutral TD patterns among divergent regions ( 35% ) , we found no functional overrepresentation of genes within those regions ( S6 Table ) ., This indicates that a variety of different genes and functions might be involved in reproductive isolation , but the current state of gene annotations does not allow drawing compelling conclusions ., Overall , the variety of molecular signatures of selection found in divergent regions suggests that different evolutionary processes shape regions of exceptional differentiation ., We acknowledge that our approach of strictly categorizing regions based on thresholds simplifies a complex situation , in which various factors most likely interact to shape genomic divergence ., However , our analysis suggests that different processes have different impacts across the genome , with selection being a probably major contributor ., Therefore , the effects of a local reduction of gene flow and local adaptation are mutually compatible and probably act in concert to shape the genomic landscape of divergence between differentiating parapatric stickleback populations ., We presented multiple lines of evidence for the role of adaptation shaping the genomic divergence patterns between lake-river populations of three-spined sticklebacks ., Aside from adaptive processes , stochastic variation in coalescent times and variable mutation rates could further contribute to the observed heterogeneity of genomic divergence 35 ., In particular , demographic history such as colonization events ( population range expansions ) might lead to a substantial variation in allele frequencies across the genome , possibly mimicking the patterns of adaptive hitchhiking 36 ., Here , we have chosen the genome-wide average as proxy of the underlying demographic history and the effect of random drift on these populations , as detailed demographic information is scarce ., Today , fish migration from the sampled rivers flowing into lake habitats is possible while migration in the opposite direction is likely constrained by physical barriers ( S1 Table ) ., However , as freshwater systems have been subject to recurrent water-level changes during de-glaciation , the spatial context at different stages of population divergence might have fluctuated over the years affecting demographic history of the populations ., Due to pronounced local differences and variable genomic patterns across the sampled continuum of genetic population differentiation we conclude that the main mode of contemporary divergence between parapatric three-spined sticklebacks is associated with population-specific local adaptation ., This is potentially partially mediated by differences in the parasite , as we also found a corresponding signature of isolation by adaptation ., Furthermore , our fine-scale examinations of molecular evolution suggest that some heterogeneity of genomic divergence is also the result of locus-specific differences in gene flow mediated by divergent selection ., Our study has taken an important step towards deciphering the underlying mechanisms responsible for the genomic patterns during speciation , one of the fundamental enigmas in evolutionary biology ., Three-spined stickleback fish were caught from five pairs of lakes and rivers in North America and Northern Europe ( S1 Table and Fig . 1 ) ., Between 12 and 17 fish were screened for macroparasites following established procedures 14 ., Both Shannon diversity indices for each population and jaccard distance between populations were estimated on the basis of 4th square root transformed parasite counts ., Muscle tissue from six sampled individuals from each location was used for DNA extraction ( using a Qiagen DNA Midi Kit following the manufacturer’s protocol for high molecular weight DNA ) and Illumina sequencing following previous methods 31 ., To capture natural variation present in the wild , we randomly picked individual fish for sequencing ( albeit targeting equal sex ratio per population and similar fish sizes across populations ) , thus without pre-selection of any particular morphological or parasitological characteristics ., For each individual , two paired-end libraries ( 100bp reads , average insert size of 140bp and 300bp ) and a mate-pair library ( 50bp reads , average insert gap of 3kb ) were produced , achieving an average depth of coverage of 26x ( S2 Table ) ., Data is deposited in the European Nucleotide Archive ( PRJEB5198 ) ., Raw sequence data was processed and filtered following previous procedures 31 and mapped against the three-spined stickleback reference genome 22 from Ensembl version 68 37 with BWA ( Burrows-Wheeler Aligner ) software 38 ., Mapped reads were further filtered and processed utilizing the Picard toolkit following previous procedures 31 ., SNPs and indels were called with GATKv1 . 6 39 , 40 using concordant SNP calls from SAMtools v0 . 1 . 18 41 for variant recalibration ., Phasing and imputation was performed with BEAGLE v3 . 1 42 ., VCFtools 43 was utilized for processing genotypes ., Positions overlapping with ‘N’s and repeat-masked regions from the Ensembl annotations ( version 68 ) were removed from the final genotype file . Furthermore , variants within 10bp of an indel or indicating copy number variation were also excluded . Copy number variable ( CNV ) regions were identified by deviations in expected read depth with the software CNVnator 44 . More details on the CNV analysis are given in a companion paper submitted by Chain et al . The following analyses were performed on the 20 autosomes , spanning 380 , 547 , 835 sites in the reference genome . After removing masked sites and CNV region and imputing genotypes across 60 individuals , 297 , 437 , 667 sites were reliably genotyped and used for estimating population genetics parameters . We used Illumina’s Golden Gate platform for cross checking genotypes from SNP sites distributed across the genome . Each chromosome held on average 9 ( range 2–21 ) markers and the total of 183 loci were mostly interspersed by at least 50 kb . We found a high overall concordance ( 98% in 12 , 041 comparable sites ) between genotype calls from the Golden Gate assay and our sequencing pipeline . The population genetics estimators of nucleotide diversity ( π and Θ ) and Tajima’s D ( TD ) were calculated with VCFtools v0 . 1 . 11 43 for each of the 10 populations ( S3 Table ) , in addition to the relative divergence ( Weir and Cockerham FST ) and absolute divergence ( Dxy 45 ) estimated for each of the 5 parapatric lake-river pairs ( S4 Table ) ., Numbers of polymorphic sites per population and per population pair are reported in S3–S4 Tables ., To illustrate the relationship amongst all sampled populations , we utilized a set of 1 , 074 , 467 intergenic autosomal polymorphic loci to estimate pairwise divergence ( Weir and Cockerham FST ) and built a neighbor joining tree ., To gain support for the tree topology we randomly down sampled this dataset 100 times to 100 , 000 loci ., For the genome scan , FST was calculated on the full dataset that was further filtered for minor allele frequencies below 25% across each pairwise comparison excluding uninformative polymorphism 46 ., This way we evaluated the divergence between parapatric population pairs on the basis of 691 , 957 to 1 , 227 , 732 sites across the 20 autosomes ., Population genetics estimators were averaged across the genome ( 20 autosomes ) in non-overlapping windows to ensure statistical independence of windows ., We used window sizes of 10 kb and 100 kb and confirmed that results are qualitatively the same ., Diversity estimates have been corrected for the number of sites for which genotypes are available ., Outlier windows were determined by combining an empirical approach with a permutation approach ., First , windows above the top 1% of the empirical distribution were identified as putative outlier windows ., Second , we applied a permutation approach in which loci across the genome were permuted 1 , 000 , 000 times and window estimates of FST were tested against permutations holding the same amount of variable sites ., Putative outlier windows from this permutation approach were identified after adjusting for a FDR of 0 . 01 ., Our final set of outlier windows consisted of those windows that were significant outliers in both approaches ., All statistical procedures and visualizations were implemented in R 47 ., Outlier window positions were compared across the five replicated lake-river comparisons ., To evaluate how many overlapping outlier windows were expected by chance , windows were permutated 10 , 000 times utilizing bedtools 48 ., To approximate the size of regions of exceptional differentiation more in detail , adjacent outlier windows were combined to form larger contiguous divergent regions of extreme differentiation ., In each resulting candidate region , the locus of maximal divergence was determined as a starting point , in which outward steps of 1 kb windows were binned to estimate barrier strength ( b , ref 25 ) ., Margins of divergent regions showing extreme differentiation were determined when b dropped below 1 ( genome-wide average ) in two consecutive 1 kb bins ., This resulted in divergent regions of exceptional differentiation with distinct sizes estimated to the nearest 1 kb ., Divergent regions with sequence coverage ( sequence information accessible , see details above ) spanning less than 50% of their length were excluded from subsequent analyses ., Average sizes of about 50 kb are independent of the initial window size used but specific values reported here are based on the 10 kb window size approach ( Table 1 ) ., We acknowledge that estimates of FST based on allele frequencies can vary depending on samples size 49 ., To reduce variation of estimates between populations we kept the samples size constant at 12 alleles per populations ., Additionally , our analysis did not rely on per site estimates but instead on averages of FST over larger regions ( see above ) ., We evaluated the effect of sample size on our power to describe genomic patterns , detect outlier windows , and define divergent regions in the three following ways ., ( i ) We tested the accuracy of our FST estimates at individual loci by comparing them to estimates based on a larger sample size ., The 183 loci used for validating the genotypes ( see above ) were also used to genotype a larger population sample ( n = 26–59 per population ) to validate allele frequencies and resulting FST estimates ., For all population pairs , the FST estimates based on the sequencing approach with 6 individuals per population ( 12 alleles ) had a significant positive correlation with the FST estimates from the Golden Gate assay using at least 26 individuals ( Pearson correlation , r = 0 . 85 , P< 10-16 , df = 241 , S4 Fig . ) ., ( ii ) We tested the consistency of window FST estimates across the whole range of potential FST values by jack-knifing samples ( S5 Fig . ) ., On average , jack-knifed values ( comparing 10 alleles per population ) had 95% confidence intervals of 0 . 039 up to a maximum of 0 . 175 ., Windows with high FST values ( >0 . 75 ) had even narrower confidence intervals ( average of 0 . 027 and maximum of 0 . 088 ) ., These results support the notion that pronounced differences ( “near-” and “post-fixation” ) can be more reliably detected using our sample sizes than more settled differences ( “pre-fixation” regime ) ., ( iii ) We tested our ability to detect known candidate genes , which highly differentiate between marine and freshwater populations ., For this we utiliz

Introduction, Results and Discussion, Materials and Methods

The patterns of genomic divergence during ecological speciation are shaped by a combination of evolutionary forces ., Processes such as genetic drift , local reduction of gene flow around genes causing reproductive isolation , hitchhiking around selected variants , variation in recombination and mutation rates are all factors that can contribute to the heterogeneity of genomic divergence ., On the basis of 60 fully sequenced three-spined stickleback genomes , we explore these different mechanisms explaining the heterogeneity of genomic divergence across five parapatric lake and river population pairs varying in their degree of genetic differentiation ., We find that divergent regions of the genome are mostly specific for each population pair , while their size and abundance are not correlated with the extent of genome-wide population differentiation ., In each pair-wise comparison , an analysis of allele frequency spectra reveals that 25–55% of the divergent regions are consistent with a local restriction of gene flow ., Another large proportion of divergent regions ( 38–75% ) appears to be mainly shaped by hitchhiking effects around positively selected variants ., We provide empirical evidence that alternative mechanisms determining the evolution of genomic patterns of divergence are not mutually exclusive , but rather act in concert to shape the genome during population differentiation , a first necessary step towards ecological speciation .

A variety of evolutionary forces influence the genomic landscape of divergence during ecological speciation ., Here we characterize the evolution of genomic divergence patterns based on 60 fully sequenced three-spined stickleback genomes , contrasting lake and river populations that differ in parasite abundance ., Our comparison of the size and abundance of divergent regions in the genomes across a continuum of population differentiation suggests that selection and the hitchhiking effect on neutral sites mainly contributes to the observed heterogeneous patterns of genomic divergence ., Additional divergent regions of the genome can be explained by a local reduction of gene flow ., Our description of genomic divergence patterns across a continuum of population differentiation combined with an analysis of molecular signatures of evolution highlights how adaptation shapes the differentiation of sticklebacks in freshwater habitats .

journal.pbio.0060108

Allele-Specific Up-Regulation of FGFR2 Increases Susceptibility to Breast Cancer

FGFR2 ( fibroblast growth factor receptor 2 ) plays a pivotal role both in mammary gland development and in cancer 1 ., The FGFR2 gene encodes a transmembrane tyrosine kinase and can function as a mitogenic , motogenic , or angiogenic factor , depending on the cell type and/or the microenvironment ., Mammary epithelial cells express FGFR2IIIb ( including alternatively spliced exon 9 ) , which binds FGF-7 and FGF-10 , which are normally expressed by surrounding mesenchymal cells ., Mouse models of mammary carcinogenesis have long established the FGF signalling pathway as a major contributor to tumorigenesis 2 , and a mouse mammary tumour virus ( MMTV ) insertional mutagenesis screen for genes involved in breast cancer has identified FGFR2 and FGF10 3 ., In human breast cancer , the expression of FGFR2 has long been known to be elevated in estrogen receptor ( ER ) –positive tumours 4 , which has been confirmed by data analysis performed with the ONCOMINE 3 . 0 array database 5 , 6 ., Likewise both FGF-7 and FGF-10 have been found to be expressed in a proportion of breast cancers 7 , 8 ., Functional studies in cell lines have implicated FGFR2 as playing a role in tumourigenesis , with an alternative splicing in the C-terminal domain of FGFR2 giving rise to a more strongly transforming isoform 9 ., However , as yet , nothing is known about the mechanism by which FGFR2 acts as a risk factor in predisposition to breast cancer ., We examined the functional implication of genetic variation in the FGFR2 haplotype associated with susceptibility to breast cancer and we demonstrate increased gene expression for the risk allele ., Two independent studies have identified FGFR2 as risk factor in breast cancer 10 , 11 ., We have shown that in Europeans , the minor disease-predisposing allele of FGFR2 is inherited as a haplotype of eight single nucleotide polymorphisms ( SNPs ) covering a region of 7 . 5 kb within intron 2 of the gene 10 ( Figure 1 ) , in a haplotype block with no linkage disequilibrium with the coding region of the gene ., Microarray gene expression analysis on the Nottingham City Hospital cohort , using both the Agilent 12 and the Illumina 13 platforms , indicated that FGFR2 is expressed at higher levels by tumours that are homozygous for the minor alleles than by those with the common alleles ( Wilcox p < 0 . 05 ) ., Analysed tumours were all diploid for this region based on array-comparative genome hybridization data 14 ., This correlation was independent of either ER expression or p53 mutation status of the cells ., Quantitative TaqMan PCR analysis confirmed a significant increase in FGFR2 expression in rare homozygotes , as compared to common homozygotes ( Wilcox p = 0 . 028 ) ( Figure 2 ) ., We also examined expression of the FGFR2 ligands FGF-7 , FGF-10 , and FGF-22 , which are usually produced by the surrounding stroma , in 45 normal breast samples as well as the microarray data on tumours , but we found no correlation with genotype ., Furthermore , FGFR2 displays a very complex splicing pattern with the most commonly expressed variants of the N terminus of the gene either including exons 1 , 2 , and 3 or including exons 1 and 2 , but lacking exon 3 ., Again , no correlation was observed between genotype and the presence or absence of exon 3 ., Thus , the risk genotype correlates with FGFR2 expression itself , rather than affecting its function through receptor-ligand interactions ., This correlation suggests that the functional SNPs map to a regulatory region within the gene , possibly by altering one or more transcription factor binding sites ., Interactions between proteins from nuclear extracts and DNA were examined for the eight most strongly disease-associated alleles ( Figure 1 ) ., Two of these candidate functional SNPs showed distinct binding patterns in electrophoretic mobility shift assays ( EMSA ) ., The common allele of rs7895676 ( FGFR2–33 ) formed strong protein–DNA complexes with nuclear extracts from the breast carcinoma cell lines HCC1954 ( Figure 3A ) and PMC42 and from HeLa cells ( unpublished data ) , whereas no binding was detected on the minor allele ., Competition studies and supershift experiments identify the bound protein as C/EBPβ ( Figure 3A ) ., We note that the FGFR2–33 sequence has considerable homology to the C/EBPβ binding site from the interleukin 6 ( IL-6 ) promoter 15 ( Figure 3C ) ., The heterogeneity of the observed protein–DNA complexes is most likely due to the presence of multiple C/EBPβ isoforms ., For rs2981578 ( FGFR2–13 ) , both alleles give rise to a strong protein–DNA complex in HCC1954 cell extracts ., However , a second more slowly migrating complex was only seen on the rarer genotype ( Figure 3B ) ., Interestingly , both alleles are able to compete for both bands , suggesting that the formation of the upper complex depends on the presence of the lower complex ., Inspection of the FGFR2 DNA indicated the presence of a perfect octamer binding site immediately adjacent to the SNP , while the SNP itself lay within a sequence with homology to Runx binding sites ( Figure 3C ) ., Competition studies and incubation with specific antisera shows that both alleles bind Oct-1 , while only the minor allele binds Oct-1 and Runx2 in HCC1954 nuclear extracts ( Figure 3B ) , as well as in PMC42 cells ( Figure S1 ) ., To establish whether or not these sites were occupied in vivo , we carried out chromatin immunoprecipitation ( ChIP ) experiments using the ER+ breast cancer cell lines HCC70 and T47D , which are homozygous for the minor and the common FGFR2 alleles , respectively ., In addition , we confirmed that these cell lines were diploid for the FGFR2 locus and only expressed the epithelial-specific isoform FGFR2IIIb 16 ., The ChIP analysis was carried out on homozygous cell lines , because the SNP overlapping the C/EBPβ site lies in a repetitive region for which the different alleles could not be distinguished reliably by TaqMan PCR ., A representative experiment is shown in Figure 3D ., After Runx2-precipitation , the FGFR2–13 site is enriched 2-fold for the minor versus the common allele , confirming the EMSA results ., Western blotting indicated that Runx2 is more abundant in T47D cells , thus confirming that differential ChIP in the two cell lines is due to the presence of the SNP ., Oct-1 precipitation did not yield enrichment of FGFR2–13 for either allele ., The Oct-1 epitope may either be sequestered within a higher-order complex or the antisera used do not work efficiently in a ChIP assay ., On the FGFR2–33 site , we observed a 1 . 7-fold enrichment of C/EBPβ binding on the common allele ., In addition , we observe that C/EBPβ can also bind to the minor allele , although less efficiently ., Both cell lines contain comparable amounts of C/EBPβ as judged by Western blotting ( unpublished data ) ., In conclusion , both the C/EBPβ and the Runx2 binding sites are occupied in vivo ., To test whether differential protein binding could alter the ability of the susceptibility alleles to activate transcription , we multimerised oligonucleotides overlapping both the Oct-1/Runx2 and the C/EBPβ binding sites , cloned these in both orientations upstream of the luciferase reporter gene in pGL3Enh ( Figure 4A ) , and assayed them in three breast cancer cell lines ( PMC42 , HCC70 , and T47D ) ., Transfections were carried out in triplicate and repeated at least twice for each cell line ., A representative transfection into HCC70 cells is shown in Figure 4B ( see Figure S2 for PMC42 and T47D ) ., In all three cell lines tested , the minor allele at the Oct-1/Runx2 site stimulated transcription 2- to 5-fold over the common allele , independent of orientation , with the average being just above a 3-fold increase ( p < 0 . 01 ) ., In contrast , the minor and common alleles of the multimerised C/EBPβ binding site did not show a consistent pattern of activation relative to each other ., It varied with the cell lines and the orientation in which constructs were tested ., Nevertheless , relative to the parental vector , the common allele always showed transcriptional activation ., Compared to the common allele , the minor allele was either not significantly different or gave rise to a smaller degree of activation ., However , in the latter case , the rare allele still activated transcription significantly above the enhancer-only construct ( p < 0 . 01 ) ., Presumably this reflects the fact that the minor allele of FGFR2–33 still binds C/EBPβ above background levels in vivo ( Figure 3D ) ., By comparing the two different sites , we found that for Oct-1/Runx2 the minor allele was more active , while for C/EBPβ , the common site yielded higher levels of transcription in the majority of experiments ., Hence their effects were opposing ., We therefore assayed a synthetic construct consisting of single sites for C/EBPβ , Oct-1 , and Runx2 ., In this arrangement , the effect of Oct-1/Runx2 clearly predominates , with the minor allele expressed at higher levels , reflecting the situation at the endogenous locus ., The data presented here lead us to conclude that the Oct-1/Runx2 binding site is the dominant determinant of differential expression between the common and minor haplotypes of FGFR2 ., Although Runx2 is a master regulator of osteoclast-specific transcription , Runx2 also plays an important role in mouse mammary gland–specific gene expression 17 , where Runx2 activity is dependent on Oct-1 18 ., It is intriguing to note that in bone cells , overexpression of constitutively active FGFR2 leads to increased levels of Runx2 mRNA 19 ., FGFR2 in turn is responsive to Runx2 in osteoclasts via the OSE2 ( osteoclast specific element 2 ) in its promoter 20 ., The description here of a Runx2 site in the FGFR2 gene that is occupied in breast cancer cells , suggests that in the presence of the minor genotype , a similar positive feedback loop could also be established in breast cells ., The role of the C/EBPβ binding site on FGFR2 expression has been harder to define ., The common allele binds C/EBPβ more tightly and activates transcription more strongly in most cases ., Yet in a composite construct the activity of the Oct-1/Runx2 site dominates ., This may be because C/EBPβ can directly bind to and synergize with Runx2 21 ., Thus , on the minor genotype , Oct-1 and Runx2 are present and able to synergize with the C/EBPβ bound ( as suggested from the ChIP experiments ) , giving rise to higher levels of transcriptional activation ., This is supported by the finding that a single copy of the C/EBPβ/Oct-1/Runx2 site gives rise to higher levels of activation than a concatemerized Oct-1/Runx2 site with six potential interaction sites ( Figure 4A ) ., A potential role for C/EBPβ in tumour etiology is supported by the observation that C/EBPβ is highly overexpressed in malignant human breast cells 22 ., In conclusion , our evidence supports Oct-1/Runx2 as the probable primary determinant of activity , with C/EBPβ contributing to the risk haplotype ., The increased risk in breast cancer conferred by the FGFR2 allele is predominant for ER+ breast tumours , while there is no significant increase in risk for ER– tumours ., Genome-wide analysis of ER binding sites has revealed three potential ER binding sites within the FGFR2 gene 23 , and ER and Oct-1/Runx2 may cooperate to increase gene expression ., This is consistent with findings that Oct and ER sites often cluster 23 ., The risk conferred by the disease-associated genotype may also depend on the signalling potential of FGFR2 in ER+ cells ., FGF-7 is over-expressed only in breast tumours that are ER+ 8 ., Elevated levels of FGFR2 may then contribute to the establishment of an autocrine signalling loop , reducing the cells propensity to undergo apoptosis 24 ., Alternatively , paracrine signalling by mesenchymally or luminally derived FGF-7 or -10 on cells overexpressing FGFR2 may also drive cell proliferation ., To our knowledge , this is the first functional study on the risk loci recently identified for breast cancer ., Our study demonstrates that SNPs identified by whole-genome scans can be used a valid starting points for studying the underlying biology of cancer ., SNPs identified in other whole-genome scans for the genetic basis of complex diseases also primarily map in intronic or intergenic regions ., Our observation that an identified SNP regulates the expression of the risk allele is therefore likely to be a common theme ., Breast cancer is one of the most common cancers in the developed world ., The FGFR2 minor allele carries only a small increase in risk and acts as part of a spectrum of risk factors ., However , it has a high minor allele frequency ( 0 . 4 ) , and FGFR2 is therefore likely to contribute to the incidence of breast cancer in many individuals ., DNA from the 170 tumour samples was genotyped using a fluorescent 5′ exonuclease assay ( TaqMan ) and the ABI PRISM 7900 Sequence Detection Sequence ( PE Biosystems ) in 384-well format ., Duplicate samples were included to assess concordance and quality of genotyping ., The genotyping assay was designed for rs2981582 , which tags the whole haplotype block associated with the disease 10 ., Analysis was performed on total RNA from breast tumour cases ., cDNA was prepared with the TaqMan Reverse Transcription Reagents kit ( Applied Biosystems ) using random hexamers , according to the manufacturers instructions ., Expression levels were determined using a TaqMan Gene Expression Assay ( Hs00240796_m1 , Applied Biosystems ) and normalized to four different housekeeping genes ., To assess whether there were significant statistical differences between the expression levels across the genotype groups we used a Wilcoxon test , fitted using the R statistical framework ., Elsewhere , Students t-tests were carried out using Microsoft Excel ., Breast cancer cell lines HCC1954 , HCC70 , T47D , and PMC42 were cultured in RPMI supplemented with 10% foetal calf serum and penicillin/streptomycin under standard conditions ., These cell lines have been characterised extensively , and karyotypes are available at the Cancer Genomics Program of the University of Cambridge ( http://www . path . cam . ac . uk/~pawefish ) ., Small-scale nuclear extracts and bandshifts were carried out as previously described 25 , except that Complete Protease Inhibitors ( Roche ) were used ., In supershift experiments , polyclonal antisera against Oct-1 ( sc-232x ) , Runx2 ( sc-10758x ) , and C/EBPβ ( sc-150x ) were obtained from Santa Cruz Biotechnology , Inc and up to 8 μl were added per reaction , unless otherwise stated ., Oligonucleotides ( Table S1 ) were annealed to complementary strands , and the resulting BamHI overhangs filled in with Klenow enzyme , using radiolabelled α32PdCTP ( GE Healthcare , UK ) ., Primers were designed using Primer Express ( Applied Biosystems ) and Lasergene ( DNA Star ) to amplify regions of up to 100 bp comprising the SNPs of interest , plus one negative control ( region of the genome not suspected to bind any of the transcription factors of interest ) ( Table S1 ) ., PCR amplification was carried out with Power SYBR Green Mastermix ( Applied Biosystems ) , using 2 μl of precipitated and purified DNA as described 23 ., The antisera were as in the EMSAs , except for C/EBPβ , which was a polyclonal serum from Abcam , UK ., The pGL3-Enhancer vector ( Promega ) was linearized with BglII and re-circularised in the presence of annealed oligonucleotides ( Table S1 ) ., All constructs were verified by sequencing ., DNA was prepared using Qiagen kits and transfected into tumour cell lines cultured in 24-well plates ., Per well , 500 ng of reporter and 100 ng CMV-β-galactosidase plasmid were tranfected using 2 μl of Fugene 6 ( Roche ) , harvested 36–48 h later and extracts prepared using 100 μl Promega lysis buffer ., Luciferase and β-galactosidase activity in 25 μl was measured using Promega reagents ., Results are given as ratios of luciferase over β-galactosidase activity .

Introduction, Results, Discussion, Materials and Methods

The recent whole-genome scan for breast cancer has revealed the FGFR2 ( fibroblast growth factor receptor 2 ) gene as a locus associated with a small , but highly significant , increase in the risk of developing breast cancer ., Using fine-scale genetic mapping of the region , it has been possible to narrow the causative locus to a haplotype of eight strongly linked single nucleotide polymorphisms ( SNPs ) spanning a region of 7 . 5 kilobases ( kb ) in the second intron of the FGFR2 gene ., Here we describe a functional analysis to define the causative SNP , and we propose a model for a disease mechanism ., Using gene expression microarray data , we observed a trend of increased FGFR2 expression in the rare homozygotes ., This trend was confirmed using real-time ( RT ) PCR , with the difference between the rare and the common homozygotes yielding a Wilcox p-value of 0 . 028 ., To elucidate which SNPs might be responsible for this difference , we examined protein–DNA interactions for the eight most strongly disease-associated SNPs in different breast cell lines ., We identify two cis-regulatory SNPs that alter binding affinity for transcription factors Oct-1/Runx2 and C/EBPβ , and we demonstrate that both sites are occupied in vivo ., In transient transfection experiments , the two SNPs can synergize giving rise to increased FGFR2 expression ., We propose a model in which the Oct-1/Runx2 and C/EBPβ binding sites in the disease-associated allele are able to lead to an increase in FGFR2 gene expression , thereby increasing the propensity for tumour formation .

Recently , a number of whole-genome association studies have identified genes that predispose individuals to common diseases such as cancer ., The challenge now is to understand how the identified risk loci contribute to disease , since the majority of these loci are located within introns ( which are discarded after transcription ) and intergenic regions , and therefore do not change the coding region of nearby genes ., This manuscript describes how two single–base pair changes in intron 2 of the FGFR2 ( fibroblast growth factor receptor 2 ) gene , “the top hit” of the breast cancer susceptibility study , exert their function ., We find that the changes alter the binding of two transcription factors and cause an increase in FGFR2 gene expression , thus providing a molecular explanation for the risk phenotype ., This is the first functional study , to our knowledge , of the risk loci identified for breast cancer in a whole-genome scan and demonstrates that these studies can be used as valid starting points for studying the underlying biology of cancer .

genetics and genomics

Recent whole-genome scans have identified novel risk genes for many common diseases, challenging researchers to determine how these genes contribute to disease. A new study provides molecular insights into a breast cancer risk factor.

journal.pcbi.1004235

Quantification of Interactions between Dynamic Cellular Network Functionalities by Cascaded Layering

Complex biochemical reaction networks serve a broad variety of different tasks within the cell ., Systems Biology researchers apply a range of systems analysis techniques to these networks to identify and model functional subsystems and their interaction structure ., In the context of biomolecular networks , the subsystems that can be identified often have biological interpretations: for example , the heat shock response and the chemotaxis pathways represent two functional subsystems within a model describing the complete biomolecular reaction network of Escherichia coli; the synthesis pathways of individual products represent distinct functional subsystems within a metabolic network; and so on ., Other functional subsystems may also have system-theoretic interpretations: for example , interacting , distributed feedback control mechanisms; or subsystems which can sense , compute , or actuate on the cell and its environment ., This tangle of different objectives within the same network leads to trade-off situations: evolutionary or synthetic changes to one functional subsystem can lead to declining performance or unexpected side effects with respect to another ., A fundamental challenge of Systems Biology is to not only establish the behaviour of each functional subsystem in isolation , but also to understand how they dynamically influence one another ., This problem is particularly acute when applying the modelling and analysis tools of Systems Biology to adapt and redesign modular biomolecular networks in Synthetic Biology 1–3 ., The dynamics of many functional subsystems , whether evolved biochemical networks or synthetic devices , often do not proceed as modelled when integrated into a cell due to their interactions with one another and the environment of their cellular host ., Possible sources of nonlinear interactions between pairs of functional subsystems and the cellular environment include retroactivity in genetic 4–6 and signalling 7 networks , crosstalk between parallel signalling pathways 8 , 9 , and the coupling of multiple transcription or translation rates through competition for shared resources 10–12 ., In each of these settings , the change in input–output behaviour of a given subsystem upon integration with its context is examined ., There are two complementary goals of this paper ., First , we investigate the behaviour of each subsystem between the two extremes of ‘isolated’ or ‘integrated’ , when it is integrated with any subset of the other subsystems ., The second goal , which is achieved as a consequence of the first , is to then systematically quantify each of the pairwise interactions between the network’s subsystems ., The approach we will take in this work is to define a functionality as a group of reactions which corresponds to an identified functional subsystem of a biomolecular network ., The reactions that determine each functionality can be selected either through biological insight , or by applying existing computational approaches such as elementary flux mode ( EFM ) analysis 13–15 ( see also Results ) ., We characterise the behaviour , or effect , of a functionality as the solution of an ordinary differential equation ( ODE ) model determined by the particular group of reactions ., This approach exploits the recently-introduced decomposition technique known as layering 16 , 17 ., As depicted in Fig 1 , such an approach is distinct from established modular approaches to network decomposition , which are characterised by identifying sets of species with a high connectivity inside the module , and significantly lower connectivity to species in other modules 18–26 ., While often many species and reactions in a given network are implicated in multiple network functions , these modular approaches generally do not allow for such a high degree of overlap between modules ., For example , if a network of two pathways responds to two external signals with a single output species , a modular decomposition of this network requires the common output species to be assigned to a module representing exactly one of the pathways , or potentially to an additional separate module ., Either way , the input–output behaviour of both pathways cannot be easily defined ., However , in the layered framework , the common output is associated with both layers , and hence the output of each layer can be defined in terms of its biological function ., Thus , in some cases , layers are preferable to modules for defining the functional subsystems of the network , since the layered framework explicitly allows for overlap in species and reaction subsets 16 , 17 , as will be illustrated further in ‘Mapping Functionalities to Layers’ below ., Most importantly , we make it explicit that the behaviour of any functionality also depends on the other functionalities with which it is integrated , to which we refer as the context of the functionality ., This contextual dependence is formalised by developing a notational framework that will unambiguously define a functionality’s behaviour in a particular context ., The resulting concept of conditional dynamics will be key to our understanding of each functionality as being defined only in the context of others , allowing us to systematically investigate the interdependence of an entire network’s functionalities ., The subsequent aim of this framework is to characterise all of the interactions between each pair of functionalities , each of which is also context-dependent ., Our approach is a formalisation and extension of previous investigations into additive ( i . e . independent ) , synergistic , or antagonistic subsystem interactions ., Examples of these phenomena include the cytokine secretion by macrophages in response to stimulation with different sets of ligands 27 , the response of bacteria to different combinations of drugs 28 , or calcium signalling responses to different stimuli 29 ., Importantly , we demonstrate how the strength and the type of interactions between functionalities depends on mediated indirect interactions with the other functionalities comprising their context ., The relationship of our approach to the concepts in 27–29 is further discussed in the section “Calculating With Layers” ., In addition to the previous literature on context-dependent dynamics , our theoretical framework is also related to steady-state methods ., For instance , the third of our examples will exploit EFMs , a technique designed to analyse the steady-state flux distribution in metabolic networks 13–15 ., Furthermore , modular and hierarchical control/response analysis is concerned with the different behaviour of subsystems both in isolation and integrated in larger systems , with particular reference to the steady-state responses of biochemical networks to parameter perturbations 6 , 30–34 ., The key distinction between our method and these is that we analyse the dynamics of kinetic models 35 , 36 represented by sets of ODEs , rather than steady states ., Moreover , our method is not based on linearisation , allowing us to adequately capture nonlinear interactions between functionalities ., Quantifying the dynamic interactions of each pair of functionalities in all possible contexts requires multiple ODE simulations; for its practical applicability , it is important to minimise the computational effort involved ., This paper is structured as follows: in the Methods section we show how to use a layered decomposition to identify the incremental effect of a functionality , making its context-dependence explicit ., We continue by defining the interdependence , or mutual dynamics , between any two functionalities ., We summarise this interdependence by the incompatibility and the cooperativity between functionalities ., In the final part of the Methods section we describe how to analyse all functionalities and their dependencies with minimal computation ., We demonstrate our method on three familiar biomolecular networks in the Results section ., The first example is of two signalling pathways with two crosstalk mechanisms , in which we use our approach to quantify the nonlinear interactions between crosstalk mechanisms ., In the second example we analyse an unstable pathway stabilised by two integral feedback loops , finding the interactions between each controller and the pathway , and also between the controllers ., Finally , we consider the glycolytic pathway in Saccharomyces cerevisiae , with functionalities defined by an EFM analysis ., We apply our approach to compare how different knock-out strategies in metabolic engineering influence the yield of ethanol , industrially relevant in biofuel production ., Consider a biochemical reaction network with NX species Xi of time-varying concentrations xi ( t ) , taking part in NR reactions R1 , … , RNR , each of which proceeds at the concentration-dependent rate vj ( x1 , … , xNX ) for j = 1 , … , NR ., Forming vectors v = ( v1 , … , vNR ) T and x = ( x1 , … , xNX ) T gives an ODE model of the system, x ˙ ( t ) = S v ( x ( t ) ) , x ( 0 ) = x 0 , ( 1 ), where the stoichiometric matrix S maps reaction rates to the rate of change of concentrations ., The layered decomposition strategy 16 , 17 defines NL new stoichiometric matrices S 1 , … , S N L such that S = S 1 + ⋯ + S N L , and defines NL associated state variables xl taking values x l ( t ) ∈ ℝ N X ( see Fig 1B ) ., Each layer’s state xl has dynamics, x ˙ l ( t ) = S l v ( x 0 + x l ( t ) + ∑ k ≠ l x k ( t ) ) , ( 2 ), from initial conditions xl ( 0 ) = 0 , for l = 1 , … , NL ., The original state’s dynamics are recovered by summing the layers’ states x ( t ) = x 0 + ∑ l = 1 N L x l ( t ) ., Denote by r = rank ( S ) the dimension of the original system , which defines the dimension of the manifold in ℝ N X in which x ( t ) evolves ., It follows that rl = rank ( Sl ) defines the dimension of the state space of each layer ., Hence , even though the state space of each layer is also embedded in ℝ N X , each layer is a lower-dimensional system than the original system if rl < r ., In our previous work , we have applied this decomposition strategy by choosing the matrices Sl to reflect timescale separation 17 , and to reflect the propagation of steady-state responses to parametric perturbations 16 ., A feature of both of these approaches was that , in the form ( 2 ) , each layer’s dynamics depend on all other layers’ states ( as in Fig 1B , for example ) ., Consequently , all layers had to be numerically integrated together , and the effect of one specific layer on all others could not be easily determined ., Also , the approach was constrained to define layers by strict partitions of the reaction set , somewhat limiting its flexibility to capture the widest possible range of functional subsystems ., In this article , we significantly extend the layering framework in two ways ., First , we introduce the concept of functionalities , which are possibly overlapping sets of reactions working together for a common purpose ., To enforce a cascade structure between the functionalities , we adapt the layered dynamics corresponding to each functionality depending on its position in the cascade ., The following section will use this cascade structure to define the incremental dynamic effect of each functionality ., Let a functionality Fi of a network be defined for i = 1 , … , NL to be a subset of N R i reactions Fi ⊆ {R1 , … , RNR} necessary to fulfil a given task of the network , where superscript integers index functionalities and their properties ., It is assumed for the remainder of this section that these subsets are given , and that all reactions take part in at least one functionality ., The question of how to choose each subset Fi ⊆ {R1 , … , RNR} remains out of the scope of this work ., Nevertheless , there are numerous non-modular decomposition strategies taken in recent related research that we can use to justify this definition of a functionality ., For example , Oishi and Klavins 37 identify control blocks as specific groups of reactions , connected by shared species ., Kurata et al . 38 identified reaction groups forming ‘flux modules’ in the Escherichia coli heat shock response system ., Similarly , the decomposition of signalling networks into component pathways exhibiting crosstalk 8 also identifies functionalities as groups of reactions ., Finally , we can also consider elementary flux modes ( EFMs ) of metabolic networks 13 as being sets of reactions with the specific ‘task’ of converting one or more substrates into given products ., Several of these examples are explored further in the Results section of this paper ., In this section , we will assume that the functionalities are ordered by their index F 1 , … , F N L . We first identify the dynamics of the isolated functionality F1 as the dynamics of a biomolecular network consisting of only the reactions associated with F1 ., We then identify the conditional dynamics of the next functionality in the cascade as the effect of extending the pre-existing network with the reactions in the new functionality ., First consider , without loss of generality , the network defined by only the subset of reactions making up functionality F1 ⊆ {R1 , … , RNR} , taken in isolation from the other reactions ., For given initial conditions x0 , we now identify the isolated dynamics of this functionality as the solution to the layer, x ˙ 1 ( t ) = S 1 v ( x 0 + x 1 ( t ) ) , x 1 ( 0 ) = 0 ., ( 3 ), Here , the stoichiometric matrix S1 is defined, S j k 1 = { S j k R k ∈ F 1 , 0 otherwise ,, by copying the columns of the original stoichiometric matrix S in ( 1 ) corresponding to the reactions in F1 and setting the other columns to zero ., We will denote this trajectory x1 = L ( F1 ) , where the notation L represents a map from the functionality F1 to the solution x1 of the dynamics ( 3 ) from initial conditions x1 ( 0 ) = 0 ., Note that L ( F1 ) depends on the specific initial condition x0 of the network ( 3 ) , which is in general distinct from the initial condition x1 ( 0 ) = 0 of the layer’s state ., To make this dependence explicit , it is sometimes helpful ( see Examples 2 and 3 ) to define a ‘zero layer’ F0 with constant trajectory L ( F0 ) = x0 ., We can then make clear that L ( F1 ) is dependent on the initial conditions by writing it as L ( F1∣F0 ) ., The layered framework also implies that the absolute concentrations in this network are modelled by the translated trajectory x0 + L ( F1∣F0 ) ., We next consider extending the functionality F1 by combining it with the reactions in F2 ., The extended network can be simulated through a similar process to the original network above , as follows ., Define S1 , 2 as, S j k 1 , 2 = { S j k R k ∈ F 1 ∪ F 2 , 0 otherwise ,, considering only the reactions in at least one of F1 or F2 ., Using S1 , 2 we can then simulate the layer corresponding to the extended network, x ˙ 1 , 2 ( t ) = S 1 , 2 v ( x 0 + x 1 , 2 ( t ) ) , x 1 , 2 ( 0 ) = 0 , ( 4 ), the solution of which can be written L ( F1 , F2∣F0 ) = x1 , 2 ., This denotes the trajectory of the combined functionalities F1 and F2 ., The fact that each of ( 3 ) and ( 4 ) are layers with states in ℝ N X implies that we can calculate the difference between each of the trajectories ., This difference is clearly interpreted as the incremental effect of extending a network made up of the initial conditions F0 and the isolated functionality F1 , by also including F2 ., We thus define, L ( F 2 | F 1 , F 0 ) = L ( F 1 , F 2 | F 0 ) - L ( F 1 | F 0 ) ( 5 ), as the conditional dynamics of F2 , given the specified context of F1 and the initial condition layer F0 ., However , rather than simulating the layer ( 4 ) representing the combined functionalities , we may further exploit the layered framework described above to directly simulate L ( F2∣F1 , F0 ) ., Suppose we already have x1 = L ( F1∣F0 ) , found as the solution to the dynamics ( 3 ) ., We now define the layer, x ˙ 2 ( t ) = S 1 , 2 v ( x 0 + x 1 ( t ) + x 2 ( t ) ) - S 1 v ( x 0 + x 1 ( t ) ) , x 2 ( 0 ) = 0 , ( 6 ), with S1 and S1 , 2 given above ., Note that this layer is downstream of ( 3 ) , since it depends on the state x1 ., It is clear from summing the vector fields in ( 3 ) and ( 6 ) that the sum ( x1 + x2 ) of the layers’ states follows exactly the same dynamics as the combined network’s state x1 , 2 in ( 4 ) ., Thus , since x2 = x1 , 2 − x1 it follows that the dynamics ( 6 ) directly simulate L ( F2∣F1 , F0 ) , with the input L ( F1∣F0 ) simulated by ( 3 ) ., We can rewrite the dynamics ( 6 ) corresponding to the simulation of L ( F2∣F1 , F0 ) as, x ˙ 2 ( t ) = S 2 v ( x 0 + x 1 ( t ) + x 2 ( t ) ) + S 1 v a l t ( x 0 + x 1 ( t ) , x 2 ( t ) ) ( 7a ), where S2 = S1 , 2 − S1 corresponds to the reactions in F2\\F1 , and, v a l t ( x 0 + x 1 , x 2 ) = v ( x 0 + x 1 + x 2 ) - v ( x 0 + x 1 ) ( 7b ), are the rates of the ‘altered reactions’: the rates of reactions in F1 which are modified by the presence of F2 ( shown as broken green arrows in Fig 1C ) ., This description allows us to see the degree to which F2 is ‘downstream’ of F1 ., For example , if valt = 0 , then we can say that the reactions in F1 are independent of those in F2 and that F2 is strictly downstream of F1 ., Note that , especially for larger networks , many of the altered reaction rates in valt are zero and can be omitted ( see Example 3 ) , simplifying simulation ., Given that the trajectory of x1 is already determined from simulating ( 3 ) , we can simulate either ( 6 ) or ( 7a ) , using x1 ( t ) as a time-dependent input to obtain the conditional dynamics L ( F2∣F1 ) ., The latter approach is taken in our examples ( see Results section ) ., The definitions above easily extend to larger combinations of functionalities ., In full generality , we can consider the network defined by the combination of n1 functionalities F 1 , … , F n 1 , and its extension through the additional n2 functionalities F n 1 + 1 , … , F n 1 + n 2 . By writing F ‾ 1 = F 1 ∪ ⋯ ∪ F n 1 and F ‾ 2 = F n 1 + 1 ∪ ⋯ ∪ F n 1 + n 2 , the definitions above can apply in the simulation of, L ( F 1 , … , F n 1 | F 0 ) = L ( F ¯ 1 | F 0 ) , L ( F n 1 + 1 , … , F n 1 + n 2 | F 0 , F 1 , … , F n ) = L ( F ¯ 2 | F 0 , F ¯ 1 ) ., Here we have defined a notation for the trajectory L ( F 1 , … , F n 1 ∣ F 0 ) of the biochemical network made up of the reactions which comprise an arbitrary combination of functionalities F 1 , … , F n 1 . We have also defined the change in trajectory L ( F n 1 + 1 , … , F n 1 + n 2 ∣ F 0 , F 1 , … , F n 1 ) incurred by extending that network with the additional reactions in F n 1 + 1 , … , F n 1 + n 2 . Finally , we have shown how to identify the dynamical systems that can simulate these trajectories ., Consider the two layers L ( F2∣F1 ) and L ( F2 ) that both describe the effect of the functionality F2 ., This effect is different depending on the presence or absence of F1 ., The difference between these two trajectories defines how the presence of F1 changes the behaviour of F2; that is , the dependence of F2 on F1 ., We will now demonstrate how our layered analysis allows us to define the interdependence between two functionalities , thereby capturing the nonlinear effects arising from modelling a biomolecular network as being constructed from a combination of functional subsystems ., In order to quantify the interactions between functionalities , we can exploit the layered formulation above ., For simplicity , from this point on we suppress the F0 notation , with the acknowledgement that all of the trajectories depend on the system’s initial conditions L ( F0 ) = x0 ., The definition of conditional dynamics in ( 5 ) implies that L ( F1 , F2 ) = L ( F2∣F1 ) + L ( F1 ) ., This represents a layered cascade , where the dynamics of an integrated network are the linear combination of the conditional dynamics of its functionalities ., There are two natural questions associated with this approach ., First , how is the contribution of functionality F2 , considered in isolation , different from the conditional dynamics of F2 when integrated with F1 ?, Secondly , how is the behaviour of the integrated F1 , F2 network different from the linear combination of the isolated functionalities ?, That is , how different is L ( F1 , F2 ) from L ( F1 ) + L ( F2 ) ?, The answers to these two questions are the same ., We denote the error incurred by approximating the integrated system as the linear combination of the isolated dynamics by the quantity M ( F1; F2 ) , defined as, M ( F 1 ; F 2 ) = L ( F 1 ) + L ( F 2 ) - L ( F 1 , F 2 ) ,, which we call mutual dynamics ., This can be interpreted as the nonlinearity that arises from integrating the two functionalities together ., Note that , since M is defined symmetrically , we can use ( 5 ) to rewrite M as, M ( F 1 ; F 2 ) = L ( F 2 ) - L ( F 2 | F 1 ) = L ( F 1 ) - L ( F 1 | F 2 ) ., ( 8 ), Therefore M measures how the function of F2 ( or F1 ) is changed when considered in the context of F1 ( or F2 ) ., Thus M ( F1; F2 ) is a symmetric measure of the interdependence between the two functionalities ., We use ( 8 ) to calculate the the mutual dynamics between two functionalities , which requires us to first obtain either both trajectories L ( F1 ) and L ( F1∣F2 ) , or alternatively both trajectories L ( F2 ) and L ( F2∣F1 ) ., These trajectories can be simulated , as described in the previous section , or calculated by the methods described in ‘Reducing Computational Burden’ below ., We have been careful to make explicit through our Bayesian-style notation that the dynamics of all functionalities are context-dependent ., It is also the case that the interdependence between any two functionalities is context-dependent ., Therefore , we need to extend the definition of mutual dynamics to consider how the interdependence between F1 and F2 is dependent on the wider context of the network , which we denote by another functionality , F3 ., Similarly to the definition above , we can define the conditional mutual dynamics between F1 and F2 , given F3 , with the formula, M ( F 1 ; F 2 | F 3 ) = L ( F 1 | F 3 ) + L ( F 2 | F 3 ) - L ( F 1 , F 2 | F 3 ) ,, to quantify the difference between the dynamics of the integrated and isolated functionalities F1 and F2 , in the context of F3 ., As before , this can also be expressed in terms of layered dynamics as, M ( F 1 ; F 2 | F 3 ) = L ( F 2 | F 3 ) - L ( F 2 | F 1 , F 3 ) = L ( F 1 | F 3 ) - L ( F 1 | F 2 , F 3 ) ,, to quantify how the effect of F2 on its context changes with the presence of F1 , and vice versa ., The geometric intuition underlying the conditional mutual dynamics can be seen in Fig 2 . The key interpretation of M is that it captures the nonlinearities that arise from combining F1 and F2 into a single network , conditioned on F3 if necessary ., The conditional mutual dynamics M ( F1; F2∣F3 ) is a time-varying , vector trajectory ., We can base on M the following time-varying scalar , which we call the incompatibility and denote I ( F1; F2∣F3 ) with formula, I ( F 1 ; F 2 | F 3 ) = ∥ M ( F 1 ; F 2 | F 3 ) ∥ ∥ L ( F 1 | F 3 ) + L ( F 2 | F 3 ) ∥ ., ( 9 ), Here , ‖ ., ‖ represents the Euclidean norm ., In this paper we use the unweighted Euclidean norm , but in certain cases it might be appropriate to introduce a weight , for example if the concentrations of the species in a network are at different orders of magnitude ., One might also decide to set the weight of certain intermediate species of limited interest to zero ( see below ) ., This incompatibility measures the relative size of the error made by approximating the integration of two functionalities as the sum of their individual behaviour ., To gain some intuition about this number , we can consider a number of special cases ., If I ( F1; F2∣F3 ) = 0 , then this indicates that the trajectory of the integrated functionalities is simply the sum of the isolated functionalities’ trajectories: L ( F1 , F2∣F3 ) = L ( F1∣F3 ) + L ( F2∣F3 ) ., If the incompatibility is nonzero but small then L ( F1 , F2∣F3 ) ≈ L ( F1∣F3 ) + L ( F2∣F3 ) is a reasonable approximation , since the incurred error is relatively small ., However , if I is of significant size , then the dynamics of the integrated functionalities can be expected to significantly differ from their individual behaviours ., Besides I , which measures the relative size of the mutual dynamics M , the direction of M is also important , since this determines if the integration of two functionalities together enhances or attenuates their individual dynamics , or causes other effects ., We define the cooperativity C as the cosine of the angle between −M and the sum of the isolated layers:, C ( F 1 ; F 2 | F 3 ) = - M ( F 1 ; F 2 | F 3 ) · ( L ( F 1 | F 3 ) + L ( F 2 | F 3 ) ) ∥ M ( F 1 ; F 2 | F 3 ) ∥ ∥ L ( F 1 | F 3 ) + L ( F 2 | F 3 ) ∥ , ( 10 ), with ⋅ denoting the scalar product ., See Fig 2 for a geometric representation of C . Note that when using a weighted norm , the scalar product should be weighted accordingly ., Again , we consider a number of special cases ., Suppose that C ( F1; F2∣F3 ) equals or is close to minus one , so that M is approximately parallel to and pointing in the same direction as L ( F1∣F3 ) + L ( F2∣F3 ) ., In this case the integrated dynamics can be approximated as an attenuation of the isolated behaviour, L ( F 1 , F 2 | F 3 ) = L ( F 1 | F 3 ) + L ( F 2 | F 3 ) - M ( F 1 ; F 2 | F 3 ) ≈ ( 1 - I ( F 1 ; F 2 | F 3 ) ) ( L ( F 1 | F 3 ) + L ( F 2 | F 3 ) ) ., Conversely , if the cooperativity equals or is close to one , the two functionalities enhance each other , in the sense that we can approximate the integrated behaviour as an amplification of the isolated behaviours, L ( F 1 , F 2 | F 3 ) ≈ ( 1 + I ( F 1 ; F 2 | F 3 ) ) ( L ( F 1 | F 3 ) + L ( F 2 | F 3 ) ) ., However , once the cooperativity C equals or is close to zero , the mutual dynamics M are orthogonal to L ( F1∣F3 ) + L ( F2∣F3 ) ., This means that when the functionalities are integrated , both isolated functionalities are maintained , but there are also additional interactions ( with an effect of strength I ) in directions orthogonal to the summed isolated dynamics ., For example , suppose that the functionalities F1 and F2 correspond to sets of reactions responsible for mediating the cellular responses to two different input signals ., By setting the influence of all but the common output ( measured ) species to zero , an incompatibility I ( F1; F2 ) close to zero corresponds to an additive interaction of the input signals ., If I ( F1; F2 ) is larger , it may correspond to either a synergetic ( for C ( F1; F2 ) = 1 ) or antagonistic ( for C ( F1; F2 ) = −1 ) interaction ( see e . g . 28 ) ., Note that for a scalar output , orthogonal dynamics are not possible ., For systems composed of many different functionalities , one might also take the time averages ⟨I⟩ and ⟨C⟩ of the incompatibility , respectively the cooperativity , over the simulation time ΔT to obtain single measures quantifying the interactions between layers:, ⟨ I ⟩ ( F 1 ; F 2 | F 3 ) = 1 Δ T ∫ 0 Δ T I ( F 1 ; F 2 | F 3 ) d t ⟨ C ⟩ ( F 1 ; F 2 | F 3 ) = 1 Δ T ∫ 0 Δ T C ( F 1 ; F 2 | F 3 ) d t, Although they are useful for obtaining a first impression of how functionalities interact , time averages should be carefully applied ., They can hide transient interactions between functionalities , including potential sign changes of the state-dependent cooperativities ( see Example 1 ) ., We have now identified how to measure the interdependence between two functionalities , which we have defined as the change in the dynamics of one functionality when the other is present ., We have made explicit how this interdependence is itself dependent on the context of the rest of the network ., In the remainder of this section we will describe how to minimise the computational burden incurred when calculating all possible interactions between functionalities ., The notation L ( F1 ) and L ( F2∣F1 ) describing the map from a functionality ( or set of functionalities ) to the resulting trajectory simplifies the calculations we may wish to carry out to understand how the functionalities combine ., Using the key definition of ‘conditional dynamics’ given by ( 5 ) , we can prove a number of rules for combining layers which appear analogous to those known from Information Theory 39 ., For example , for two random variables X and Y , the well-known quantities of joint entropy H ( X , Y ) = H ( X ) + H ( Y∣X ) and mutual information I ( X; Y ) = H ( X ) − H ( X∣Y ) are each definitions of the same form as those given above of conditional dynamics ( 5 ) and mutual dynamics ( 8 ) respectively ., However , it is important to note that this similarity is only superficial , and any intuition gained by seeking analogies between our work and information theoretic concepts should be applied carefully ., This caveat applies in particular to the two results below ., Two lemmas allowing the quick combination of layer dynamics can be easily proved directly from the definitions of L ( F1 ) and L ( F2∣F1 ) , their extensions to larger combinations of functionalities , and Eq ( 5 ) ., The first is an analogue of Bayes’ Rule , given by, L ( F 1 | F 2 , F 3 ) = L ( F 2 | F 1 , F 3 ) + L ( F 1 | F 3 ) - L ( F 2 | F 3 ) ., ( 11 ), We will demonstrate how this rule can be used for quickly deducing the incremental effects of layers when combined in a different order ., This is fundamental , since a natural ordering of the layers is generally not given ., A second rule , which is analogous to Bayes’ Factor , is given by, L ( F 3 | F 1 ) - L ( F 2 | F 1 ) ︸ Posterior Dynamics = L ( F 1 | F 3 ) - L ( F 1 | F 2 ) ︸ Bayes Factor B 32 | 1 + L ( F 3 ) - L ( F 2 ) ︸ Prior dynamics ., ( 12 ), This rule applies when we have a choice between integrating two functionalities to the F1-only network ., The difference in their effects is decomposed into the difference L ( F3 ) − L ( F2 ) between their isolated behaviours , summed with the difference L ( F1∣F3 ) − L ( F1∣F2 ) in the incremental effect of F1 on each ., We can use these rules to compute all possible functionality combinations with a minimal amount of simulation ., In order to answer particular biological questions , we may be interested in the incremental effect of a given functionality on a specific ‘base’ network , such as those described in Examples 1 and 2 in the Results section ., In other cases we may be interested in all possible interactions between the functionalities , such as the situation in Example 3 . In a biochemical network whose reactions are decomposed into NL functionalities , the latter case suggests that we must simulate all NL layers for each of the NL ! different orderings of the functionalities , resulting in ( NL + 1 ) !, − NL ! layers to be numerically solved ., This burden can be significantly reduced using the calculation rules deduced above ., The cascaded layers representing all possible orderings of functionalities can be arranged in an acyclic directed graph , shown in Fig 3 . Each node represents the trajectory arising from the incremental addition of a new functionality , given those already present ., The graph is organised into levels , corresponding to the position of the new layer in the sequence ., The root of the layering graph ( referred to as Level 0 ) represents the given initial conditions x0 ., Each of the subsequent levels l = 1 , … , NL consists of ( NL−l+1 ) ( NLl−1 ) nodes ., Each node represents the dynamics of a functionality Fi conditioned on a subset of size l − 1 of the remaining functionalities Fj , j ≠ i ., A directed edge from a node in Level l to a node in Level l + 1 exists if the node in Level l + 1 is conditioned on all functionalities taking part in the node in Level l ., Each directed path from Level 0 to Level NL ( i . e . from the root to a leaf ) represents one of the NL ! possible orderings of functionalities ., By adding up the layer dynamics corresponding to the nodes in each path , the trajectory of the complete network is obtained ., Each node in Level 1

Introduction, Methods, Results, Discussion

Large , naturally evolved biomolecular networks typically fulfil multiple functions ., When modelling or redesigning such systems , functional subsystems are often analysed independently first , before subsequent integration into larger-scale computational models ., In the design and analysis process , it is therefore important to quantitatively analyse and predict the dynamics of the interactions between integrated subsystems; in particular , how the incremental effect of integrating a subsystem into a network depends on the existing dynamics of that network ., In this paper we present a framework for simulating the contribution of any given functional subsystem when integrated together with one or more other subsystems ., This is achieved through a cascaded layering of a network into functional subsystems , where each layer is defined by an appropriate subset of the reactions ., We exploit symmetries in our formulation to exhaustively quantify each subsystem’s incremental effects with minimal computational effort ., When combining subsystems , their isolated behaviour may be amplified , attenuated , or be subject to more complicated effects ., We propose the concept of mutual dynamics to quantify such nonlinear phenomena , thereby defining the incompatibility and cooperativity between all pairs of subsystems when integrated into any larger network ., We exemplify our theoretical framework by analysing diverse behaviours in three dynamic models of signalling and metabolic pathways: the effect of crosstalk mechanisms on the dynamics of parallel signal transduction pathways; reciprocal side-effects between several integral feedback mechanisms and the subsystems they stabilise; and consequences of nonlinear interactions between elementary flux modes in glycolysis for metabolic engineering strategies ., Our analysis shows that it is not sufficient to just specify subsystems and analyse their pairwise interactions; the environment in which the interaction takes place must also be explicitly defined ., Our framework provides a natural representation of nonlinear interaction phenomena , and will therefore be an important tool for modelling large-scale evolved or synthetic biomolecular networks .

To better understand the dynamic behaviour of cells and their interaction with the environment , mathematical models describing the interplay between proteins , metabolites or signalling molecules are used extensively in Systems Biology ., Typically , such models focus on single functional subsystems and neglect the rest of the biochemical reaction network ., However , the behaviour of multiple functional subsystems when integrated together can differ significantly from each subsystem’s isolated behaviour ., In this article we describe a methodology for assessing the nonlinear effects of combining multiple functional subsystems of a biological system ., This is key for answering questions related to Systems and Synthetic Biology as well as Metabolic Engineering ., For example , if we can identify the isolated behaviours of two subsystems , we can determine if they persist when the subsystems interact ., Similarly , we can show how modifications to single functional subsystems ( such as increasing particular metabolic yields ) have different effects in the context of the integrated system .

journal.pcbi.1007124

Transient crosslinking kinetics optimize gene cluster interactions

The 4D Nucleome Project 1 proposes the integration of diverse approaches: increasingly powerful chromosome conformation capture techniques including high-throughput chromosome conformation capture ( Hi-C ) ; statistical and topological analyses of these massive Hi-C datasets; 3-dimensional ( 3D ) and 4D super-resolution imaging datasets; and computational modeling approaches , both constrained by and independent of Hi-C datasets ., The project aims to gain mechanistic understanding of 3D structure and dynamics of the genome within the nucleus , and to learn how the active chromosome architecture facilitates nuclear functions ., In this paper , we contribute to these aims by combining three approaches:, ( i ) live cell microscopy for experiments studying the effect on gene clustering for normal and condensin-modified mutant cell strains;, ( ii ) first-principles-based , computational modeling based on the statistical physics of chromosome polymers coupled with transient gene-gene crosslinks formed by condensin proteins; and, ( iii ) analysis of the dynamic chromosome architecture with temporal network community detection algorithms applied to 4D modeling datasets across four decades of crosslinking timescales ., Our present understanding of basic principles that govern high-order genome organization can be attributed to incorporation of the physical properties of long-chain polymers 2–7 ., The fluctuations of long-chain polymers , numerically simulated with Rouse-like bead-spring chain models of chromosomes confined to the nucleus , capture the tendency of chromosomes to self-associate and occupy territories 8–11 In addition , these models make predictions with regard to the spatial and dynamic timescales of inter-chromosomal interactions , a dynamic analog of topologically associated domains ., The convergence of robust physical models with high-throughput biological data reveals the fractal nature of chromosome organization , namely an apparently self-similar cascade of loops within loops , or structure within structure , as one examines chromosomes at higher and higher resolution 12–14 ., De novo stochastic bead-spring polymer models of the dynamics and conformation of “live” chromosomes , plus the action on top of the genome by transient binding interactions of structural maintenance of chromosome ( SMC ) proteins , e . g . condensin , provide complementary information to chromosome conformation capture ( 3C ) techniques , genome-wide high-throughput ( Hi-C ) techniques , and restraint-based modeling 12 , 15–20 ., 3C and Hi-C experiments rely on population averages of gene-gene proximity on all chromosomes over many thousands of dead cells whose chromosomes have been permanently crosslinked by formaldehyde; the restraint-based modeling approach then explores 3D chromosome architecture that optimizes agreement with the experimental data on gene-gene frequency and proximity across the genome ., Many powerful inferences have been drawn from both Hi-C and polymer modeling approaches , using analyses of empirical and synthetic datasets encoding maps related to the pairwise distances between genes ., A common major limitation for existing polymer models and whole-genome contact maps in mammalian cells is in mapping two essential regions of the chromosome , namely the centromere and the nucleolus ., The centromere , essential for chromosome segregation , and the nucleolus , the sub-nuclear domain of ribosomal DNA , are comprised of megabases of repeated DNA ( centromere satellites and nucleolus rDNA ) ., Furthermore , these regions are not captured in methods used for generating contact maps ., We have used single live cell imaging of the nucleolus in budding yeast coupled with whole genome polymer modeling to explore the minimal requirements for sub-compartmentalization ., Implementation of protein-mediated cross-linking within the nucleolus is sufficient to partition this region of the genome from the remaining chromosomes ., Furthermore , stochastic polymer models reveal that the relative timescales of crosslinking kinetics and fluctuations of the chromosome chains have a profound influence on nucleolar morphology 21 ., In single cells , the positional fluctuations of tagged DNA sequences on specific chromosomes 22–24 through the lac operator/lac repressor reporter system validate the bead-spring models ., Chromosomes fluctuate as predicted for the conformational dynamics of idealized Rouse chains 25 ., Polymer simulations over the entire genome have revealed the ability of relatively fast binding and unbinding , and thereby short-lived ( fraction of a second ) protein crosslinks to concentrate the rDNA chain sequence in a smaller volume and increase the simulated fluorescent signal intensity variance when the model datasets were convolved with a point spread function to create two-dimensional , maximum intensity projections 21 ., Visualizations of the monomers in the simulations revealed that the fastest kinetics explored , or shortest-lived crosslinks ( ∼ . 09s ) , generated several clusters of high polymer density , and overall compaction of the nucleolus ., In contrast , much slower kinetics ( decades longer-lived protein crosslinks ( ∼ 90s ) ) tended to homogenize the fluorescent signal intensity as evidenced in the decrease in simulated fluorescent signal intensity variance ., These model visualizations were consistent with experimental results on live budding yeast ., There is a growing interest to analyze Hi-C datasets and model chromosome interactions using network models 26–28 , which has opened the door to study chromosomal datasets using network-based algorithms including centrality analysis 29 , 30 and community detection 31 , 32 ., In this context , a ‘gene cluster’ is a set of genes that are in close physical proximity , and it is represented in a network by a community of nodes ( i . e . , a set of nodes between which there is a prevalence of edges ) ., These detection algorithms perform an unbiased search for robust structures ( communities or clusters ) at the scale they exist in an automated manner , quantifying how chromosome conformational changes can precede changes to transcription factors and gene expression 33 , 34 and leading to new approaches for cellular reprogramming 29 , 35 ., Here , we apply temporal community detection algorithms including multilayer modularity 36 , 37 to simulated 4D datasets over four decades of SMC-binding kinetic timescales ., This approach integrates both temporal and spatial information so that each community now represents a set of genes that are not only nearby one another , but they remain in close proximity for some duration ., This approach allows us to detect , track and label transient gene communities ( clusters ) in the nucleolus ., Simultaneously , we record summary statistics on the sizes and numbers as well as persistence times of communities , and the frequencies of community interactions leading to gene exchanges ., We likewise record standard bead-bead summary statistics ., In doing so , we detect spatial and temporal organization at the scales they exist , beyond two-point ( gene-gene ) spatial proximity statistics ., We identify the timescales over which spatial organization persists , linking the timescales to the cluster identification algorithm ., Since clusters can deform through the flux of genes into and out of clusters , we further are able to identify crosslink timescales for which spatial clustering persists over extended timescales , and whether individual clusters are relatively permanent or experience frequent interactions and gene exchanges ., Perhaps the most striking prediction of our modeling and data analysis is that specific gene organization tasks ( amplified below ) are optimized at a relatively short crosslink timescale , on the order of . 19 sec ., With these network tools applied to physics-based 4D nucleome simulated datasets , we explore the mechanistic basis for the experimentally observed variance in nucleolar morphology ., From a high-resolution sampling of the timescales for crosslinking of 5k base pair ( bp ) domains , 4D model simulations of the yeast genome reveal the nucleolus on Chromosome XII undergoes a stark transition in dynamics and structure , and does so within a narrow “mean on” crosslink timescale range of . 09 − 1 . 6 sec ., A highly stable clustering regime exists with relatively short-lived crosslinks ( . 09 sec ) , with relatively few cluster interactions and gene exchanges , as reported previously in 21 ., At slightly longer-lived ( . 19 sec ) timescales , a novel “flexible” behavior is revealed ., Gene clusters continue to self-organize , yet clusters are more mobile , frequently interact , and exchange genes ., Indeed , there is a peak timescale , marked by highly mobile gene clusters , at which both pairwise and community-scale gene interactions are maximized ., As the binding affinity of crosslinker proteins increases only slightly longer ( 1 . 6 sec ) , the community-scale structure has dissolved , with no identifiable nucleolar sub-substructure ., See Fig 1 ., From a methods perspective , our analysis of the 4D simulated datasets is based on network modeling and a temporal community-detection algorithm known as multilayer modularity 37 ., From a biological perspective , this tunable dynamic self-organization reflects a powerful mechanism to coordinate gene regulation and the coalescence of non-contiguous genes into identifiable clusters ( substructures ) ., The transition shown in Fig 1 occurs within such a narrow crosslinker timescale regime ( . 09 − 1 . 6 sec ) , suggesting a relatively simple mechanism to control dynamic sub-organization of the genome; indeed we performed and report experiments below to support this prediction ., Finally , we emphasize the counter-intuitive nature of this mechanism: clustering is most often associated with segregation , however we observe that the dynamic element of flexible clusters facilitates an overall increase in global gene interactions in the nucleolus ., We first focus in the relatively short crosslink timescale regime , extending the simulations of 21 at discrete values μ = 0 . 09 , 0 . 9 , 90 ., These will establish a basic understanding of how the kinetic timescale μ for crosslinking sensitively affects the organization of the nucleolus and the dynamics of the architecture ., From our refined simulations across the above four decades , the essence of the story can be told with results for three selected values μ ∈ {0 . 09 , 0 . 19 , 1 . 6} ., In Fig 1 ( A ) –1 ( C ) , we present visualizations , i . e . , “snapshots , ” of the beads’ 3D positions during the simulations ., The nucleolus on Chromosome XII is highlighted in blue and all remaining chromosome arms are colored gray ., In Fig 1 ( D ) –1 ( F ) , we show only the nucleolar beads , which are colored according to the network community detection analyses that we describe in the following sections ., We also show videos of the time evolution of the beads , along with a simulated microscope projection , for each timescale in S1 , S2 and S3 Videos ., Based on Fig 1 ( A ) –1 ( F ) and the videos , we identify three qualitative regimes for nucleolus clustering: We will continue to use this terminology when referring to these three clustering regimes ., We note that 21 discovered the two extreme regimes: robust clusters for μ = 0 . 09 , and the lack of clusters for μ = 90 ., As they did not finely sample the decades of timescales in between , they did not discover that the transition from robust to no clustering is in fact non-monotone with respect to gene-gene interactions , nor that the transition is essentially complete already at μ = 1 . 6 , and that the most biologically interesting and relevant regime occurs at μ = 0 . 19 ., Furthermore , without automated structure detection algorithms , they would not have been able to detect and dynamically track clusters of genes and their interactions that explain the peak in gene-gene interactions at μ = 0 . 19 ., This transition behavior and the optimal properties that arise will be the focus of several sections to follow ., In Fig 1 ( G ) –1 ( I ) , we show heatmaps of the bead-bead distances associated with the bead positions of the snapshots in ( D ) – ( F ) , identical to those in ( A ) – ( C ) ; construction of heatmaps is described in the Methods Section: Pairwise-distance maps for high-throughput chromosome conformation capture ( Hi-C ) ., Heatmaps are widely used in Hi-C to depict population averages of pairwise gene-gene proximity data 16 , 39–43 and in simulated data from polymer bead-spring models , both from 3D snapshots and time averages 9 , 11 , 21 ., Comparing the second and third columns of Fig 1 , we note the difficulty ( false negatives and false positives ) in detecting the presence of structure and sub-organization in column 2 from visual examination of heatmaps in column, 3 . As shown in 21 , the time average of 4D simulated datasets , even in the strong clustering regime , wipes out the sub-structure of snapshots when averaging over the entire G1 phase ., An alternative approach has been to use polymer modeling to generate chromosome conformations , and to select those conformations that best match Hi-C data , so-called restraint-based polymer modeling 1 ., Simultaneously , there have been efforts to develop methodologies to identify gene clusters in a rigorous and automated way from Hi-C data 26–28 ., Our conclusion is that there is a need for a more reliable and objective method to study the clustering of chromosome domains in the nucleolus , especially spatio-temporal methods that take into account how bead positions and sub-organization change with time , weighing both spatial proximity and temporal coherence in the detection method ., In the following sections , we present a scalable and automated technique to identify and track the dynamics of clusters ., First , however , we will present new experiments that provide empirical evidence for clustering in the nucleolus ., We conducted experiments to qualitatively compare image-based cluster analysis between our model and empirical measurements obtained from live cell microscopy and demonstrate the effect that SMC protein mutation can have on clustering in the nucleolus , extending the results previously reported in 21 ., Here , we study three yeast strains: wild-type ( WT ) , fob1 and hmo1 ., Importantly , fob1Δ and hmo1Δ are mutations that lack key proteins reported to crosslink or loop segments of rDNA within the nucleolus ., Fob1Δ is required for maintenance of the rDNA copy number and regulates the association of condensin with rDNA repeats 44 , 45 ., The replication fork barrier within the rDNA is a binding site for Fob1Δ that , together with several other components ( Tof1 , Csm1 and Lrs4 ) , are responsible for the concentration of condensin within the nucleolus 44 ., Hmo1 is an abundant high mobility group protein that localizes to the nucleolus and has been proposed to share functions with UBF1 , which is involved in rDNA transcriptional regulation within the nucleolus 46 , 47 ., Fob1 and Hmo1 are non-essential genes and were deleted from the genome to allow us to study their effect on nucleolus morphology due to functional modifications of crosslinking ., See Methods Section: Yeast strains for experiment for further details ., In Fig 2 , we present images and analyses of nucleoli of these strains using fluorescent , live-cell microscopy ., To visualize nucleoli , we fused Cdc14 protein phosphatase to green fluorescent protein ( GFP ) 21 ., Nucleolar protein fusions occupy a distinct region of the nucleus that is adjacent to the nuclear envelope and ( typically ) opposed to the spindle pole body ., We describe the image acquisition and processing steps in Methods Section: Image acquisition and baseline processing , and we highlight a few details here ., Following image acquisition , we construct maximum intensity projections ( MIP ) centered on the nucleolus ., See top row of Fig 2 ( A ) ., Due to potential variation in CDC14-GFP protein copy number and nucleolar/rDNA size from cell to cell , we normalized the nucleolar CDC14-GFP signal after excluding all intensity values below an intensity threshold ., To this end , we first selected a threshold using Otsu’s method 48 , which we implemented using the MATLAB function multithresh ., One can interpret the threshold as a binary mask , as shown in the second row of Fig 2 ( A ) ., After applying the mask , we normalized the nucleolar signal by subtracting all intensities by the minimum value and then dividing them by the new maximum intensity that is obtained after subtraction ., The third row of Fig 2 ( A ) depicts normalized images ., Mutations of Hmo1 and Fob1 were found to alter the area and signal intensity of nucleoli labeled with CDC14-GFP across a range of intensity thresholds , which we surmise is due to alterations in the architecture of , i . e . , clustering within , the nucleolus ., In Fig 2 ( B ) and 2 ( C ) , we provide results for an analysis of nucleolar morphology: ( B ) the area of nucleolar signal; and ( C ) the standard deviation of the normalized signal ., This analysis was implemented using the numerical algorithms presented in 21 , which we further describe in Methods Section: Image analysis ., As shown in Fig 2 ( B ) , null mutations of hmo1Δ significantly altered the area of the nucleolar signal , whereas null mutations of fob1Δ did not ., This was assessed by a Student’s two-tailed T-test , which yielded p = 3 × 10−8 for the former and p = 0 . 07 for the latter ., As shown in Fig 2 ( C ) , the standard deviations of the normalized images were significantly lowered for the fob1Δ null mutation , but this did not occur for the hmo1Δ null mutation p = 0 . 01 versus p = 0 . 2 ) ., The non-significant changes are labeled ‘NS’ in the figure ., The error bars indicate standard errors across n cells , where n = 84 , 70 and 77 for the WT , fob1Δ and hmo1Δ strains , respectively ., We note that 21 also studied the area and variance of the nucleolus using experimental and simulated images ., They found , for example , that the distribution of areas occupied by the nucleolus displays a lognormal distribution for WT cells in G1 ., Also , recall that we implemented thresholding based on Otsu’s method; in contrast , 21 explored a range of threshold values and found qualitatively similar results to be consistent across a range of threshold values ., They did not , however , explore the area and variance for simulated images for a wide range of μ , which is the focus of our next experiment ., To explore whether varying the kinetic timescale μ for our simulations yields similar changes as those arising under the fob1Δ and hmo1Δ mutations , we applied the microscope simulator of 21 to our 4D simulated data and analyzed the images using the same image analyses as described in Fig, 2 . First , we converted our 4D simulated data into a timelapse sequence with 22 time points , i . e . , snapshots ., Each nucleolus bead was convolved with a point spread function and a maximum intensity projection was created for each timepoint ., We depict 11 such images in Fig 3 ( A ) ., In panel ( B ) , we plot the area of the nucleolar signal ( computed using Otsu’s threshold ) versus μ ., Note that the nucleolus area increases as μ increases ., In panel ( C ) , we plot the standard deviation of nucleolar signal versus μ , which has the opposite trend ., In Fig 3 ( D ) , we plot the standard deviation of images obtained after a normalization step that is identical to that implemented for the experimental images ( see discussion for Fig, 2 . Interestingly , the dependence on μ of the signal’s standard deviation drastically changes depending on whether or not it is normalized ., Given that normalization is required to control for cell-to-cell differences in CDC14-GFP and in nucleolar/rDNA size , we sought develop a metric to measure clustering in the CDC14-GFP signal that was independent of the absolute values of the intensities ., Our final experiment studies cluster formation in the nucleolus and compares clustering observed in the experimental and simulated microscopy images ., We developed a cluster detection algorithm written with MATLAB ( see Methods Section: Image analysis ) and applied it to both the experimental and simulated images ., We have made the code available at 49 ., In Fig 4 ( A ) , we depict images of maximum intensity projections for WT , fob1Δ and hmo1Δ strains with ( top row ) and without ( bottom row ) visualizations of detected clusters , which are represented by green circles ., In panel ( B ) , we depict identical information as in panel ( A ) except we show simulated images for three values of μ ., In Fig 4 ( C ) and 4 ( D ) , we show the number of clusters for the experimental and simulated images , respectively ., For the experimental images , we give results for WT , fob1Δ and hmo1Δ , whereas for the simulated images we present results for μ ∈ . 09 , 90 ., We observe that the number of clusters was significantly decreased in the hmo1Δ null mutation , but not the fob1Δ null mutation ( p = 0 . 04 for hmo1Δ versus p = 0 . 3 for fob1Δ ) ., We also observe that increasing μ yielded a general trend in which there were fewer clusters ., Taken together , these data suggest that gene clustering can directly impact the size and shape of the nucleolus ., This underscores the need for robust and objective tools for identifying gene clusters ., A simple and previously used method for analyzing distances between beads is to create a histogram of all bead-bead pairwise distances ., As explored in 21 , this two-point statistic can provide evidence of clustering and can be used to query simple properties such as whether or not the clusters change over time ., In this section , we repeat this analysis on our data and extend it by showing what effects averaging over time and averaging over populations have on the results ., We show that averaging one cell over time prevents observing the flexible clustering through pairwise distances , and averaging over populations prevents observing any sort of clustering ., We provide further details on the computation of these distances in Methods Section: Pairwise-distance maps for high-throughput chromosome conformation capture ( Hi-C ) ., In Fig 5 ( A ) , we plot the distribution of all pairwise distances { d i j ( t ) } at a single time t for three kinetic timescales , given by the same values μ = 0 . 09 , 0 . 19 , 1 . 6 as shown in Fig, 1 . For μ = 0 . 09 , the pairwise distance distribution is clearly a multimodal distribution 21 ., The peak near d ≈ 50 represents a large number of very short pairwise distances between beads in the same cluster ., For slightly larger d , the density drops to zeros , indicating a separation distance between clusters ., Interestingly , we observe two more peaks near d ≈ 300 and d ≈ 600 ., The clarity of these peaks suggests that the clusters themselves are regularly spaced from one another , reminiscent of a lattice structure ., This shows the three layers of the multiscale structure of the nucleolus for μ = 0 . 09: its existence as a dense , secluded section of the nucleus , the self-organization of intra-nucleolar clusters , and the individual beads within each cluster ., For μ = 0 . 19 , one can also observe in Fig 5 ( A ) three peaks in the empirical probability density for bead-bead distances , but these peaks are much less pronounced ., This shows a gradual transition in the degree of clustering as we increase μ ., There is also a smaller gap between peaks ., Together , these observations recapitulate our observations in Fig 1 ( E ) , wherein the clusters can be observed to be less compact ., Finally , for μ = 1 . 6 , there is no multimodal structure in the bead-bead distance plot ., This is consistent with our expectation that there is no clustering structure present for this range of μ ., The rigid and flexible clustering cases differ not only in how strong the clustering is at any given time , but also in how stable the structure is in time ., We investigate this by considering how averaging pairwise-distances either across across time ( Fig 5 ( B ) ) or over multiple simulations ( Fig 5 ( C ) ) influences pairwise-distance probability densities ., In Fig 5 ( B ) , we plot the empirical probability densities for pairwise distances averaged across our 20 minute simulations ., Note for μ = 0 . 19 that the density is no longer multimodal , implying that aggregating the data across a large time range inhibits the detection of flexible clusters , which by definition change with time ., Note that the rigid clusters , which are very stable across time , remain discernable as the pairwise probability density remains multimodal ., Unsurprisingly , the slow crosslinking appears qualitatively very similar in the long time average , as there was no apparent structure in the first place ., In Fig 5 ( C ) , we plot the empirical probability densities for pairwise distances at a single time but averaged across 10 simulations with different random initial conditions ., Note for all μ that there is no longer any multimodal structure for these densities , highlighting that averaging across heterogeneous cell populations obscures the detection of clusters ., Next , we study how the kinetic time scale μ ( i . e . , and thus the presence of clusters ) affects the properties of pairwise gene interactions ., A pair of beads is said to be interacting if they are in very close proximity and the distance between them drops below d* ., As discussed in Fig 5 , we choose d* = 100nm unless otherwise noted ., In the following experiment , we show that increasing μ not only inhibits the formation of clusters , but that there exists a particular range of μ that optimizes gene mixing , or the overall interaction frequency of all pairs of genes ., These experiments illustrate how clustering—which inherently describes multi-way relationships—can be studied through pairwise distances—which inherently describe two-way relationships — , and how there remain important open problems related to the time series signal processing of 4D chromosome conformation datasets ., We study the following summary statistics for gene mixing: In Fig 6 ( A ) –6 ( D ) , we plot these summary statistics across a wide range of μ ., We identify three regimes that optimize different attributes ., μ ≈ 0 . 1 yields a self-organized structure that maximizes the number and the duration of gene-gene interactions ( see panels ( B ) and ( D ) ) ., Recall from Video 1 that μ = 0 . 09 yields many large clusters that are stable ( i . e . , do not change ) over time ., This is reflected in a high number of interactions with beads in the same cluster and low number of interactions with beads not in the same cluster ., With 0 . 15 ⪅ μ ⪅ 1 , we see flexible clustering behavior from Video, 2 . Notably , we find here that this flexible clustering has interesting properties beyond simply being a weaker version of the strong clustering from the rigid clustering regime ., Namely , Fig 6 ( A ) shows that these μ values maximize the fraction of pairs of beads that interact at least once over the simulation , and Fig 6 ( C ) shows that these values minimize the waiting times between subsequent interactions ., Thus , we can say that flexible clustering promotes the number of both simultaneous and overall distinct pairwise gene interactions in the nucleolus ., This behavior arises from a balance between the number of intra-cluster gene-gene interactions , which is still elevated due to the moderate clustering as shown in ( B ) , and the ability for genes to frequently switch between clusters during cluster interactions , as indicated by the reduced waiting time in ( C ) ., SMC proteins with such crosslinking timescales will thereby promote collective interactions among all active genes ., These circumstances could accelerate a homology search , for example , to facilitate DNA repair , if the sister chromosomes were suddenly activated by a family of SMC proteins whose binding affinity was near this “sweet spot” ., Finally , μ ⪆ 1 . 5 is associated with a non-clustering regime , as shown in Video, 3 . The lack of clustering is reflected by a low number of gene-gene interactions , and the freely diffusing nature of the beads is reflected by short interaction duration and high interaction fraction ., Having found that flexible clustering maximizes interesting properties of gene interaction , we seek to develop tools to identify and label the spatiotemporal clusters ., In the rigid clustering regime , the clusters are so well-defined that any reasonable algorithm will detect them , but this is not the case for the flexible clustering ., To detect and track flexible clustering , we utilize both spatial and temporal information to identify and track clusters ., While we have access to 4D bead position time series data , we begin by transforming this into a multilayer network problem as described in Methods Section: Gene-interaction networks from pairwise-distance data ., This is motivated by the fact that the most similar data available in biology , the Hi-C dataset , does not measure true distances between genome regions , but rather a notion of similarity based on average proximity 12 ., The result of this transformation is a time sequence of weighted , undirected networks whose edge weights represent how near two beads tend to be to each other at that point in time ., We refer to this sequence of networks as a temporal network ., Given a gene-interaction network , we identify communities using an approach based on multilayer modularity 37 ., See 31 , 32 for examples where community detection was applied to network models derived from Hi-C data ., We present the algorithm in detail in Methods Section: Spatiotemporal gene clusters revealed by community detection in temporal networks , and we briefly describe it here ., The modularity measure was originally introduced 36 to detect communities in a single , non-temporal network; it is a scalar that quantifies—as compared to a null-model lacking communities—the extent to which a network’s nodes can be partitioned into disjoint sets ( i . e . communities ) so that there is a prevalence of edges between nodes in the same community and relatively few edges between nodes in different communities ., By searching over different possible ways to partition nodes into communities , one seeks to find an optimal partition that maximizes the modularity score 38 ., Because each community contains a prevalence of edges , and edges only exist between pairs of genes that are in close proximity , a modularity-optimizing partition equivalently assigns genes into disjoint clusters so that each gene is nearer to genes in its cluster than to genes in other clusters ., Our analysis is primarily based on an extended version of modularity that allows one to detect time-varying communities in temporal networks and is called the multilayer modularity measure 37 ., In contrast to a community in a time-independent network ( which is defined by a set of nodes ) , to specify a time-varying community one must also identify for each node the time-steps for which it is in the community ., Using a variational technique 38 , we optimize the multilayer modularity measure to simultaneously assign every node to a community at every time step ., Each time-varying community in the network corresponds to a time-varying gene cluster , which is a set of genes that remain in close proximity for some duration ., A key feature of the multilayer-modularity approach for community detection is that the framework involves two parameters , γ and ω , which provide “tuning knobs” 37 , 50 to identify , respectively , the appropriate sizes and temporal coherence of communities/clusters ., Parameter γ is a resolution parameter 38 and allows one to select whether modularity-optimizing partitions involve many small communities or just a few , very large communities ., Similarly , ω is a couplin

Introduction, Results, Discussion, Materials and methods

Our understanding of how chromosomes structurally organize and dynamically interact has been revolutionized through the lens of long-chain polymer physics ., Major protein contributors to chromosome structure and dynamics are condensin and cohesin that stochastically generate loops within and between chains , and entrap proximal strands of sister chromatids ., In this paper , we explore the ability of transient , protein-mediated , gene-gene crosslinks to induce clusters of genes , thereby dynamic architecture , within the highly repeated ribosomal DNA that comprises the nucleolus of budding yeast ., We implement three approaches: live cell microscopy; computational modeling of the full genome during G1 in budding yeast , exploring four decades of timescales for transient crosslinks between 5kbp domains ( genes ) in the nucleolus on Chromosome XII; and , temporal network models with automated community ( cluster ) detection algorithms applied to the full range of 4D modeling datasets ., The data analysis tools detect and track gene clusters , their size , number , persistence time , and their plasticity ( deformation ) ., Of biological significance , our analysis reveals an optimal mean crosslink lifetime that promotes pairwise and cluster gene interactions through “flexible” clustering ., In this state , large gene clusters self-assemble yet frequently interact ( merge and separate ) , marked by gene exchanges between clusters , which in turn maximizes global gene interactions in the nucleolus ., This regime stands between two limiting cases each with far less global gene interactions: with shorter crosslink lifetimes , “rigid” clustering emerges with clusters that interact infrequently; with longer crosslink lifetimes , there is a dissolution of clusters ., These observations are compared with imaging experiments on a normal yeast strain and two condensin-modified mutant cell strains ., We apply the same image analysis pipeline to the experimental and simulated datasets , providing support for the modeling predictions .

The spatiotemporal organization of the genome plays an important role in cellular processes involving DNA , but remains poorly understood , especially in the nucleolus , which does not facilitate conventional techniques ., Polymer chain models have shown ability in recent years to make accurate predictions of the dynamics of the genome ., We consider a polymer bead-chain model of the full yeast genome during the interphase portion of the cell cycle , featuring special dynamic crosslinking to model the effects of structural maintenance proteins in the nucleolus , and investigate how the kinetic timescale on which the crosslinks bind and unbind affects the resulting dynamics inside the nucleolus ., It was previously known that when this timescale is sufficiently short , large , stable clusters appear , but when it is long , there is no resulting structure ., We find that there additionally exists a range of timescales for which flexible clusters appear , in which beads frequently enter and leave clusters ., Furthermore , we demonstrate that these flexible clusters maximize the cross-communication between beads in the nucleolus ., Finally , we apply network temporal community detection algorithms to identify what beads are in what communities at what times , in a way that is more robust and objective than conventional visual-based methods .

chemical bonding, chromosome structure and function, applied mathematics, simulation and modeling, algorithms, chromosome mapping, mathematics, materials science, network analysis, genome analysis, molecular biology techniques, cell nucleus, macromolecules, cellular structures and organelles, nucleolus, physical chemistry, research and analysis methods, polymers, polymer chemistry, computer and information sciences, genomics, gene mapping, chromosome biology, chemistry, cross-linking, molecular biology, cell biology, genetics, biology and life sciences, physical sciences, materials, computational biology, chromosomes

journal.pgen.1003452

Sensory Neuron-Derived Eph Regulates Glomerular Arbors and Modulatory Function of a Central Serotonergic Neuron

Serotonin , 5-hydroxytryptamine ( 5-HT ) , an evolutionarily ancient monoamine , plays diverse roles in the brain 1 , 2 , 3 ., In the mammalian brain , serotonin is implicated in the regulation of behavioural arousal and control of motor output 4 , 5 with a proposed phylogenetically ancient function in modulating a drive to withdraw from dangerous and aversive environments and seek contentment 6 ., In the fruitfly , Drosophila melanogaster , serotonin regulates diverse aspects of behaviour such as aggression , sleep , circadian rhythm , learning and memory 7 , 8 , 9 , 10 , 11 ., It is estimated that there is one serotonergic neuron per million in the mammalian central nervous system , yet , when axon terminals are examined in the rat cortex , as many as 1/500 are serotonergic 2 , suggesting that a small set of neurons may act through their broad arborization pattern to play roles in modulating many brain circuits ., Understanding how serotonin and other neuromodulators function to modify intrinsic dynamic properties of neuronal circuits and thereby alter animal behaviour , is a daunting task ., An iconic preparation in which this has been carried out is the circuit that drives pyloric rhythm in the crab/lobster stomatogastric system 12 , 13 ., Such studies have led to the view that understanding the function of brain circuits not only requires a characterization of intrinsic dynamic properties of constituent neurons and their connectivity but also an understanding of how specific neurotransmitters and neuromodulators impinge on the circuit 14 ., Functional imaging and electrophysiology suggests that serotonergic modulation of olfactory information is an important conserved feature 15 , 16 , 17 ., In the Drosophila antennal lobe ( AL ) , innervated by ∼2500 olfactory sensory neurons ( OSNs ) , ∼150 projection neurons ( PNs ) , and ∼200 local interneurons ( LNs ) , the CSDn is the sole serotonergic neuron 18 , 19 , 20 ., This and its accessibility to genetic manipulation 18 , 21 allow the development of the capacity for serotonergic modulation to be studied in the context of the well-characterized olfactory glomerular system ., While the CSDns axonal terminals spread over multiple glomeruli in the adult AL 18 , it also exhibits glomerular-specific differences in innervation pattern ( this study ) ., Such wide-field arborizations , with variations in specific glomeruli , are seen in multi-glomerular olfactory LNs 22 , 23 , but the underlying mechanisms that regulate these arborizations have not been studied ., This is in contrast with the many elegant studies that have led to significant understanding of mechanisms underlying targeting of the uni-glomerular OSNs and PNs 24 , 25 , 26 ., The glomerular-specific pattern of wide-field interneurons is also likely to be important for their function as context- specific modulators of olfactory information , a hypothesis that has not been tested ., Serotonergic neurons have been suggested to act in a paracrine manner: serotonin-containing varicosities release serotonin that can diffuse away and act on extra-synaptically located receptors 27 ., While the arbors of such diffuse neuromodulatory neurons are suggested to be distributed to optimize efficient coverage of brain regions , the heterogeneous distribution of the terminal arbors of the CSDn in the AL suggests the possibility that arborization in a specific glomeruli is an important functional feature and could be behaviourally relevant , a view which we test and show to be valid ., In searching for the mechanistic underpinning of the CSDns terminal aroborization pattern we homed in on Eph-ephrin signaling as a likely candidate ., Eph receptors ( Eph ) form the largest family of receptor tyrosine kinases ( RTKs ) and mediate contact-dependent bidirectional communication between cells through short-range interactions 28 , ., Such short-range interactions between axonal arbors and their target cells could be relevant for emergence of regional differences in the arborization pattern of neurons in the CNS ., We find that an Eph/ephrin signaling-mediated repulsion plays a key role in glomerular-specific positioning of axonal terminals of the CSDn ., Sensory neurons differentially express Eph , which interacts with Ephrin on the CSDn to establish glomerular-specific innervation pattern of the CSDn axonal terminals ., Further , we show that this glomerular-specific innervation pattern of the CSDn allows it to modulate olfactory behaviour in an odor-specific manner ., We have determined the function of the CSDn in modulating odor-guided behaviour and shown that its glomerular-specific modulatory properties are dependent on the developmental regulation of its terminal arborization ., Since the CSDn is the only serotonergic neuron in the AL , our study behaviourally dissects out the role of this important neuromodulator in the olfactory system and shows , for the first time , how its function is developmentally put in place ., Our results also point to how sensory neurons , which are targeted to specific glomeruli , could locally regulate terminal arbors of other wide- field neurons ., Finally , we examine Eph-ephrin signaling at the resolution of a single neuron , for the first time , to show how short-range signaling can sculpt local pattern , and thereby , function ., We had earlier characterized the development the CSDn in Drosophila 18 , 21 ., In these studies , the CSDn 18 is labeled using a combination of cis-FRT/FLP and Gal4/UAS method 31 , 32 ., This method can result in activation of CD8::GFP reporter protein expression in the CSDn in one antennal lobe , while the neuron on the contralateral side remains unlabeled , thereby allowing the examination of its arbors without the pattern being obscured by its homolog in the other hemisegment ., Although the CSDns terminal arbors in the contralateral AL innervate all glomeruli 18 , a closer examination showed clear glomerular-specific differences in the innervation pattern ( Figure 1A , 1E ) ., We focused on glomeruli whose function in olfactory perception is well established in behavioural assays allowing us to correlate connectivity of the CSDn with its function in modulating behaviour ., We therefore analyzed the VA1d , DA1 , VA1l/m , DL3 , which respond to fly- derived odors 33 ., Of these , sensory neurons innervating DA1 and DL3 respond to the pheromone cis-vaccenyl acetate - cVA 33 , 34 , 35 ., We also examined the V glomerulus , which responds to Carbon dioxide ( CO2 ) 36 , 37 ., Quantification of axonal branch tip number of the CSDn in these glomeruli demonstrated prominent glomerular-specific differences in its innervation pattern: VA1d and V were innervated by many arbors while DA1 , VA1l/m and DL3 received fewer inputs from the CSDn ( Figure 1A , 1E; Figure S1 and Table S1 ) ., In order to understand the cellular and molecular mechanism ( s ) underlying such differences in innervation pattern of the wide-field neuron we analyzed the possible role of signaling molecules and observed a clear disruption of this pattern in Ephrin hypomorphs ( Figure 1B , 1F; Figure S1 and Table S1 ) ., Axonal branch tip number increased dramatically in DA1 , VA1l/m and DL3 glomeruli of Ephrin hypomorphs while innervations to glomeruli VA1d and V is comparable to controls ( Figure 1B , 1F; Figure S1 and Table S1 ) : The glomeruli that normally had fewer arbors of the CSDn ( DA1 , VA1l/m and DL3 ) were densely innervated in Ephrin hypomorphs , whereas arbors in densely innervated glomeruli ( VA1d and V ) remained unchanged in this mutant ., Further , CSDn-specific expression of Ephrin rescued glomerular-specific innervation pattern defects observed in Ephrin hypomorphs ( Figure 1C , 1D , 1G; Figure S1 and Table S1 ) suggesting that Ephrin is required autonomously in the CSDn although it is widely expressed in the developing AL ( Figure 1H–1L ) ., Overexpression of Ephrin in the CSDn did not change overall pattern of axonal branch tip distribution although a small decrease in final branch tip number was observed ( Figure 1C , 1G; Figure S1 and Table S1 ) ., This reduction in the overall branch tip number could either be due to increased Eph-mediated repulsion or due to other as yet unknown molecular interactions within the AL ., While Ephrin was required in the CSDn for positioning its terminal arbors in a glomerular-specific manner ( Figure 1A–1D and 1F–1G ) , expression analysis showed that it is uniformly distributed in the developing AL ( Figure 1H–1L ) and thus may not provide the positional information for glomerular-specific branching ., We therefore examined the expression of Eph , the receptor for Ephrin , in the developing AL ., Interestingly , Eph expression , as revealed by an Ephrin-Fc probe 38 , was detected in a small subset of glomeruli within the developing AL from 50 h after puparium formation ( 50 hAPF; Figure 2A–2D ) ., Most prominent Eph expression was detected in DA1 , VA1l/m and DL3 glomeruli ., These are the same glomeruli that receive fewer arbors of the CSDn in control animals and show substantial increase in innervation by the CSDn in Ephrin hypomorphs ., The observation of commissural expression of Eph ( arrow in Figure 2C and 2E ) along with the above glomerular specific pattern suggests that the OSNs are the source of Eph ., Consistent with this interpretation , targeted expression of EphRNAi in sensory neurons ( pebbled-Gal4/+; UAS EphRNAi/+ ) abolished Eph expression in the AL ( Figure 2E–2F; Figure S2 ) ., Targeted misexpression of Eph in sensory neurons ( pebbled-Gal4/+; UAS Eph/+ ) lead to Ephrin-Fc labeling in the whole AL , further validating the specificity of the Ephrin-Fc probe ( Figure S2 ) ., Targeted expression of the EphRNAi in the projection neurons or in the local interneurons did not affect glomerular-specific Eph expression ( data not shown ) ., Furthermore , in amos mutant animals , of the genotype amos1/Df ( 2L ) M36F-S6 39 , which lack most OSNs , the AL expression of Eph is also substantially reduced ( Figure 2G–2H ) ., Taken together , we conclude that Eph is expressed by a small set of sensory neurons and enriched in cognate glomeruli that received reduced arbors of the CSDn compared to other glomeruli where Eph levels are low ., Ephrin expressed by the CSDn may initiate repulsive interactions upon encountering high levels of Eph on sensory neurons ., This hypothesis predicts that high levels of Ephrin ectopically expressed in other interneurons in these glomeruli would result in their arbors being repelled by high Eph expression ., To test this hypothesis , we overexpressed Ephrin in PNs and focused our analysis on their arbors in the high Eph-expressing VA1l/m glomerulus , visualized using the Or47b::rCD2 strain ( Gal4-GH146 , UASmCD8::GFP; Or47b::rCD2; UASEphrin ) ., Indeed , targeted overexpression of Ephrin in PNs resulted in a drastic reduction of PN innervations in the VA1l/m glomerulus ( Figure 2I–2J ) , consistent with the view that Eph-ephrin signaling mediates a repulsive interaction within the developing AL ., Similar effect of Ephrin misexpression on PN arborization was observed in other high-Eph expressing glomeruli , DL3 and DA1 ( Figure S3 ) ., This suggests that under normal circumstances , CSDn-derived Ephrin could interact with sensory neuron-derived Eph to appropriately position terminals of the CSDn in a glomerular-specific manner ., In order to directly assess the role of sensory neuron-derived Eph , we used a combination of Gal4/UAS and LexA/lexAOp dual expression system ., We generated RN2flp , tub>stop>LexA::VP16; lexAOpCD2GFP line which labels the CSDn ( Figure 3A ) and showed a clear glomerular-specific arborization pattern similar to that seen in the GAL4 reporter ( Figure 3B , 3D ) ., OSN-specific knockdown of Eph , achieved by targeted expression of EphRNAi in OSNs driven by the pebbled-Gal4 , leads to increased innervation of CSDn in DA1 , VA1l/m glomerulus ( Figure 3C , 3D; Figure S1 ) similar to the phenotype that we observed in Ephrin hypomorphs ( Figure 1 ) ., Such a change was also seen for DL3 glomerulus ( Figure 3D , Figure S1 ) ., These results implicate OSNs in a previously unknown role in the development of a central neuron through their regulated expression of Eph ., OSN terminals enter the lobe at 22 h APF and are key components of glomerular development 40 ., OSN expression of Eph in the developing antennal lobe becomes prominent after 50 hAPF ( Figure 2A–2D ) ., To further validate the role of OSNs in CSDn patterning , we examined the CSDn arborization pattern in animals developing without antennae 41 and thus without the antennal OSNs ( Figure 3F ) or in animals in which antennae are transformed to legs ( Figure 3H ) ., In both the cases , the innervation pattern of CSDn in the antennal lobe was uniform ( Figure 3F , 3H ) , unlike control animals where axonal terminals exhibited glomerular-specific differences in the innervation pattern ( Figure 3E , 3G ) ., Taken together , these data substantiate a role for OSNs in providing positional cues necessary for glomerular-specific arborization patterning of an identified central serotonergic neuron ., Eph-ephrin interactions can lead to diverse outcomes in terms of attraction , repulsion and cell adhesion in a context-dependent manner ., High affinity Eph/ephrin signaling is known to initiate contact-dependent repulsion while low level signaling can lead to attraction and directed neuronal branch extension 42 , 43 , 44 , 45 ., We further investigated how Eph/ephrin signaling levels could control the final arborization pattern of the CSDn ., To achieve a complete loss of Eph-ephrin signaling we utilized an allele EphX652 in where Eph expression is completely abolished 38; Figure S4 ., Since Eph is expressed in the OSNs , we first tested the role of Eph during the development of OSNs and projection neurons ( PNs ) , the primary synaptic partners of the OSNs ., Terminals of OSNs ( Figure 4A–4F ) and uniglomerular PNs ( Figure 4I–4J ) develop normally in Eph null animals suggesting that Eph is not necessary for development of these components of the AL circuit , which have uniglomerular projections ., Next , we asked if misexpression of Eph in the majority of the OSNs during a time window when Eph is expressed in very few glomeruli would affect OSN patterning in the AL ., To this end , we used Or83bGal4 46 , which drives Gal4 expression in ∼80% of the OSNs starting from mid-metamorphosis ., Misexpression of Eph using Or83bGal4 did not affect OSN patterning in the AL ( Figure 4G , 4H ) ., Overall , these observations allow us to argue that Eph signaling does not play any obvious role in OSN/uniglomerular PN patterning within the AL ., Surprisingly , terminal innervations of CSDn were reduced in animals homozygous for EphX652 to all the glomeruli examined ( Figure 5B , 5H and Table S1 ) ., This was in marked contrast to the situation where Eph-ephrin signaling was not completely abolished but only reduced in the Ephrin hypomorphs ( Figure 1B ) or where Eph was knocked down specifically in the OSNs ( Figure 3C ) ., The CSDn innervation pattern was differentially affected in the latter cases and glomeruli with normally less innervations showed a substantial increase , leaving the densely innervated glomeruli unaffected ., These differences in phenotypes indicate a requirement of Eph signaling at multiple stages of the CSDn development ., Complete loss of Eph throughout development might influence overall branching and hence we observed reduced arborization of the CSDn in Eph null ., On the other hand , OSN-derived Eph controls glomerular-specifc innervation of the CSDn during pupal stages ., In any event , our observations suggest a key role for Eph/ephrin pathway in patterning axonal terminals of the CSDn ., To further test this , we ectopically expressed Eph in the CSDn ., Targeted ectopic expression of Eph in the CSDn resulted in striking reversal of axonal branch tip distribution in the glomeruli ( Figure 5C , 5I and Table S1 ) ., Axonal terminals of Eph-expressing CSDn preferentially innervated glomeruli with high Eph and completely avoided VA1d glomerulus , which expresses low Eph ( Figure 5C , 5I ) ., This exquisite mistargeting further strengthens the suggestion that levels of Eph/ephrin signaling control glomerular-specific innervation of this serotonergic neuron ., One possibility is that preferential targeting to high Eph-expressing glomeruli could be due to attractive homotypic interactions between Eph expressing neurons ., Eph-mediated homotypic interactions have been shown to promote cell adhesion between Eph-expressing cells during rhombomere-boundary formation in zebrafish 47 ., Another possibility , not excluding the first , is that Eph-ephrin interaction within CSDn could result in ‘cis inhibition’ 28 , 48 of the signaling pathway due to simultaneous presence of Eph and ephrin in the same cell , which in turn could reduce repulsive interaction and increase the attractive one ., We next examined if the developmental timing of the CSDn arborization is consistent with OSN derived Eph playing a role in the process ., Glomerular-specific innervation of the CSDn involves directed growth of terminals to the target glomeruli ., At 50 h after puparium formation ( APF ) , very few arbors of CSDn were seen in the regions of the antennal lobe where VA1l/m , VA1d , DA1 and DL3 glomeruli were developing ( Figure 5D ) ., An adult-like pattern was seen by 70 h APF without an intermediate stage where excess arbors were seen ( Figure 5E ) ., Terminals of the CSDn failed to innervate these glomeruli in Eph null animals ( Figure 5F–5G ) ., The time course of the development of glomerular-specific arborization of the CSDn coincided with the expression profile of Eph , described above and is consistent with a role for Eph/ephrin pathway as regulators of this process ., These observations demonstrate that the final arborization of the CSDn is not an outcome of excess growth in every glomerulus , followed by pruning but is an outcome of the repulsive signaling operating in high-Eph expressing glomeruli , which restrict the growth of CSDn terminals during development ., We next examined if the extent of glomerular-specific arborizations of the CSDn has functional implications in behaving animals ., To address this , an understanding of the role of the CSDn in odor-guided behaviours in Drosophila is first required ., The CSDn is the only identified source of serotonin in the Drosophila AL 18 , 49 suggesting an important role for this neuron in modulating olfactory perception ., Although functional imaging studies have demonstrated that serotonin can change response properties of neurons in the AL 16 , a direct demonstration of behavioural requirement of this neuron is lacking ., We used the R60F02Gal4 strain 50 which consistently labels the CSDn bilaterally in the adult brain ( Figure 6A ) , providing an advantage over the cis-FRT/FLP method , for behavioural analysis ., R60F02Gal4s restricted expression in the central brain , with prominent expression in the CSDn and only a few arborizations in the suboesophageal ganglion provides an excellent reagent for behavioural experiments ( Figure 6A ) ., We validated that R60F02Gal4 indeed labels the CSDn in two ways ., Firstly , the anatomy of its projections ( Figure 6A , 6Ai and 6Aii ) was similar to the described characteristic anatomy of the CSDn 18 ., Furthermore by examining serotonin immunoreactivity in a genetic background where R60F02Gal4 expresses GFP , it was found that the only serotonin positive neuron in the AL co-localized with the GFP ( Figure 6Aiii–vi ) confirming that the Gal4 indeed specifically labels the CSDn ., For behavioural analysis , we selected two odorants; CO2 ( perceived by the low Eph-expressing V glomerulus ) and cVA ( perceived by the high Eph-expressing DA1 and DL3 glomeruli ) as innervations of the CSDn in the cognate glomeruli have been characterized by us ., The behavioural response of wild-type adult Drosophila towards these odorants and the underlying neural circuitry is understood in good detail 34 , 36 ., CO2 is a repulsive stress pheromone in flies and is sensed by the V glomerulus 36 ., Blocking evoked neurotransmitter release from the CSDn by targeted expression of tetanus neurotoxin light chain TNTG; 51 rendered animals behaviourally more sensitive towards CO2 and these animals exhibited increased repulsion to CO2 compared to controls ( Figure 6B , p\u200a=\u200a0 . 017 ) ., Further , suppressing excitability of the CSDn by ectopic expression of an inward rectifying human K+ channel , Kir2 . 1 52 in the neuron resulted in an increased CO2 avoidance behaviour ( Figure 6C , p<0 . 01 ) ., Perturbation of neuronal activity during development has known consequences on the dendritic pattern of the CSDn 18 , 21 and could be argued that this affects the behaviour ., In order to circumvent the behavioural effects deriving from a developmental requirement of neural activity we manipulated the CSDn activity only during adulthood by using the temperature-sensitive Gal80 repressor of Gal4 ( Gal80ts ) 53 ., Adult-specific suppression of the CSDn excitability by overexpression of the Kir2 . 1 in adult flies lead to increased CO2 sensitivity ( Figure 6D ) suggesting that the CSDn function in modulating olfactory behaviour is required during adulthood ., In order to further validate the view that behavioural defects are indeed through serotonin signaling , we analyzed the expression pattern and function of serotonin receptors in the AL and then manipulated them ., A Gal4 reporter line for serotonin receptor 5-HT1BDro 9 labels a small set of local interneurons in the adult AL ( Figure 6E ) suggesting that these neurons could be possible downstream target of serotonin released by the CSDn ., RNAi-mediated knock down of 5-HT1BDro 9 in 5-HT1BDro expression domain lead to an increase in CO2 avoidance behaviour ( Figure 6F ) ., However , 5-HT1BDro is also expressed in the mushroom body neurons 9 , which are a crucial component of the olfactory circuit underlying olfactory learning and memory 54 , 55 ., In order to define better , the domain of 5HT1BDro expression relevant in mediating CO2 avoidance behaviour , 5HT1BDro levels were ‘knocked-down’ using an RNAi construct 9 driven by the 5-HT1BDro-Gal4 driver in a context where Gal80 repressor of Gal4 is expressed under a mushroom-body promoter 56 ., These animals will have normal 5HT1BDro in the mushroom body neurons , due to Gal80 repressing GAL4 expression in this tissue , but lowered expression in the olfactory local interneurons due to RNAi ., Behavioural experiments show that these animals exhibit an increased CO2 avoidance behaviour ., Taken together , these observations suggest that the CSDn releases serotonin as a neuromodulatory transmitter and serotonergic receptor-expressing local interneurons play an important role in CO2 sensitivity ., Next , we tested the role of the CSDn in cVA-dependent courtship behaviour ., cVA , a male pheromone , is transferred to females during mating and renders them less attractive to other males in subsequent encounters ., Virgin males therefore , show reduced courtship towards cVA-treated females 35 ., The males sense the presence of cVA through OSNs that target to DA1 and DL3 glomeruli 34 , 35 ., Blocking neurotransmitter release from the CSDn by targeted expression of tetanus neurotoxin light chain in the CSDn resulted in reduced behavioural sensitivity towards cVA and these males exhibited increased courtship towards cVA-treated females compared to controls ( Figure 7A , p\u200a=\u200a0 . 028 ) ., Taken together , these experiments demonstrate a role for the CSDn in modulating olfactory perception of behaving animals in an odor-dependent manner ., Having established a role for the CSDn function in odor-response modulation , we examined the basis for this modulation ., Modulation could be achieved in a variety of ways , such as the differential expression of serotonin receptors in the AL or/and by the differential arborization ( as observed in the present study , Figure 1A ) , which in turn may result in differential levels of local serotonin release by the CSDn ., Suppressing the function of the CSDn causes reduced behavioural sensitivity towards cVA , indicating that serotonin release is important for enhanced sensitivity towards cVA ( Figure 7A ) ., The level of serotonin release in the cVA-specific glomeruli ( DA1 and DL3 ) is likely to be more in cases where there is an increase in the innervations of the CSDn to these glomeruli ., Innervations in these glomeruli increase heavily in Ephrin hypomorphs compared to control ( Figure 1 ) predicting that Ephrin hypomorphs should be much more sensitive to cVA ., This was indeed the case; Ephrin hypomorphs showed a remarkable behavioural sensitivity to cVA and thus showed highly reduced courtship towards cVA-treated females ( Figure 7B , p<0 . 001 ) ., If increased behavioural sensitivity in Ephrin hypomorphs is indeed due to increased DA1/DL3 innervations by the CSDn then rescuing the CSDn branching pattern to control levels should show a rescue of the behavioural phenotype ., Targeted expression of Ephrin in the CSDn in Ephrin hypomorphs leads to a partial rescue of the behavioural sensitivity of Ephrin hypomorphs towards cVA ( Figure 7C , p\u200a=\u200a0 . 004 compared to Ephrin hypomorphs ) ., This suggests that the increased sensitivity to cVA in the Ephrin hypomorphs is indeed due to the increased innervations of the CSDn ., However , the absence of a complete rescue of the behavioural phenotype suggests the possibility that the terminals of other interneurons are defective in the relevant glomeruli in Ephrin hypomorphs ., As mentioned earlier , the widespread expression of Ephrin in the lobe indicates that other interneurons may also require the molecule ., Nevertheless , a partial behavioural rescue by Ephrin expression in the CSDn in Ephrin hypomorphs suggests that Eph/ephrin signaling has a role in development of the pheromone modulatory circuit and regulates correct positioning of neuronal arbors in a manner relevant for behaviour ., Normal courtship in Ephrin hypomrphs ( male courtship index\u200a=\u200a0 . 72±0 . 032; n\u200a=\u200a33 ) is comparable to controls ( male courtship index\u200a=\u200a0 . 76±0 . 037; n\u200a=\u200a36 ) suggesting that these animals dont display a defect in courtship behaviour ., A similar analysis could not be performed for Eph null animals as these showed severely reduced normal courtship ( data not shown ) ., We next checked whether this is true for the other odor we have examined , CO2 ., The CSDn innervations in the V glomerulus of Ephrin hypomorphs are comparable to controls ( Figure 1 ) and their response towards CO2 is also comparable to control animals ( Figure 7D , p\u200a=\u200a0 . 98 ) ., However , EphX652 null animals , which have reduced innervations of the CSDn in the V glomerulus show an increased repulsion to CO2 when compared to controls ( Figure 7E , p\u200a=\u200a0 . 003 ) ., This phenotype is comparable to what we observed upon silencing or blocking neurotransmitter release from the CSDn ( Figure 6 ) ., Thus , the olfactory sensitivity towards CO2 changes only in the contexts where the CSDn branching has been affected in V glomerulus ., Taken together , our data suggests that the serotonergic CSDn has a modulatory effect in olfactory behavioural sensitivity and glomerular positioning of its terminals during development is essential for its function in the adult ., Several studies on the OSN and PN targeting in the Drosophila antennal lobe have led to a view that PNs organize the coarse map of the antennal lobe and thus provide spatial information necessary for appropriate fine-targeting of other lobe neurons 26 , 62 ., Glomerular organization of the antennal lobe is complete by ∼48 hAPF and synaptogenesis starts between 48 and 72 hAPF 63 ., Our work shows that this developmental time window is not only relevant for synaptogenesis in the antennal lobe but also for OSN-driven patterning of wide-field interneurons ., A small set of OSNs start to express Eph at the onset of the synaptogenic time window and provide spatial information to growing axonal terminals of the CSDn ., Eph expressed by OSNs may not influence gross targeting of PNs as PN targeting occurs much earlier in the AL ., However , OSN-derived Eph may regulate patterning of axonal terminals of other interneurons , which elaborate their branches during late metamorphosis ., It will be interesting to see if selective Eph expression in the OSNs during the phase of synaptogenesis requires olfactory co-receptor expression or neuronal activity ., That the CSDn is a modified larval neuron 18 and that the glomerular-specific terminal pattern is set-up during pupal development both raise the possibility that serotonin release from this neuron has a role in antennal lobe development and plasticity ., This possibility emerges from the very elegant set of studies from the Beltz laboratory 64 , 65 , which demonstrate a role of serotonin through its receptors , in adult neurogenesis in decapod crustaceans ., One possibility is that the CSDn acts in the Drosophila larva to influence neurogenesis in the adult , during larval or pupal life , by regulating the specific LN and PN stem-cell linages and their neuronal morphogenesis in the antennal lobe 22 , 23 , 66 ., Another possibility , not excluding the first , is that serotonin is relevant to experience dependent changes in the glomeruli , such as observed in Drosophila 67 ., We see no obvious alteration in the size of the antennal lobe in contexts where the CSDn function is blocked ( data not shown ) and , a detailed developmental role for serotonin is outside the scope of the current study ., Nevertheless , the CSDns singular presence in the antennal lobe make studying the developmental role of serotonin an attractive direction and an area that will surely be embarked on soon ., In most brain regions closely studied , each neuromodulatory transmitter is usually released by more than one neuron and co-expression with other neurotransmitters is not uncommon 68 ., The Drosophila antennal lobe is likely to be similar and a recent study using mass- spectrometry and genetic tools suggests presence of a large number of neuromodulators in the AL 69 ., This makes the linking of the development of identified neurons with their role in behaviour difficult to tease apart ., The CSD neuron is special in that it is the only serotonergic neuron that innervates the AL and does not appear to have a co-transmitter ., The CSD neuron preparation is thus valuable in that it allows the examination of neuromodulation from development of its anatomy to the role of this anatomy in behaviour ., While there may well be a matrix of neuromodulators which together function in the behavioural paradigms we have tested , our results on ablating the function of the CSDn suggest that this neuron is likely to be a key player ., Serotonergic neurons usually have a diffuse neuromodulatory role in the CNS ., In such contexts , serotonin is able to diffuse from the release site in order to act upon extra-synaptic receptors and serotonergic neurons often branch in a manner to have complete coverage of the neuropil 2 , 14 , 27 , 70 ., The context-dependent response to serotonin is mediated by multiple serotonin receptors , which initiate diverse intracellular signal transduction pathways and also differ in their expression pattern in the central nervous system 71 , 72 ., Our analysis at the resolution of an identified neuron suggests that contextual specificity is also regulated at the level of innervation pattern and connectivity of serotonergic neurons ., Our data points to a general mechanism underlying the emergence of contextual specificity in neuromodulation: peripheral neurons developmentally regulate the extent of innervation by modulatory neurons , which in turn , regulate the extent of neuromodulation of specifi

Introduction, Results, Discussion, Materials and Methods

Olfactory sensory neurons connect to the antennal lobe of the fly to create the primary units for processing odor cues , the glomeruli ., Unique amongst antennal-lobe neurons is an identified wide-field serotonergic neuron , the contralaterally-projecting , serotonin-immunoreactive deutocerebral neuron ( CSDn ) ., The CSDn spreads its termini all over the contralateral antennal lobe , suggesting a diffuse neuromodulatory role ., A closer examination , however , reveals a restricted pattern of the CSDn arborization in some glomeruli ., We show that sensory neuron-derived Eph interacts with Ephrin in the CSDn , to regulate these arborizations ., Behavioural analysis of animals with altered Eph-ephrin signaling and with consequent arborization defects suggests that neuromodulation requires local glomerular-specific patterning of the CSDn termini ., Our results show the importance of developmental regulation of terminal arborization of even the diffuse modulatory neurons to allow them to route sensory-inputs according to the behavioural contexts .

Serotonin , a major neuromodulatory transmitter , regulates diverse behaviours ., Serotonergic dysfunction is implicated in various neuropsychological disorders , such as anxiety and depression , as well as in neurodegenerative disorders ., In the central nervous systems , across taxa , serotonergic neurons are often small in number but connect to and act upon multiple brain circuits through their wide-field arborization pattern ., We set out to decipher mechanisms by which wide-field serotonergic neurons differentially innervate their target-field to modulate behavior in a context-dependent manner ., We took advantage of the sophisticated antennal lobe circuitry , the primary olfactory centre in the adult fruitfly Drosophila melanogaster ., Olfactory sensory neurons and projection neurons connect in a partner-specific manner to create glomerular units in the antennal lobe for processing the sense of smell ., Our analysis at a single-cell resolution reveals that a wide-field serotonergic neuron connects to all the glomeruli in the antennal lobe but exhibits the glomerular-specific differences in its innervation pattern ., Our key finding is that Eph from sensory neurons regulates the glomerular-specific innervation pattern of the central serotonergic neuron , which in turn is essential for modulation of odor-guided behaviours in an odor-specific manner .

developmental biology, developmental neuroscience, cellular neuroscience, behavioral neuroscience, neuronal morphology, genetics, biology, neuroscience, neural circuit formation

journal.pntd.0002677

A Multi-species Bait for Chagas Disease Vectors

The flagellate parasite Trypanosoma cruzi ( Chagas ) , the etiological agent of Chagas disease , is transmitted to humans by vectors of the subfamily Triatominae ., In part as a result of the distribution of vectors , Chagas disease occurs exclusively in Latin America where it is estimated that 90 million people are at risk of transmission , while 12 million are already infected 1 ., The primary vector in the Southern Cone of South America is Triatoma infestans 2 , 3 , however other species such as Panstrongylus megistus and Triatoma brasiliensis , play an important role in transmission in some regions of Brazil 4 , 5 ., Control of the sylvatic vector species at domiciliary ecotopes is particularly difficult as they readily invade households from wild ecotopes where they cannot be controlled with available methods 4 , 5 ., Triatomine bugs are obligate haematophagous insects , which feed mostly on the blood of birds and mammals ., During daylight hours , these insects are usually found aggregated inside protected shelters which they leave after dusk in search of food 6 ., Triatomine aggregation behavior inside shelters is well documented and is mediated by chemical signals and thigmotaxis 7 ., Previously two chemical signals have been implicated in the aggregation of T . infestans inside the shelters: one released from feces 7–9 and another present in their cuticle 10 , however no definitive chemical identifications were carried out ., Pheromones are substances used by organisms to transfer information between two or more members of the same species ., These can be single chemical compounds or blends of several components ., To date , the pheromone blend emitted by triatomine feces has not been fully described for any species ., Aggregation mediated by a fecal pheromone was demonstrated in several triatomine species , including P . megistus and T . brasiliensis 11–16 ., T . infestans deposit their feces in and around shelters and the volatiles emitted from fecal depositions act as chemical landmarks helping the bugs to locate their refuges 7 ., It has been demonstrated that only dry feces of T . infestans promote aggregation , while fresh feces induces rejection 9 ., Feces become attractive three hours after being deposited and attract bugs for up to 10 days 17 ., Triatoma infestans , Rhodnius prolixus and Triatoma mazzoti showed changes in their responses to the fecal aggregation signal depending on their nutritional status 8 , 17 , 18 ., In fact , recently fed T . infestans do not aggregate in response to this signal until 8–10 hours after feeding 17 ., Four compounds were previously identified in polar solvent extracts of feces of T . infestans and T . mazzottii 19; however no behavioral response was reported ., Subsequently it was demonstrated that adult females and fifth instar larvae of T . infestans were attracted to blends of the fecal compounds 4-methylquinazoline and 2 , 4-dimethylquinazoline 20 ., In addition it has been reported that ammonia from humidified feces attracts larvae in bioassays 21 ., To the best of our knowledge no previous studies have demonstrated aggregation behavior in response to synthetic compounds ., Several authors have reported the occurrence of cross-aggregation responses to feces in diverse triatomine species 11–16 ., The characterization of a common aggregation signal may allow for the development of chemical baits for monitoring multiple vector species ., In the present report , we first aimed to identify common and readily obtainable compounds that promote cross species aggregation in triatomines ., For this , we identified volatile compounds present in the feces of T . infestans , P . megistus and T . brasiliensis ., The behavior-modifying capacity of the volatiles common to all species was subsequently tested with larvae of each of the three species ., Finally , we evaluated the potential of artificial shelters baited with a blend of these fecal volatiles for promoting the aggregation of larvae of each species ., We show that a synthetic blend of substances is capable of recruiting bugs into shelters , mimicking the effects of the natural aggregation signals ., Panstrongylus megistus and T . infestans colonies originated from insects captured at domestic and peridomestic refuges in Minas Gerais state , while that of T . brasiliensis came from insects from similar ecotopes in Ceará state , Brazil ., These colonies were kept for at least ten years in a rearing chamber ( 4 . 5×1 . 6 mt ) with controlled temperature and a 12∶12 hour light:dark illumination cycle provided by artificial lights ( 4 fluorescent lamps , cold white light , 6400K , 40W ) ., As previous reports have shown that all developmental stages of triatomines make use of fecal aggregation pheromones 9 , 11 , we chose to use immature bugs both for our chemical and behavioral experiments because they are readily available ., Triatomine larvae also have the additional benefit of not emitting alarm or sexual pheromones which could interfere in case experiments were performed with adult insects 22 ., Feces from third and fourth instar larvae of T . infestans , P . megistus and T . brasiliensis were collected separately in 2 ml glass vials ( 12×32 mm standard autosampler vials , Sigma-Aldrich ) by gently pressing the abdomens of bugs with forceps ., A solid phase microextraction ( SPME ) fiber ( PDMS/DVB Stableflex , 65 µm , Supelco , Sigma-Aldrich ) was exposed to the headspace of the samples for 30 min at room temperature prior to analysis by gas chromatography with mass-spectrometric detection ( GC-MS Shimadzu 17A coupled to a Shimadzu 5050A ) ., The desorption time in the injection port of the GC was 1 min ., Helium was used as carrier gas ( 30 cm/s ) ., GC injector and transfer line temperatures were 240°C and 250°C , respectively ., The ionization energy was 70 eV ., The oven program was 80°C for 1 min and 5°C/min to 240°C using a SupelcoWax-10 column ( 30 m×0 . 25 mm i . d . ×0 . 25 µm film; Supelco , USA ) ., Tentative identification of volatiles was based on the comparison of retention indices and mass spectra with data from the literature and a spectral library ( NIST-02 ) ., All tentative identifications were confirmed by co-injections with authentic standards ., Samples of volatiles of T . infestans and P . megistus were first obtained three hours after the collection of feces ( i . e . , from fresh feces ) and afterwards , every 24 hours during the subsequent five days ( dry feces ) ., The samples of T . brasiliensis were obtained during days 0 ( fresh feces ) , 1 , 3 and 5 ( dry feces ) ., In all cases , samples of feces were collected from bugs that had been fed one week earlier ., The vials with samples of feces were left open in a closed room during the study , with each species sampled separately ., The relative abundance of each compound was determined using the area under the peak of the chromatogram ., Empty vials without feces , placed in the same room , were used as blanks ., The common substances identified in samples of feces of T . infestans , P . megistus and T . brasiliensis were selected for behavioral tests ., Standards of acetamide ( Fluka ) , 2 , 3-butanediol ( Supelco ) , acetic acid ( Fluka ) , 3-methylbutyric acid and hexanoic acid ( Sigma-Aldrich ) were at least 98% pure ., Individual compounds were tested with each vector species in decadic steps in dose-response assays ranging from 1 pg to 100 µg ., Solutions of 2 , 3-butanediol , acetic acid , 3-methylbutyric acid and hexanoic acid were prepared in dichloromethane ( Nanograde , Mallinckrodt ) and acetamide in distilled water ., All experiments were made at 25±2°C and 65±10% R . H . using a circular glass arena ( 14 cm ∅ ) where the bottom was lined with filter paper ., Two pieces of flat filter paper ( 1×1 . 5 cm ) were placed on opposite halves of the arena , one impregnated with the test solution ( test ) and the other impregnated with pure solvent ( control ) ., These two papers were positioned 1 cm from the edge of the arena and separated by approximately 10 cm 9 , 13 , 14 ., Two control series of tests ( 32 assays per series ) were performed for each species ., i ) two pieces of clean filter paper on opposite sides of the arena , in order to test for environmental asymmetries ., ii ) one clean filter paper against a filter paper impregnated with dichloromethane or distilled water , in order to test for solvent effects on behavior ., For each species , we performed 32 tests with each dose of the five compounds tested ., In these experiments , individual insects , third instar larvae starved for 11±4 days post ecdysis , were used per test and discarded afterwards ., All insects were tested during the light phase of their daily cycle , maximizing their tendency to respond to chemical signals related to shelter search ., The results were analyzed by means of a binomial test ., An individual insect was placed in the center of the arena using a small inverted plastic container that avoided it to climb due to its smooth surface ., After 10 minutes , the insect was released by means of a string that allowed raising the container from a distance without disturbing it ., One hour later , the position of the insect was recorded ., Triatomines are typically found motionless when displaying aggregation inside shelters 7 and given that we aimed to evaluate the potential role of these substances for promoting aggregation , only motionless insects were considered in our analyses ., A square glass arena ( 1 m2 ) lined with filter paper was used for these tests ( Figure 1A ) ., Two artificial shelters made from a piece of corrugated cardboard ( 20×10 cm ) , folded to generate a 10 cm2 shelter with two lateral slits of approximately 0 . 5 cm in height 7 , were placed on opposite sides of the arena ( Figure 1 ) ., In one of the shelters , a piece of filter paper ( 4×6 cm ) impregnated with a blend of the five compounds was introduced , while the other shelter contained a piece of filter paper treated with solvent as control ., This shelter design ( Figure 1B ) is proven to successfully recruit triatomine bugs 6 , 7 , 23 , 24 , which tend to enter the shelters due to their strong thigmotaxis and intense negative phototaxis 25 ., Fifth instar larvae starved for 11±4 days post ecdysis were used in this experiment ., A group of 30 insects was released in the center of the arena two hours before the beginning of the scotophase ., The illumination of the experimental room was set to a 12∶12 LL/DD regime ., Two hours after the end of the scotophase , the shelters were carefully removed from the arena and the number of bugs inside each of them was recorded ., Three doses of each of the five compounds ( 16 ng , 160 ng , and 1 . 6 µg ) were applied in the test shelter in three separate series of assays ., These tests were performed separately for T . infestans , P . megistus and T . brasiliensis ., Eight replicates were performed for each dose tested with each of the three species ., Since our goal was to compare the aggregation inside baited vs unbaited shelters , insects found outside refuges were excluded from the statistical analysis because they were not aggregating ., It is important to highlight that in triatomines the decision to aggregate or to remain dispersed is influenced by factors such as thigmotaxis and phototaxis , i . e . factors other than the presence of a bait inside a shelter ., Therefore , an additional phenomenon would have been evaluated if all insects present in the arena were included in the analysis ., The effect of the different doses of compounds on the distribution of insects in baited and unbaited refuges was analyzed by means of a Generalized Linear Model ( GLM ) using a logistic regression adapted to the binomial nature of the response variable ., This analysis was followed by a Wilcoxon signed rank test with continuity correction in order to compare shelter choice data obtained with each concentration and species against a random choice between shelters ., Both tests were performed in R software 3 . 0 . 2 ( R Core Team , 2013 ) ., Five compounds were common to the feces of all species studied: acetamide , 2 , 3-butanediol , acetic acid , 3-methylbutyric acid and hexanoic acid ., The relative abundance and proportion of these five compounds varied during the five days of sampling for the three species: T . infestans ( Figure 2A ) , P . megistus ( Figure 2B ) and T . brasiliensis ( Figure 2C ) ., In T . infestans the abundance of acetic acid and 2 , 3-butanediol increased markedly 24 h after sample collection before decreasing between the first and second day ., 3-Methylbutyric acid was initially present in fresh feces and its abundance decreased until only traces were detectable after 72 h ., A relatively constant low abundance of hexanoic acid was detected across the five days of sampling ., Only traces of acetamide were detected ( Figure 2A ) ., In samples of feces of P . megistus acetic acid was more abundant at the first and second days of sampling , a result which was consistent with our findings in T . infestans samples ., This was the most abundant compound during all five days ., 3-Methylbutyric acid was the second most abundant compound ., For P . megistus , hexanoic acid , 2 , 3-butanediol and acetamide were detected as traces ( Figure 2B ) ., In T . brasiliensis all compounds were consistently detected in very low amounts over all five days , except 2 , 3-butanediol , for which the abundance decreased steadily from the first to the fifth day ( Figure 2C ) ., Each compound was tested individually for each species ., All substances were capable of attracting the three species studied with the exception of 2 , 3-butanediol which did not induce any effect on P . megistus bugs ( Table 1 ) ., Control tests evaluating two clean filter papers or a solvent control vs a clean filter paper presented a random distribution ( Binomial test , N . S . ) ., The proportion of insects of each species found inside blend baited and control shelters is presented in Figure 3 ., The proportion of insects remaining outside refuges at the end of the experiments varied according to the species ., The GLM with binomial error revealed a significant effect of the bait dose on the choice for a refuge by the three species ( T . infestans: z\u200a=\u200a3 . 11 , P\u200a=\u200a0 . 00187 , residual deviance\u200a=\u200a12 . 940 on 22 degrees of freedom; T . brasiliensis: z\u200a=\u200a2 . 403 , P\u200a=\u200a0 . 0162 , residual deviance\u200a=\u200a25 . 084 on 22 degrees of freedom and P . megistus: z\u200a=\u200a3 . 653 , P\u200a=\u200a0 . 000259 , residual deviance\u200a=\u200a13 . 669 on 22 degrees of freedom ) ., Refuges associated with a mixture of 16 ng of each compound promoted the aggregation of only as many triatomines as clean shelters ( Wilcoxon signed rank test , NS , Figure 3 ) ., Conversely , shelters associated with mixtures of 160 ng or 1 . 6 µg of each compound promoted a significantly stronger aggregation on T . infestans ( 160 ng V\u200a=\u200a36; P\u200a=\u200a0 . 0078; 1 . 6 µg , V\u200a=\u200a36 , P\u200a=\u200a0 . 014 ) , T . brasiliensis ( 160 ng , V\u200a=\u200a36 , P\u200a=\u200a0 . 0078; 1 . 6 µg , V\u200a=\u200a28 , P\u200a=\u200a0 . 022 ) and P . megistus ( 160 ng , V\u200a=\u200a28 , P\u200a=\u200a0 . 022; V\u200a=\u200a36; P\u200a=\u200a0 . 014 ) larvae , than clean shelters ., As initially hypothesized , the present study identified volatile compounds common to feces of three species of triatomine vectors ., We combined these five substances in a blend that was capable of attracting bugs of the three species into shelters ., In contrast to previous studies we focused on the identification of compounds that were readily available and common to all species ., This approach was favored as it is expected to reduce the production cost of chemical baits ., We found that the presence of five common compounds was consistent , but their abundance was highly variable throughout the sampling period in all cases ., Previously it has been demonstrated that feces , despite this changing proportion of volatiles over time , are attractive for up to 10 days 17 ., The low , dynamic abundance and high volatility of the fecal compounds warranted SPME as the sampling method ., Out of the five substances selected for behavioral testing , three have not previously been identified in triatomine feces: 3-methylbutyric acid , hexanoic acid and 2 , 3-butanediol ., Acetic acid and acetamide have been identified in the fecal samples of T . infestans and T . mazzotii 19 , however the biological activity of these compounds was not assessed ., The results from our behavioral experiments suggest that they are all constituents of the aggregation signal from triatomine feces and reinforce the hypothesis that this aggregation signal is involved in the marking of shelters by triatomines 7–9 ., Whether substances other than those reported here play a role in triatomine aggregation remains unclear; nevertheless this is the first report of a chemical blend identified in triatomine feces acting as an aggregation signal for Chagas disease vectors ., The results obtained in shelter experiments with all the species studied in this work were very similar , which is consistent with the proposed low specificity of aggregation signals from feces of triatomines 11–14 , 16 ., Therefore , we suggest that our five substance blend could be applied as a general triatomine bait suitable for areas with several sympatric species ., Currently , the detection of domiciliary infestations in control programs is performed by manual search for triatomine bugs and/or colonization signals , such as feces , eggs and exuviae 26 ., In cases of low infestation , the use of chemical dislodging agents , e . g . , 0 . 2% tetramethrin , has been introduced to induce insects to abandon their shelters and become exposed 27 ., In Argentina , regularly monitored cardboard boxes that offer shelter to bugs in the walls of houses or in their peridomestic structures , have been used 28 , 29 ., This type of un-baited refuge does not contain glue or insecticide in order to capture or kill visiting insects; instead the bugs find them by chance and generally choose them as a shelter due to their physical properties ., The association of these devices with baits , such as the volatile mixture developed in this work , may significantly increase their detection sensitivity ., Furthermore , addition of glue 30 , insecticide or pathogenic microorganisms 31 may allow transforming these devices into sensitive control tools representing a detection/capture device highly specific for triatomines ., One particularly relevant application is the detection of dispersing individuals of sylvatic species that frequently re-invade houses from wild environments after insecticide spraying 32–38 ., It is important to highlight that under the low density of current infestations in most geographic locations , detection of triatomines is extremely difficult ., This limitation may affect the utilization of our mixture that needs to be evaluated under field conditions ., We suggest that a long-lasting formulation which would allow a cumulative sampling of bug presence may increase the chances of effective use ., In a broader context , chemical baits based on pheromones or host odors have been proposed as cost effective and environmentally benign alternative tools for detection and control of several pest insects 39 , 40 , 41 ., For triatomines , an odor mixture luring the bugs into detection devices may similarly constitute a practical , economical and environmentally friendly method to monitor infestations by Chagas disease vectors ., Further experiments should allow the development of a slow-release formulation for the blend , as well as demonstrate its effectiveness under field conditions ., Ultimately , our blend could be developed into more advanced control tools for Chagas disease vectors , which is especially relevant where colonies have developed resistance to current insecticides 30 , 31 .

Introduction, Materials and Methods, Results, Discussion

Triatomine bugs are the insect vectors of Trypanosoma cruzi , the etiological agent of Chagas disease ., These insects are known to aggregate inside shelters during daylight hours and it has been demonstrated that within shelters , the aggregation is induced by volatiles emitted from bug feces ., These signals promote inter-species aggregation among most species studied , but the chemical composition is unknown ., In the present work , feces from larvae of the three species were obtained and volatile compounds were identified by solid phase microextraction-gas chromatography-mass spectrometry ( SPME-GC-MS ) ., We identified five compounds , all present in feces of all of the three species: Triatoma infestans , Panstrongylus megistus and Triatoma brasiliensis ., These substances were tested for attractivity and ability to recruit insects into shelters ., Behaviorally active doses of the five substances were obtained for all three triatomine species ., The bugs were significantly attracted to shelters baited with blends of 160 ng or 1 . 6 µg of each substance ., Common compounds were found in the feces of vectors of Chagas disease that actively recruited insects into shelters , which suggests that this blend of compounds could be used for the development of baits for early detection of reinfestation with triatomine bugs .

Chagas disease is a parasitic infection affecting approximately 12 million people , and is considered to be one of the most severe burdens for public health in Latin America ., Control of the disease is based on attempted elimination of domestic populations of triatomine bugs , the insects transmitting the disease to humans , by means of insecticide spraying ., Currently , vigilance programs monitoring triatomine reinfestation processes in houses are performed by manual search for bugs ., Effective and sustainable new methods allowing continuous monitoring of domestic triatomine populations are required ., Based on the fact that the insects hide in dark refuges that are marked by volatile signals emitted in their feces , we screened the feces of three species for volatile compounds common to these prominent vectors ., The potential for these odors to promote triatomine aggregation was evaluated and we present evidence that a synthetic blend of these substances is capable of recruiting bugs into shelters , mimicking the natural pheromone ., This blend may be used to develop a bait to monitor triatomine reinfestation processes in a similar manner as is used commonly for the monitoring of agricultural pests .

chemical ecology, animal behavior, zoology, ecology, entomology, olfactory system, sensory systems, biology, neuroethology, neuroscience, parasitology

journal.pgen.1006823

Allelic variants of OsHKT1;1 underlie the divergence between indica and japonica subspecies of rice (Oryza sativa) for root sodium content

Salinity is a widespread limitation for agricultural productivity , especially for irrigated agriculture and coastal lowlands prone to seawater ingress 1 , 2 ., By definition , salinity occurs when there is a high concentration of soluble salts in soil 3 ., More than 800 million hectares worldwide is affected by salt , which accounts for 6% of the total land area 3 ., Besides natural causes such as rising sea levels during the dry and wet cropping seasons , the poor quality of irrigation water and improper drainage , also collectively increases soluble salt concentration in the root zone 2 , 4 ., Rice ( Oryza sativa L . ) is one of the most important crop species and is a staple food for more than half of the world’s population ., Salinity is a major impediment to increasing production in many rice growing regions , including temperate and tropical environments , around the world 5 , 6 ., Rice is the most salt-sensitive species among major cereal crops 3 ., The susceptibility of rice to salinity stress varies with growth stages 7 , 8 ., Rice is less sensitive to saline conditions at germination , active tillering and maturity stage 7 , 9 , 10 ., Vegetative growth during the early seedling stage is highly sensitive to saline conditions , and often translates to reduced stand density in salt-affected fields 11 , 12 ., Some rice varieties are most sensitive to salt stress during early tillering and panicle initiation stages of growth 8 ., This developmentally-dependent salt-sensitivity , in context of yield reduction , was associated with a significant decrease in tiller number per plant , spikelet number per panicle , fertility , panicle length and primary branches per panicle 7 , 8 , 13 , 14 ., Despite the overall high salt-sensitivity of rice , several studies have demonstrated that considerable natural variation for salinity tolerance exists in rice germplasm 15 , 16 ., Traditional landraces or cultivars such as ‘Pokkali’ , ‘Nona Bokra’ , ‘Cheriviruppu’ and ‘SR26B’ have originated or have been selected in coastal regions and are more tolerant to saline conditions 5 , 12 , 17 , 18 ., Quantitative trait loci ( QTL ) underlying salinity tolerance have undergone intensive investigations 16 , 18–24 ., Although many QTL have been identified across the rice genome , the most well-characterized QTL is Saltol/SKC1 , which harbors HKT1;5 , on the short arm of chromosome 1 18 , 20–22 ., The SKC1 gene ( HKT1;5 ) was subsequently cloned from a salt-tolerant indica landrace , ‘Nona Bokra’ , and encodes a Na+ transporter that regulates shoot Na+:K+ homeostasis during salt stress 25 ., Salinity tolerance is a complex polygenic trait , and several physiological mechanisms , including tissue tolerance , sodium exclusion , osmotic stress tolerance , and tissue-specific sodium sequestration can be utilized for improving salinity tolerance 3 ., While many QTL have been reported for salinity tolerance in rice , few studies have identified the causal genes and confirmed the importance of these resources for improving salinity tolerance ., Hence , the genetic resources ( QTL and genes ) available to rice breeders for improving salt tolerance are limited ., Identification of loci that regulate salt accumulation and/or distribution will enable the introgression of favorable genic combinations and greatly accelerate the development of robust salt-tolerant rice varieties ., Genetic variation within the rice germplasm collection can be utilized to identify important loci controlling variation for salinity tolerance through genome-wide association ( GWA ) analysis , which provides greater mapping resolution and evaluates greater allelic diversity compared to linkage mapping strategies 16 , 24 , 26 , 27 ., In this study , we used GWA to investigate the genetic architecture of salinity tolerance using the Rice Diversity Panel 1 ( RDP1 ) 28–30 ., RDP1 is comprised of 421 accessions collected from 85 countries and was developed to identify alleles associated with morphological , physiological and agronomic traits 28–30 ., RDP1 captures much of the diversity in the rice germplasm collection worldwide 28–30 ., We used several quantitative measures to characterize the rice diversity panel for physiological and morphological responses to salinity stress ., Here , we show that allelic variants of a sodium transporter ( HKT1;1 ) underlie natural variation for root Na+ content in rice ., Using a multifaceted approach , we demonstrate that variants within HKT1;1 alter Na+ transport and can explain the basis of divergence in root Na+ content between the indica and japonica subspecies of cultivated rice ., To examine the relationships between each of the eight traits , Pearson correlation analysis was performed across all accessions ., No significant relationship was observed between shoot biomass and ion traits ( S3 Table ) ., Moreover , root and shoot ion content showed no significant relationship when the analysis was performed with all accessions ( S3 Table ) ., Due to the deep population structure in rice , correlation analysis was also performed for each of the five major subpopulations in RDP1 ( here , admixed accessions were considered a separate subpopulation; S4–S8 Tables ) 29 , 31 ., Root growth response ( the ratio of root biomass in salt to control ) showed a weak negative correlation with shoot Na+:K+ in admix and tropical japonica ( trj ) accessions ( S4 and S5 Tables , respectively ) ., In trj , aus , and tej subpopulations significant , albeit weak , positive correlations were observed between shoot Na+ and root Na+:K+ ( S5–S7 Tables ) ., Comparisons between each of the subpopulations showed significant differences for shoot and root Na+ , K+ and Na+:K+ ( Fig 1 ) ., Indica accessions exhibited significantly higher root Na+ content and Na+:K+ compared to the other four subpopulations ( Fig 1A and 1B ) ., Significantly lower shoot Na+ and Na+:K+ were observed in indica and aus subpopulations compared to temperate japonica ( tej ) , tropical japonica ( trj ) and admix accessions ( Fig 1C and 1D ) ., These results suggest that there are inherent differences in root and shoot ion homeostasis between subpopulations , with indica accessions generally displaying higher root Na+ and Na+:K+ , and indica and aus accessions exhibiting lower shoot Na+ content and Na+:K+ ., To identify loci associated with salt tolerance-related phenotypes , GWA mapping was conducted using 397 , 812 SNPs and eight salinity-related phenotypes collected on 365 rice accessions ( Fig 2; S1–S4 Figs ) 32 ., A linear mixed model implemented in EMMA was used for the association analysis 33 ., A total of 90 highly significant QTL ( 245 SNPs; p < 10−5 ) were identified for salinity-related traits with the strongest associations detected for root Na+ content followed by root Na+:K+ ( Fig 2A and 2B , respectively ) ., A region located at ~30 . 6 Mb on chromosome 4 was found to have the largest effect and explained 15% of the phenotypic variation beyond that explained by population structure for root Na+:K+ ( S2 File ) ., An additional 25% of the phenotypic variance for root Na+:K+ was explained by population structure suggesting that this trait may be heavily influenced by the differences between the major subpopulations in rice ., For each trait , the number of significant QTL ranged from 3–24 , with the highest number of QTL identified for root biomass ratio ( 24 QTL ) ., Many of these QTL had small effects , explaining ~4 . 7–7 . 5% of phenotypic variation for root growth response ., These results indicate a polygenetic architecture for root growth responses to salinity ., A large number of QTL with minor effects ( explaining < 7% phenotypic variation ) were identified for shoot Na+ content and Na+:K+ , suggesting a polygenic architecture for these traits in rice ., This trend was observed for all traits , with the exception of root Na+ and Na+:K+ , suggesting that salinity tolerance in terms of growth and shoot ion homeostasis in rice is regulated by many loci with small effects ., Twenty QTL were commonly detected for two or more traits ., Shoot Na+ and Na+:K+ showed the largest number of shared QTL ( 12 QTL ) , however much of this similarity is likely driven by the strong phenotypic and genetic correlation observed between these traits within tissues ( Table 1; S2 Table ) ., The most significant QTL for root Na+ content and Na+:K+ , named Root Na+ Content 4 ( RNC4 ) , spans a region of ~575 Kb ( 30 , 481 , 871–31 , 057 , 205 ) on chromosome 4 ( Fig 3A ) ., To characterize this region further and identify candidate genes that may be underlying natural variation for this trait , this region was segmented into haplotype blocks and the contributions of each block to root Na+ content and Na+:K+ were determined using ANOVA ., A total of 36 blocks were identified in this 575 Kb region ( S5–S8 Figs ) ., A single 9 . 7 Kb block from 30 , 727 , 920–30 , 737 , 580 bp was found to have the largest contribution to root Na+ and Na+:K+ , with approximately 16% of phenotypic variation explained for root Na+ and 17 . 5% explained for root Na+:K+ ( Fig 3B; Table 2 ) ., The region spanning from the 5’ boundary of block 2 to the 5’ boundary of block 3 harbored only two genes , both of which were annotated as sodium transporters , HKT1;1 and HKT1;4 ( LOC_Os04g51820 and LOC_Os04g51830 respectively; Fig 3C ) ., To further characterize HKT1;1 and HKT1;4 , the expression patterns of both genes were examined in twelve tissues at three developmental time points ( early seedling , early tillering and anthesis ) ., The expression of both HKT1;1 and HKT1;4 were higher in leaf tissue compared to root tissue during the seedling stage ( Fig 4 , S9 Fig ) ., However , the expression of HKT1;1 and HKT1;4 within aerial tissues differed across developmental stages ., HKT1;1 was highly expressed in the leaf blade and leaf sheath during the early seedling stage ( Fig 4A ) ., HKT1;4 , on the other hand displayed the highest expression during reproductive stage , specifically in culm tissue at ~7 days after anthesis ( Fig 4B ) ., To examine whether transcript abundance may be a component of the phenotypic differences observed between allelic groups at RNC4 , RNA sequencing was performed on shoot tissue of 32 accessions in control and saline conditions , and the expression of both genes was compared between allelic groups at RNC4 ( Fig 4C and 4D; S3 File ) ., For both genes , accessions that showed higher root Na+ content ( T allele at SNP-4-30535352 ) , also showed higher expression in both control and saline conditions compared to accessions with low root Na+ content ( G allele at SNP-4-30535352 ) ., The expression of HKT1;1 was approximately 92% higher in high root Na+ lines in control conditions compared to low root Na+ lines , while a 44% higher expression was observed in saline conditions ( Fig 4C ) ., While the overall expression level was much lower for HKT1;4 compared to HKT1;1 , a similar trend in gene expression was also observed between the two allelic groups of HKT1;4 ( Fig 4D ) ., A 46% and 57% higher expression was observed in lines with high root Na+ content compared to lines with low root Na+ content in control and saline conditions , respectively ( Fig 4D ) ., These results suggest that differences in expression of HKT1;1 and/or HKT1;4 may be a component underlying variation in root Na+ content at RNC4 ., To determine if these two HKTs within RNC4 regulate Na+ content during salinity stress at the early tillering stage , three independent RNA-interference ( RNAi ) lines were generated for both genes ., Transcript levels in the leaf tissue was reduced by approximately 2 . 9–6 . 2 and 2–2 . 2 fold in HKT1;4RNAi and HKT1;1RNAi lines compared to wild-type ( WT ) ‘Kitaake’ , respectively ( S10 Fig ) ., A 9 dS·m-1 ( ~90 mM NaCl ) was gradually imposed at 10 DAT for 14 days to replicate the stress treatment for the large-scale screening ., Reduced expression of HKT1;1 had severe phenotypic effects on shoot and root ion homeostasis as well as shoot and root growth under salinity ., Shoot Na+ and Na+:K+ were 31–41% and 27–41% higher , respectively , in HKT1;1RNAi lines compared to WT ( p < 0 . 0001 , p < 0 . 05 respectively; Fig 5A–5C ) ., A 21–27% reduction in root Na+ was observed in HKT1;1RNAi and 31–33% lower root Na+:K+ was observed in HKT1;1RNAi compared to WT ( p < 0 . 05 and p < 0 . 0001 , respectively; Fig 5D–5F ) ., In RNAi plants , shoot and root growth was reduced by 44–55% and 78–72% respectively in salt treated plants relative to those in control conditions , while in WT a 26% and 45% reduction in shoot and root growth , respectively was observed in WT plants ( S11 Fig ) ., No differences were observed between HKT1;4RNAi and WT plants ( Fig 5 , S11 Fig ) ., These results suggest that HKT1;1 may influence the shoot and root Na+ content during the early tillering stage , and is likely the causal gene underlying RNC4 ., To determine whether there were sequence differences between allelic groups at RNC4 , sequencing data was mined for variants in HKT1;1 ( S4 File ) ., Nine variants were detected in the coding region of HKT1;1 with four SNPs resulting in non-synonymous amino acid substitutions in HKT1;1 ( Fig 6; S12 Fig ) 34 ., Of the nine variants , only M4 displayed a significant deviation from the expected frequency in the minor allelic group , indicating that it is unlikely to be important for the high root Na+ phenotype exhibited by accessions in the minor allelic group ( Pearson’s chi squared test , p < 1 . 26 x 10−5 ) ., The remaining three non-synonymous mutations ( M3 , M5 and M8 ) were detected in thirteen accessions all belonging to the minor allelic group , which is characterized by high root Na+ content , at the most significant SNP for root Na+ content ( SNP-4-30535352 ) ., The higher frequency of these three non-synonymous mutations observed in minor allele accessions ( T ) suggests that allelic variation in HKT1;1 could be a component in the genetic basis of the observed difference in root Na+ content between major and minor alleles ., No sequence differences in HKT1;4 were observed between allelic groups at RNC4 ., To characterize the biophysical properties of the two major isoforms identified between allelic groups at RNC4 , HKT1;1 was isolated from two representative accessions , ‘Nipponbare’ and ‘Zhenshan 2’ , which have the reference and the three non-synonymous mutations at the three locations ( M3 , M5 and M8 ) , respectively ( S12 Fig ) ., At the transporter structure level , two non-synonymous SNPs ( M8 and M5 ) lead to amino acid substitutions in cytosolic regions of HKT1;1: proline to leucine within the N-terminal cytosolic region , phenylalanine to serine in the cytosolic loop between the first and second transmembrane segment-pore region-transmembrane segment ( MPM ) domains ( Fig 6A and 6B; S12 Fig ) ., The third non-synonymous SNP results in an asparagine to serine substitution in the external part of the pore-forming region of the second MPM ( Fig 6A and 6B; S12 Fig ) ., Functional analysis was performed by voltage-clamp electrophysiology using Xenopus oocytes for the two variants of HKT1;1 ( Fig 6 ) ., The amount of expressed transporters targeted to the oocyte membrane was similar for the two variants , as indicated by the mean GFP fluorescence intensity emitted by either of the tagged transporters at the membrane ( Fig 6C and 6D ) ., In agreement with previous reports , both isoforms of HKT1;1 displayed low affinity , high Na+ versus K+ selectivity , inward rectifying activity and no time-dependent kinetics ( Fig 6E; S13 and S14 Figs ) 34 ., However , the two allelic variants displayed considerable differences in Na+ transport activity ., The variant from the accessions with high root Na+ , HKT1;1-Zh , exhibited higher inward ( negative ) currents compared to that from ‘Nipponbare’ , HKT1;1-Ni ( Fig 6E and 6F ) , essentially due to a less negative voltage threshold of inward rectifying current activation by 20–25 mV in all ionic conditions ( Fig 6E and 6F , S13 and S14 Figs ) ., This latter feature was especially expected to favor transport activity of HKT1;1-Zh compared to HKT1;1-Ni during salinity stress where the high concentration of Na+ in the apoplast results in a depolarization of the plasma membrane 35 , 36 ., Thus , at a weak negative voltage , the current could be more than six-fold higher in HKT1;1-Zh , compared to HKT1;1-Ni ( Fig 6G ) ., To determine if these differences in transport activity have physiological effects in vivo , native overexpression lines were generated for each variant ( HKT1;1Ni , HKT1;1Zh ) ., A ~4 . 3 kb genomic region was isolated from ‘Nipponbare’ and ‘Zhenshan 2’ , which included the entire CDS of HKT1;1 and a 1 . 9 kb promoter , and was expressed in ‘Kitaake’ ., The endogenous HKT1;1 in ‘Kitaake’ , at the protein level , is identical to HKT1;1-Ni , and thus lacks the three non-synonymous variants ., Two independent transformants for each variant ( HKT1;1Ni , HKT1;1Zh ) , each containing only a single copy of the transgene , were evaluated under a 9 dS·m-1 salt stress for a period of two-weeks ., The expression of HKT1;1Zh resulted in an increase in root Na+ and Na+:K+ compared to HKT1;1Ni , while no differences were observed between variants for root K+ ( Fig 7 ) ., A considerable increase in both root Na+ and Na+:K+ , as well as a reduction in root K+ was observed in both native overexpression lines ( HKT1;1Ni and HKT1;1Zh ) compared to ‘Kitaake’ , which is opposite to the root phenotype observed in the HKT1;1RNAi lines ( Fig 7 ) ., However , expression under the native promoter had no effects on shoot Na+ or Na+:K+ ( Fig 7 ) ., Together , these results provide further evidence that HKT1;1 is responsible for the higher root Na+ phenotype , and that the difference in Na+ content between the allelic groups at RNC4 is likely due to functional differences in Na+ transport by HKT1;1 alleles , with the three non-synonymous SNPs in HKT1;1-Zh resulting in higher Na+ transport activity ., A difference in allele frequencies of the three non-synonymous mutations in HKT1;1 was observed between the major subpopulations in the 32 sequenced accessions of RDP1 ., However , since it was difficult to examine subpopulation differentiation with this small of a sample size , the differences were explored in more depth using resequencing data from a larger diversity panel of 3 , 024 accessions 37 ., A total of 206 SNPs spanning a ~38 Kb region around HKT1;1 was used for haplotype analysis ., In agreement with the allele frequency observed in the 32 accessions of RDP1 , a clear differentiation could be observed between indica ( ind1A , ind1B , ind2 , ind3 and indx ) and japonica ( temp , trop1 , trop2 and japx ) subspecies in the larger diversity panel ( Fig 8 ) ., Haplotypes H1 , H5 and H8 harbored the three non-synonymous alleles and were found in nearly 85% of the indica accessions ., The sequence similarity between high root Na+ haplotypes was very high , ranging from ~88–94% identity ., Haplotypes containing high root Na+ alleles of HKT1;1 were also found in the japonica ( temp , trop1 , trop2 and japx ) , aus and aromatic subpopulations , albeit at a much lower frequency ( 0–3% ) ., In contrast , haplotypes H2 , H3 , H4 , and H7 were found predominately in the japonica accessions and lacked the high root Na+ allelic form of HKT1;1 ., Within the low root Na+ group , haplotypes exhibited high sequence similarity ( ~65–94% ) ., Given the clear divergence between indica and japonica for HKT1;1 haplotypes and the effects of HKT1;1 isoforms on root Na+ content , collectively these results strongly suggest that a significant proportion of the difference between rice subpopulations in root Na+ in RDP1 is due to differences in frequency of HKT1;1 variants ., Given the contrasting haplotype frequencies of high and low root Na+ variants of HKT1;1 between subpopulations of cultivated rice , we explored the origins of these haplotypes by examining their frequencies in a collection of 446 Oryza rufipogon accessions collected throughout South and Southeast Asia 38 ., These accessions represent three major populations ( Or-I , Or-II and Or-III ) and provide an adequate representation of the ancestral populations of cultivated rice 38 ., Two haplotypes ( H1 and H5 ) were identified that harbored the high root Na+ variants of HKT1;1 , and were found in nearly 70% of the O . rufipogon accessions ., The H1 haplotype displayed the highest frequency in the Or-II clade and was also found in the majority of indica accessions , suggesting that the indica allele is likely derived from Or-II ., In contrast , two haplotypes ( H2 and H6 ) were identified with the low root Na+ variant and were present in only 19% of the O . rufipogon accessions ., The H6 haplotype was the most frequent and present in 18% of the O . rufipogon accessions , but absent from the japonica cultivated rice accessions ., In contrast , H2 occurred at high frequency ( 44% ) in cultivated japonica , particularly the tropical japonica subpopulation , suggesting that H2 is potentially the ancestral haplotype for the japonica subspecies ., Interestingly , the haplotypes found at high frequencies in the japonica subspecies were present at considerably lower frequencies in wild rice accessions ( the highest frequency observed was 0 . 16 ) , indicating that these haplotypes in japonica subspecies may be derived from a relatively small population of wild progenitors ., RNC4 harbors two Na+ transporter genes , HKT1;1 and HKT1;4 ., HKTs are well-known components of salinity tolerance in several plant species including rice ( HKT1;5 is likely the causal gene in the SalTol QTL ) , wheat and Arabidopsis 25 , 43–51 ., Although both HKT1;1 and HKT1;4 displayed significant differences in expression between allelic groups at RNC4 , several key findings suggest that HKT1;1 is more important for root Na+ content during the early tillering stage and for the salinity level imposed in our experimental set-up ., First , the genes are expressed at different developmental stages ., HKT1;1 was expressed at the highest levels in blade and leaf sheath tissues of seedlings , while HKT1;4 showed the highest expression in culms of mature plants ( Fig 4B ) ., Second , reduced expression of HKT1;1 in transgenic RNAi lines resulted in a greater sensitivity to salinity compared to WT , while HKT1;4RNAi and WT plants displayed similar phenotypes under salinity ( Fig 5 ) ., In a recent report , Suzuki et al showed that HKT1;4 is primarily expressed in peduncles during flowering ( 14 week old plants ) and , through RNAi , showed that HKT1;4 is primarily involved in Na+ homeostasis only during the reproductive phase 51 ., Since the current study was conducted during the early tillering stage ( < 1 month old plants ) , it is unlikely that this gene would have an impact on salinity tolerance in this developmental window ., Finally , increased expression of HKT1;1 with the native promoter resulted in higher Na+ in root tissue , which is identical to the phenotype associated with RNC4 ., Together , these data suggests that HKT1;1 is the causal gene underlying RNC4 and contributes to root Na+ content during the early tillering stage ., The differences in Na+ content observed between allelic groups at RNC4 is likely due to functional differences in Na+ transport by HKT1;1 alleles , with the three non-synonymous SNPs in HKT1;1-Zh resulting in higher Na+ transport activity ., Na+ transport occurred at less negative voltages in the isoform found in accessions with high root Na+ compared to that isolated from accessions with low root Na+ ., During salt stress , the accumulation of Na+ in the apoplastic space increases HKT1;1 Na+ transport activity , the apparent affinity for Na+ of this transporter type is particularly low ( Km ~ 80 mM; S12 Fig ) , but in the meantime , uptake of Na+ from the apoplast results in membrane depolarization , which reduces HKT1;1 conductance due to inward rectification property 34 ., In the high root Na+ isoform , a higher ( less negative ) voltage threshold of current activation was observed , for instance in the presence of 10 mM external Na+ noticeable Na+ transport was observed between -75 and -90 mV , while in the low root Na+ isoforms , activation occurred at more negative voltages ( Fig 6C and 6D ) ., Thus , lower Na+ concentrations are required to induce Na+ uptake in the high root Na+ isoform of HKT1;1 ., In summary , the enhanced ability to transport Na+ in accessions harboring the high root Na+ isoform of HKT1;1 is likely due to the early activation of Na+ transport ., Indica varieties have long been recognized to as a source of salt tolerance , largely due to Na+ exclusion from leaf tissue ., The most widely used QTL , SalTol , was identified by Lin et al . using a biparental population derived from the salt tolerant indica landrace ‘Nona Bokra’ and sensitive japonica variety ‘Koshihikari’ 21 ., Tolerance mediated by SalTol is associated with the exclusion of Na+ from shoot tissue , through the removal of Na+ from the xylem and sequestration in xylem parenchyma cells in the root tissue 18 , 25 ., While several studies have demonstrated that the indica subspecies harbors many varieties exhibiting high shoot Na+ exclusion ability , tolerant alleles in SalTol have only been utilized from a few indica landraces , and it is likely that other loci are contributing to Na+ exclusion in the indica subspecies 12 , 52 ., In agreement with previous studies , a considerable difference among the five subpopulations was observed in root and shoot Na+ content and Na+:K+ , with indica accessions generally displaying higher root Na+ content and Na+:K+ , as well as slightly lower shoot Na+ and Na+:K+ ., The relationship between root and shoot ion traits ( specifically Na+ and Na+:K+ ) differed considerably within each of the subpopulations ., For instance , positive correlations were observed between tissues for Na+ and Na+:K+ in the tej , trj and aus subpopulations ., However , in the indica and admix subpopulations no relationships were observed between tissues for Na+ and Na+:K+ ., The moderate positive genetic correlation observed between tissues across all accessions of RDP1 indicates that these traits may be regulated in part by common genes ., However , this may be highly dependent on the subpopulation ., The high frequency of the Na+ accumulating isoform for of HKT1;1 in the indica and admix subpopulations may “uncouple” the relationship between tissues for Na+ and Na+:K+ ., The contrasting root Na+ content observed between indica and japonica accessions of RDP1 is consistent with the differences in transport activity and the frequencies of the high and low root Na+ isoforms of HKT1;1 ., The haplotypes of HKT1;1 could be clearly separated into two distinct groups , corresponding to the japonica ( H2 , H3 , H4 and H7 ) and indica predominate forms ( H1 and H5 ) ., The high root Na+ haplotypes ( H1 , H5 and H8 ) were most frequent in Oryza rufipogon , while the low root Na+ haplotypes were identified in only ~31% of the Oryza rufipogon accessions and were nearly fixed in japonica accessions ., The two major subspecies of Oryza sativa were domesticated from two geographically isolated populations of Oryza rufipogon 38 , 53 ., The low diversity in japonica germplasm reported by several studies is consistent with a bottleneck during domestication , and suggests that the japonica subspecies may be derived from a relatively small founding population of Oryza rufipogon 38 , 54–56 ( S15 Fig ) ., Although the high root Na+ isoform was found in ~30% of the Or-III subpopulation , the founding subpopulation of Oryza rufipogon , it is plausible that the bottleneck experienced during domestication may have resulted in the loss of the high root Na+ HKT1;1 variant from japonica subspecies ., Like many other HKT members , HKT1;1 is well-expressed in the vascular tissue of the shoot , and to a lesser extent in the root 34 , 48 , 50 ., In the current study , HKT1;1RNAi lines were more sensitive to salt stress , and exhibited higher shoot Na+ content and lower root Na+ content compared to WT plants ., The expression patterns of HKT1;1 , as well as the phenotypes exhibited by HKT1;1RNAi lines are in agreement with those reported by Mäser et al for AtHKT1;1 in Arabidopsis , suggesting that the genes may have similar physiological functions 57 ., Like HKT1;1RNAi , athkt1;1 knockout mutants are hypersensitive to salt stress and exhibit higher shoot Na+ and lower root Na+ 57 , 58 ., In rice , Wang et al showed that hkt1;1 knockout mutants accumulate Na+ in xylem sap and display a reduction in Na+ in phloem sap compared to WT 50 ., These observations together with the observed accumulation of Na+ in shoot tissue prompted Wang et al to suggest that HKT1;1 may regulate sodium exclusion from the shoot of seedlings possibly through xylem-to-phloem or parenchyma-to-xylem transfer of Na+ 50 ., Such xylem-to-phloem transfer of Na+ by a HKT member has been debated in Arabidopsis 44 , 45 , 58 ., In agreement with hkt1;1 mutant phenotype reported by Wang et al , athkt1;1 knockout mutants also exhibit higher xylem Na+ and lower phloem Na+ 44 , 45 , 50 ., Although AtHKT1;1 was initially proposed to function in the recirculation of Na+ from the root to the shoot ( via loading of Na+ into the phloem in the shoots ) , Sunarpi et al later proposed that AtHKT1;1 functions primarily in the removal of Na+ from the xylem sap and eventually to the phloem through symplastic diffusion 44 , 45 ., However , a later study showed that AtHKT1;1 was primarily involved in the retrieval of Na+ from the xylem in root tissue , and suggested that the function of AtHKT1;1 in shoot tissue may be dependent on the experimental conditions ( discussed in 3 ) 58 ., For the case of HKT1;1 in rice , further studies ( outside the scope of this manuscript ) are required to provide the exact mechanism for the regulation of root Na+ content and/or shoot Na+ exclusion ., Given the phenotypes exhibited by HKT1;1RNAi lines , as well as the proposed function described by Wang et al . , the absence of an association of HKT1;1 with shoot Na+ or Na+:K+ is surprising 50 ., If HKT1;1 regulates retrieval of Na+ from the parenchyma or xylem in shoot tissues , one would expect that the high root Na+ allele would also have a large impact on shoot Na+ content ., However , the concentration of Na+ in shoot tissue is likely more dependent on the amount of Na+ loaded into the xylem , and thus mechanisms which limit the delivery of Na+ to xylem stream would likely be more effective mechanism for shoot Na+ exclusion 3 ., Without an effective mechanism to limit Na+ entry into the xylem stream in the root , very high expression of HKT1;1 , or a highly active variant of HKT1;1 would likely be necessary to reduce shoot Na+ content ., While the indica ( high root Na+ content ) variant of HKT1;1 displayed higher transport activity compared to japonica variant ( low root Na content ) , it is likely that these biophysical differences are not sufficient to have an impact on shoot Na+ content ., Other members of the HKT family have been identified that are expressed in the vascular tissue of the root , and primarily function to remove Na+ from the xylem to limit the delivery of Na+ to the shoot ., In rice , this function is largely achieved through the action of HKT1;5 25 , 59 ., In contrast to HKT1;1 , HKT1;5 is mostly expressed in the root and therefore is essentially involved in xylem sap desalinization 25 ., In the current study , the SalTol QTL that harbors SKC1/HKT1;5 explained only a small portion of phenotypic variation for shoot Na+ and shoot Na+:K+ ( ~6%; SNP-1 . 11472400 ) ., Several studies have identified alleles within SKC1/HKT1;5 that are associated with Na+ exclusion and salt tolerance , but it is unclear whether the effects of these alleles are as strong as those reported by Gregorio et al . and Bonilla et al . 18 , 20 , 21 , 23 , 25 , 60 ., Given the small effect of this QTL in the current study , as well as the large number of QTL identified for shoot Na+ and Na+:K+ , it

Introduction, Results, Discussion, Materials and methods

Salinity is a major factor limiting crop productivity ., Rice ( Oryza sativa ) , a staple crop for the majority of the world , is highly sensitive to salinity stress ., To discover novel sources of genetic variation for salt tolerance-related traits in rice , we screened 390 diverse accessions under 14 days of moderate ( 9 dS·m-1 ) salinity ., In this study , shoot growth responses to moderate levels of salinity were independent of tissue Na+ content ., A significant difference in root Na+ content was observed between the major subpopulations of rice , with indica accessions displaying higher root Na+ and japonica accessions exhibiting lower root Na+ content ., The genetic basis of the observed variation in phenotypes was elucidated through genome-wide association ( GWA ) ., The strongest associations were identified for root Na+:K+ ratio and root Na+ content in a region spanning ~575 Kb on chromosome 4 , named Root Na+ Content 4 ( RNC4 ) ., Two Na+ transporters , HKT1;1 and HKT1;4 were identified as candidates for RNC4 ., Reduced expression of both HKT1;1 and HKT1;4 through RNA interference indicated that HKT1;1 regulates shoot and root Na+ content , and is likely the causal gene underlying RNC4 ., Three non-synonymous mutations within HKT1;1 were present at higher frequency in the indica subpopulation ., When expressed in Xenopus oocytes the indica-predominant isoform exhibited higher inward ( negative ) currents and a less negative voltage threshold of inward rectifying current activation compared to the japonica-predominant isoform ., The introduction of a 4 . 5kb fragment containing the HKT1;1 promoter and CDS from an indica variety into a japonica background , resulted in a phenotype similar to the indica subpopulation , with higher root Na+ and Na+:K+ ., This study provides evidence that HKT1;1 regulates root Na+ content , and underlies the divergence in root Na+ content between the two major subspecies in rice .

Despite intensive research , few genes have been identified that underlie natural variation for salinity responses in rice ., In this study , we used a rice diversity panel for genome wide association mapping to identify HKT1;1 as a factor regulating Na+ distribution ., Within the rice diversity panel we observed higher Na+ levels in root tissue in the indica subpopulation compared to japonica accessions ., Three non-synonymous variants were identified within HKT1;1 that were associated with altered Na+ accumulation in root tissue , and displayed contrasting frequencies between indica and japonica subspecies ., The introduction of HKT1;1 from an indica accession that contained the three non-synonymous variants into a japonica background resulted in a phenotype similar to that exhibited by the indica subpopulation ., This work suggests that these allelic variants are likely responsible for the higher root Na+ observed in indica accessions ., This study has identified a genetic resource for modifying Na+ content rice , and provides evidence that HKT1;1 underlies the divergence between indica and japonica subspecies in root Na+ content .

oryza, quantitative trait loci, plant growth and development, vertebrates, animals, genetic mapping, xenopus, animal models, developmental biology, plant science, rice, model organisms, amphibians, experimental organism systems, plants, chemical properties, physical chemistry, research and analysis methods, salinity, grasses, chemistry, genetic loci, xenopus oocytes, haplotypes, plant and algal models, root growth, phenotypes, heredity, genetics, biology and life sciences, physical sciences, frogs, organisms

journal.pgen.1004288

Genome-Wide Profiling of Yeast DNA:RNA Hybrid Prone Sites with DRIP-Chip

Elevated DNA:RNA hybrid formation due to defects in RNA processing pathways leads to genome instability and replication stress across species 1–7 ., R loops threaten genome stability and often form under abnormal conditions where nascent mRNA is improperly processed or RNA half-life is increased , resulting in RNA that can hybridize with template DNA , displacing the non-transcribed DNA strand 8 ., A recent study also found that hybrid formation can occur in trans via Rad51-mediated DNA-RNA strand exchange 9 ., Persistent R loops pose a major threat to genome stability through two mechanisms ., First , the exposed non-transcribed strand is susceptible to endogenous DNA damage due to the increased exposure of chemically reactive groups ., The second , more widespread mechanism , identified in Escherichia coli , Saccharomyces cerevisiae , Caenorhabditis elegans and human cells , involves the R loops and associated stalled transcription complexes , which block DNA replication fork progression 3 , 4 , 8 , 10 , 11 ., R loop-mediated instability is an area of great interest primarily because genome instability is considered an enabling characteristic of tumor formation 12 ., Moreover , mutations in RNA splicing/processing factors are frequently found in human cancer , heritable diseases like Aicardi-Goutieres syndrome , and a degenerative ataxia associated with Senataxin mutations 13–17 ., To avoid the deleterious effects of R loops , cells express enzymes for the removal of abnormally formed DNA:RNA hybrids ., In S . cerevisiae , RNH1 and RNH201 , each encoding RNase H are responsible for one of the best characterized mechanisms for reducing R loop formation by enzymatically degrading the RNA in DNA:RNA hybrids 8 ., Another extensively studied anti-hybrid factor is the THO/TREX complex which functions to suppress hybrid formation at the level of transcription termination and mRNA packaging 4 , 11 , 18 , 19 ., In addition , the Senataxin helicase , yeast Sen1 , plays an important role in facilitating replication fork progress through transcribed regions and unwinding RNA in hybrids to mitigate R loop formation and RNA polymerase II transcription-associated genome instability 5 , 20 ., Several additional anti-hybrid mechanisms have also been identified including topoisomerases and other RNA processing factors 2 , 6 , 7 , 9 , 21–23 ., To add to the complexity of DNA:RNA hybrid management in the cell , hybrids also occur naturally and have important biological functions 24 ., In human cells , R loop formation facilitates immunoglobulin class switching , protects against DNA methylation at CpG island promoters and plays a key role in pause site-dependent transcription termination 25–28 ., Transcription of telomeres by RNA polymerase II also produces telomeric repeat-containing RNAs ( TERRA ) , which associate with telomeres and inhibit telomere elongation in a DNA:RNA hybrid-dependent fashion 29–31 ., Noncoding ( nc ) RNA such as antisense transcripts , perform a regulatory role in the expression of sense transcripts that may involve R loops 32 ., The proposed mechanisms of antisense transcription regulation are not clearly understood and involve different modes of action specific to each locus ., Current models include chromatin modification resulting from antisense-associated transcription , antisense transcription modulation of transcription regulators , collision of sense and antisense transcription machineries and antisense transcripts expressed in trans interacting with the promoter for sense transcription 32–40 ., More recently , studies in Arabidopsis thaliana found an antisense transcript that forms R loops , which can be differentially stabilized to modulate gene regulation 41 ., Similarly , in mouse cells the stabilization of an R loop was shown to inhibit antisense transcription 42 ., Here we describe , for the first time , a genome-wide profile of DNA:RNA hybrid prone loci in S . cerevisiae by DNA:RNA immunoprecipitation followed by hybridization on tiling microarrays ( DRIP-chip ) ., We found that DNA:RNA hybrids occurred at highly transcribed regions in wild type cells , including some identified in previous studies ., Remarkably , we observed that DNA:RNA hybrids were significantly associated with genes that have corresponding antisense transcripts , suggesting a role for hybrid formation at these loci in gene regulation ., Consistently , we found that genes whose expression was altered by overexpression of RNase H were also significantly associated with antisense transcripts ., A small-scale cytological screen found that diverse RNA processing mutants had increased hybrid formation and additional DRIP-chip studies revealed specific hybrid-site biases in the RNase H , Sen1 and THO complex subunit Hpr1 mutants ., These genome-wide analyses enhance our understanding of DNA:RNA hybrid-forming regions in vivo , highlight the role of cellular RNA processing activities in suppressing hybrid formation , and implicate DNA:RNA hybrids in control of a subset of antisense regulated loci ., DNA:RNA hybrids have been previously immunoprecipitated at specific genomic sites such as rDNA , selected endogenous loci , and reporter constructs 2 , 5 ., Subsequently , DRIP coupled with deep sequencing in human cells has demonstrated the prevalence of R loops at CpG island promoters with high GC skew 26 ., To investigate the global profile of DNA:RNA hybrid prone loci in a tractable model , we performed genome-wide DRIP-chip analysis of wild type S . cerevisiae ( ArrayExpress E-MTAB-2388 ) using the S9 . 6 monoclonal antibody which specifically binds DNA:RNA hybrids , as characterized previously 43 , 44 ., DRIP-chip profiles were generated in duplicate ( spearmans ρ\u200a=\u200a0 . 78 when comparing each of over 2 million probes after normalization and data smoothing , Supplementary Figure S1 ) and normalized to a no antibody control ., Overall , our DRIP-chip profiles identified several previously reported DNA:RNA hybrid prone sites including the rDNA locus and telomeric repeat regions ( Figure 1 , Supplementary Tables S1 , S2 ) 2 , 29–31 ., DNA:RNA hybrids were also observed at 1217 open reading frames ( ORFs ) ( containing greater than 50% of probes above the threshold of 1 . 5 and found in both wild type replicates ) ( Supplementary Table S3 ) ., These were generally shorter in length than average ( p\u200a=\u200a4 . 29e−58 ) , highly transcribed ( Wilcoxon rank sum test p\u200a=\u200a2 . 21e−6 ) , and had higher GC content ( p\u200a=\u200a2 . 52e−50 ) ( Figure 2A , 2B and 2C , Supplementary Figure S2 ) ., Importantly , despite the correlation between DNA:RNA hybrid association and transcriptional frequency , the wild type DRIP-chip profiles compared to the localization profile of the RNA polymerase II subunit Rpb3 revealed very low correlation ( ρ\u200a=\u200a0 . 0097; 45 ) ., This suggests that the DRIP-chip method was not unduly biased towards the short DNA:RNA hybrids that could theoretically have been captured within active transcription bubbles ., Importantly , because genes with high GC content also have high transcriptional frequencies ( Supplementary Figure S3 ) , it is not clear from our findings whether GC content or transcriptional frequency contributed more to DNA:RNA hybrid forming potential ., Furthermore , we observe that DNA:RNA hybrid prone loci do not encode for mRNA transcripts with particularly long half-lives ( Supplementary Figure S2D ) , suggesting that the act of transcription is vital to DNA:RNA hybrid formation and supporting the notion of co-transcriptional hybrid formation as the major source of endogenous DNA:RNA hybrids ., Our data also revealed DNA:RNA hybrids highly associated with Ty1 and Ty2 subclasses of retrotransposons ( Figure 2E , Supplementary Table S4 ) ., Consistent with our findings at ORFs , the levels of DNA:RNA hybrids correspond well with the known levels of expression of these elements ., In general , Ty1 which constitutes one of the most abundant transcripts in the cell has the highest levels of DNA:RNA hybrids ., Ty3 and Ty4 that are only slightly expressed have much lower levels of hybrids , and the lone Ty5 retrotransposon which is transcriptionally silent is not enriched for DNA:RNA hybrids ( Figure 2E ) ( 46–48 ) ., In contrast to the trends observed with ORFs , GC content in retrotransposons is not highly correlated with the levels of expression , suggesting that expression is the main contributor to DNA:RNA hybrid formation ., Specifically , Ty3 retrotransposons have the highest GC content but have only modest levels of expression and DNA:RNA hybrids ., Certain DNA:RNA hybrid enriched regions identified by our DRIP-chip analysis such as rDNA and retrotransposons are associated with antisense transcripts 49 , 50 ., Therefore , we checked if this was a common feature of DNA:RNA prone sites by comparing our list of DNA:RNA prone loci to a list of antisense-associated genes ( 51 ) ., Because the expression of antisense-associated transcripts may be highly dependent on environmental conditions , we based our analysis on a list of transcripts identified in S288c yeast grown to mid-log phase in rich media which most closely mirrors the growth conditions of our cultures analyzed by DRIP-chip ( 51 ) ., DNA:RNA hybrid enriched genes significantly overlapped with antisense-associated genes , suggesting that DNA:RNA hybrids may play a role in antisense transcript-mediated regulation of gene expression ( Fishers exact test p\u200a=\u200a1 . 03e−12 ) ( Figure 3A , 3B and 3C , Supplementary Table S5 ) ., RNase H overexpression reduces detectable levels of DNA:RNA hybrids in cytological screens and suppresses genomic instability associated with R loop formation presumably through the degradation of DNA:RNA hybrids 7 , 52 , 53 ., To test for a potential role of DNA:RNA hybrids in antisense-mediated gene regulation , we performed gene expression microarray analysis of an RNase H overexpression strain compared to an empty vector control ( GEO GSE46652 ) ., This identified genes that had increased mRNA levels ( upregulated n\u200a=\u200a212 ) or decreased mRNA levels ( downregulated n\u200a=\u200a88 ) as a result of RNase H overexpression ., A significant portion of the genes with increased mRNA levels were antisense-associated ( Fisher exact test p\u200a=\u200a2 . 9e−7 ) ( Figure 3D , Supplementary Table S5 ) and tended to have high GC content , similar to DNA:RNA hybrid enriched genes in wild type ( Supplementary Figure S4 ) ., However , the genes with increased mRNA levels under RNase H overexpression and the antisense-associated genes enriched for DNA:RNA hybrids in our DRIP experiment both tended towards lower transcriptional frequencies ( Figure 3E ) ., These findings suggest that antisense-associated DNA:RNA hybrids moderate the levels of gene expression ., Indeed , genes that were both modulated by RNase H overexpression and enriched for DNA:RNA hybrids were all found to be antisense-associated ( Figure 3F ) ., The mechanism underlying altered gene expression in cells overexpressing RNase H remains unclear ., While the association with antisense transcription is compelling , alternative models exist ., One possibility is that the stress of RNase H overexpression triggers gene expression programs that coincidentally are antisense regulated ., We analyzed gene ontology ( GO ) terms enriched among genes whose expression was changed by RNase H overexpression ., Consistent with previous work , genes for iron uptake and incorporation were strongly activated by RNase H overexpression ( p\u200a=\u200a2 . 21e−12 ) ( Figure 4A , Supplementary Table S6 ) and several of these iron transport genes ( i . e . FRE4 , FRE2 , FRE3 , FET3 , FET4 ) are antisense-associated ( 51 , 54 ) suggesting that overexpression of RNase H activates transcription of these genes by perturbing antisense-mediated regulation ., Alternatively , changes in RNase H levels may increase the cellular iron requirements since sensitivity to low iron concentration is associated with DNA damage and repair 55 ., To test this alternative hypothesis , we tested the RNase H deletion and sen1-1 mutants for sensitivity to low iron conditions compared to a fet3Δ positive control ( Figure 4B ) ., The sen1-1 mutant , RNase H depletion or overexpression did not induce sensitivity to low iron ruling out the possibility that the transcriptional response in cells overexpressing RNase H was a result of cellular iron requirement ., Collectively , our DRIP-chip and microarray analysis suggest that DNA:RNA hybrids may be an important player in antisense-mediated gene regulation ., Transcription-coupled DNA:RNA hybrids have been shown to accumulate in a diverse set of transcription and RNA processing mutants involved in a wide range of transcription related processes ( Table 1 ) ., To gain a broader understanding of factors involved in R loop formation , we performed a cytological screen of RNA processing , transcription and chromatin modification mutants for DNA:RNA hybrids using the S9 . 6 antibody ., Importantly , previous work in our lab has shown that all of the mutants screened exhibit chromosome instability ( CIN ) , which would be consistent with increased hybrid formation 53 ., Significantly elevated hybrid levels were found in 22 of the 40 mutants tested compared to wild type , including a SUB2 mutant which has been previously linked to R loop formation ( Figure 5 , 4 ) ., We also assayed some of the well-characterized R-loop forming mutants , RNase H , Sen1 and Hpr1 , as positive controls for elevated DNA:RNA hybrid levels ( Figure 5 ) ., In our screen , we detected hybrids in mutants affecting several pathways linked to DNA:RNA hybrid formation such as transcription , nuclear export and the exosome ( Figure 5 , Table 1 ) ., Consistent with findings in metazoan cells , we also observed hybrid formation in some splicing mutants ( Figure 5 , Table 1; 56 ) ., Several rRNA processing mutants were enriched for DNA:RNA hybrids ( 7 out of the 22 positive hits ) , likely due to DNA:RNA hybrid accumulation at rDNA genes , a sensitized hybrid formation site ( Figure 1; 2 ) ., It is possible that , as seen in mRNA cleavage and polyadenylation mutants , DNA:RNA hybrid formation may contribute to their CIN phenotypes 6 ., Currently , there are 52 yeast genes whose disruptions have been found to lead to DNA:RNA hybrid accumulation , 21 of which were newly identified by our screen ( Table 1 ) ., The success of this small-scale screen suggests that most RNA processing pathways suppress hybrid formation to some degree and that many DNA:RNA hybrid forming mutants remain undiscovered ., To better understand the mechanism by which cells regulate DNA:RNA hybrids , we performed DRIP-chip analysis of rnh1Δrnh201Δ , hpr1Δ , and sen1-1 mutants in order to determine if these contribute differentially to the DNA:RNA hybrid genomic profile ., The rnh1Δrnh201Δ , hpr1Δ , and sen1-1 mutants are particularly interesting because they have well established roles in the regulation of transcription dependent DNA:RNA hybrid formation ., Our DRIP-chip profiles revealed that , similar to wild type profiles , the mutant profiles were enriched for DNA:RNA hybrids at rDNA , telomeres , and retrotransposons ( Figure 6 , Supplementary Tables S1 , S2 , S3 ) ., The rnh1Δrnh201Δ , hpr1Δ , and sen1-1 mutants also exhibited DNA:RNA hybrid enrichment in 1206 , 1490 and 1424 ORFs respectively compared to the 1217 DNA:RNA hybrid enriched ORFs identified in wild type ( Supplementary Table S4 ) ., Interestingly , in addition to the similarities described above , our profiles also identified differential effects of the mutants on the levels of DNA:RNA hybrids ., In particular , we observed that deletion of HPR1 resulted in higher levels of DNA:RNA hybrids along the length of most ORFs with a preference for longer genes compared to wild type ( Figure 7A , 7B and 7C ) ., This observation is consistent with Hpr1s role in bridging transcription elongation to mRNA export and its localization at actively transcribed genes ( 4 , 57–59 ) ., In contrast , mutating SEN1 resulted in higher levels of DNA:RNA hybrids at shorter genes ( Figure 7A and 7B ) , which is consistent with Sen1s role in transcription termination particularly for short protein-coding genes ( 5 , 60 , 61 ) ., The rnh1Δrnh201Δ mutant revealed higher levels of DNA:RNA hybrids at highly transcribed and longer genes ( Figure 7A and 7B ) which is supported by a wealth of evidence of RNase Hs role in suppressing R loops in long genes to prevent collisions between transcription and replication machineries ( 8 , 62 ) ., Further inspection of our profiles also revealed that rnh1Δrnh201Δ and sen1-1 mutants but not the hpr1Δ mutant had increased DNA:RNA hybrids at tRNA genes ( two tailed unpaired Wilcox test p\u200a=\u200a1 . 56e−19 in the rnh1Δrnh201Δ mutant and 1 . 68e−15 in the sen1-1 mutant ) ( Figure 8A , 8B and 8C , Supplementary Table S7 ) and this was confirmed by DRIP-quantitative PCR ( qPCR ) of two tRNA genes in wild type and rnh1Δrnh201Δ ( Supplementary Figure S5 ) ., Because tRNAs are transcribed by RNA polymerase III , this observation indicates that Hpr1 is primarily involved in the regulation of RNA polymerase II specific DNA:RNA hybrids while RNase H and Sen1 have roles in a wider range of transcripts ., Mutation of SEN1 also led to increased levels DNA:RNA hybrids at snoRNA ( two tailed unpaired Wilcox test p\u200a=\u200a1 . 81e−6 ) ( Figure 8D , 8E and 8F , Supplementary Table S8 ) consistent with its role in 3′ end processing of snoRNAs ( 63 ) ., Identifying the landscape of genomic loci predisposed to DNA:RNA hybrids is of fundamental importance to delineating mechanisms of hybrid formation and the contributions of various cellular pathways ., Although our profiles depend on the specificity of the anti-DNA:RNA hybrid S9 . 6 monoclonal antibody , this aspect has been well characterized 44 and several of our observations are consistent with what has been reported in the literature ., Locus specific tests showed that DNA:RNA hybrids occur more frequently at genes with high transcriptional frequency and GC content 4 , 5 , 18 ., Moreover , in rnh201Δ cells , there is an inverse relationship between GC content and gene expression levels , suggesting that DNA:RNA hybrids accumulate at regions of high GC content and block transcription in the absence of RNase H 64 ., Our work extends the knowledge of DNA:RNA hybrids from a few locus-specific observations to show that , in wild type , there are potentially hundreds of hybrid prone genes that tend to be shorter in length , frequently transcribed and high in GC content 2 , 4 , 56 ., The latter is consistent with recent studies in human cells that demonstrated that genomic regions with high GC skew are prone to R loop formation , which plays a regulatory role in DNA methylation 26 , 27 ., However , while we determined the relationship between GC content and DNA:RNA hybrid formation , we were unable to do the same analysis for GC skew , likely due to the low level of GC skew and lack of DNA methylation in Saccharomyces ., This is unsurprising since the best characterized functional element associated with GC skew , CpG island promoters 26 , , are not found in yeast ., Importantly , our findings at retrotransposons support the notion that expression levels and not GC content contribute more to DNA:RNA hybrid forming potential ., Additionally , DRIP-chip analysis of wild type cells identified hybrid enrichment at rDNA , retrotransposons , and telomeric regions ., Along with previous studies , our DRIP-chip analysis confirms that rDNA is a hybrid prone genomic site and suggests that many factors of rRNA processing and ribosome assembly suppress potentially damaging rDNA:rRNA hybrid formation 2 , 7 ., The presence of TERRA-DNA hybrids at telomeres is supported by our observation of significant hybrid signal at telomeric repeat regions across all DRIP-chip experiments ., The DRIP-chip dataset is a resource for future studies seeking to elucidate the localization of DNA:RNA hybrids across antisense-associated regions and the impact of DNA:RNA hybrid removal on genome-wide transcription ., We observed that genes associated with antisense transcripts were significantly enriched for DNA:RNA hybrids and modulated at the transcript level by RNase H overexpression ., Antisense regulation has been reported at mammalian rDNA and yeast Ty1 retrotransposons , loci that were also enriched for DNA:RNA hybrids in our DRIP-chip 49 , 50 ., The role of DNA:RNA hybrids and RNase H in antisense regulation is currently unclear ., However , there are several non-exclusive models of antisense gene regulation ., One model proposes that the physical presence of the antisense transcripts is crucial to antisense gene regulation ., For instance , trans-acting antisense transcripts have been shown to control Ty1 retrotransposon transcription , reverse transcription and retrotransposition 65 ., Another study has further shown that trans-acting antisense transcripts that only overlap with the sense strand promoter can block sense transcription , potentially by hybridizing with the non-template DNA strand 33 ., These suggest that antisense transcription in cis is not necessary as long as the antisense transcript is present ., It is possible that DNA:RNA hybrids may be formed by the antisense or the sense transcript with genomic DNA ., Moreover , DNA:RNA hybrids may play a functional role in antisense transcription regulation as shown by antisense-associated genes both enriched for DNA:RNA hybrids and affected transcriptionally by RNase H overexpression ., Experiments comparing the ratio of antisense versus sense transcripts and determining the amount of DNA:RNA hybrid formation by either transcript under conditions known to regulate the particular gene will further elucidate the role of RNase H and DNA:RNA hybrids in antisense regulation ., Our investigation of mutant-specific DNA:RNA hybrid formation sites is consistent with the existing literature on Hpr1 , Sen1 and RNase H . Significantly , the hpr1Δ and rnh1Δrnh201Δ mutants exhibited increased DNA:RNA hybrid levels along the length of long genes , while the sen1-1 mutant exhibited increased DNA:RNA hybrid levels along the length of short genes ( Figure 7A ) ., This coheres with Hpr1s function in transcription elongation and mRNA export , and RNase Hs role in preventing transcription apparatus and replication fork collisions , which carry greater consequence for long genes ( 4 , 57–59 , 62 ) ., In contrast , Sen1 is particularly important for transcription termination at short genes ( 61 ) ., In addition , the RNase H deletion and sen1-1 mutants had increased hybrids at tRNA genes , suggesting that they are both required to prevent tRNA:DNA hybrid accumulation ., Interestingly , a recent study found that the mRNA levels of genes encoding RNA polymerase III and proteins that modify tRNA are increased in an rnh1Δrnh201Δ mutant 64 , which may be in response to a lack of properly processed tRNA transcripts ., The finding that both tRNA and snoRNA genes were enriched for hybrids in sen1-1 highlights the role of Sen1 in RNA polymerase I , II and III transcription termination and transcript maturation 60 , 63 , 66 ., More broadly , our data and the literature support the notion that transcripts from RNA polymerases I , II and III can be subject to DNA:RNA hybrid formation especially in RNA processing mutant backgrounds ., Factors regulating ectopic , genome destabilizing DNA:RNA hybrids are best characterized in yeast , although less is known about the functions of native R loop structures ., The genome-wide maps of DNA:RNA hybrids presented here recapitulate the known sites of hybrid formation but also add important new insights to potential functions of R loops ., Most importantly , we demonstrate the usefulness of DRIP profiling for detecting biologically meaningful differences in mutant strains ., Therefore , DRIP profiling of yeast genomes in various mutant backgrounds will be key to understanding the causes and consequences of inappropriate R loop formation and how these are modulated by other cellular pathways ., All strains are listed in Supplementary Table S9 ., For RNase H overexpression experiments , recombinant human RNase H1 was expressed from plasmid p425-GPD-RNase H1 ( 2μ , LEU2 , GPDpr-RNase H1 ) and compared to an empty control plasmid p425-GPD ( 2μ , LEU2 , GPDpr ) 7 ., Briefly , cells were grown overnight , diluted to 0 . 15 OD600 and grown to 0 . 7 OD600 ., Crosslinking was done with 1% formaldehyde for 20 minutes ., Chromatin was purified as described previously 67 and sonicated to yield approximately 500 bp fragments ., 40 µg of the anti-DNA:RNA hybrid monoclonal mouse antibody S9 . 6 ( gift from Stephen Leppla ) was coupled to 60 µL of protein A magnetic beads ( Invitrogen ) ., For ChIP-qPCR , crosslinking reversal and DNA purification were followed by qPCR analysis of the immunoprecipitated and input DNA ., DNA was analyzed using a Rotor-Gene 600 ( Corbett Research ) and PerfeCTa SYBR green FastMix ( Quanta Biosciences ) ., Samples were analyzed in triplicate on three independent DRIP samples for wild type and rnh1Δrnh201Δ ., Primers are listed in Supplementary Table S11 ., For DRIP-chip , precipitated DNA was amplified via two rounds of T7 RNA polymerase amplification ( 68 ) , biotin labeled and hybridized to Affymetrix 1 . 0R S . cerevisiae microarrays ., Samples were normalized to a no antibody control sample ( mock ) using the rMAT software and relative occupancy scores were calculated for all probes using a 300 bp sliding window ., All profiles were generated in duplicate and replicates were quantile normalized and averaged ., Spearman correlation scores between replicates are listed in Supplementary Table S10 ., Coordinates of enriched regions are available in Dataset S1/S2/S3/S4/S5/S6/S7/S8 ., DRIP-chip data is available at ArrayExpress E-MTAB-2388 ., Enriched features had at least 50% of the probes contained in the feature above the threshold of 1 . 5 ., Only features enriched in both replicates were reported ., Transcriptional frequency 69 , GC content ( 70 ) and gene length were compared using the Wilcoxon rank sum test ., Antisense association was analyzed by the Fishers exact test using R . Statistical analysis of genomic feature enrichment was performed using a Monte Carlo simulation , which randomly generates start positions for the particular set of features and calculates the proportion of that feature that would be enriched in a given DRIP-chip profile if the feature were distributed at random 67 ., 500 simulations were run per feature for each DRIP-chip replicate to obtain mean and standard deviation values ., These values were used to calculate the cumulative probability ( P ) on a normal distribution of seeing a score lower than the observed value by chance ., CHROMATRA plots were generated as described previously ( 71 ) ., Relative occupancy scores for each transcript were binned into segments of 150 bp ., Transcripts were sorted by their length , transcriptional frequency or GC content and aligned by their Transcription Start Sites ( TSS ) ., For transcriptional frequency transcripts were grouped into five classes according to their transcriptional frequency described by Holstege et al 1998 ., For GC content transcripts were grouped into four classes according to their GC content obtained from BioMart ( 70 ) ., Average gene , tRNA or snoRNA profiles were generated by averaging all the probes that were encompassed by the features of interest ., For averaging ORFs , corresponding probes were split into 40 bins while 1500 bp of UTRs and their probes were split into 20 bins ., For smaller features like tRNAs and snoRNAs corresponding probes were split into only 3 bins ., Average enrichment scores were calculated using in house scripts that average the score of all the probes encompassed by the feature ., Gene expression microarray data is available at GEO GSE46652 ., Strains harboring the RNase H1 over-expression plasmid or empty vector were grown in SC-Leucine at 30°C ., All profiles were generated in duplicate ., Total RNA was isolated from 1 OD600 of yeast cells using a RiboPure Yeast kit ( A&B Applied Biosystems ) , amplified , labeled , fragmented using a Message-Amp III RNA Amplification Kit ( A&B Applied Biosystems ) and hybridized to a GeneChIP Yeast Genome 2 . 0 microarray using the GeneChip Hybridization , Wash , and Stain Kit ( Affymetrix ) ., Arrays were scanned by the Gene Chip Scanner 3000 7G and expression data was extracted using Expression Console Software ( Affymetrix ) with the MAS5 . 0 statistical algorithm ., All arrays were scaled to a median target intensity of 500 ., A minimum cut off of p-value of 0 . 05 and signal strength of 100 across all samples were implemented and only transcripts that had over a 2-fold change in the RNase H over-expression strain compared to wild type were considered significant ., The correlation between duplicate biological samples was: control ( r\u200a=\u200a0 . 9955 ) , RNase H over-expression ( r\u200a=\u200a0 . 9719 ) ., For statistical analysis , GC content , transcription frequencies and antisense association were analyzed as for DRIP-chip analysis ., Cells were grown to mid-log phase in YEPD rich media at 30°C and washed in spheroplasting solution ( 1 . 2 M sorbitol , 0 . 1 M potassium phosphate , 0 . 5 M MgCl2 , pH 7 ) and digested in spheroplasting solution with 10 mM DTT and 150 µg/mL Zymolase 20T at 37°C for 20 minutes similar to previously described ( 72 ) ., The digestion was halted by addition of ice-cold stop solution ( 0 . 1 M MES , 1 M sorbital , 1 mM EDTA , 0 . 5 mM MgCl2 , pH 6 . 4 ) and spheroplasts were lysed with 1% vol/vol Lipsol and fixed on slides using 4% wt/vol paraformaldehyde/3 . 4% wt/vol sucrose ( 73 ) ., Chromosome spread slides were incubated with the mouse monoclonal antibody S9 . 6 ( 1 µg/mL in blocking buffer of 5% BSA , 0 . 2% milk and 1× PBS ) ., The slides were further incubated with a secondary Cy3-conjugated goat anti-mouse antibody ( Jackson Laboratories , #115-165-003 , diluted 1∶1000 in blocking buffer ) ., For each replicate , at least 100 nuclei were visualized and manually counted to obtain the fraction with detectable DNA:RNA hybrids ., Each mutant was assayed in triplicate ., Mutants were compared to wild type by the Fishers exact test ., To correct for multiple hypothesis testing , we implemented a cut off of p<0 . 01 divided by the total number of mutants compared to wild type , meaning mutants with p<0 . 00024 were considered significantly different from wild type ., 10-fold serial dilutions of each strain was spotted on 90 µM BPS plates with FeSO4 concentrations of 0 , 2 . 5 , 20 or 100 µM and grown at 30°C for 3 days 55 ., A summary of this paper was presented at the 26th International Conference on Yeast Genetics and Molecular Biology , August 2013 74 .

Introduction, Results, Discussion, Methods

DNA:RNA hybrid formation is emerging as a significant cause of genome instability in biological systems ranging from bacteria to mammals ., Here we describe the genome-wide distribution of DNA:RNA hybrid prone loci in Saccharomyces cerevisiae by DNA:RNA immunoprecipitation ( DRIP ) followed by hybridization on tiling microarray ., These profiles show that DNA:RNA hybrids preferentially accumulated at rDNA , Ty1 and Ty2 transposons , telomeric repeat regions and a subset of open reading frames ( ORFs ) ., The latter are generally highly transcribed and have high GC content ., Interestingly , significant DNA:RNA hybrid enrichment was also detected at genes associated with antisense transcripts ., The expression of antisense-associated genes was also significantly altered upon overexpression of RNase H , which degrades the RNA in hybrids ., Finally , we uncover mutant-specific differences in the DRIP profiles of a Sen1 helicase mutant , RNase H deletion mutant and Hpr1 THO complex mutant compared to wild type , suggesting different roles for these proteins in DNA:RNA hybrid biology ., Our profiles of DNA:RNA hybrid prone loci provide a resource for understanding the properties of hybrid-forming regions in vivo , extend our knowledge of hybrid-mitigating enzymes , and contribute to models of antisense-mediated gene regulation ., A summary of this paper was presented at the 26th International Conference on Yeast Genetics and Molecular Biology , August 2013 .

RNA processing factors are mutated in human cancers , inherited developmental disorders and neurodegenerative syndromes ., Defects in RNA processing have been associated with increased levels of mutations and DNA damage in part via the formation of DNA:RNA hybrids ., Although it is likely that specific regions of the genome are more prone to DNA:RNA hybrid formation , a map of hybrid-prone regions is not available ., In this study , we describe the genome-wide distribution of DNA:RNA hybrids in both normal and mutant Saccharomyces cerevisiae cells ., The resulting profiles contribute to both our understanding of the general properties of hybrid-forming loci and to our knowledge of hybrid-mitigating enzymes ., Interestingly , significant DNA:RNA hybrid enrichment was detected at genes associated with antisense transcription ., We show that overexpression of RNase H , which degrades the RNA in hybrids , significantly affects the expression of genes associated with antisense transcripts ., These findings support a role for DNA:RNA hybrids in regulation of gene expression by antisense transcripts .

model organisms, cell biology, genetics, biology and life sciences, genomics, molecular cell biology, research and analysis methods

journal.ppat.1005916

The Tax-Inducible Actin-Bundling Protein Fascin Is Crucial for Release and Cell-to-Cell Transmission of Human T-Cell Leukemia Virus Type 1 (HTLV-1)

Human T-cell leukemia virus type 1 ( HTLV-1 ) , which infects approximately 5–10 million people worldwide 1 , is the only human retrovirus causing cancer: adult T-cell leukemia/lymphoma ( ATL ) , a fatal neoplasia of CD4+ T-cells 2–4 ., Further , HTLV-1 is the causative agent of a neurodegenerative , inflammatory disease , HTLV-1-associated myelopathy/tropical spastic paraparesis ( HAM/TSP ) 5 , 6 ., Both diseases can develop as a consequence of prolonged viral persistence in T-cells after a clinical latency of decades in 1–5% ( ATL ) or 3–5% ( HAM/TSP ) of infected individuals 7 , 8 ., Activated CD4+ T-cells are the main and preferential target for HTLV-1 infection , but the virus is also present in very low amounts in other cell types including CD8+ T-cells , monocytes , and dendritic cells ( DC ) 9 ., After binding to its receptor , which is composed of the glucose transporter GLUT-1 , neuropilin-1 ( NRP-1 ) and heparan sulfate proteoglycans ( HSPGs ) 10–12 , HTLV-1 integrates into the host cell genome ., The virus is mainly maintained in its provirus form ( 9 . 1 kb ) , which is flanked by long terminal repeats ( LTR ) in both the 5’ and 3’ region ., In addition to structural proteins and enzymes common for retroviruses , HTLV-1 encodes regulatory ( Tax , Rex ) and accessory ( p12/p8 , p13 , p30 , HBZ ) proteins 13 ., HTLV-1 replicates either by infecting new cells or by mitotic division and clonal expansion of infected T-cells 14–16 ., Efficient infection of CD4+ T-cells requires cell-cell contacts and coordinated steps of the virus infectious cycle with events in the cell-cell adhesion process ., Thus , transmission of HTLV-1 occurs via breast feeding , sexual intercourse , and cell-containing blood products 9 , 17 ., Unlike human immunodeficiency virus ( HIV ) or murine leukemia virus ( MLV ) , cell-free transmission of HTLV-1 to T-cells is inefficient ., Therefore , only a limited amount of poorly infectious viral particles is produced from infected lymphocytes and free virions can hardly be detected in infected individuals 18–20 ., Thus far , two types of cell-cell contacts have been described to be critical for HTLV-1 transmission , tight cell-cell contacts and cellular conduits 21 , 22 ., For transmission at tight cell-cell contacts , two non-exclusive mechanisms of virus transmission at the virological synapse ( VS ) , a virus-induced specialized cell-cell contact 23 , have been proposed 17: ( 1 ) polarized budding of HTLV-1 into synaptic clefts 21 , and ( 2 ) cell surface transfer of viral biofilms 24 ., The latter consist of extracellular , concentrated viral assemblies that are surrounded by components of the extracellular matrix and cellular lectins 24 ., Beyond , transmission via biofilms seems to be a major route of transmission since removal of biofilms by heparin treatment impairs cell-to-cell transmission by 80% in vitro 24 ., Independent of the route of HTLV-1 transmission , viral particles are thought to be transmitted in confined areas protected from the immune response of the host in vivo ., Moreover , cytoskeletal remodeling and cell-cell contacts are a prerequisite for all routes of virus transmission 21 , 25 ., Although it is known that the viral protein Tax and polarization of the host cell cytoskeleton are crucial for formation of the VS and for HTLV-1 transmission ( for details see: 17 , 23 ) , only little is known about the link between Tax and remodeling of the cytoskeleton to foster viral spread ., The regulatory protein Tax is essential for viral replication due to strong enhancement of viral mRNA synthesis by transactivating the HTLV-1 LTR ( U3R ) promoter ., Beyond , Tax is a potent transactivator of cellular transcription and important for initiating oncogenic transformation 13 ., Tax is also critical for HTLV-1 transmission since Tax cooperates with intercellular adhesion molecule 1 ( ICAM-1 ) , thereby inducing polarization of the microtubule organizing center ( MTOC ) at the VS 26 and thus , enhancing HTLV-1 cell-to-cell transfer ., Furthermore , Tax enhances both actin- and tubulin-dependent transmission of virus-like particles ( VLPs; 25 ) ., However , only few host cell factors with a role in Tax-induced virus transmission have been characterized ., Among those is ICAM-1 , which is induced by Tax and cooperates with Tax in VS formation 26 , 27 ., The Tax-induced small GTP-binding protein GEM enhances cellular migration , conjugate formation , and thus , is required for viral transmission 28 ., In our search of novel target genes of Tax with a putative role in virus transmission , we have previously identified the evolutionary conserved actin-bundling protein and tumor marker Fascin as a new host cell factor strongly induced by Tax 29 ., Fascin cross-links filamentous actin and stabilizes cellular protrusions , filopodia , and invadopodia 30 ., Recent work shows that Fascin also interacts with microtubules to regulate adhesion dynamics and cell migration 31 ., Fascin has evolved as a therapeutic target in several types of cancer since Fascin expression is associated with metastasis in malignant tumors and it correlates with clinical aggressiveness of some tumors 30 ., In hematopoietic cells , Fascin is expressed in mature DC where it is important for stability of dendrites and for formation of the immunological synapse 32 , while no expression of Fascin can be detected in unstimulated human T-cells 33 ., We found that expression of Fascin is a common feature of chronically HTLV-1-infected T-cell lines ., Fascin colocalizes with actin in the cytoplasm and at the membrane of HTLV-1-infected cells ., Furthermore , knockdown of Fascin reduces the invasive capacity of HTLV-1-infected ATL-derived T-cells into extracellular matrix 29 ., Since expression of Tax is sufficient to induce expression of Fascin 29 , 34 and Tax enhances actin-dependent virus transmission 25 , we now asked whether Fascin affects HTLV-1 cell-to-cell transfer ., Here , we report that Fascin is crucial for release and cell-to-cell transmission of HTLV-1 in different cell model systems ., While T-cell conjugate formation is Fascin-independent , cell adhesion of infected cells in co-culture with uninfected cells is impaired upon repression of Fascin ., Imaging of Fascin and the viral gag protein at cell-cell contacts suggests a role of Fascin in transmission potentially by redirecting viral proteins to budding sites ., Thus , Fascin as a major contributor to HTLV-1 transmission provides a link between Tax’s activity and virus transmission ., 293T cells ( kindly provided by Ralph Grassmann ( deceased ) , FAU , Erlangen , Germany ) were cultured in DMEM containing 10% fetal calf serum ( FCS ) , L-glutamine ( 0 . 35g/l ) and penicillin/streptomycin ( Pen/Strep; 0 . 12g/l each ) ., For selection of stable 293T cells carrying shRNAs , 4μg/ml puromycin was added to the media ., The CD4+ T-cell line Jurkat ( ATCC , LGC Standards GmbH , Wesel , Germany ) from acute lymphoblastic leukemia was cultured in RPMI 1640M , Panserin , 10% FCS , L-glutamine and Pen/Strep 35 ., The human Epstein-Barr virus ( EBV ) -positive B-cell line Raji derived from Burkitt’s lymphoma containing the surface receptor CD4 ( Raji/CD4+ ) was a kind gift from Vineet N . Kewal Ramani ( NIH , Frederick , Maryland , USA ) and was cultured in RPMI 1640M , Panserin , 10% FCS , L-glutamine and Pen/Strep containing 500μg/ml geneticin to ensure retainment of the CD4 receptor 36 ., The HTLV-1 in vitro transformed CD4+ T-cell line MT-2 3 and the ATL-derived CD4+ T-cell line HuT-102 2 , 37 were kindly provided by Ralph Grassmann ( deceased , FAU , Erlangen , Germany ) and were cultured in RPMI 1640M , 10% FCS and Pen/Strep ., The HTLV-1 in vitro transformed T-cell line MS-9 ( containing a single , full-length provirus ) 38 was a kind gift from Charles Bangham ( Imperial College , London , UK ) and was cultured in RPMI 1640M , Panserin , 20% FCS , Pen/Strep and 100U/ml interleukin 2 ( IL-2 ) ., All cell lines were checked for integrity by DNA profiling of eight different and highly polymorphic short tandem repeat loci ( DSMZ , Braunschweig , Germany ) ., In general , 107 Jurkat T-cells were transiently transfected by electroporation using the Gene Pulser X Electroporation System ( BioRad , Munich , Germany ) at 290V and 1500μF ., Cells were transfected using a total of 50 or 100μg of DNA ., 5x105 293T cells or stable 293T cell lines that carry shNonsense , shFascin5 or shFascin4 were seeded in 6-well plates 24h prior to transfection ., Cells were transfected with GeneJuice reagent ( Merck Millipore , Darmstadt , Germany ) according to the manufacturer’s protocol using a total amount of 2μg DNA ., HuT-102 cells stably transduced with shRNAs targeting Fascin ( shFascin5 ) or a control ( shNonsense ) were co-cultured with Jurkat T-cells that had been transfected 24h earlier with the luciferase reporter plasmid pGL3-U3R-Luc carrying the luc gene under control of the HTLV-1 core promoter U3R 44 , or with the control plasmid pGL3-Basic ( Promega , Mannheim , Germany ) ., After 48h of co-culture at 37°C , luciferase reporter gene assays were performed ., Relative light units ( RLU ) were normalized on protein content and on background activity of controls ( pGL3-Basic ) ., Values obtained in control cells were set as 100% and at least three independent experiments each performed in triplicate were executed ., Cells were washed once with PBS ( without Ca2+ and Mg2+ ) and then lysed in 100μl lysis buffer ( 100mM Tris/HCl ( pH 7 . 8 ) , 1M dithiotreitol ( DTT ) , 0 . 18mM DCTA , 0 . 2% Triton X-100 , 20% glycerol ) ., After shaking for 30min at 30°C , samples were centrifuged ( 14 . 000rpm , 15min , 4°C ) and supernatants were kept ., Luciferase activities were measured according to the manufacturer’s instructions ( Orion luminometer ) using assay buffer ( 100mM KPO4 , 15mM MgSO4 , 4mM ATP ) and D-luciferin ( 0 . 26 mg/ml; Roche Diagnostics , Indianapolis , IN , USA ) dissolved in assay buffer ., 5x105 of the respective cells were seeded and incubated for 48h at 37°C ., Cells were centrifuged ( 1200rpm , 5min , 25°C ) , pellets were used for western blot analysis , and supernatants of MT-2 cells or of co-cultures from experiments using the single-cycle replication-dependent reporter vectors ( see Infection assays ) were sterile filtrated , and virus release was measured using gag p19 ELISA according to the manufacturer’s protocol ( ZeptoMetrix Corporation , Buffalo , NY , USA ) ., MT-2 cells were either left untreated or treated with DMSO ( solvent control ) , cytochalasin D or nocodazole ( 5μM each ) 48h prior to harvest ., Additionally , MT-2 cells that stably carry shRNAs ( shNonsense or shFascin5 ) were analyzed ., Data were obtained using Softmax Pro Version 5 . 3 software ( MDS Analytical Technologies , Sunnyvale , California , USA ) ., At least , four independent experiments were performed ., Cells were washed once with PBS and protein lysates were obtained by lysis of cells in 100μl lysis buffer ( 150mM NaCl , 10mM Tris/HCl ( pH 7 . 0 ) , 10mM EDTA , 1% Triton X-100 , 2mM DTT and protease inhibitors leupeptin , aprotinin ( 20μg/ml each ) and 1mM phenylmethylsulfonyl fluoride ( PMSF; 1mM ) ) ., After repeated freeze-and-thaw cycles , lysates were centrifuged ( 14 . 000rpm , 15min , 4°C ) ., For detection of Tax , samples were sonicated three times for 20sec before centrifugation ., Equal amounts of protein ( 50μg ) were denatured for 5min at 95°C in sodium dodecyl sulfate ( SDS ) loading dye ( 10mM Tris/HCl ( pH 6 . 8 ) , 10% glycerin , 2% SDS , 0 . 1% bromophenol blue , 5% β-mercaptoethanol ) ., After SDS-PAGE and immunoblotting on nitrocellulose transfer membranes ( Whatmann , Protran , Whatmann GmbH , Dassel , Germany ) , proteins were detected using the following antibodies: rabbit polyclonal antibodies anti-V5 ( Sigma ) , mouse monoclonal antibodies anti-Fascin ( 55K-2; Dako Deutschland GmbH , Hamburg , Germany ) , anti-β-actin ( ACTB; Sigma ) , anti-Hsp90 α/β ( F-8; Santa Cruz Biotechnology , Heidelberg , Germany ) , anti-HTLV-1 gag p19 ( ZeptoMetrix Corporation ) , and anti-GFP ( Sigma ) , and mouse antibodies to Tax , which were derived from the hybridoma cell line 168B17-46-34 ( provided by B . Langton through the AIDS Research and Reference Reagent Program , Division of AIDS , NIAID , NIH; 45 ) ., Secondary antibodies conjugated with horseradish peroxidase ( HRP; GE Healthcare , Little Chalfont , UK ) were used ., Peroxidase activity was detected by enhanced chemiluminescence ( ECL; 98 . 9% ECL A , 1% ECL B , 0 . 031% H2O2 ) using a CCD camera ( Kodak Image Station 4000MM Pro camera , Kodak or Fujifilm LAS-1000 Intelligent Dark Box; Fujifilm ) ., One of at least three independent western blots per experiment is shown ., Intensities of specific bands were quantitated using Advanced Image Data Analyser ( AIDA Version 4 . 22 . 034 , Raytest Isotopenmessgeräte GmbH , Straubenhardt , Germany ) , and values were normalized on those of the housekeeping gene Hsp90 α/β ., 107 Jurkat T-cells were transfected with 50μg pEFTax or pEF ., 5x105 293T cells were transfected with 2μg of pEFTax or pEF ( see Transfections ) ., 48h later , total cellular RNA was isolated from transfected Jurkat or 293T cells ( RNA isolation Kit II , Macherey-Nagel , Düren , Germany ) and reversely transcribed to cDNA using SuperScript II and random hexamer primers ( both Life Technologies GmbH ) ., 200ng of cDNA and SensiMix II Probe Kit ( BioLine GmbH , Luckenwalde , Germany ) were used according to the manufacturer’s instructions for quantitative real-time RT-PCR ( qPCR ) in an ABI Prism 7500 Sequence Analyzer ( Applied Biosystems , Foster City , CA , USA ) ., Primers and FAM ( 6-carboxyfluorescein ) / TAMRA ( tetramethylrhodamine ) -labeled probes for detection of β-actin ( ACTB ) and Tax transcripts have been described before 46 ., A TaqMan Gene Expression Assay ( Hs00979631_g1; Applied Biosystems ) was used for quantitation of Fascin transcripts ., Expression levels were computed by interpolation from standard curves generated from plasmids carrying the respective target sequences and calculation of the mean of triplicated samples ., Relative copy numbers ( rcn ) were determined by normalizing copy numbers on those of ß-actin ( ACTB ) ., At least , three independent experiments were performed ., Microsoft Office Excel software was used for statistical analysis using the t-test ( unpaired ) ., P<0 . 05 was considered to be significant ., To assess the role of Fascin on HTLV-1 transmission , we made use of a single-cycle replication-dependent reporter system that is transfected into donor cells and allows monitoring of reporter gene activity in newly infected target cells only 25 ., Briefly , a virus packaging plasmid encoding all HTLV-1 genes ( wildtype; wt ) and a replication-dependent HTLV-1 reporter vector containing a CMV-promoter driven luciferase ( luc ) gene were co-transfected into 293T cells ., Alternatively , an HTLV-1 packaging plasmid carrying a mutation in the envelope ( env ) gene , and a VSV-G-expression plasmid were co-transfected instead of wt ., The luc gene is oriented in antisense and is interrupted by an intron oriented in sense , therefore translation of the reporter mRNA in transfected cells is precluded ., The vector mRNA is spliced and packaged into VLPs ., After infection and replication , a provirus that lacks the intron is generated , and reporter gene expression ( luc activity ) can be measured in the target cell 25 ., We previously used this system to assess the role of cellular restriction factors on HTLV-1 49 ., To analyze whether Fascin is important for transmission of these HTLV-1 reporter vectors , stable 293T cells with a knockdown of Fascin were generated ., For this purpose , cells were transfected with two different shRNA constructs carrying an IRES-EGFP expression cassette and shRNAs targeting Fascin ( shFascin5 , shFascin4; 29 , 42 ) or a control ( shNonsense ) , and cells were selected with puromycin ., Flow cytometry monitoring GFP-expression revealed that approximately 90% of cells carried the shRNA-constructs ( S2A Fig ) ., Beyond , vitality of stable cell lines was unaffected by the presence of shRNAs as detected by live/dead staining ( S2B Fig ) ., Cell lines were transfected with single-cycle replication-dependent HTLV-1 reporter vectors ( inluc ) , a viral packaging plasmid ( Δenv or wt ) , and as indicated with VSV-G for pseudotyping ( Fig 1A ) ., sh293T cells ( shNonsense ) transfected with inluc and Δenv served as negative control ( control , Fig 1B ) for both wt env-carrying and VSV-G-pseudotyped viral particles ., After 24h , media were changed , and after another 24h , cells were harvested to measure cell-to-cell transmission in luciferase assays ( Fig 1B ) , virus release by gag p19 ELISA ( Fig 1C ) and protein expression by western blot analysis ( Fig 1D ) ., Making use of single-cycle replication-dependent HTLV-1 reporter vectors revealed that stable repression of endogenous Fascin by shRNAs leads to a significant reduction of reporter gene activity ( Fig 1B ) ., While shFascin5 resulted in a strong reduction of reporter gene activity by more than 70% ( Fig 1B ) and , in parallel , of Fascin protein ( Fig 1D ) , the influence of shFascin4 on reporter gene activity ( Fig 1B; by 30% ) and on Fascin protein expression ( Fig 1D ) was less pronounced ., Overexpression of Tax ( black bars ) did not enhance transmission of VSV-G-pseudotyped HTLV-1 ( Fig 1B , left part of left panel ) , confirming earlier observations 25 ., However , overexpressed Tax enhanced cell-to-cell-transmission of HTLV-1 reporters packaged with wt env ( Fig 1B , right part of left panel and enlargement in right panel ) contrary to previous observations 25 ., The latter suggests that Tax and wt env cooperate in cell-to-cell transmission ., The relative infectivity of VSV-G pseudotyped reporter vectors was about 7-fold higher than that of wt env-pseudotyped reporter vectors ( Fig 1B ) and undetectable , if no envelope was added ( control ) ., However , independent of the envelope type used , repression of Fascin significantly reduced the relative infectivity of both wt env-carrying and VSV-G-pseudotyped reporter vectors ., Moreover , independent of the experimental condition , Tax could not further induce expression of Fascin protein in 293T cells ( Fig 1D ) ., While Tax led to a robust induction of Fascin mRNA ( S4A Fig ) and protein ( S4B Fig ) in Jurkat T-cells confirming our previous work 29 , Tax did not further modulate Fascin expression in 293T cells , which already exhibit high amounts of endogenous Fascin ( S4A and S4B Fig ) ., Yet , reporter gene activity as a measure of HTLV-1 cell-to-cell transmission was Fascin-dependent in presence of overexpressed Tax , too ( Fig 1B ) , suggesting that Fascin is also important for HTLV-1 cell-to-cell transmission in cells which express high amounts of endogenous Fascin ., To analyze whether Fascin also impairs virus release , the viral gag p19 protein was measured by ELISA in cells that had been transfected with HTLV-1 reporters packaged with wt env ., Overexpression of Tax ( Fig 1C , black bars ) did not further enhance gag p19 levels in the supernatants of 293T cells and repression of Fascin led to approximately 40% reduction of virus release only when Tax was supplemented suggesting that Tax and Fascin cooperate in processes that are important for virus release ., Since repression of Fascin led to a severe defect of cell-to-cell transmission also in absence of supplemented Tax ( Fig 1B , grey bars ) , but not to a decrease in virus release ( Fig 1C , grey bars ) , this suggests that Fascin’s role in cell-to-cell transmission dominates over its role on virus release in this experimental setup ., However , results obtained by the reporter system only provide a signal upon productive infection of a target cell , while the gag p19 ELISA also quantifies non-infectious VLPs ., To exclude that virus production in the cell is already impaired by repression of Fascin , western blots detecting gag were performed ( Fig 1D ) ., Overall levels of cell-associated gag p55 were comparable between different experimental conditions ., Beyond , Fascin was strongly repressed in presence of shFascin5 and only moderately repressed in presence of shFascin4 ., We could also detect Tax expressed from the packaging plasmids as tiny band , and an increased expression of Tax upon supplementing a Tax expression plasmid ., Taken together , our data indicate that Fascin is important for transmission of HTLV-1 reporter vectors independent of the envelope type in 293T cells ., To strengthen our results , we made use of Fascin-specific nanobodies that had been developed and characterized previously 43 ., Briefly , nanobodies are antigen-binding domains of camelid heavy-chain antibodies ., The employed Fascin-specific nanobodies contain a mitochondrial outer membrane ( MOM ) signal that leads to targeted subcellular delocalization of Fascin to the MOM 43 ., Use of these nanobodies allowed us to trigger Fascin protein loss of function without changing its expression ., Upon transfection of HTLV-1 reporter vectors and expression plasmids encoding Fascin-specific nanobodies into 293T cells , luciferase assays ( Fig 2A ) , gag p19 ELISA ( Fig 2B ) , western blot analysis ( Fig 2C ) and immunofluorescence stains followed by confocal laser scanning microscopy ( Fig 2D–2F ) were performed ., We found that Fascin nanobody 5 ( FASNb5 ) significantly reduced reporter gene activity in a dose-dependent manner compared to a control nanobody ( GFPNb ) and to FASNb2 ( Fig 2A ) ., In gag p19 ELISA ( Fig 2B ) , we measured a dose-dependent decrease of released gag p19 into the supernatants in presence of FASNb5 , suggesting that this nanobody also impairs release of HTLV-1 ., Expression of the nanobodies and the unaltered expression of Fascin were confirmed by western blot analysis ( Fig 2C ) ., Additionally , immunofluorescence was performed confirming earlier studies 43 showing co-localizations of V5-tagged and MOM-expressing nanobodies ( GFPNb , FASNb2 , FASNb5 ) with mitochondria ( Fig 2D ) ., Next , we checked whether Fascin-specific nanobodies lead to efficient delocalization of Fascin by staining V5-tagged nanobodies and Fascin ., Immunofluorescence analysis revealed that FASNb5 lead to a more efficient delocalization of Fascin ( 90 . 2% of Fascin delocalized ) compared to FASNb2 ( 71 . 1%; Fig 2E ) , which mirrors the different impact of FASNb5 and FASNb2 on cell-to-cell transmission ( Fig 2B ) and virus release ( Fig 2C ) ., However , since FASNb5 also impairs Fascin-mediated actin-bundling compared to FASNb2 43 , these data also suggest that Fascin’s actin-bundling activity could be required for transmission of HTLV-1 ., Contrary to delocalizing Fascin , FASNb5 did not delocalize gag to the mitochondria ( Fig 2F ) , suggesting that Fascin and gag do not directly interact , or , if they interact , the interaction is not sustained during delocalization ., Moreover , the impact of FASNb5 on virus release as measured by gag p19 ELISA may be indirect ( Fig 2B ) , e . g . by impairing the transport of gag to budding sites or by impairing budding itself ., Summed up , not only repression , but also delocalization of Fascin in the cell interferes with HTLV-1 cell-to-cell transmission ., Our results obtained thus far do not exclude that repression of Fascin impairs viral entry since we used a one-step transfection/infection co-culture system , where transfected cells produce VLPs that infect neighboring cells 25 , which also harbor a knockdown of Fascin , or which could be impaired by Fascin-specific nanobodies ., Further , since HTLV-1 predominantly infects CD4+ T-cells in vivo , we switched to a more physiological system and analyzed the role of Tax and Fascin on HTLV-1 transmission in CD4+ Jurkat T-cells in co-culture with Raji/CD4+ B-cells , a co-culture system that had been described earlier to allow monitoring of HTLV-1 transmission with single-cycle replication-dependent reporter vectors 25 ., Upon co-transfection of Jurkat T-cells with HTLV-1 reporters ( inluc ) , packaging plasmids ( wt ) , Tax expression plasmids and shRNAs targeting Fascin ( Fig 3A ) , media were changed at 24h , and Jurkat T-cells were co-cultured for another 48h with Raji/CD4+ B-cells ., Measurement of luciferase reporter gene activity reflecting cell-to-cell transmission revealed that repression of Fascin did not affect basal HTLV-1 cell-to-cell transmission ( Fig 3B , grey bars ) ., However , upon overexpression of Tax ( black bars ) , reporter gene activity significantly increased confirming earlier observations in this cell type 25 ., Interestingly , repression of Fascin led to a reduction of Tax-induced reporter gene activity suggesting that Fascin is a major contributor of Tax-induced cell-to-cell transmission ., To exclude and to confirm that the measured reporter gene activity is not due to cell-free virus transmission , we incubated Raji/CD4+ B-cells with supernatants of Jurkat T-cells that had been transfected with the reporter system , which did not result in detectable luciferase signals ( S3 Fig; 25 ) ., Measuring of gag p19 release by ELISA mirrored the results obtained by luciferase assays and showed that Tax-enhanced virus release in Jurkat T-cells occurs Fascin-dependently ( Fig 3C ) ., Knockdown of Fascin in presence of overexpressed Tax led to a reduction of gag p19 release nearly reaching those levels measured without supplemented Tax ., Western blot analysis showed that , contrary to 293T cells ( Fig 1D , S4 Fig ) , Tax is a potent inducer of Fascin transcript and protein expression in Jurkat T-cells ( Fig 3D , S4 Fig ) confirming our previous results 29 , 34 ., Thus , too low levels of endogenous Fascin in Jurkat T-cells without overexpressed Tax ( S4 Fig ) could be a potential explanation for the Tax-dependency of the Fascin-effect in this cell type ., Concomitant with our findings obtained in 293T cells ( Fig 1D ) , western blot analysis revealed that the levels of cell-associated gag p55 were comparable between different experimental conditions also in Jurkat T-cells ( Fig 3D ) ., Contrary to 293T cells , Fascin was induced by Tax in Jurkat T-cells ., Further , Fascin was strongly repressed in presence of shFascin5 and moderately repressed in presence of shFascin4 ., Tax expressed from the packaging plasmids could be detected as a tiny band , and an increased expression of Tax was detectable upon supplementing a Tax expression plasmid ., Thus , Tax enhances virus release , and augments cell-to-cell transmission from Jurkat T-cells to Raji/CD4+ B-cells dependent on Fascin ., To substantiate these findings , we tested Fascin-specific nanobodies instead of shRNAs in the Jurkat-Raji/CD4+ co-culture model ., Compared to the control nanobody GFPNb and to FASNb2 , FASNb5 led to a significant reduction ( by 61% ) of Tax-induced HTLV-1 reporter gene activity ( Fig 3E ) ., This suggests that delocalization of Fascin without changing its expression ( Fig 3F ) and inhibition of Fascin’s actin-bundling activity by FASNb5 43 impair HTLV-1 cell-to-cell transmission in different cell types ( Figs 2 , 3E and 3F ) ., Taken together , use of single-cycle replication-dependent HTLV-1 reporter vectors revealed that stable repression of endogenous Fascin ( 293T cells ) , or of Tax-induced Fascin ( Jurkat T-cells ) by shRNAs and inhibition of Fascin using specific nanobodies impair both gag p19 release and HTLV-1 cell-to-cell transmission ., Next , we asked whether Fascin contributes to HTLV-1 cell-to-cell transmission also in chronically HTLV-1-infected T-cells , which express high amounts of Fascin protein ., For this purpose , ATL-derived HTLV-1-infected HuT-102 cells were stably transduced with lentiviral vectors expressing either a shRNA targeting Fascin ( shFascin5 ) or a nonsense shRNA ( shNonsense ) 29 ., According to a published protocol 50 , HuT-102 cells were co-cultured with Jurkat T-cells that had been transfected with an HTLV-1-LTR ( U3R ) -dependent luc gene reporter system ( pGL3-U3R; Fig 4A ) ., Upon infection of Jurkat T-cells , the viral Tax protein should activate expression of the HTLV-1 U3R resulting in enhanced luciferase activity ., After 24h of co-culture , luciferase activity was measured and normalized on protein content and on transactivation of a mock luciferase construct ( Fig 4B ) ., Transactivation of the reporter in Jurkat T-cells was diminished by more than 50% when Fascin was knocked down in the co-cultured HTLV-1-infected HuT-102 cells hinting at a role of Fascin for HTLV-1-mediated cell-to-cell transfer ., In parallel , knockdown of Fascin in HuT-102 was verified in immunoblots and shown to be Fascin-specific since Tax protein and the housekeeping gene ß-actin ( ACTB ) were not affected by shFascin5 ( Fig 4C ) ., Further , we excluded detrimental effects of the shRNA on cell vitality by measuring apoptosis and cell death in stable cell lines compared to cells treated with 15μM etoposide , which is known to induce cell death ( S5 Fig ) ., Since results from co-culture assays may also reflect cell fusion events or the transfer of Tax-containing exosomes 51 , we decided to measure HTLV-1 infection also directly by a flow cytometry based assay that allows monitoring of newly infected cells ., For this purpose , we first stably transduced the chronically HTLV-1-infected T-cell line MT-2 cells with lentiviral vectors expressing either shFascin5 or shNonsense ., According to an established protocol 28 , MT-2 cells were then co-cultured with uninfected Jurkat T-cells for 1h and the number of newly infected Jurkat T-cells was detected by flow cytometry by measuring the amount of the viral matrix protein gag p19 in these cells ( Fig 4D and 4E ) ., For this purpose , co-cultures were permeabilized and stained with antibodies targeting HTLV-1 gag p19 and the IL-2 receptor alpha chain CD25 , which is present on MT-2 cells , but not on Jurkat T-cells 52 ., Flow cytometry revealed that repression of Fascin led to a significant reduction of newly infected , gag p19-positive Jurkat T-cells to 68% compared to the control ( Fig 4E ) ., Beyond , western blot analysis confirmed a robust reduction of Fascin protein in MT-2 cells carrying shFascin5 ( Fig 4F ) , while Tax and ACTB were unaffected ., Further , cell vitality was also unaffected by repression of Fascin as indicated by live/dead stainings ( S5 Fig ) ., We also measured release of gag p19 into culture supernatants and found that knockdown of Fascin not only reduced infection of co-cultured Jurkat T-cells , but also diminished the release of gag p19 ( Fig 4G ) ., Similar results were obtained with MT-2 cells treated with cytochalasin D or nocodazole ( 5μM each ) , which interfere with actin or tubulin polymerization , respectively ( Fig 4H ) ., Overall , our observations are in line with the data we obtained with the HTLV-1 reporter vectors ( Figs 1–3 ) suggesting that , independent of the cell and test system used , repression of Fascin impairs release and cell-to-cell transmission of HTLV-1 ., Immunoblot analysis revealed that cell-associated gag and processing of the gag p55 precursor into gag p19 and gag p27 was unaffected by knockdown of Fascin ( Fig 4I ) ., However , treatment of MT-2 cells with the compounds cytochalasin D and nocodazole interfered with processing of gag p55 suggesting that chemical interference with the cytoskeleton acts differently on virus production than Fascin repression ., Taken together , we found that Fascin is critical for release and cell-to-cell transmission of HTLV-1 reporter vectors , and for transactivation and infection of co-cultured T-cells indicating an important role of Fascin in HTLV-1 cell-to-cell transmission ., To shed light on the mechanism of Fascin’s role during HTLV-1 transmission , we asked whether Fascin enhances conjugate formation between infected and uninfected T-cells similar to the small GTP-binding protein GEM 28 ., For this purpose , we performed a flow cytometry-based conjugate formation

Introduction, Materials and Methods, Results, Discussion

The delta-retrovirus Human T-cell leukemia virus type 1 ( HTLV-1 ) preferentially infects CD4+ T-cells via cell-to-cell transmission ., Viruses are transmitted by polarized budding and by transfer of viral biofilms at the virological synapse ( VS ) ., Formation of the VS requires the viral Tax protein and polarization of the host cytoskeleton , however , molecular mechanisms of HTLV-1 cell-to-cell transmission remain incompletely understood ., Recently , we could show Tax-dependent upregulation of the actin-bundling protein Fascin ( FSCN-1 ) in HTLV-1-infected T-cells ., Here , we report that Fascin contributes to HTLV-1 transmission ., Using single-cycle replication-dependent HTLV-1 reporter vectors , we found that repression of endogenous Fascin by short hairpin RNAs and by Fascin-specific nanobodies impaired gag p19 release and cell-to-cell transmission in 293T cells ., In Jurkat T-cells , Tax-induced Fascin expression enhanced virus release and Fascin-dependently augmented cell-to-cell transmission to Raji/CD4+ B-cells ., Repression of Fascin in HTLV-1-infected T-cells diminished virus release and gag p19 transfer to co-cultured T-cells ., Spotting the mechanism , flow cytometry and automatic image analysis showed that Tax-induced T-cell conjugate formation occurred Fascin-independently ., However , adhesion of HTLV-1-infected MT-2 cells in co-culture with Jurkat T-cells was reduced upon knockdown of Fascin , suggesting that Fascin contributes to dissemination of infected T-cells ., Imaging of chronically infected MS-9 T-cells in co-culture with Jurkat T-cells revealed that Fascin’s localization at tight cell-cell contacts is accompanied by gag polarization suggesting that Fascin directly affects the distribution of gag to budding sites , and therefore , indirectly viral transmission ., In detail , we found gag clusters that are interspersed with Fascin clusters , suggesting that Fascin makes room for gag in viral biofilms ., Moreover , we observed short , Fascin-containing membrane extensions surrounding gag clusters and clutching uninfected T-cells ., Finally , we detected Fascin and gag in long-distance cellular protrusions ., Taken together , we show for the first time that HTLV-1 usurps the host cell factor Fascin to foster virus release and cell-to-cell transmission .

Human T-cell leukemia virus type 1 ( HTLV-1 ) is the only human retrovirus causing cancer and is transmitted via breast feeding , sexual intercourse , and cell-containing blood products ., Efficient infection of CD4+ T-cells occurs via polarized budding of virions or via cell surface transfer of viral biofilms at a tight , specialized cell-cell contact , the virological synapse ( VS ) ., The viral protein Tax and polarization of the host cell cytoskeleton are crucial for formation of the VS , however , only little is known about the link between Tax and remodeling of the cytoskeleton to foster viral spread ., The actin-bundling protein Fascin has evolved as a therapeutic target in several types of cancer ., Here , we show that Fascin is also crucial for release and transmission of the tumorvirus HTLV-1 ., Since Fascin is a transcriptional target gene of Tax in T-cells , our work provides a link between Tax’s activity and virus transmission ., Visualization of cell-cell contacts between infected and uninfected T-cells suggests a role of Fascin in viral transmission potentially by facilitating the transport of viral proteins to budding sites ., Thus , Fascin is not only crucial for metastasis of tumors , but also for transmission of HTLV-1 and is a new cellular target to counteract HTLV-1 .

blood cells, flow cytometry, medicine and health sciences, immune cells, pathology and laboratory medicine, 293t cells, pathogens, immunology, biological cultures, microbiology, plasmid construction, retroviruses, viruses, rna viruses, dna construction, molecular biology techniques, infectious disease control, research and analysis methods, specimen preparation and treatment, staining, infectious diseases, white blood cells, spectrum analysis techniques, animal cells, medical microbiology, htlv-1, t cells, microbial pathogens, cell lines, molecular biology, spectrophotometry, antibody-producing cells, cytophotometry, cell staining, cell biology, b cells, viral pathogens, biology and life sciences, cellular types, organisms

journal.ppat.1005270

PD-L1 Blockade Differentially Impacts Regulatory T Cells from HIV-Infected Individuals Depending on Plasma Viremia

Inhibiting programmed cell death 1 ( PD-1 ) signalling has a potential therapeutic value for treating cancers and persistent viral infections ( reviewed in 1–5 ) ., PD-1 is a co-inhibitory receptor that plays a major role in exhaustion , a dysfunctional state of effector cells caused by antigen persistence 6 ., Exhausted T cells present defects in effector function including impaired proliferation , cytotoxic capacity and cytokine production ., These defects can be partially restored by blocking the interaction between PD-1 and its ligand programmed death ligand-1 ( PD-L1 ) , which notably reduces viral loads in several animal infection models 7–10 ., This observation has also been extended to important persistent human infections such as the human immunodeficiency virus ( HIV ) infection , both in vitro 11–14 and in vivo in HIV-infected humanized mice 15 , 16 ., Since the HIV load is directly correlated with disease progression 17 , an augmentation of antiviral immune responses by blocking the PD-1/PD-L1 pathway might help to control viral replication and slow down pathogenesis ., Furthermore , it may facilitate clearance of latently infected cells , and thus may represent a promising strategy to reach a functional cure of HIV infection 18 , 19 ., PD-1 and PD-L1 are expressed on several cell types including regulatory T cells ( Treg cells ) 20 ., Treg cells are a suppressive T cell subset mediating self-tolerance and immune homeostasis ( reviewed in 21 , 22 ) ., During HIV-infection , Treg cells have both , beneficial and detrimental roles ( reviewed in 23–25 ) ., For example , Treg cells control excessive immune activation that limits immunopathology and the availability of HIV target cells ., On the contrary , Treg cells contribute to the destruction of the lymphatic tissue architecture , and inhibit HIV-specific immune responses promoting virus persistence ., Therefore , any therapeutic alteration of Treg cell numbers and function may directly influence the balance between immunopathology and viral control ., PD-L1 blockade therapy in HIV-infected individuals is expected to affect their Treg cells ., Indeed , several roles of the PD-1/PD-L1 pathway are already described for this cell subset ., For example , PD-1/PD-L1 pathway is essential in the induction of Treg cells in the periphery 26–28 and the maintenance of their suppressive capacity 28–33 ., PD-1 is also described as a negative regulator of Treg cells in hepatitis C virus infection 34 ., Likewise , in vivo blockade of PD-L1 increased the numbers of Treg cells in the Friend virus mice model 35 ., In the context of a HIV infection , PD-1 was found up-regulated in Treg cells compared with healthy controls 36–38 ., Nonetheless , as most reports have focused on effector cells , possible effects from PD-L1 blockade on Treg cells have been neglected ., In light of the upcoming therapeutic trials blocking PD-L1 in HIV-infected patients , it is important to understand its consequences for Treg cells , as they are essential players balancing immunopathology and antiviral effector responses ., The purpose of this report was to investigate the impact of ex vivo PD-L1 blockade on Treg cells from HIV-infected individuals ., We found that PD-L1 blockade had no effect on Treg cells’ suppressive capacity ., However , PD-L1 blockade increased the proliferative capacity of Treg cells from viremic individuals but had no significant effect on Treg cells from individuals that control viremia ., Interestingly , we found that PD-L1 blockade in peripheral blood mononuclear cells ( PBMC ) from viremic individuals increased virus production ., This increase was related with increased Treg cell frequencies , suggesting that the inhibitory function of Treg cells may play a role in virus expansion upon PD-L1 blockade ., In contrast to the differential effect from PD-L1 blockade on Treg cell proliferation , we observed an increase in the proliferation of effector T cells in all groups of HIV–infected individuals ., Therefore , manipulating PD-L1 in vivo is expected to influence the net gain of effector function depending on the subject’s plasma viremia ., Previous studies have shown that, ( i ) PD-1 is overexpressed on CD4- and CD8- T cells in several persistent infections and cancers , and, ( ii ) that this overexpression plays a key role in the exhausted phenotype of these cells ( reviewed in 39 ) ., To first evaluate the expression of PD-1 and its ligand PD-L1 on Treg cells from HIV-infected individuals , we used the gating strategy of Miyara and colleagues 40 ., It distinguishes between effector Treg cells ( eTreg , CD4+CD45RA-FOXP3hi ) and resting Treg cells ( rTreg , CD4+CD45RA+FOXP3lo ) ( Figs 1 , S1 and S2 ) ., The advantage over the traditional Treg cell characterization by CD4+CD25hiCD127loFOXP3+ is that conventional CD4 T cells with an up-regulated CD25 and FOXP3 expression due to the generalized , infection-related immune activation are excluded from the analysis 41 ., PBMC from HIV-infected individuals and healthy controls were isolated , stained with fluorescence-labelled antibodies and characterized by flow cytometry ., A significantly higher percentage of PD-1+ Treg cells was observed for HIV-infected individuals ( 8 . 2% ± 0 . 8 SEM ) compared with controls ( 3 . 0% ± 0 . 4 SEM ) ( Fig 1B ) ., This difference in PD-1 expression was due to PD-1 on effector Treg cells ( 13 . 6% ± 1 . 2 SEM ) since very little PD-1 was expressed on resting Treg cells ( 1 . 5% ± 0 . 2 SEM ) ., These observations are concordant with previous data 36–38 and fit to the current understanding of PD-1 upregulation induced by T cell stimulation 42 ., We also found a higher percentage of PD-L1+ Treg cells for HIV-infected individuals ( 6 . 59% ± 1 . 1 SEM ) compared with controls ( 1 . 75% ± 0 . 33 SEM ) ( Fig 1B ) ., However , in contrast to the expression pattern of PD-1 , an increased percentage of PD-L1+ cells was observed for both , effector and resting Treg cells ., Furthermore , the expression of PD-1 and PD-L1 on Treg cells from HIV-infected individuals correlated positively ( Fig 1C ) ., The differential distribution of PD-1 and PD-L1 on resting and effector Treg cells from HIV-infected individuals and healthy controls suggested that the virus itself could induce PD-L1 upregulation on Treg cells ., To test this hypothesis , PBMC from healthy controls were isolated and exposed to HIV-1Bal containing supernatants without additional stimuli or additional interleukin-2 ., As controls , we used supernatants from non-infected PBMC that have been cultured under similar conditions as the virus-exposed cells ., While culture supernatants from non-infected PBMC increased the frequency of PD-L1+ effector and resting Treg cells , virus exposure dramatically augmented this effect in a dose-dependent manner ( Figs 1D and S3A ) ., In contrast , virus exposure had no effect on PD-1 expression ( Fig 1D ) ., To evaluate whether PD-L1 upregulation occurred without infection , we cultured PBMC with infectious HIV-1Bal in the presence or absence of the HIV entry inhibitor T20 ., PD-L1 upregulation on Treg cells occurred after virus exposure even in the presence of T20 ( S3B Fig ) ., These data are in line with previous studies that demonstrate a PD-L1 upregulation upon HIV exposure in different cell populations including monocytes , macrophages , dendritic cells , neutrophils and CCR5+T cells 43–46 ., In addition , we observed that the HIV-envelope protein gp120 induced PD-L1 upregulation in Treg cells providing an extra mechanistic insight into how HIV can induce PD-L1 ( S3C Fig ) ., When taken together , our results show an upregulation of PD-1 and PD-L1 on Treg cells of HIV-infected individuals that may be mediated by different routes and suggest that the Treg cell compartment is likely to be influenced by immunotherapy targeting the PD-1/PD-L1 pathway ., PD-1 expression on CD4- and CD8- T cells correlates with HIV disease progression 11 , 13 , 47 ., To test whether the same is true for Treg cells , we analysed PD-1 expression on these cells from HIV-infected individuals categorized into 4 groups according to CD4 T cell counts and viral load ( S2 Table ) ., The highest percentage of PD-1-expressing Treg cells was found in the HIV study group with the lowest CD4 T cell counts and highest viral loads ( Fig 2B ) ., As many as 13 . 6% ± 2 . 3 SEM of Treg cells where PD-1+ in this group whereas only 4 . 7% ± 0 . 4 SEM of Treg cells where PD-1+ in the group of individuals under combination antiretroviral therapy ( cART ) ., PD-1 expression on Treg cells paralleled PD-1 on total CD4-T cells but differed from CD8-T cells ( S4A Fig ) ., Within viremic individuals , the percentage of PD-1+ CD8 T cells was high irrespective of the CD4 T cell counts , whereas the percentage of PD-1+ Treg and PD-1+ total CD4 T cells was higher in individuals with low CD4 T cell counts ( <500 CD4/μL ) ., Consistently , PD-1 on Treg cells correlated positively with viral load and negatively with CD4 T cell counts ( Fig 2C ) ., This is concordant with previous observations made for CD4 T cells and CD8 T cells from HIV-infected patients 11 , 13 , 47 and reproduced here with individuals of our study groups ( S4B Fig ) ., To further substantiate the relation between PD-1 expression on Treg cells and antigen exposure , we followed 5 patients before and after antiretroviral treatment interruptions ., Samples were collected from the same patient at 4 time-points , ( 1 ) before starting treatment , ( 2 ) during treatment , ( 3 ) upon interruption of treatment , and ( 4 ) after restarting treatment ., As can be seen in Fig 3 , the percentage of PD-1 expressing Treg cells followed viremia and cART reduced PD-1 expression on Treg cells ., PD-1 is a negative regulator of the proliferative capacity in effector T cells ., To characterize the relationship between PD-1 expression and the proliferative capacity of Treg cells , PBMC from individuals of the different HIV study groups were labelled with Carboxyfluorescein succinimidyl ester ( CFSE ) , stimulated with HIV Gag peptides and analysed for proliferation by CFSE dilution via flow cytometry ., As shown in Fig 4B , the proliferative capacity of Treg cells was strikingly impaired in non-treated individuals ., It correlated positively with CD4 T cell counts and negatively with viral loads ( Fig 4C and 4D ) and PD-1 expression on Treg cells prior to stimulation ( Fig 4E ) ., These observations parallel those reported for effector T cells 11 and suggest a negative role of PD-1 for Treg cell proliferation ., To analyse the impact of a PD-L1 blockade on the proliferative capacity of Treg cells from HIV-infected individuals , CFSE-labelled PBMC were cultured in the presence of HIV Gag peptides and PD-L1 blocking antibody or an isotype control antibody ., Cell proliferation was quantified by CFSE dilution via flow cytometry ., A significant gain on the proliferative capacity of Treg cells , as well as that of effector CD4- and CD8- T cells , is shown in Fig 5 as fold change in proliferation relative to the isotype antibody control stimulations ( p <0 . 0001 ) ., PD-L1 blockade led to a roughly 2 fold mean increase in the percentage of proliferating Treg cells , comparable to that of effector CD4- and CD8- T cells ( Fig 5B ) ( p = 0 . 668 ) ., The range of responses was broad ., The increase in Treg cell proliferation positively correlated with the viral load of the analysed individuals ( Fig 5C ) ., To further analyse the functional consequences that a PD-L1 blockade may have on Treg cell function after an antigenic stimulation , we analysed the increase of Treg cells expressing effector molecules such as CD39 and CTLA4 as well as their suppressive capacity ., Upon PD-L1 blockade a slight but significant increase in the frequency of CD39- and CTLA4- expressing Treg cells was observed relative to the control condition ( p = 0 . 026 and 0 . 039 , respectively ) as well as to the fold change in the percentage of Helios-expressing Treg cells ( p = 0 . 027 ) ( Figs 5E and S6A ) ., The latter is a transcription factor suggested to identify thymic Treg cells , and used as a control ., As expected , the frequency of Helios-expressing Treg cells did not increase upon PD-L1 blockade ( p = 0 . 862 ) ., To test the capacity of expanded Treg cells to suppress CD8 T cell proliferation , Treg cells were isolated from PBMCs after a 6-day-culture in the presence of PD-L1 blocking antibody or an isotype control antibody , and co-cultured with CFSE-labelled PBMCs in the presence of anti-CD3/anti-CD28 and interleukin-2 ., Proliferation of CD8 T cells was quantified by analysing CFSE profiles by flow cytometry ., A dose-dependent inhibition of CD8 T cell proliferation was observed that was not significantly different from that of isolated Treg cells expanded under control conditions ( Fig 5F ) ., Although the in vitro study of the suppressive capacity of Treg cells might not always be predictive of in vivo function , the presented data suggest that the relief of the PD-1/PD-L1 interaction during expansion does not alter the suppressive capacity of Treg cells on a per cell basis ., To analyse whether the PD-L1 blockade-mediated restoration of the proliferative capacity of Treg , CD4- and CD8- T cells as shown in Fig 5B was dependent on the HIV infection stage of the host , the respective data points were grouped according to viremia , CD4 T cells counts and antiretroviral treatment ( patient grouping as of S2 Table ) ., The PD-L1 blockade significantly increased the proliferation of Treg cells from patients with high viremia irrespective of their CD4 T cell counts ( Figs 6A and S6B ) ., Treg cells from patients that controlled viremia ( either spontaneously or by cART ) showed no significant proliferation increase compared with their isotype antibody control stimulation ., In contrast , the increase of CD8 T cell proliferation mediated by the PD-L1 blockade was significant with respect to the isotype antibody control for all 4 patient groups ( Figs 6A and S6B ) ., Importantly , the PD-L1 blockade affected the Treg cells from the high viremic groups more than the respective CD8 T cells ., The inverse was true in the groups with controlled viremia ., To confirm that PD-L1 blockade differentially impacts Treg cells and CD8 T cells depending on the plasma viremia of the host , fold changes in proliferation upon PD-L1 blockade were measured longitudinally in 7 additional individuals before and after antiretroviral treatment ( pre-cART and on-cART , respectively ) ., As can be seen in Fig 6B , in samples from pre-cART treatment , PD-L1 blockade increased the proliferative capacity of Treg cells by approximately 2 fold , which is comparable to that of effector CD4- and CD8- T cells ., However , in samples from the same patients after a cART period , PD-L1 blockade had no significant effect on Treg cell proliferation ., In contrast , the proliferative capacity of effector T cells was increased 2 fold upon PD-L1 blockade comparable to that of pre-cART samples ( Fig 6B ) ., As shown in Fig 6C , PD-L1 blockade preferentially increased the proliferative capacity of effector T cells over regulatory T cells in samples from individuals under cART treatment ., Thus the net gain of T cell effector function after PD-L1 blockade may critically depend on plasma viremia ., To analyse the consequences of the PD-L1 blockade for ex vivo HIV reactivation , supernatants of the above-described PBMC cultures were collected and tested for viral production by HIV p24 antigen determination ., HIV production was readily detectable in most PBMC cultures from the viremic patient groups ( 16 from 19 samples ) ( Fig 7A ) ., The blockade of PD-L1 consistently increased viral production relative to the isotype antibody control stimulations ., This increase of HIV production correlated positively with the increase in CD4 T cell proliferation ( Fig 7B ) and the fraction of Treg cells in the lymphocyte population ( Fig 7C ) but not with the increase in Treg cell proliferation ( S7 Fig ) ., Furthermore , the increase in HIV production correlated negatively with the fold changes of CD4 T cell/Treg cell as well as CD8 T cell/Treg cell ratios respectively ( Fig 7D and 7E ) ., Together this suggests that the inhibitory function of the Treg cells rather than their capacity of being an HIV target cell may play a role in virus expansion under these conditions ., Perhaps the most striking finding of this work is the observation that PD-L1 blockade restores the proliferative capacity of regulatory T cells ( Treg cells ) from HIV-infected individuals differentially depending on plasma viremia ., Treg cells from viremic patients show the largest fold increase in proliferation while the cell´s suppressive capacity is maintained ., As Treg cells contribute to maintain exhaustion 48 , 49 , therapeutic interventions aiming to disrupt T cell exhaustion by means of blocking the PD-1 signalling pathway should first reduce the HIV load by antiretroviral drugs ., Only this may guarantee the biggest possible net gain of effector T cell function and subsequent better immunological control over HIV ., Relative to healthy controls , the frequency of PD-1-expressing Treg cells was significantly increased in all four groups of HIV-infected individuals and correlated positively with markers of disease progression such as virus load and reduction of CD4 T cells ., With this , Treg cells follow the same trend as total CD4- and CD8- T cells ( S4 Fig ) as previously described 11 , 13 , 47 ., This is intriguing , as a coordinated up-regulation of a negative signalling receptor on both effector T cells and suppressor T cells seems counterintuitive ., However , it is consistent with the model of dynamic co-evolution of memory and regulatory T cells at sites of infection 50 , 51 and data from subsequent studies of Treg cells in chronic hepatitis C virus infection 34 , 52 ., Accordingly , an expanding virus triggers an effector T cell response with concomitant Treg cell generation ., The killing of infected cells by effector T cells then promotes tissue injury ., This is dampened via PD-1 signalling on effector cells as well as expanding Treg cells ., To limit exaggerated suppression and maintain homeostasis , Treg cell expansion is also controlled by PD-1 on Treg cells ., While the data from our cross-sectional study do not enable the analysis of the temporal appearance of the PD-1-expressing T cell subsets in the HIV-infected individuals studied , they are concordant with this model of an antigen-driven coordinated response in order to balance, ( i ) virus reduction by effector T cells and, ( ii ) reduction of immunopathology by Treg cells with, ( iii ) maintenance of the adaptability of T cell responses to subsequent viral bursts ., The direct correlation of PD-1 on Treg cells with patient’s viral load is consistent with the idea that persistent antigen exposure is a main trigger of PD-1 expression ., However , exposure of PBMCs to HIV under non-stimulating conditions did not induce PD-1 on effector or on resting Treg cells significantly thus suggesting that additional activation signals are required ., Interestingly , this exposure to HIV was sufficient to massively up-regulate PD-L1 on both eTreg and rTreg cells in a virus dose-dependent manner ., As PD-L1 can participate in Treg cell induction 26–28 as well as PD-1-mediated suppression 29 , 30 , 32 , the role of both Treg cell states in virus sensing and subsequent signal conversion merits further investigation ., Upon PD-L1 blockade , the fold change of Treg cell proliferation was the highest in viremic individuals and correlated with viral loads ., However the fold change of Treg cell proliferation was not significantly correlated with PD-1 expression itself although PD-1 expression was correlated to viral loads ., Thus there might be additional factors that have participated in the observed gain in proliferation ., One good candidate for this is IL-2 that is produced upon T cell stimulation but may have been limiting with increasing PD-1 expression and negative signalling on CD4 T cells , the main IL-2 producers ., Consistent with this are previous observations that, ( i ) the PD-1/PD-L1 pathway negatively regulates Treg cell proliferation by inhibiting the IL-2 signalling cascade 34 and, ( ii ) exogenous IL-2 can overcome PD-1/PD-L1-mediated inhibition of proliferation 53 ., The PD-L1 blockade of PBMC from viremic HIV-infected individuals under stimulating conditions commonly led to increased reactivation of HIV ex vivo ., The observed correlations between the fold change in virus p24 with the fold change in proliferating CD4 T cells , the percentage of Treg cells as well as the negative correlations with the fold changes of CD4 and CD8 to Treg cell ratios respectively , may indicate that an increase of Treg cells in relation to CD8 T cells promotes virus expansion ., These observations are compatible with the characteristics of HIV biology and stress the importance of applying antiviral treatment in addition to the PD-L1 blockade therapy ., In summary , this ex-vivo study of Treg cell behaviour from different HIV-infected patient groups demonstrates, ( i ) an up-regulation of PD-1 and PD-L1 that correlates with markers of disease progression and, ( ii ) a differential and plasma viremia-dependent gain of Treg cell proliferation and overall suppressive function upon PD-L1 blockade ., This has direct consequences for patient selection to enter clinical trials targeting the PD-1/PD-L1 signalling pathway and treatment modalities ., While the translation of our results to HIV infection in vivo is complex , and many aspects about the weight of the possible PD-1/PD-L1 roles for HIV expansion at different anatomical sites are not completely defined , the upcoming clinical trials will definitely increase our knowledge on Treg cell biology and provide a clear picture on the pros and cons of immune checkpoint modification in HIV infection ., Blood was obtained from healthy , HIV-uninfected volunteers and HIV-infected individuals at the Hospital Clinic and the Hospital del Mar , both in Barcelona , Spain ., HIV-infected individuals were categorized into 4 groups: ( 1 ) fewer than 500 CD4/μL and more than 2000 RNA copies/mL blood; ( 2 ) more than 500 CD4/μL and more than 2000 RNA copies/mL blood; ( 3 ) more than 500 CD4/μL and fewer than 2000 RNA copies/mL blood; ( 4 ) HIV-infected patients under successful antiretroviral treatment ( cART ) for at least 2 years with more than 500 CD4/μL blood and viral loads below the limit of detection ( 40 RNA copies/mL blood ) ., HIV-infected individuals from the cross-sectional study with the exception of the cART group were naïve to antiretroviral therapy at the time of testing and were not in the primary infection phase ( S1 and S2 Tables ) ., In addition , we studied two subgroups of patients longitudinally ., A group of 5 individuals were followed before starting cART , during cART , upon interruption of cART , and after restarting cART ( S3 Table , group A ) ., A second group of 7 individuals were followed before cART and after 2 years receiving cART ( S3 Table , group B ) ., Peripheral blood mononuclear cells ( PBMC ) were isolated by Ficoll density centrifugation ( Invitrogen ) and frozen for subsequent analyses ., Ethical committee approval and written informed consent from all subjects , in accordance with the Declaration of Helsinki , were obtained prior to study initiation ., The study was approved by the institutions’ ethical committees: CEIC- Parc de Salut Mar , Barcelona , Spain ( Protocol approval number: 2013/5422/I ) and Comitè étic dinvestigació clínica , Hospital Clinic , Barcelona , Spain ( Protocol approval numbers: 2013/8671 and 2008/4575 amendment version 1 . 0 from 13/03/2013 ) ., PBMC were stained with the Live/Dead fixable violet dye ( Invitrogen ) and the following fluorochrome-labelled monoclonal antibodies: CD3-BV605 ( clone SK7 ) , CD3-PerCPCy5 . 5 ( clone SK7 ) , CD4-PECy7 ( clone SK4 ) , CD4-APCCy7 ( clone SK3 ) , CD8-PE ( clone RPA-T8 ) , CD8-APCCy7 ( clone SK1 ) , CD25-APCH7 ( clone M-A251 ) , CD127-PE ( clone HIL-7R-M21 ) , CD45RA-FITC ( clone HI100 ) , CD45RA-PerCPCy5 . 5 ( clone HI100 ) , CD45RA-eFluor605 ( eBioscience , clone HI100 ) , PD-1-PerCPCy5 . 5 ( clone EH12 . 1 ) , PD-L1-PE ( clone MIH1 ) , CD39-PECy7 ( eBioscience , clone eBiosA1 ) and CTLA4-PE ( clone BNI3 ) , FOXP3-Alexa647 ( clone 259D/C7 ) , Helios-PerCPCy5 . 5 ( clone 22F6 ) and Ki67-PE ( clone B56 ) ., For intracellular detection of FOXP3 , Helios and Ki67 , cells were fixed and permeabilized using the FOXP3 staining kit ( eBioscience ) according to manufacturer’s instructions ., All antibodies were from BD Biosciences unless otherwise stated ., Flow cytometry data were collected on a LSR Fortessa ( BD biosciences ) and analysed with Flow Jo software ( Tree Star ) ., Panels containing the corresponding isotype controls were collected to set PD-1 , PD-L1 , CTLA4 and Ki-67 gates ., Treg cells were identified as a joint population of effector Treg cells ( CD4+CD45RA-FOXP3hi ) ( eTreg ) and resting Treg cells ( CD4+CD45RA+FOXP3lo ) ( rTreg ) , in which the cut offs for FOXP3 were set manually in relation to CD45RA expression as previously described 40 ( Figs 1 and S1A ) ., To verify that the CD4+CD45RA+FOXP3lo & CD4+CD45RA-FOXP3hi cell populations define Treg cells after a 6-day culture , both rTreg cells ( CD4+ CD127lo CD25+ CD45RA+ ) and conventional CD4 T cells ( CD4+ CD127+ CD25- ) were isolated from HIV-infected individuals by flow cytometry , labelled with CFSE and cultured in the presence of autologous , unlabelled PBMC and Gag peptides as described below ., For this , PBMC samples were enriched for CD4 T cells using magnetic beads ( Miltenyi Biotec ) and sorted by an ARIA SORP ( BD biosciences ) ( S1B Fig ) ., HIV-1Bal was obtained from the Centre for AIDS Reagents NIBSC ( repository reference: ARP118 ) and propagated in PHA-stimulated PBMC in RPMI-1640 media ( Gibco ) supplemented with 20% FBS ( Sigma ) , 1% penicillin/ streptomycin ( Gibco ) and 10U/mL rhIL-2 ( R&D Systems ) for 7 days ., Supernatants were collected , titrated on TZM-bl cells and frozen at -80°C until use ., Supernatants from non-infected PBMC cultured under identical conditions were collected as mock control ., PBMC from healthy controls were cultured in the presence of HIV-1Bal ( or mock ) at a multiplicity of infection ( MOI ) of 0 . 3 and 0 . 03 ., To discard that PD-L1 upregulation requires HIV infection , PBMC were infected at MOI 0 . 3 in the presence or absence of 5 μM T20 HIV entry inhibitor ., In parallel , PBMC were also cultured with HIV-1 Bal gp120 ( NIH reagent program catalogue number 4961 ) at 0 . 01ng/mL and 1ng/mL ., After 4 hours , HIV-exposed cells were washed twice while gp120-exposed cells were left as such ., 0 . 5·106 PBMC/well were cultured in 48-well plates in RPMI-1640 media ( Gibco ) supplemented with 10% FBS ( Sigma ) , 1% penicillin/ streptomycin ( Gibco ) ., After 3 days , cells were harvested and stained to analyse PD-1 and PD-L1 expression on Treg cells by flow cytometry ., The efficacy of HIV-1Bal inhibition by T20 treatment was controlled by stimulating the virus-exposed cells from above with 5μg/ml PHA at day 3 and culturing them for further 7 days ., The presence or absence of virus production was determined by a p24 HIV core antigen ELISA kit ( Innogenetics ) ., T20 treatment completely blocked HIV-1Bal infection under these conditions ., 2·106 PBMC/well were cultured in 24-well plates in RPMI-1640 media ( Gibco ) supplemented with 10% FBS ( Sigma ) , 1% penicillin/ streptomycin ( Gibco ) and 1μg/mL anti-CD28 and anti-CD49d antibodies ., Cells were either left unstimulated or incubated with 1μg/mL Gag pool of overlapping peptides ( Gag peptides; NIH reagent program catalogue number 8117 and 8118 , in part kindly provided by Anja Germann and Hagen von Briesen , Fraunhofer IBMT , Germany ) plus 5μg/mL anti-PD-L1 blocking or isotype control antibodies ( eBioscience ) ., For proliferation assays , PBMC were stained with Carboxyfluorescein succinimidyl ester ( CFSE ) ( Invitrogen ) as described in ( Quah et al . , Nature Protocols , 2007 ) ., After 6-day-culture cells were harvested and stained to analyse proliferation of Treg , CD4- and CD8- T cell subsets ., Alternatively , to analyse proliferation after a 6-day culture in longitudinal samples for which cell numbers were limited , non-CFSE stained PBMC were cultured as previously described , and stained with Ki67 or isotype control antibodies ., Fold change in proliferation ( FC proliferation ) was calculated as a ratio of proliferation under PD-L1 blockade condition divided by proliferation under control condition ., PBMC were stimulated with Gag peptides in the presence of anti-PD-L1 blocking antibody or isotype control antibody as described above ., After 6-day culture , Treg cells where isolated by magnetic beads using the CD4+CD25+CD127dim/- regulatory T cell isolation kit II ( Mitenyi Biotec; Treg cell purity shown in S5 Fig ) ., Purified Treg cells were co-cultured with 50 . 000 CFSE-labelled PBMC at different ratios and stimulated with 0 . 5μg/mL anti-CD3 and 1μg/mL anti-CD28 antibodies ., Cells were cultured in 96-U bottom well plates ( Greiner bio-one ) in RPMI-1640 media ( Gibco ) supplemented with 10% FBS ( Sigma ) , 1% penicillin/ streptomycin ( Gibco ) , 50U/mL rhIL-2 ( R&D Systems ) and 1mM sodium pyruvate ( Sigma ) ., After a 3-day-culture , cells were harvested and stained to analyse proliferation of CD8 T cells ., The Treg cell suppressive capacity was determined by the percentage of inhibition of CD8 proliferation , calculated as: ( CD8 proliferation − CD8 proliferation in presence of Treg cells ) / CD8 proliferation x 100 ., Culture supernatants after a 4-day culture in the presence of Gag peptides and anti-PD-L1 blocking or isotype control antibodies ( as described in previous section “Cell culture and proliferation assay” ) were collected for p24 HIV core antigen quantification by an ELISA kit ( Innogenetics ) ., Comparisons between two groups were performed using Mann-Whitney U test , between more than two groups using Kruskal-Wallis test and within the same patient using Wilcoxon matched pairs test ., Bonferroni correction was applied to adjust significance for multiple comparisons ., Correlation coefficients ( r ) were calculated using the Spearman rank correlation test ., Categorical variables between study groups were compared using Chi-squared and Fisher’s exact test ., Statistical analyses were performed using GraphPad Prism 5 . 0 ( San Diego , CA , USA ) and SPSS 15 . 0 statistical software ( Chicago , IL , USA ) ., p-values ( P ) below 0 . 05 were considered significant and were indicated by asterisks: * p<0 . 05; ** p<0 . 01; *** p<0 . 001 ., Non-significant differences were indicated as “ns” ., Accession numbers in Uniprot database for proteins mentioned in the text are: PD-1 ( Q15116 ) , PD-L1 ( Q9NZQ7 ) , CD3 ( P07766 ) , CD4 ( P01730 ) , CD8 ( P01732 ) , CD25 ( P01589 ) , CD45RA ( P08575 ) , FOXP3 ( Q9BZS1 ) , CD127 ( P16871 ) , CD39 ( P49961 ) , CTLA4 ( P16410 ) , Helios ( Q9UKS7 ) , Ki67 ( P46013 ) , Gag ( Q73367 ) , IL-2 ( P60568 ) , CD28 ( P10747 ) and CD49d ( P13612 ) .

Introduction, Results, Discussion, Materials and Methods

Blocking the PD-1/PD-L1 pathway has emerged as a potential therapy to restore impaired immune responses in human immunodeficiency virus ( HIV ) -infected individuals ., Most reports have studied the impact of the PD-L1 blockade on effector cells and neglected possible effects on regulatory T cells ( Treg cells ) , which play an essential role in balancing immunopathology and antiviral effector responses ., The aim of this study was to define the consequences of ex vivo PD-L1 blockade on Treg cells from HIV-infected individuals ., We observed that HIV infection led to an increase in PD-1+ and PD-L1+ Treg cells ., This upregulation correlated with disease progression and decreased under antiretroviral treatment ., Treg cells from viremic individuals had a particularly high PD-1 expression and impaired proliferative capacity in comparison with Treg cells from individuals under antiretroviral treatment ., PD-L1 blockade restored the proliferative capacity of Treg cells from viremic individuals but had no effect on its suppressive capacity ., Moreover , it increased the viral production in cell cultures from viremic individuals ., This increase in viral production correlated with an increase in Treg cell percentage and a reduction in the CD4/Treg and CD8/Treg cell ratios ., In contrast to the effect of the PD-L1 blockade on Treg cells from viremic individuals , we did not observe a significant effect on the proliferative capacity of Treg cells from individuals in whom viremia was controlled ( either spontaneously or by antiretroviral treatment ) ., However , PD-L1 blockade resulted in an increased proliferative capacity of HIV-specific-CD8 T cells in all subjects ., Taken together , our findings suggest that manipulating PD-L1 in vivo can be expected to influence the net gain of effector function depending on the subject’s plasma viremia .

HIV infection causes a progressive impairment of effector immune responses , contributing to virus persistence ., The restoration of these responses is essential to achieve a drug-free control over HIV ., One strategy that could restore effector immune responses is the relief of the inhibitory signal displayed by the PD-1/PD-L1 pathway on effector cells ., However , the PD-1/PD-L1 pathway also plays a role in the biology of regulatory T cells , which in turn suppress effector responses ., Here we show that ex vivo PD-L1 blockade on peripheral blood mononuclear cells from HIV-infected individuals differentially increases the proliferative capacity of regulatory- and effector- T cells depending on the subject’s plasma viremia ., Our results suggest that PD-L1 blockade will skew the effector-to-regulatory T cell ratio in favour of effector cells only in patients in whom viremia is controlled ., In patients with uncontrolled viremia , PD-L1 blockade will not favour effector- T cells over regulatory- T cells , and might also boost virus reactivation ., Our findings support the rationale to combine a PD-L1 blockade with antiretroviral treatment to restore effector responses in HIV-infected individuals .

journal.pntd.0001442

Novel Structural Components of the Ventral Disc and Lateral Crest in Giardia intestinalis

Giardia intestinalis is a widespread zoonotic parasitic protist ., Infection with this parasite results in giardiasis , a common protozoan intestinal disease ., Both chronic and acute giardiasis contribute to high morbidity rates in developed 1 and developing countries 2 ., Due to a continuing lack of concerted research efforts into the basic biology and mechanisms of pathogenesis of Giardia , giardiasis has been designated a World Health Organization ( WHO ) neglected disease 2 ., The growing need for identification of alternative anti-giardial compounds is underscored by recent evidence of resistance to the widely used anti-giardial drug metronidazole 3 , 4 , 5 ., Giardia has a two-stage life cycle characterized by an infectious “cyst” form that persists in the environment 6 , 7 and a flagellated “trophozoite” form that colonizes the small intestine , causing the characteristic symptoms of giardiasis ., Attachment is essential for pathogenesis 8 ., Giardiasis remains a serious concern worldwide in areas that lack proper sanitation because of contamination of potable water by giardial cysts 9 ., When ingested , giardial cysts begin to “excyst” in the stomachs of their mammalian hosts ., In the small intestine , motile trophozoites attach non-invasively and colonize the intestinal epithelium using a specialized cytoskeletal organelle termed the ventral disc 10 ., Unattached trophozoites enter the large intestine , “encyst” , and are eventually passed on into the environment ., To proliferate and colonize the small intestine of the host , trophozoites find suitable sites for attachment using flagellar motility 11 , and must then remain attached to avoid peristalsis ., Giardial attachment via the ventral disc , either to biological surfaces or to inert laboratory surfaces such as culture tubes or slides , is a rapid , stepwise process that occurs in seconds 12 ., We have recently shown that flagellar motility is not directly required to maintain attachment 12 , invalidating the most widely held model of giardial attachment , the “hydrodynamic model” 13 ., Alternative mechanisms for giardial attachment could include overall conformational changes in the ventral disc that could be directly or indirectly responsible for attachment to surfaces ., These disc conformational changes could be sufficient to generate suction for in vitro attachment or could result in “grasping” of the intestinal epithelium in vivo ., Alternatively , the rigid structure of the ventral disc may indirectly contribute to attachment by maintaining a negative pressure differential underneath the disc that is created by some other unknown mechanism 10 , 14 , 15 ., Conflicting biophysical data 13–20 , and a lack of knowledge of molecular components comprising the ventral disc 7 have made it challenging to evaluate any proposed attachment mechanism at the molecular level ., The ventral disc is a highly ordered and complex spiral microtubule array ( ∼150–400 nm thick ) with elaborated structures that protrude dorsally into the cell body 10 , 21–25 ., The “bare area” region , lacking MTs , is located in the center of the array , ventral to the flagellar basal bodies 7; this structure contains numerous membrane-bound vacuoles 10 ., The ventral disc is comprised of three primary structural elements:, 1 ) a right-handed spiral sheet of uniformly spaced MTs ( ∼250–300 nm apart ) ;, 2 ) trilaminar “microribbons” extending dorsally along the entire length of the MT spiral 24 , 25; and, 3 ) regularly spaced “crossbridge” structures linking adjacent microribbons 24 ., The ventral disc MT spiral is physically linked to the ventral plasma membrane by small MT-associated structures termed “sidearms” 24 ., The composition and function of the trilaminar microribbons , microribbon-connecting crossbridges , and MT-associated sidearm structures are unknown ., The periphery of the ventral disc is surrounded by another highly ordered structure of unknown composition , the lateral crest 26 , which has purported , yet unconfirmed , contractile function 21 ., We have recently shown using Total Internal Reflection Fluorescence Microscopy ( TIRFM ) that the lateral crest region contacts the attachment surface , forming a seal during attachment 12 ., Finally , a partial left-handed MT spiral array , the supernumerary MT array , lies either dorsal or ventral to the main ventral disc structure and may also possess partially-formed microribbons 24 ., The function of the supernumerary MTs in attachment or disc biogenesis is unknown ., In summary , the ventral disc MT spiral with associated microribbons and sidearms , the lateral crest , and the supernumerary MTs all comprise the complex structure of the ventral disc required for giardial attachment 27 ., Disc-associated proteins were initially termed “giardins” ., Three separate gene families of giardins are now known to localize to the ventral disc: three annexins , or alpha-giardins 28–31; three striated fiber ( SF ) –assemblins , including beta-giardin , delta-giardin , and SALP-1 32; and one novel protein , gamma-giardin 33 ., Several disc-associated proteins have cell cycle-specific localization , including an ERK1 kinase homolog that localizes to the disc during encystation 34 ., Recently , aurora kinase was shown to localize to the ventral disc during cell division 35 , yet the localization of aurora kinase to specific structural elements within the disc , and its role , if any , in interphase remains unknown ., Two other putatively cell cycle-specific disc-associated Nek kinases were recently identified in a screen for basal body-associated proteins 36 ., Thus , while fifteen proteins are now known to localize to the ventral disc at some point in the cell cycle 28–36 , the composition of each of the primary ventral disc structures ( e . g . , microribbons , crossbridges , lateral crest ) remains to be determined ., Here we used a “shotgun” proteomic strategy 37 with a detergent-extracted , isolated ventral disc preparation to discover novel ventral disc and lateral crest proteins ., Candidate disc-associated proteins ( or “DAPs” ) were identified through peptide sequence analysis and comparisons to the completed Giardia genome 38 ., Candidate DAPs were then verified by construction of C-terminal DAP::GFP fusions using a high throughput cloning strategy we recently modified for use in Giardia 39 ., Transformation of Giardia with the GFP fusion constructs allowed us to assess the localization of over 50 putative DAPs , and to confirm previously identified 32 , 33 , 40 and novel DAPs using GFP-tagging and live imaging of GFP fusion proteins in trophozoites ., Live imaging of GFP-tagged DAPs also enabled us to distinguish between stable and dynamic pools of representative DAPs using Fluorescence Recovery After Photobleaching ( FRAP ) 39 ., Ultimately , functional analyses of these novel structural DAPs and of any as-yet-unidentified , but potentially dynamic or regulatory , DAPs will be central toward understanding the mechanism of ventral disc-mediated attachment and testing attachment hypotheses ., Giardia intestinalis strain WBC6 ( ATCC 50803 ) trophozoites were maintained at 37°C in modified TYI-S-33 medium with bovine bile 41 in 16 ml screw cap tubes ( Fisher Scientific ) ., The primary goal in the isolation of intact ventral discs for proteomic analysis was the maintenance of microtubule-associated proteins , by removing radicals and metal ions that could damage disc structure , and by stabilizing microtubules using drugs like Taxol ., We modified a cytoskeletal preparation from Holberton et al . 42 to isolate disc and flagellar cytoskeletons from Giardia ., First , TYI-S-33 medium was decanted from one confluent 12 ml culture of trophozoites ., Cells were demembranated and cytoskeletons were extracted by adding 1 ml of 1% Triton X-100 in 1X PHEM plus Taxol ( 60 mM PIPES , 25 mM HEPES , 10 mM EGTA , 1 mM MgCl2 , pH 7 . 4 , 1 mM DTT , 10 µM paclitaxel ( Sigma ) ) and vortexing continuously at the highest setting for 3 minutes ., To prevent proteolysis , protease inhibitors ( Roche ) were added to the preparation ., Ventral disc cytoskeletons were then pelleted by centrifugation at 16 , 000×g , and the pellets were washed four times in 1X PHEM+Taxol lacking 1% Triton X-100 ., Sufficient extraction of cytoskeletons was confirmed by wet mount using phase contrast or DIC microscopy ( see Figure 1 ) ., We identified the proteins present in the ventral disc preparation using liquid chromatography tandem mass spectrometry ( LC-MS/MS LTQ ) 37 ., All MS/MS samples were analyzed using X !, Tandem ( www . thegpm . org; version TORNADO ( 2008 . 02 . 01 . 2 ) ) ., X !, Tandem was set up to search protein sequences downloaded from Genbank ( Giardia intestinalis ) assuming the digestion enzyme trypsin ., X !, Tandem was searched with a fragment ion mass tolerance of 0 . 40 Da and a parent ion tolerance of 1 . 8 Da ., Iodoacetamide derivative of cysteine was specified in X !, Tandem as a fixed modification ., Deamidation of asparagine , oxidation of methionine , sulphone of methionine , tryptophan oxidation to formylkynurenin of tryptophan and acetylation of the N-terminus were specified in X !, Tandem as variable modifications ., Scaffold ( version Scaffold_2_03_01 , Proteome Software Inc . , Portland , OR ) was used to validate MS/MS based peptide and protein identifications ., Peptide identifications were accepted if they could be established at greater than 80 . 0% probability as specified by the Peptide Prophet algorithm 43 ., Protein identifications were accepted if they could be established at greater than 95 . 0% probability and contained at least one identified peptide ., Protein probabilities were assigned by the Protein Prophet algorithm 44 ., Proteins that contained similar peptides and could not be differentiated based on MS/MS analysis alone were grouped to satisfy the principles of parsimony ., Fifty-eight of the 102 candidate DAPs identified in the proteomic survey were chosen for localization; candidates that appeared to be metabolic , flagellar-associated or chromatin-associated proteins were excluded ., All candidate DAP PCR forward primers ( see Table S1 ) were designed to bind approximately 200–250 bp upstream of the gene to include the Giardia native promoter and contained the sequence CACC at the 5′ end to facilitate directional cloning ., Blunt-ended PCR amplicons were generated by PCR using PfuTurbo Hotstart PCR Mastermix ( Stratagene ) with Giardia intestinalis strain WBC6 genomic DNA ., The candidate DAP PCR amplicons were subsequently subcloned into the Invitrogen pENTR/D-TOPO backbone to generate Gateway entry clones ., Inserts in entry clones were sequenced to confirm the identity and correct orientation of the gene ., To construct DAP::GFP fusions , positive entry clones were then recombined , via LR reaction , with a 1-fragment GFP tagging E . coli/Giardia shuttle destination vector ( pcGFP1F . pac , 39 ) using LR Clonase II Plus ( Invitrogen ) ., LR reactions were performed using 100 ng pcGFP1F . pac and 150 ng of DAP entry clone plasmid DNA ., Positive clones were screened by digestion with AscI , and bulk plasmid DNA was prepared using Qiagens Endofree Plasmid Maxi Kit ., To create C-terminal GFP-tagged candidate DAP strains , Giardia intestinalis strain WBC6 was electroporated with roughly 20 µg of plasmid DNA ( above ) using the GenePulserXL ( BioRad ) under previously described conditions 45 ., Episomal DAP::GFP constructs were maintained in transformants using antibiotic selection ( 50 µg/ml puromycin ) 46 ., Immunostaining of the GFP-tagged DAP strains was performed as previously described 45 ., To confirm disc localization , Metamorph image acquisition software ( MDS Technologies ) was used to collect 3D images using a Leica DMI 6000 wide-field inverted fluorescence microscope with a PlanApo 100X , NA 1 . 40 oil immersion objective ., Serial sections of DAP::GFP strains were acquired at 0 . 2 µm intervals , and deconvolved using Huygens Professional deconvolution software ( SVI ) ., Two dimensional maximum intensity projections were created from the 3D data sets for presentation purposes ., We used laser fluorescence photobleaching of specific regions to measure the movement and steady state turnover of the new DAPs in Giardia , a technique that has been used extensively in other organisms 47 ., Three DAP::GFP-expressing strains ( whole disc , DAP5374; lateral crest , DAP13981; and disc plus axonemes , DAP17090 ) were selected as representative examples of different ventral disc localizations ., The media in a confluent 12 ml culture was replaced with 1X HBS for 1 hour at 37°C ., The culture was then iced for 15 minutes to detach cells , and 3 ml of the cell suspension were transferred to a coverslip placed in an 8-well plastic plate ., Cells were incubated for 30 minutes at 37°C under nitrogen gas to allow them to attach to the coverslip ., After 1 hour , 1 µl of CellMask orange ( Invitrogen ) was added to the cell suspension ., Stained cells were incubated for 5 minutes at 37°C then rinsed twice with warmed 1X HBS ., The edges of the coverslip were blotted and the coverslip was inverted onto a slide with double stick tape ., Warmed 2% low-melt agarose ( Sigma ) in 1X HBS was added under the coverslip to embed the attached cells and the prep was sealed on all sides with VALAP ., An Olympus FV1000 scanning laser confocal microscope equipped with a four channel PMT was used for imaging and simultaneous 405 nm bleaching ., The pre-bleach image of the cell was acquired using a 60× , 1 . 42 NA objective and a 488 nm laser ( at 5% with 4 µs/pixel scan speed ) ., To photobleach a specific region of DAP localization , we used the 405 nm laser ( 90% power for 200 milliseconds ) ., Fluorescence recovery in the GFP-tagged DAP strains was assessed by imaging once every minute , for up to 10 minutes , using the 488 nm low power laser excitation ., Normalized GFP fluorescence recovery was calculated by subtracting the PMT background noise from the ROI intensity measurement; the background-subtracted intensity measurement was then divided by a fluorescent control ROI intensity measurement to normalize for photobleaching due to imaging ., Detergent extracted cytoskeletons containing GFP-tagged DAPs were isolated ( see above ) ., Immunolabeling and negative staining of the isolated discs was performed as previously described 48 in 1 . 5 ml Eppendorf tubes with gentle shaking at room temperature ., Cytoskeletons were placed in a blocking buffer of 3% nonfat dry milk in PHEM buffer ( 60 mM PIPES , 25 mM HEPES , 10 mM EGTA , 2 mM MgCl2 ) for 1 hour ., Cytoskeletons were then labeled with an anti-GFP antibody in blocking buffer for 1 . 5 hours , and then rinsed 3 times , for 15 minutes each , in PHEM ., Pelleting between steps was done at 2 , 000×g for 5 minutes with a short vortexing step for resuspension ., Samples were incubated with 5 nm goat-anti-rabbit F ( ab′ ) 2 IgG antibody ( BB International ) in blocking buffer for 1 hour , then rinsed 3 times in PHEM for 15 minutes ., For negative controls ( not shown ) , we used secondary antibody only ., For negative staining of DAP::GFP fusion strains , 300 mesh copper grids ( EMS ) were Formvar-coated , carbon-coated , then glow-discharged to make them more hydrophilic ., A 5 µl droplet of cytoskeleton solution was placed on the grid , blotted , and then negative stained with a 5 µl droplet of 2% aqueous uranyl acetate ( Ted Pella ) and blotted ., Grids were imaged with an AMT digital camera in a CM100 ( FEI ) transmission electron microscope operating at 80 kV ., The detergent extracted disc preparation for proteomic analysis yielded 102 candidate disc-associated proteins ., The list of candidate disc proteins and their GiardiaDB ( www . giardiadb . org ) accession numbers are shown in Table S2 ., Based on protein functional predictions and motif analysis , we classified these proteins into three categories: putatively disc-associated ( 57 total ) , putatively flagellar ( 17 total ) , or metabolic or chromatin-associated ( 28 total ) ., Flagellar proteins were likely present due to the presence of axonemes in the disc preparation; only a few were confirmed to localize to the flagella by GFP-tagging ( Table S2 ) ., Metabolic or chromatin-associated proteins were deemed contaminants unlikely to be associated with the ventral disc ., We confirmed the presence and localization of known disc-associated proteins identified by our proteomic analysis ( Figure 2 and Table 1 ) , including beta-giardin 22 , delta-giardin 40 , gamma-giardin 33 and SALP-1 32 ., We also identified three annexins ( alpha-2 , alpha-6 , and alpha-17 ) in the disc proteome ., We have previously shown that alpha-2 annexin localizes to the ventral disc and ventral flagella 12 ., We did not identify several annexins 28 , aurora kinase 35 , or several other proteins previously reported to localize to the ventral disc 34 , 36 , 49 , possibly due to slight differences in sample preparation ., We created C-terminal GFP fusions to 58 giardial proteins using a high-throughput Gateway-based cloning strategy for Giardia 39 followed by transformation into trophozoites ( see Materials and Methods ) ., Ventral disc or lateral crest localization was confirmed for 18 novel DAPs of the 57 candidates using 3D deconvolution microcopy ( Table 1 , and Figure 3 , Figure 4 and Figure 5 ) ., Candidate proteins that were found to localize to the ventral disc were not necessarily those for which the greatest number of mass spectra were obtained ( Table S2 ) ; of the 18 new DAPs , a relatively large number of spectra was obtained only for DAP16343 ., Non-disc localizations included the basal bodies , the median body , regions of various flagellar pairs , or the cytoplasm ., We also noted a lack of GFP expression in some interphase trophozoites ( Table S2 ) , which could indicate cell cycle-specific expression ., Each confirmed DAP has a conserved homolog in the other two sequenced Giardia genomes 50 , 51 ., Two novel disc-associated proteins ( DAP16424 and DAP16343 ) had no conserved motifs or known homologs in other organisms ., Many other DAPs had conserved motifs including: ankyrin repeats ( 13 DAPs ) , Nek kinase domains ( 3 DAPs ) , and SAM domains ( DAP17090 ) ( see Table 1 ) ; some had more than one of these motifs ., The disc-associated kinases ( DAP24321 , DAP17231 , and DAP13981 ) may be non-functional pseudokinases 52 , as they lack highly conserved catalytic residues ( Figure S1 ) ., Two of the newly identified DAPs are putatively microtubule-associated , including DAP5374 , a CAP-Gly motif containing protein 53 , and DAP16263 , a DIP13 homolog 54 with a conserved MT-binding motif ( see Figure S2 ) ., Finally , we observed that the “median body protein” ( DAP16343 ) 55 has an obvious disc localization with occasional localization to the median body ( Figure 3 ) ., We categorized the eighteen novel disc associated proteins into three general types of localization: to the entire disc spiral ( Figure 3 ) , to the disc edge or lateral crest ( Figure 4 ) , or to the supernumerary MTs ( Figure 5 ) ., In some cases , we observed additional localization to the basal bodies , to regions of the axonemes ( Figure 5 ) , or to the median body ( Figures 2 and 3 ) ., Eight ventral disc-associated proteins have previously been localized using specific antibodies ., We show that GFP-tagged DAP localization concurs with the localizations previously described using immunostaining ( Table 1 and Figure 2 ) with the exception of alpha17-annexin ( Table S2 ) ., We confirmed localization to the entire ventral disc spiral for the SF-assemblin homologs ( e . g . , beta-giardin , delta-giardin , and SALP-1 ) and gamma-giardin ( Figure 2 ) ., Notably , we also find some localization of these DAPs to the median body , primarily during prophase ., Using anti-GFP immunogold labeling with negative staining 25 , we demonstrate that the SF-assemblin homologs and gamma giardin localize to the microribbons 22 that are bound to the spiral microtubule array and project into the cytoplasm ( Figure S3 ) ., Six novel disc-associated proteins localized to the entire disc ( Figure 3 ) , including three ankyrin repeat proteins ( DAP13766 , DAP103807 and DAP17053 ) , the median body protein ( MBP; DAP16343 ) , one CAP-Gly protein ( DAP5374 ) and one Nek kinase ( DAP24321 ) ., The ankyrin repeat protein DAP17053 has a unique localization in that it is present in throughout the ventral disc spiral , but is completely absent in the ventrolateral flange region ( see Figure 3 ) ., The Nek kinase DAP24321 is present throughout the cell and ventral disc but localizes most strongly to the posterior regions of the disc ., DAP16343 ( MBP ) localizes throughout the disc spiral and the lateral crest , and DAP5374 localizes throughout the disc spiral and the lateral crest ., The ventral disc spiral is surrounded by a putatively contractile repetitive structure termed the “lateral crest” ( Figure 1 and 26 ) ., Ten of the novel DAPs localize primarily to the disc perimeter , presumptively in the region of the lateral crest or along the outside edge of the ventral disc spiral ( Figure 4 ) ., Two ( DAP13981 and DAP17231 ) are Nek kinases with ankyrin repeat domains , another ( DAP16424 ) is novel with no homology to known proteins or domains , and the remaining seven are ankyrin repeat proteins ( Table 1 and Figure 4 ) ., The Nek kinase DAP13981 was specifically localized to the lateral crest using negative staining with anti-GFP immunogold labeling ( Figure 4 ) ., One ankyrin repeat protein ( DAP103810 ) is notable in that it also localizes to the inner perimeter of the disc , near the “bare area” while all others localize only to the outer disc edge ., DAP16424 has a regularly spaced , punctate localization around the disc edge , and is also present at the basal bodies and cytoplasmic regions of the anterior axonemes ., In addition to localizing to the disc perimeter , DAP17097 also localizes to the median body in many cells ., DAP17096 , DAP17097 , and DAP16424 also localize to the basal bodies during interphase ., We identified two novel DAPs that localize specifically to the supernumerary MTs ( Figure 1 ) and to axonemes , yet not to the entire ventral disc structure ( see schematic in Figure 5 ) ., DAP17090 , a novel protein containing a SAM motif , localizes to the supernumerary MTs , the ventral flagella axonemes , the cytoplasmic regions of the caudal and anterior axonemes , and to the basal bodies ( Figure 5 and Video S1 ) ., DAP16263 , a DIP13 homolog , has a similar localization and also localizes faintly to the lateral crest ( Figure 5 and Video S2 ) ., To assess whether DAPs localizing to the ventral disc are stable or dynamic , we used FRAP to examine protein turnover in representative GFP-tagged DAP strains ( for the lateral crest , DAP13981; for the entire disc , DAP4410 ( SALP-1 ) ; and for the supernumerary MTs , DAP17090 ) ., For each of these DAP::GFP fusions , we observed no recovery of GFP fluorescence to the ventral disc in interphase trophozoites for more than 10 minutes post photobleaching ( Figure 6 ) ., In the DAP17090::GFP strain with localization to the basal bodies and/or axonemes ( 26% , n\u200a=\u200a100 ) as well as to the disc , GFP fluorescence did recover , but only to the non-disc structures ., We noted a partial recovery ( 47% ) of fluorescence at the axonemes of the DAP17090::GFP strain within 8 minutes ., Similarly , the axonemes of the lateral crest strain DAP13981::GFP recovered to 80% within 7 minutes ( Figure 6 ) ., However , DAP13981 localization to the axonemes was visible only in DAP13981::GFP trophozoites undergoing cytokinesis ., We did not observe ventral disc recovery for DAP4410 , which only localizes to the ventral disc and not to other cytoskeletal elements ( Figure 2 ) ., We confirmed that several previously identified DAPs are associated with the ventral disc microribbons ( Figure S3 ) ., The microribbons extend from the spiral MT array into the cytoplasm ( Figure 2 ) and consist of two sheets of globular subunits , separated by a fibrous inner core , forming a structure about 25 nm thick 25 ., Beta-giardin and the other previously identified SF-assemblin homologs , including delta-giardin and SALP-1 32 likely form the structural basis of the microribbons upon which other microribbon-associated proteins assemble 26 ., We confirmed the microribbon localization of beta-giardin using GFP-tagging and negative staining ( Figure S3 ) , and have also confirmed the microribbon association of delta-giardin , SALP-1 , and gamma-giardin , a protein that lacks conserved domains and is unique to Giardia ., Like beta-giardin 39 , the microribbon-associated protein SALP-1 ( DAP4410 ) also does not turn over following photobleaching ., Thus we believe that the ventral disc is a relatively stable structure ., We hypothesize that microribbon-associated proteins likely assemble into the ventral disc prior cell division , and that MTs of the disc do not undergo rapid polymerization dynamics as was previously proposed 58 ., Some DAPs associate with the entire ventral disc structure ( Figure 3 ) , yet other DAPs localize to specific structural components in other regions of the ventral disc , including the lateral crest , the ventrolateral flange , and the supernumerary MTs ( Figure 4–5 ) ., A putative role for some of these disc proteins can be inferred from the conserved motifs they contain ., Two of the novel disc-associated proteins , DAP5374 and DAP16263 , have microtubule binding motifs ., In general , microtubule-associated proteins mediate dynamic processes of microtubules ., They include proteins that promote polymerization or depolymerization dynamics , microtubule end- or side-binding proteins , enzymes that modify tubulin , and microtubule motors such as kinesins and dyneins that generate cellular forces ., Many of these have been identified in the Giardia genome 38 ., DAP5374 is a CAP-Gly protein 53 that also contains an N-terminal ubiquitination ( UBQ ) motif , indicating that it might target the parental ventral disc for cell cycle-specific degradation and disassembly via a proteasome-dependent pathway ., CAP-Gly proteins interact with tubulin monomers , dimers , and/or MTs , regulate microtubule dynamics and organization , and are involved in intracellular signaling and the distribution of cellular organelles ., DAP16263 is a homolog of DIP13 , a 13 kDa Chlamydomonas protein that defines a new , and likely ancient , MT-associated protein family conserved in diverse protists , plants , and animals that have flagellated cell stages 54 , 59 ., In Chlamydomonas , DIP13 localizes to the centrioles and to cytoplasmic and flagellar MTs , and is purported to either stabilize or connect MTs to other cellular structures 54 ., DIP13 homologs contain a conserved MT binding motif – “KREE” – that directly binds MTs 54 ., Because the giardial DIP13 homolog ( DAP16263 ) lacks this motif ( see Figure S2 ) , it is unclear whether DAP16263 can directly bind MTs ., Additionally , antisense RNA knockdown of DIP13 in Chlamydomonas results in severe cytological and cell division defects , including improper flagellar number and orientation ., We observed localization of DAP16263 to the caudal and the ventral flagella , as well as to the basal bodies and the supernumerary MT spiral ., As many components of the ventral disc are MT-associated proteins ( CAP-Gly or DIP13 domains ) or are related to flagellar root structures ( SF-assemblins 60 ) , the ventral disc may be derived from ancestral flagellar structures ., Several DAPs are NIMA-Related Kinases ( Neks ) ., These kinases are ancient members of the large serine/threonine kinase family with putative roles in the cell cycle and in ciliary function 61 ., Over seventy Nek kinases are present in the Giardia genome 62 ., We identified one that localizes to the ventral disc ( DAP24321 ) and two that localize to the lateral crest ( DAP13981 and DAP17231 ) ., These disc-associated Nek kinases appear to be pseudokinases as they lack conserved catalytic residues ( 52 and Figure S1 ) ., Two other giardial Nek kinases were also shown to localize to the ventral disc in a recent survey of giardial basal body-associated proteins 36 ., Nek pseudokinases could simply contribute to disc and lateral crest structure; however , the lack of functional catalytic sites in pseudokinases does not always result in a lack of kinase activity 63 , and some pseudokinases have been assigned roles in the regulation of other kinases 52 ., Thus , disc-associated Neks lacking functional catalytic sites might still perform regulatory functions required for attachment dynamics , dorsal daughter disc biogenesis or parental disc disassembly during cell division 64 ., Many disc-associated proteins ( Table 1 ) possess conserved ankyrin repeat domains , roughly 33 amino acid protein-protein interaction motifs often present in tandem arrays in diverse proteins in diverse eukaryotes 65 ., Ankyrin repeat domain-containing proteins are very abundant ( up to 3 . 6% of ORFs ) in the Giardia genome 38 , 66 ., Ankyrin repeat domain-containing DAPs could interact with the microtubules or tubulin 67 , could be associated with the microribbons , crossbridges or sidearm structures , or possibly , may connect the ventral disc structure to other cytoskeletal or membrane proteins ., Finally , one intriguing protein of the disc proteome that lacks homology to known proteins is the 101 kDa “median body protein” , or MBP ( DAP16343 ) ., MBP is clearly an abundant disc protein that may have localization to the median body 68 at specific points in the cell cycle , especially prior to mitosis ., Beta-giardin , gamma giardin , and several other newly identified DAPs ( DAP17090 , DAP16263 , DAP10796 , DAP17097 , DAP16424 ) also have occasional localization to the median body , basal bodies or axonemes as well as to the ventral disc , primarily in prophase ., Components of the ventral disc and lateral crest should include both stable , structural elements and dynamic or regulatory elements ., Stable structural components of the ventral disc would be expected to exhibit little protein turnover , whereas dynamic components of the disc , such as regulatory components or MT motors , would be expected to turn over at a faster rate ., The lack of protein turnover observed in the live imaging of representative DAP::GFP strains using FRAP ( Figure 6 ) implies that the DAPs identified here are likely structural components of the disc , rather than transiently associated or regulatory elements ., This may also indicate that the repair of the ventral disc structure is minimal during interphase ., In contrast , dynamic DAPs may be only loosely associated with the ventral disc structure and could have either regulatory or cell cycle-specific functions ., Loosely or transiently associated proteins like these could have been lost in our disc preparation ., This might explain why we did not identify several annexins and protein kinases or phosphatases reported to localize to the ventral disc 28 , 34 , 36 , 49 ., Other regulatory or dynamic disc-associated proteins may have similar transient associations with the ventral disc characterized by rapid turnover , leading to them to elude identification by the methods employed here ., For example , we have recently shown that alpha-2 annexin is only transiently associated with the ventral disc , as it recovers after photobleaching 12 ., Although we did not observe protein turnover of disc-localizing DAPs , we did observe some turnover of DAPs when they also localized to non-disc structures ., For example , when DAP13981 or DAP17090 localized to the basal bodies or axonemes , fluorescence recovered within several minutes ( Figure 6 ) ., This localization was only observed in a small proportion of cells , and thus may be cell cycl

Introduction, Materials and Methods, Results, Discussion

Giardia intestinalis is a ubiquitous parasitic protist that is the causative agent of giardiasis , one of the most common protozoan diarrheal diseases in the world ., Giardia trophozoites attach to the intestinal epithelium using a specialized and elaborate microtubule structure , the ventral disc ., Surrounding the ventral disc is a less characterized putatively contractile structure , the lateral crest , which forms a continuous perimeter seal with the substrate ., A better understanding of ventral disc and lateral crest structure , conformational dynamics , and biogenesis is critical for understanding the mechanism of giardial attachment to the host ., To determine the components comprising the ventral disc and lateral crest , we used shotgun proteomics to identify proteins in a preparation of isolated ventral discs ., Candidate disc-associated proteins , or DAPs , were GFP-tagged using a ligation-independent high-throughput cloning method ., Based on disc localization , we identified eighteen novel DAPs , which more than doubles the number of known disc-associated proteins ., Ten of the novel DAPs are associated with the lateral crest or outer edge of the disc , and are the first confirmed components of this structure ., Using Fluorescence Recovery After Photobleaching ( FRAP ) with representative novel DAP::GFP strains we found that the newly identified DAPs tested did not recover after photobleaching and are therefore structural components of the ventral disc or lateral crest ., Functional analyses of the novel DAPs will be central toward understanding the mechanism of ventral disc-mediated attachment and the mechanism of disc biogenesis during cell division ., Since attachment of Giardia to the intestine via the ventral disc is essential for pathogenesis , it is possible that some proteins comprising the disc could be potential drug targets if their loss or disruption interfered with disc biogenesis or function , preventing attachment .

Giardia is a unicellular intestinal parasite that infects millions of people worldwide each year ., Colonization of the small intestine is a critical stage in Giardia infection ., Giardia colonizes the intestinal wall using a specialized suction cup-like structure , the ventral disc ., Little is known about the protein composition of the disc or about how the disc functions during attachment ., We identified and confirmed eighteen new ventral disc proteins in a preparation of isolated discs using modern genomic methods for analyzing protein composition ., We imaged these disc proteins in Giardia cells by labeling the proteins with fluorescent tags ., A number of these proteins were present on the rim of the ventral disc , a region that appears necessary for the disc to form a seal during attachment to the host ., These new ventral disc proteins form the building blocks of the ventral disc structure ., Future studies of the roles of the ventral disc proteins either in the assembly of the ventral disc during cell division , or in the functioning of the disc during attachment will enable a better understanding of Giardias colonization of the host .

biology, microbiology, parasitology

journal.pgen.1003245

The Role of Autophagy in Genome Stability through Suppression of Abnormal Mitosis under Starvation

Autophagy is a protein degradation pathway that is conserved from yeast to mammals ., Previous studies have reported the functions and molecular mechanisms of autophagy , including the identification of autophagy-related genes 1 , 2 , the characterization of the molecular mechanisms of each autophagy stage 3 , 4 , and the detection of selective autophagy 5 , 6 ., Autophagy is induced in response to nutrient starvation ., In addition , homeostatic autophagy occurs at a low level in mammalian cells under nutrient-rich conditions 7 , 8 , and is regulated in a cell cycle-dependent manner 9 , 10 ., Furthermore , autophagy is induced during developmental stages 11 ., In budding yeast , cell division and cell growth are precisely regulated by an intrinsic mechanism , and are tightly modulated by nutrient conditions ., Target of rapamycin ( TOR ) is a phosphatidylinositol kinase-related Ser/Thr kinase and a critical regulator of cell growth , which senses the nutrient conditions 12 ., TOR forms two distinct complexes , TOR complex 1 ( TORC1 ) and TOR complex 2 ( TORC2 ) 13 ., Yeast TORC1 consists of either of the two yeast TOR homologs , Tor1 or Tor2 , together with co-factors Kog1 , Lst8 , and Tco89 , which are sensitive to inhibition by rapamycin ., Conversely , yeast TORC2 , which is produced from Tor2 , Avo1–3 , and Lst8 , is not sensitive to rapamycin ., While TORC2 regulates spatial aspects of cell growth , such as the organization of the actin cytoskeleton , TORC1 regulates temporal aspects of cell growth , including protein synthesis , gene transcription , ribosome biogenesis , amino acid uptake , and induction of autophagy 13–16 ., Under nutrient-rich conditions , TORC1 is active and inhibits the induction of autophagy through inhibitory phosphorylation of Atg13 17 ., In contrast , TORC1 is inactive during starvation , thereby inducing autophagy 18 ., Inhibition of TORC1 by rapamycin or by nutrient starvation leads to cell cycle arrest in the G1 phase 19 , demonstrating that TORC1 plays a crucial role linking cell growth and cell cycle progression ., In addition to its role in the regulation of cell cycle progression in the G1 phase , we recently found that TORC1 is involved in the G2/M transition in budding yeast 20 ., Reduced TORC1 activity , caused by nutrient starvation or by temperature-sensitive mutation in KOG1 , which is an essential component of TORC1 , leads to cell cycle arrest at G2/M ., Two recent reports have shown that nutrient starvation also blocks the onset of mitosis in mammalian cells and the fission yeast Schizosaccharomyces pombe 21 , 22 ., The regulation is achieved by TORC1-mediated activation of Cdc2/cyclin B in mammalian cells 21 , whereas it is suggested to be mediated by TORC2 as well as Sty1 MAPK signaling in the fission yeast 22 , 23 ., Although the underlying mechanisms may differ among species , the transient suppression of mitotic entry in response to nutrient starvation is conserved throughout evolution ., However , it is well-known that cells arrest in the G1 phase and enter G0 during starvation 24 ., Thus , the mechanism as to how G2/M-delayed cells progress through the cell cycle to return to the G1 phase remains unclear ., In this study , we elucidated the molecular mechanism underlying cell cycle progression under nutrient-limited conditions ., We show that the cell cycle delay at G2/M is rescued in an autophagy-dependent manner ., Regulation of the cell cycle by autophagy during starvation is believed to be involved in genome integrity by coupling cell growth with cell division ., TORC1 is thought to be a nutrient sensor that controls the cell cycle and autophagy directly 12 , 18 ., We first examined fluctuation in TORC1 activity under starvation conditions by an immunoblot of Atg13 , a TORC1 substrate , and RT-qPCR of genes whose expression is a product of TORC1 function ., TORC1 directly phosphorylates Atg13 17 , thus , inactivation of TORC1 causes dephosphorylation of Atg13 ., In addition , TORC1 induces the transcription of RPS26A , RPL9A , and NOG1 but suppresses that of MEP2 and GAP1 16 , 25 , 26 ., When exponentially growing wild-type ( WT ) cells were released into nitrogen-depleted medium , TORC1 activity was immediately decreased; Atg13 became dephosphorylated ( Figure 1A ) , resulting in a concomitant decrease in the expression of RPS26A , RPL9A , and NOG1 and an increase in the expression of MEP2 and GAP1 ( Figure 1B ) ., We found that phosphorylation of Atg13 gradually recovered after 2–18 h in nitrogen-depleted medium ( Figure 1A ) , suggesting that TORC1 activity was restored in cells ., The expressions of RPS26A , RPL9A , and NOG1 were increased , but those of MEP2 and GAP1 remained at high levels 18 h after release into nitrogen-depleted medium ( Figure 1B ) ., The lack of a correlation between the recovery of Atg13 phosphorylation and the levels of MEP2 and GAP1 suggests that TORC1 activity did not completely recover during this process ., Next , we investigated whether autophagy is involved in the recovery of TORC1 , and assessed the fluctuation in TORC1 activity in autophagy-deficient Δatg2 cells ., When Δatg2 cells were placed in nitrogen-depleted medium , TORC1 activity decreased , similar to WT cells up to 4 h ( Figure 1A and 1B ) ., Interestingly , TORC1 activity was not restored in Δatg2 cells 18 h after being released into the starvation medium ( Figure 1A and 1B , left panel ) ., Re-phosphorylation of Atg13 and increased expression of RPS26A and RPL9A under starvation conditions required other autophagy-related genes , such as ATG1 and ATG7 , which are essential for autophagosome formation , and PEP4 encoding vacuolar proteinase A , which is responsible for autophagic degradation of proteins accompanied by recycling of amino acids ( Figure 1C and 1D , left panel ) ., In contrast , recovery of TORC1 activity was not affected by deletion of ATG11 , which is essential in selective autophagy and dispensable for starvation-induced autophagy ( Figure 1C ) ., Therefore , these results suggest that the partial recovery of TORC1 activity is induced in a non-selective and starvation-induced autophagy-dependent manner ., The cell cycle is arrested in the G1 phase under starvation conditions 19 , and we previously reported that TORC1 inactivation caused by rapamycin treatment and nitrogen starvation induces a G2/M delay 20 ., Therefore , we postulated that an unknown mechanism may exist that regulates cell cycle re-progression from G2/M under nutrient starvation ., We carefully examined the cell cycle profiles of nitrogen-starved cells , particularly the relationship between cell cycle progression and TORC1 activity ., First , WT and Δatg2 cells were synchronized in the G1 phase by treatment with α-factor , and released into SCD medium ., When the majority of cells progressed to the S phase , they were released into nitrogen-depleted medium ., During this time course , cells were collected at intervals and the DNA content of cells at each time point was examined by FACS analysis ., As previously reported , WT and Δatg2 cells remained arrested at 2C DNA content for 2–4 h after α-factor release , demonstrating that cell cycle progression was delayed at G2/M ( Figure 2A ) ., In WT cells , the delay in cell cycle progression was overcome after 5 h , and most cells reached the G1 phase after 25 h ( Figure 2A ) ., In contrast , a signification portion of Δatg2 cells , as well as Δatg1 cells , remained arrested at 2C DNA content after 25 h ( Figure 2A and Figure S2 ) ., The difference in the cell cycle profiles between WT and atg2 mutant cells was confirmed by Clb2 levels; G2/M cyclin Clb2 decreased 4–5 h after α-factor release in WT cells , indicating that the cells entered mitosis , whereas Clb2 was consistently present in Δatg2 cells ( Figure 2B ) ., These results show that autophagy contributes to re-progression at G2/M during starvation ., Next , we investigated whether re-activation of TORC1 is correlated with re-progression at G2/M ., We analyzed time-dependent changes in gene expression downstream of TORC1 ., In WT cells , transcription of RPS26A and RPL9A decreased in response to starvation ( 2 h ) , but was gradually restored after 5–25 h ( Figure 2C ) ., Mitotic entry correlated with the re-activation of TORC1 ( Figure 2A and 2C ) ., Conversely , transcription of RPS26A and RPL9A remained low in Δatg2 cells , and the cells remained arrested at G2/M ( Figure 2A and 2C ) ., These results show again that TORC1 is partially restored in an autophagy-dependent manner in nitrogen-starved cells , and suggest that a correlation between the partial recovery of TORC1 activity and cell cycle re-progression at G2/M ., To further address the mechanism of autophagy-dependent cell cycle re-progression during starvation , we next focused on the role of amino acid pools ., Autophagy contributes to the maintenance of amino acid pools in yeast; during the first two hours of nitrogen starvation , the intracellular amino acid level decreases rapidly , and is then partially recovered in an autophagy-dependent manner 27 , 28 ., Since the amino acids produced by enhanced autophagy are utilized for new protein synthesis 27 , 28 , we investigated whether specific amino acids produced by autophagy contribute to cell cycle re-progression after a G2/M delay ., The Δatg2 strain AMY250 possesses his3 , trp1 , and ura3 mutations causing auxotrophies for histidine , tryptophan , and uracil , respectively ., We postulated that these auxotrophic mutations would affect the intracellular pools of the corresponding amino acids especially upon starvation ., As hypothesized , the G2/M-delayed Δatg2 mutant returned to G1 phase similarly to wild-type ( WT ) cells when tryptophan was added to the nitrogen-starved medium ( Figure 3A ) ., In contrast , the addition of histidine did not affect the suppression of the prolonged G2/M delay in the Δatg2 mutant , nor did the addition of glutamine , the level of which is believed to be involved in TORC1 activity in yeast 29 ( Figure 3A ) ., Recent studies have shown that TORC1 activity is regulated by the availability of some amino acid species including leucine; this is dependent on the editing function of aminoacyl-tRNA transferase 30 , 31 ., Using another Δatg2 mutant ( AMY296 ) , which is congenic to AMY250 with adenine and leucine auxotrophies , we further examined the effect of amino acid supplementation to a starvation medium on autophagy-deficient cells ., As shown in Figure 3B , the addition of tryptophan alone was insufficient for the cell cycle progression of AMY296 ., However , the simultaneous addition of tryptophan and leucine efficiently rendered the cells to a G1 arrest ., These results show that cell cycle perturbations caused by a deficiency in autophagy may be suppressed by the addition of amino acids specific to the strains , suggesting that autophagy contributes to cell cycle progression by allocating some , but not all , amino acids that are present in a limited intracellular amount ., Next , we examined whether the addition of specific amino acids would upregulate the starvation-repressed TORC1 activity ., Interestingly , while the addition of amino acids mimicked cell cycle progression in the autophagy-proficient cells , it was not associated with an increase in transcription of TORC1 downstream RPS26A and RPL9A ( Figure 3C and 3D ) ., This result raises the possibility that TORC1 recovery is not essential for cell cycle re-progression after a G2/M delay , although TORC1 re-activation correlates with re-progression ., To examine the causal relationship between the recovery of TORC1 activity and cell cycle re-progression , we examined cell cycle progression during starvation when TORC1 activity was inhibited by the addition of rapamycin ., As reported previously 20 , rapamycin treatment transiently delayed cell cycle progression at G2/M in nutrient-rich SCD medium , and most cells returned to G1 irrespective of their autophagic competency ( Figure 4A ) ., In contrast , rapamycin-treated WT cells showed a G2/M delay and proceeded to G1 when rapamycin was added to the nitrogen-starved medium ( Figure 4A ) ., TORC1 activity , monitored by RPS26A expression , was persistently lower in rapamycin-treated cells than that in those without rapamycin treatment during starvation ( Figure 4B ) , supporting the hypothesis that TORC1 recovery is dispensable for cell cycle re-progression after a G2/M delay ., Cell cycle progression to G1 phase in rapamycin-treated starved cells was significantly faster than in those that without rapamycin treatment ( Figure 4A ) ., However , the effect of rapamycin was not observed in Δatg2 cells , in which the cell cycle remained arrested in the G2/M transition in nitrogen-starved medium , even with the addition of rapamycin ( Figure 4A ) ., Again , these results show that autophagy is critical for cell cycle progression from G2/M to G1 under starvation conditions , and suggest the involvement of autophagy-dependent supply of amino acid pools in the accelerated cell cycle progression by rapamycin treatment ., We further examined the role of TORC1 activity in cell cycle re-progression using a temperature-sensitive KOG1mutant ( kog1-105 ) 20 , encoding an essential component of TORC1 ., Since the kog1-105 mutant showed a G2/M arrest phenotype at the restrictive temperature , the experiment was performed at 25°C , a permissive temperature at which the mutant is able to proliferate ., When cells were cultured under nitrogen-starved conditions , the G2/M delay was prolonged in kog1-105 cells compared to WT cells ( Figure S3A ) ., During the course of the experiment , TORC1 activity monitored by RPS26A expression was slightly lower in the kog1-105 mutant than in WT cells , both of which were exposed to nutrient-rich condition and nitrogen-starved conditions ( Figure S3B ) ., Moreover , the delay in cell cycle re-progression in the kog1-105 cells was relieved by the addition of rapamycin ( Figure S3C ) ., This result shows that further inhibition of TORC1 activity in kog1-105 cells by rapamycin is sufficient for cell cycle re-progression at G2/M , suggesting that the delay in cell cycle re-progression of the kog1-105 mutant is not likely due to the reduction in global TORC1 activity ., Rather , kog1-105 may cause a defect in specific pathways downstream of TORC1 , and the residual activity of TORC1 in the kog1-105 mutant , which can be inhibited by rapamycin , may prevent cell cycle recovery after a G2/M delay ., We investigated how amino acids produced by autophagy contribute to re-progression of the cell cycle from a G2/M delay under starvation conditions ., We first observed the morphology of WT and Δatg2 cells cultured under starvation conditions ., As shown in Figure 5 , the daughter cell of Δatg2 cells was smaller than that of WT cells 4 h after α-factor release , suggesting that amino acids produced by autophagy are important for sufficient cell growth during nutrient starvation ., The status of the bud ( daughter cell ) and timing of nuclear division are strictly regulated by the Swe1-dependent checkpoint mechanism in budding yeast 32–35 ., Therefore , we further analyzed cell division using Δswe1 and Δatg2 Δswe1 cells by DAPI staining to examine the potential relationship between the checkpoint mechanism and autophagy ., As shown in Figure 6A and 6B , loss of the Swe1 function in both WT and Δatg2 cells caused an increase in premature mitosis at early time points ( 2–3 h after α-factor release; type 4 in Figure 7A ) ., Interestingly , some of the Δswe1 cells arrested at 2C DNA content underwent normal cell division and returned to the G1 phase after 25 h ( Figure S4 ) , suggesting that premature mitosis caused by Swe1 dysfunction was rescued ., This result is likely due to the spindle orientation checkpoint 36 , which can correct an abnormal nuclear position at later time points ., Since the Swe1-dependent checkpoint delays entry into mitosis by regulating Swe1 degradation 37 , we examined the amount of Swe1 in response to nitrogen starvation ., During normal cell cycling under nutrient-rich conditions , Swe1 was accumulated 40 min after α-factor release , and degraded after 1 . 5 h ( Figure 6C ) ., Under starvation conditions , Swe1 was more stable , and degraded only after 3 h ( Figure 6C ) ., These results support the hypothesis that the starvation-induced delay of mitosis is mediated by Swe1 ., This notion is consistent with our previous result that TORC1 regulates G2/M progression through budding yeast polo-like kinase Cdc5 , a negative regulator of Swe1 in the initiation of mitosis 20 ., In Δatg2 cells , degradation of Swe1 occurred normally in nutrient-rich medium , and was delayed under starvation conditions ( Figure 6C ) ., Indeed , in Δatg2 cells , cells that underwent premature mitosis were scarcely observed 2 h after α-factor release ( Figure 6A and 6B ) ., Degradation of Swe1 was delayed in Δatg2 cells , and a portion of Swe1 still remained 5 h after α-factor release ( Figure 6C ) , showing that autophagy is important for Swe1 degradation in nutrient-starved cells ., However , we noted that Swe1 decreased gradually in Δatg2 cells under starvation conditions ., Consistently , after 2–3 h , approximately half of the Δatg2 cells were arrested before mitosis , but at later time points , the majority of the cells appeared to enter mitosis , as indicated by a reduction in the population of cells with one nucleus ( type 2 in Figure 7A ) ., Thus , autophagy is important for efficient recovery from the Swe1-dependent checkpoint under starvation conditions , although autophagy-deficient cells might , at least partly , execute nuclear division after a prolonged cell cycle delay ., Nonetheless , cells cannot complete cell division without autophagy under starvation conditions because , unlike Δswe1 cells , the majority of Δatg2 Δswe1 cells remained arrested at 2C DNA content even 25 h after α-factor release ( Figure S4 ) ., Indeed , cells that passed nuclear division but not cytokinesis ( type 3 in Figure 7A ) also accumulated transiently in WT cells 2–4 h after α-factor release , and such cells were consistently observed even at later time points in Δatg2 cells ( Figure 7A ) ., To examine whether autophagy is involved in cell cycle progression after nuclear division , we used an anti-microtubule drug nocodazole to synchronize cell cycle at metaphase ., WT cells were first synchronized in the G1 phase by treatment with α-factor , released into SCD medium , and then transferred into nitrogen-depleted medium containing nocodazole ., After 3 h when the majority of cells was arrested at metaphase , cells were collected and re-released into SD-N medium containing 1 mM PMSF that specifically inhibited autophagic degradation under nutrient-starved conditions 38 ., As shown in Figure 7B–7D , nocodazole-arrested cells showed a delay in completion of cytokinesis when autophagy was inhibited by the addition of PMSF ., Therefore , autophagy contributes not only to nuclear division , but also to cytokinesis , or cell separation , under nutrient-starved conditions ., Finally , we investigated the physiological significance of autophagy-dependent cell cycle re-progression from G2/M to G1 during starvation conditions ., Δatg2 cells cultured in nitrogen-starved conditions for 24 h showed an increased frequency of aberrant mitosis , in which two nuclei were present in a mother cell ( type 4 in Figure 7A; Figure 6A and 6B ) ., This was likely due to the leaky recovery from the Swe1-dependent mitotic delay ., Moreover , when cells were replenished with a nitrogen source and cultured for 2 h , cells with unusually high DNA content ( 3C DNA content ) appeared in Δatg2 and Δatg1 mutants but not in WT cells ( Figure 2A and Figure S2 ) ., In addition , analysis by a genetic system using intrachromosomal recombination 39 revealed that Δatg2 cells cultured in starvation medium displayed an increased frequency of aneuploidy ( 2 . 3-fold and 6 . 6-fold higher than that of WT cells after starvation for 24 and 48 h , respectively; Figure 8 ) ., These results indicate that cell cycle arrest in the G1 phase under starvation conditions is critical for the normal progression of mitosis after restoration of nutrient conditions , and highlight the importance of autophagy-dependent cell cycle re-progression in genome stability ., Although nutrient starvation reduces TORC1 activity and subsequently induces cell cycle arrest in the G1 phase 19 , cell cycle progression at the G2/M boundary was blocked under starvation conditions 20–23 ., Autophagy has been implicated in cellular physiologies under starvation conditions , and the data presented herein uncover another aspect of autophagic functions , namely , its contribution to cell cycle regulation and genome integrity in nutrient-starved yeast cells through the regulation of mitosis progression ., Nuclear division and cytokinesis are two critical events during the cell cycle , both of which require protein synthesis 40 ., Here we showed that , in addition to nuclear division , cytokinesis is another step of the cell cycle with limitations , whose entry is blocked or slowed by nutrient starvation , and that starvation-induced autophagy is required to overcome the cell cycle delay at both steps ( Figure 6 and Figure 7 ) ., Previous reports have revealed that amino acids produced by autophagy are used for protein synthesis 27 , 28 ., In the present study , we showed that the defects in cell cycle re-progression caused by a lack of autophagy was suppressed by the addition of particular amino acids ( Figure 3 ) , and that autophagy promoted the growth of daughter cells in starvation medium ( Figure 5 ) ., These results clearly show that the amino acid supply through autophagy contributes to sufficient cell growth during nutrient starvation ., We found that this phenotype was associated with auxotrophies to specific amino acids , including leucine and tryptophan , but not histidine or nucleobases , including uracil and adenine ( Figure 3 ) ., The composition of the amino acid pool in budding yeast cells is influenced by nutrient conditions; cells using NH4+ as the sole nitrogen source accumulate glutamic acid and arginine , but contain amino acids such as tyrosine , leucine , tryptophan , and phenylalanine at low levels 41 ., Thus , we assume that , at least under our experimental conditions , leucine and tryptophan are limiting amino acids whose pools cannot support protein synthesis without autophagy upon nitrogen starvation ., It will be interesting to examine the potential difference in amino acid requirements for the suppression of autophagy deficiency by varying the composition of amino acid pools using alternate nitrogen sources ., In cultured human cells , the cellular pools of glutamine and glutamic acid are maintained at high levels , whereas those of tryptophan , cysteine , and arginine are maintained at low levels 42 ., It is important to note that the latter three are essential amino acids that cannot be synthesized de novo in humans; thus , autophagy may be involved in maintaining the pools of these amino acids in human cells ., A recent paper reported that TORC1 activity was partially reactivated in an autophagy-dependent manner in ongoing starvation conditions , which played a role in the attenuation of autophagy 43 ., Consistent with their findings , we demonstrated that starvation-induced autophagy leads to partial re-activation of TORC1 activity and that the timing of the re-activation correlates with that of cell cycle re-progression at G2/M ( Figure 2 ) ., Although a reduction in TORC1 activity appears to contribute to the transient cell cycle delay at G2/M , we showed that the cell cycle delay was relieved under TORC1-repressed conditions ( Figure 4 ) ., These results argue against the positive role of TORC1 re-activation during a rescue from the delay ., Although TORC1 activity is involved in translation 19 , previous studies have shown that the synthesis of specific proteins continues under TORC1-inhibited conditions 44 ., Thus , even in case that TORC1 activity is not re-activated , protein synthesis supported by the autophagy-mediated amino acid pools may be sufficient for starvation-adapted cells to complete a final round of the cell cycle ., Although the inhibition of TORC1 by rapamycin did not abolish cell cycle re-progression from a G2/M delay , the temperature-sensitive kog1-105 mutation did affect this process ( Figure 4 and Figure S3 ) ., There are three possible models to explain the cell cycle-specific phenotype of kog1-105 cells ., In the first model , kog1-105 cells may be defective in only a part of multiple TORC1 functions , and the function affected by kog1-105 is necessary for cell cycle progression after a G2/M delay ., In the second model , TORC1 may be significantly affected in kog1-105 and the level of TORC1 activity in the mutant is below that caused by nutrient starvation or rapamycin treatment in WT cells ., The last model , based on the possibility that kog1-105 is a gain-of-function mutation , suggests that Kog1-105 protein fulfills a function in addition to the TORC1 function , which is not performed by the normal Kog1 protein ., Our results showing that rapamycin suppresses the kog1-105 defect contradict the latter two models; therefore , it is likely that the kog1-105 mutation blocks only a portion of the pathways downstream of TORC1 ., This scenario is consistent with our previous findings that kog1-105 is not defective in the progression of G1 and does not induce autophagy under nutrient-rich conditions at a restrictive temperature 20 ., We also found that the kog1-105 mutation affected the interaction between Kog1 and TORC1 20 ., Raptor , the mammalian ortholog of Kog1 , contributes to the inhibition of mTOR activity upon nutrient depletion through stabilization of the mTOR-Raptor complex 45 ., Although it is unknown whether Kog1 inhibits TORC1 under starvation conditions , a decreased interaction with TORC1 could contribute to the phenotypes observed in kog1-105 , which are distinct from those caused by rapamycin treatment ., There are several examples of rapamycin-insensitive , TORC1-dependent processes in mammalian cells 46 ., Although the presence of such processes is elusive in yeast , it is possible that rapamycin-insensitive TORC1 activity is involved in the recovery from a G2/M delay ., However , under these circumstances , it is difficult to explain why the addition of rapamycin to nitrogen-depleted medium leads to an early recovery from a cell cycle delay ( Figure 4A ) ., It is interesting to note that the addition of rapamycin to fission yeast cells leads to early mitotic onset in nutrient-rich medium , resembling the cell cycle behavior caused by a reduction in the quality of the nutrient source 47 ., In budding yeast , cellular responses to rapamycin are not identical to those induced by nitrogen starvation 48 , 49; rapamycin treatment rapidly activates the quality-sensitive nitrogen discrimination pathway , which is distinct from the nitrogen starvation pathway , to facilitate use of poor nitrogen sources ., Such a difference may induce rapid adaptation to nitrogen starvation in rapamycin-treated cells ., Indeed , the addition of rapamycin facilitates adaptation to an environment containing a low-quality nitrogen source in budding yeast 50 ., We tested the involvement of the nitrogen discrimination pathway in the rapamycin-induced early cell cycle recovery by deleting GLN3 , which encodes the key transcriptional regulator in the nitrogen discrimination pathway , and found that rapamycin treatment accelerated recovery from the G2/M delay even in Δgln3 mutants ( data not shown ) ., Therefore , we can at least conclude that rapamycin-induced transcription through Gln3 is not essential for this phenotype ., It is known that TORC1 regulates a variety of cellular events including transcription , translation , and post-translation 51 ., Note that autophagy is essential for rapamycin-mediated early cell cycle recovery ( Figure 4A ) and that protein synthesis supported by the amino acid pool appears to be involved in this mechanism ., Further studies are required to specify the pathway responsible for the early recovery of cell cycle by rapamycin ., In addition , it would be interesting to determine physiological conditions that induce the early cell cycle recovery phenotype in a similar manner to that caused by rapamycin treatment; such approaches will help clarify the significance of cell cycle regulation in response to acute inhibition of TORC1 activity ., We have shown that the G2/M specific function of TORC1 is mediated by the polo-like kinase Cdc5 20 , an upstream regulator of Swe1 52 ., Swe1 is involved in a checkpoint mechanism to ensure accuracy of cell division by monitoring daughter cells 32–35 ., When this checkpoint mechanism is inhibited by the swe1 mutation , or overexpression of its negative regulators , HSL1 and HSL7 , nuclear division occurs prematurely , even if the growth of the bud is suppressed by a mutation in CDC24 53 ., We observed a similar premature mitosis of Δswe1 cells under nitrogen starvation ( Figure 5A and 5C ) , indicating that the Swe1-dependent checkpoint , probably activated by insufficient bud growth , contributes to the G2/M delay phenotype induced by starvation ., We found that a deficiency in SWE1 also increased the fraction of cells that cannot return to the G1 phase normally ( Figure S4 ) , indicating that both the Swe1-dependent cell cycle delay and the autophagy-dependent recovery are critical for the integrity of mitosis ., Thus , timely regulation of cell cycle progression is of significance under starvation conditions ., Although autophagy is required for the growth of daughter cells during starvation conditions ( Figure 5 ) , in our experimental conditions , more than 80% of the autophagy-deficient cells ultimately proceeded to nuclear division after delay at the G2/M boundary ., This may be because cells can produce a limited pool of amino acids that are independent of autophagy that support maturation of the bud to overcome the Swe1-dependent checkpoint ., Otherwise , long-term starvation might cause an imbalance in the amount of proteins regulating M phase entry , thereby initiating nuclear division , since the Swe1-dependent checkpoint does not appear to monitor bud size per se , but detects an accumulation of mitotic cyclins in the bud 54 ., In either case , cell division is not completed before cytokinesis without autophagy ., Cytokinesis may serve as an additional control gate for the fulfillment of daughter cell maturation under nutrient-limited conditions ., Cell cycle-dependent regulation of constitutive autophagy has been shown in mammalian cells 7–10 ., However , this study established the direct involvement of autophagy in cell cycle regulation under starvation conditions; autophagy ensures the accomplishment of the final round of cell cycle progression to nutritionally starved cells ., Without autophagy , prolonged treatment with nitrogen starvation caused perturbation of the cell cycle , including premature mitosis , and caused an increased frequency of aneuploidy in budding yeast ( Figure 6A , 6B and Figure 8 ) ., Thus , in addition to the developmental significance of returning to the G1 phase under starvation conditions ( i . e . , meiotic division is only initiated from the G1 phase in diploid yeast cells ) , our results indicate that returning to the G1 phase is critical for maintaining genome integrity after restoration of the nutrient condition from starvation ., Notably , a previous study reported that compromised autophagy promotes genomic instability , such as increased DNA damage , gene amplification , and aneuploidy in mammalian cells 55 , consistent with the tumor-suppressive activity of autophagy that was previously reported 56–59 ., Our results demonstrate that autophagy allows a final round of cell cycle progression in budding yeast cells by supplying amino acids during nutrient

Introduction, Results, Discussion, Materials and Methods

The coordination of subcellular processes during adaptation to environmental change is a key feature of biological systems ., Starvation of essential nutrients slows cell cycling and ultimately causes G1 arrest , and nitrogen starvation delays G2/M progression ., Here , we show that budding yeast cells can be efficiently returned to the G1 phase under starvation conditions in an autophagy-dependent manner ., Starvation attenuates TORC1 activity , causing a G2/M delay in a Swe1-dependent checkpoint mechanism , and starvation-induced autophagy assists in the recovery from a G2/M delay by supplying amino acids required for cell growth ., Persistent delay of the cell cycle by a deficiency in autophagy causes aberrant nuclear division without sufficient cell growth , leading to an increased frequency in aneuploidy after refeeding the nitrogen source ., Our data establish the role of autophagy in genome stability through modulation of cell division under conditions that repress cell growth .

A nutrient stress such as nitrogen depletion induces pleiotropic responses in eukaryotic cells ., For example , nutrient starvation slows cell cycling and ultimately causes G1 arrest ., In addition , it is known that nitrogen starvation delays G2/M progression ., However , the mechanism as to how G2/M-delayed cells progress through the cell cycle to return to the G1 phase remains unclear ., Cells subjected to a nutrient stress induce autophagy , a bulk degradation system within lysosomes/vacuoles , to reconstitute cellular components ., In this study , we show that an autophagy-dependent supply of amino acid pools is critical for completion of cell cycle under starvation conditions in the budding yeast Saccharomyces cerevisiae ., Autophagy deficiency causes a defect in cell growth and leads to abnormal mitosis associated with a higher incidence of aneuploidy ., Thus , our data establish the role of autophagy in genome stability through modulation of cell division under conditions that repress cell growth , which provides a possible mechanism of tumor suppression by autophagy shown in mammalian cells .

cellular stress responses, genetic mutation, microbiology, gene function, model organisms, cell division, cell growth, molecular genetics, chromosome biology, biology, cell biology, genetics, yeast and fungal models, molecular cell biology, genetics and genomics

journal.pcbi.1006016

Particle-based simulations of polarity establishment reveal stochastic promotion of Turing pattern formation

Cell polarity refers to the localization of signaling molecules to specific regions of the plasma membrane , and is required for fundamental cellular processes such as migration , directed growth , and differentiation ., In the yeast Saccharomyces cerevisiae , polarization is required for directed growth during budding and mating ., Because of its experimental tractability , yeast represents a powerful model organism for studying polarity establishment ., Normally , yeast polarization involves internal or external spatial cues such as bud scars and pheromone gradients ., However , polarization still occurs if these cues are removed 1 ., Mathematical models have been used to explain spontaneous pattern formation by biochemical systems since the 1950s 2 , 3 ., These models use diffusion-driven instabilities to generate symmetry breaking without relying on mechanisms such as diffusional barriers , directed transport , and molecular cross-linking ., Instead , these systems require: ( 1 ) positive feedback to amplify local fluctuations; ( 2 ) chemical species that diffuse at different rates; and ( 3 ) a mechanism for limiting the growth of the polarity site ., Models in which patterning can be induced by an arbitrarily weak perturbation ( e . g . molecular-level fluctuations ) are called Turing-type ., Goryachev and Polkhilko were the earliest to use a Turing-type model to study yeast polarization 4 ., Other , non-Turing type models of polarity require perturbations of finite strength to induce pattern formation 5 ., A common approach to modeling the spatiotemporal dynamics of polarizing biochemical systems is to use reaction-diffusion equations ( RDEs ) in the form of non-linear partial differential equations ( PDEs ) ., RDEs are deterministic and ignore stochastic effects intrinsic to chemical reactions and thermal diffusion ., In some systems , stochastic effects have been shown to expand the parameter space that leads to patterning and accelerate pattern formation 6 , 7 ., Many modeling approaches are used to study stochastic effects in biological signaling networks , including stochastic differential equations , such as chemical Langevin equations 8 , 9; spatially discretized , temporally-continuous approaches , such as the spatial Gillespie algorithm 10–12; exact Brownian dynamics , such as Green’s function reaction dynamics 13 , 14; and direct particle-based simulations , as implemented in Smoldyn and MCell 15 , 16 ., We cannot adequately cover the full spectrum of approaches and computational tools here , but refer the reader to excellent reviews that describe the theoretical underpinnings and software implementations of these methods 17–20 ., In Table 1 , we describe advantages and limitations for some of the more common methods ., Hybrid approaches , such as the method described in 21 , that mix particle simulations with a deterministic partial differential equation solver are most similar to the approach we take here ., The effects of noise in non-Turing models of yeast polarization have been investigated using a variety of stochastic methods 21 , 23–25 ., Typically , these models make simplifying assumptions to reduce the complexity of the polarity system ., Some models , such as the neutral drift polarity model , used particle-based approaches; others , like models based on wave-pinning , used Gillespie or stochastic PDE-based approaches 26 , 27 ., Other investigations of stochasticity in polarization with more detailed signaling models , including the Turing-type Goryachev–Pokhilko model , leveraged Gillespie and stochastic PDE approaches 28 , 29 ., Here , we present particle-based simulations of the Goryachev–Pokhilko model , and compare them to RDE simulations of the same system to evaluate stochastic effects on polarization ., In this model , reactions occur between either two reactants on the membrane , or between a reactant on the membrane and a reactant in the cytoplasm ., Exchange can occur between the membrane and cytoplasm ., We design our simulations to explicitly track molecules at and near the cell membrane , where polarization occurs , and implicitly handle molecules away from the membrane ., We consider two different scenarios ., In the first , we treat the cell membrane and the nearby cytoplasm as purely two-dimensional ( 2D ) and ignore the remaining bulk cytoplasm ., In the second , we approximate the bulk cytoplasm by attaching a molecular reservoir in which we only track molecular abundances ., Molecules are stochastically exchanged between the 2D particle-based domain and the reservoir with rates determined by diffusion , thus creating a quasi-three dimensional system ( Fig 1 ) ., An outline of our paper is as follows ., First , we demonstrate that our particle-based simulations generate results consistent with deterministic rate equations in the well-stirred limit ., We then show how deviations from this idealized behavior occur as spatial effects become important ., These deviations occur when the 2D reactions become diffusion-influenced , and it is no longer possible to describe the kinetics of second-order reactions with a single macroscopic rate constant 30 , 31 ., Interestingly , existing models of the yeast polarity system contain second-order rate constants that appear to fall within this diffusion-influenced regime , calling into question the validity of the model equations ( Fig C in S1 Text ) ., However , existing theories for computing second-order rate constants from microscopic parameters do not take into account the situation in which chemical species can transition between different diffusional states , e . g . membrane versus cytosolic ., Therefore , we empirically determined second-order rate constants by fitting rate equations to results from particle-based simulations ., Each potentially diffusion-influenced bimolecular reaction was simulated in combination with the relevant membrane-cytoplasm exchange reactions ., Our results demonstrate that in many cases the chemical kinetics of this expanded system can be well-approximated using a single second-order rate constant ., This empirical mapping between the microscopic and macroscopic regimes allows us to compare the polarization results from particle-based simulations to solutions of the corresponding RDEs ., We show that molecular fluctuations increase the rate at which polarization occurs in a purely 2D system , lacking the cytoplasmic reservoir ., Polarity also occurs over a broader range of Cdc42 concentrations ., These observations are consistent with previous reports in other systems where Turing patterning was enhanced due to either particle-based fluctuations 6 , 7 or sufficiently strong perturbations 32 , 33 ., We also show that stochastic effects inherent to particle-based simulations can generate large scale variability in polarization dynamics and metastable multi-patch states ., This is in agreement with theoretical and experimental 4 , 34 , 35 demonstrations of emergent , competing multi-polar states ., Moving on to particle-based simulations with the quasi-3D molecular reservoir , we find that the particle-based simulations still exhibit enhanced polarization compared to the deterministic RDEs within parameters representative of a typical yeast cell ., In the quasi-3D particle-based simulations , the resolution of multi-patch states takes place on a timescale of minutes , consistent with experimental measurements 34 , 35 ., To our knowledge , our work represents the first particle-based simulations of a model for yeast polarization that is based on a Turing mechanism ., Our simulations underscore important effects of stochasticity on polarity establishment , including more rapid competition between polarity sites and increased robustness to changes in molecular abundances ., We first considered a purely 2D computational domain representing molecules in the cell membrane and a thin volume of cytoplasm adjacent to the membrane ., Molecules in the membrane or cytoplasmic layer were differentiated by their diffusivity and reactivity ., We neglect the rest of the cytoplasm until later ( see subsection “A quasi-3D approach to full cell simulations” ) ., The spatial coordinates of molecules were treated as continuous variables , while time was discretized in intervals of Δt ., Thermal diffusion was handled using the Euler-Maruyama method 36 ., First-order or unimolecular reactions were assigned probabilities of occurring in Δt given by Pi = 1 –exp ( -kiΔt ) , where ki was the rate constant for the i-th reaction ., If the first-order reaction involved the dissociation of two molecules , then the two products were placed a distance of σ¯ apart , with one of the molecules located at the position of the complex , and the orientation angle chosen at random from a uniform distribution ., For second-order or bimolecular reactions , we assumed that two molecules react with probability Pλ = λΔt when they are within a distance ϱ¯ ., Thus , if the two reactants are within a reactive range ϱ¯ , they react with an average rate λ ., This approach is based on the Doi method 37 ., It is distinct from the classic diffusion-limited Smoluchowski approach , where molecules react upon finding one another for the first time and molecular radii are adjusted to reach the desired kinetics 38 ., Investigating the role of molecular fluctuations in polarity establishment requires a way to compare particle-based simulation results to the deterministic behavior of the system in the macroscopic limit , where the spatiotemporal dynamics of biochemical concentrations are governed by reaction-diffusion equations ( RDEs ) ., Therefore , we needed a way to relate microscopic parameters to macroscopic rate constants in two dimensions ., For first-order reactions , this is trivial , and follows the relation noted above ., The situation is more complicated for second-order reactions ., In chemical kinetic theory , there are two limiting regimes for second-order reactions ., The first is the diffusion limit , in which two particles react when they encounter one another for the first time ., The diffusion limit represents the maximum rate at which a second-order reaction can proceed ., In 3D , it is possible to define a macroscopic rate constant in the diffusion limit by considering the diffusional flux through an absorbing sphere of radius ϱ¯ located at the origin , when the concentration C of the reactant is held fixed at infinity 39 , 40 ., The flux into the sphere is given by J=4πDϱ¯C , where D is the sum of the diffusion coefficients of the reactants ., From this expression , the second-order rate constant in the 3D diffusion limit is defined to be k=4πDϱ¯ ., In 2D , diffusion-limited second-order rate constants are not well-defined 11 , 30 ., However , we were able to estimate a time scale by computing the flux through an absorbing circle of radius ϱ¯ when the computational domain remains finite ( see Appendix A in S1 Text for details ) ., In this case , the flux is given by J=2πDC/ln\u2061 ( rmax/ϱ¯ ) , where rmax characterizes the size of the computational domain ., In contrast to the 3D case , in the limit rmax → ∞ , the 2D flux goes to zero ., Thus , we used the flux on a finite domain to estimate a time scale for second-order diffusion-limited reactions , kDL=2πD/ln\u2061 ( rmax/ϱ¯ ) , which has the units of a 2D second-order rate constant ., This expression is represented by the red curve in Fig 2 ., The other regime for second-order reactions is the reaction limit ., In this limit , multiple encounters on average are required before the reaction occurs ., We computed a second-order rate constant in the reaction limit by assuming the reactants are uniformly distributed ., In 2D , this produces an overall reaction rate of ( πϱ¯2/A ) λNANB , where πϱ¯2 is the capture area , A is the area of the system , λ is the microscopic reaction rate , and NA and NB are the particle numbers for the two reactants ., This leads to a second-order rate constant of kRL=πϱ¯2λ ., The reaction limit is illustrated by the black dashed line in Fig 2 ., In 3D , the λ−ϱ¯ theory of Erban , Chapman and co-workers can be used to compute macroscopic rate constants from the underlying microscopic parameters ( λ , D , and ϱ¯ ) regardless of the reactants diffusion coefficients 39 , 40 ., The theory assumes the two reactants have a summed diffusivity D , and that reactions proceed with rate λ if the two reactants are within ϱ¯ of one another ., In general , the λ−ϱ¯ theory cannot be extended to 2D , because in the diffusion limit , the rate at which a 2D second-order reaction proceeds cannot be described using a single rate constant 30 ., Despite that , we used the λ−ϱ¯ formulism to compute 2D rate constants ( Appendix A in S1 Text , Figs A and B in S1 Text ) ., Comparing calculations using the λ−ϱ¯ formulism ( green dashed line , left panel , Fig 2 ) to the results based on particle simulations ( yellow diamonds , left panel , Fig 2 ) , we find they are accurate predictions of reaction kinetics if the system is not too far from the reaction limit ., We then attempted to estimate λ values from rate constants used in published models of yeast polarity establishment ., In doing so , we discovered that several published second-order rate constants appeared to be larger than our estimate for the diffusion limit , kDL ( Fig C in S1 Text ) ., However , the considerations discussed above do not take into account the fact that molecules involved in polarity can transition between the cell membrane and cytoplasm ., As discussed next , the different diffusion coefficients associated with these different cellular compartments further complicates the mapping between microscopic and macroscopic parameters ., In the biochemical network that drives polarity , reactive chemical species can exchange between the membrane , where diffusion is relatively slow , and the cytoplasm , where diffusion is relatively fast ., The reactivity of these species also changes depending upon whether they are in the membrane or cytoplasm ., Existing methods to estimate 2D macroscopic rate constants from microscopic parameters under diffusion-limited conditions 30 do not consider the effect of membrane-cytoplasm exchange ., Here , we were able to overcome this issue by empirically estimating effective rate constants by fitting chemical kinetic equations to results from our particle-based simulations ., We conducted particle-based simulations of reversible second-order association reactions that accounted for mass exchange between the cytoplasm and membrane in a purely 2D system ., Briefly , for each parameter set , we started with previously published rate constants from RDE models for polarity establishment 28 , 35 , 41 , used the λ−ϱ¯ formalism to estimate λ’s , then performed particle simulations and fit rate equations to the simulation results to compute the macroscopic rate constants ., For purely 2D simulations , significant changes to the published parameter values were made to facilitate polarization for benchmarking purposes ., For whole cell , quasi-3D simulations , parameters were held close to published values with exceptions for the bimolecular reactions obtained from the fitting procedure ., We present the fits for the purely 2D and quasi-3D cases ( Fig 3; Figs . I and M in S1 Text ) , as well as the rate constants ( Table 2 , Table A in S1 Text ) ., Fitting the simulation results to appropriate chemical rate equations produced good estimates for the quasi-3D case ( Fig 3 , bottom row ) and reasonable ones for the purely 2D case ( Fig 3 , top row ) ., Additional analyses of the polarity network , discussed below , further supported the validity of the mapping ., We compared stochastic particle-based and deterministic reaction-diffusion-based simulations ., First , we focus on our results in the purely 2D system ., Our initial conditions for the particle-based simulations had all molecules in the cytoplasm in inactive and uncomplexed states ., As expected , stochastic fluctuations permitted escape from this spatially homogeneous initial state , ultimately leading to polarization ( Fig 4A , S1 Movie ) ., To fairly compare particle simulation results with solutions to the RDEs , molecular distributions from particle-based simulations at t = 1 second were used as initial conditions for the RDEs ( see Models and Methods and Fig D in S1 Text ) ., The two simulation methods generated similar polarized distributions ( Fig 4B ) ., To quantitatively compare polarization between the two approaches , we used the function H ( r ) , which measures the deviation of a particle distribution from a uniform distribution based on the pairwise distance distribution ( see Models and Methods and Fig 5C ) ., H ( r ) and the related metric , Ripleys K-function , have been used frequently to study clustering in biology 42 , 43 ., Positive values of H ( r ) correspond to increased density of the distribution at distances r relative to a uniform distribution ., A maximum in H ( r ) denotes a characteristic size ., We use this measure of spatial heterogeneity to quantify polarity ., At steady-state , polarized distributions from the particle-based and RDE simulations had essentially identical H ( r ) curves ( Fig E in S1 Text ) , suggesting the two systems were equivalently parameterized ., We calculated H ( r ) over time for simulations using different parameter values to quantitatively compare polarization dynamics from particle-based and RDE simulations ., Rather than choosing the r that maximizes H ( r ) under different conditions , we chose r = 0 . 5 μm for our analyses ., This value allowed comparisons across all data sets , including those where the simulation domain size was varied ., Qualitative features of our results do not depend on our choice of r , nor on the particle-based time point used to initialize the RDEs ( Figs F and G in S1 Text ) ., To account for variability in polarization dynamics , we considered multiple realizations of single simulation conditions ( Fig 5 ) ., In several cases , metastable multi-polar states emerged from initially unpolarized distributions , consistent with prior theoretical and experimental 4 , 34 , 35 work ., Though it is not possible to identify multi-polar states from looking at H ( r ) alone , if the system goes through a slow phase of competition wherein metastable patches exist , then the time course of H ( r ) temporarily plateaus ., For one realization , resolution into a single polarity site did not occur by 200 seconds ( Fig 5 , Simulation 3 ) ., For other realizations of the same parameter set , the simulation yielded a unique polarity site in half the time ., Overall , the particle-based simulations polarized more rapidly than the RDEs , which were completely unpolarized at t = 200s ., This indicates that molecular fluctuations increased the rate at which polarity establishment occurred ., The RDEs did not exhibit transient plateaus in H ( r ) , indicating metastable multi-patch states did not emerge , which is a direct consequence of the initial conditions ( see also Fig H in S1 Text ) ., It has been demonstrated that sufficiently strong fluctuations can allow polarization outside of the Turing unstable regime 6 , 7 , 32 , 33 ., These investigations relied on simplified models or phenomenological methods for introducing noise into the system ., To test if intrinsic fluctuations are sufficient to produce “noise-induced” polarity , we examined 2D polarity establishment as a function of Cdc42 concentration , Bem1-Cdc24 ( GEF ) concentration , and total particle number at fixed concentration , generating bifurcation diagrams for these parameters ., We used linear stability analysis of the RDEs to determine the bifurcation point at which the spatially homogenous solution goes through a Turing instability as molecular abundances and system size were varied ( see Models and Methods and Fig J in S1 Text ) ., This analysis established threshold values at which the RDEs no longer polarize , i . e . the homogeneous stable regime ., The bifurcation plots are shown in Fig 6 ., Across all parameters tested , none of the RDE simulations polarized to a measurable degree after 200 seconds ., In contrast , most particle-based simulations exhibited polarity by then ., Within the Turing unstable regime , the RDE simulations show similar levels of polarization around 600 seconds compared to the particle-based simulations ., However , near the bifurcation point within the Turing unstable regime , the RDEs did not polarize even after 600s , consistent with the slowed patterning expected from bifurcation theory ., In this parameter regime , the particle-based simulations still clearly exhibited polarity within 200 seconds ., Furthermore , for the Cdc42 and GEF bifurcation diagrams , the particle-based simulations showed polarization below the critical point , in the Turing stable regime , showing that molecular fluctuations can increase the range over which polarity establishment occurs ., Together , our observations reveal that stochastic effects facilitate polarization in this 2D instance of the Turing-type model by decreasing time to polarize and expanding the parameter space in which polarity can occur ., We next expanded our approach to approximate a whole cell by introducing a molecular reservoir to account for contributions from the bulk cytoplasm , yielding a quasi-3D approach ( Fig 7A and 7B ) ., The cytoplasmic reservoir was treated implicitly: we only tracked the number of molecules in the reservoir , instead of the dynamics of individual particles ., To simulate stochastic exchange between the explicitly-modeled and implicitly-modeled regions of the cytoplasm , we took a similar approach as described in 44 , using diffusional probability distributions to determine the number of molecules injected into ( ninj ) and ejected from ( nejc ) the explicitly-modeled cytoplasm at each time step ., Diffusional probability densities were integrated to obtain Pinj and Pejc , which correspond to the probability that a single molecule at a depth z diffuses the distance required to enter ( zimpl—z ) or exit ( z–zimpl ) the explicit simulation region ( see Appendix D in S1 Text for derivation ) ., Pinj ( z ) =12erf\u2061 ( zmax−z4DΔt ) −erf\u2061 ( zimpl−z4DΔt ) , Pejc ( z ) =12erf\u2061 ( z4DΔt ) −erf\u2061 ( − ( zimpl−z ) 4DΔt ) , where zmax is the total height of the implicit and explicit domains , and zimpl is the height of the implicit domain ., Pinj ( z ) and Pejc ( z ) are approximations , since the probability densities in the derivation correspond to a freely diffusing particle on an infinite domain ., Next , to calculate the mean number of particles that are injected and ejected , we integrated the injection and ejection probability densities over the appropriate domain , and multiplied by the 3D concentration and the surface area ., Finally , to approximate the stochastic fluctuations introduced by particles diffusing in and out of the explicit simulation domain , we sampled from Poisson distributions at each time step with means ⟨ninj⟩ and ⟨nejc⟩ ., Coupling this reservoir to the cytoplasmic layer of the 2D particle-based method yielded our quasi-3D full-cell particle-based approach ., Comparisons between this approximate method and Brownian dynamics simulations of diffusing particles showed that our molecular reservoir approach was consistent with both the mean and standard deviation for particle number over time ( Fig 7C and 7D ) ., We performed quasi-3D simulations of a whole yeast cell by combining our 2D particle-based approach with stochastic exchange to and from a molecular reservoir representing the bulk cytoplasm ., Empirical estimation of rate constants was again performed by fitting rate equations to the particle-based simulations ( Fig M in S1 Text ) ., We conducted simulations using 0 . 050~0 . 3 μM Cdc42 , and 0 . 06 μM BemGEF ( NCdc42 = 1 , 970~11 , 820 and NBemGEF = 2364 , assuming a volume corresponding to a spherical cell with a 5 μm diameter ) ., Quantitative Western blotting experiments by Watson et al . support 5 , 000–10 , 000 Cdc42 copies per cell , consistent with our choice for concentration range 45 , while previous models assumed Cdc42 ranging from 19 . 3 nM 46 to 5 μM 28 ., Other models specify BemGEF ranging from 0 . 017 μM 4 , 28 to 0 . 06 μM 41 ., Since prior experimental work showed that multi-polar states can resolve within 2 minutes 34 , 35 , we initially limited our particle-based simulations to 200 seconds ., This simulation time was insufficient for complete polarization , as multiple or misshapen patches were observed at t = 200s ( Fig 8 middle; Fig N in S1 Text ) ., Before performing longer particle-based simulations , we determined the bifurcation point as a function of Cdc42 concentration in the analogous quasi-3D RDEs ( see Models ) ., Rather than perform linear stability analysis of these equations , we generated pre-polarized distributions and examined whether they decayed towards homogeneity to estimate the bifurcation point ( Fig P in S1 Text ) ., We found that Cdc42 ≥ 0 . 055 μM was sufficient for polarization , but Cdc42 = 0 . 050 μM could not sustain polarity ., Therefore , we chose to extend simulations with Cdc42 = 0 . 050 , 0 . 055 , 0 . 060 , 0 . 150 , and 0 . 155 μM for another 400 seconds ., This simulation time was sufficient to tighten misshapen polarity sites if no competitor patch existed ( Fig 9 , Simulations 2 and 3; see also Fig N in S1 Text ) ., In one case , two co-existing patches resolved into one within the 400s extension period ( Fig 8 ) ., In another case , the patches did not resolve ( Fig 9 , Simulation 1 ) ., The capacity to resolve competition within the 400s window suggests that biologically relevant competition time scales can be obtained purely through stochastic molecular fluctuations ., The time scale for competition observed here is consistent with Wu et al . ’s theoretical work on this signaling model , where about 5 minutes was needed to resolve two-patch competition in the context of an RDE with Gaussian noise added 28 ., To compare with the deterministic case , we ran quasi-3D RDE simulations for 1800s total , initialized with molecular distributions from t = 1 s of the quasi-3D particle-based simulations ., Polarization dynamics were quantified using H ( r = 2 μm ) , which matched the size of a fully-formed polarity site ., Similar to the purely 2D case , we found that fully polarized particle-based simulations were quantitatively consistent with fully polarized RDE simulations , and that the RDE simulations took much longer to polarize than the corresponding particle-based simulations ( Fig 9 ) ., No multi-patch states emerged in the RDEs , but we expect multi-patch states to compete even more slowly , supporting the importance of molecular fluctuations in using a Turing-type model to capture appropriate polarization timescales ., Finally , to examine the robustness of this behavior over realistic concentration regimes , we compared polarization in the quasi-3D particle-based and quasi-3D RDE systems as a function of Cdc42 concentration ., Our observations here were consistent with the purely 2D results ., Particle-based simulations at t = 600s exhibited clear polarization , even at Cdc42 = 0 . 050 μM , outside the deterministically non-polarizing region ( Fig 10 , S2 Movie ) ., At the highest concentration , quasi-3D RDE simulations exhibited partial polarization at t = 600s , but by t = 1800s , most of the RDEs beyond the bifurcation exhibited measurable polarization ., The macroscopic system we studied here represents a 3-compartment model ( membrane , near-membrane , and bulk cytoplasm ) ., Though Wu et al . reported a similar competition time scale , they utilized a volume-adjusted , two-compartment model of the RDEs 28 ., To facilitate comparison , we also performed particle-based simulations to examine the volume-adjusted , two-compartment systems bifurcation diagram with respect to Cdc42 concentration ., There is qualitatively no change in our results , and linear stability analysis of the volume-adjusted , two-compartment system is consistent with the numerically determined bifurcation point for the q3D-RDEs ( Fig P in S1 Text ) ., Strong positive feedback to amplify heterogeneities in molecular distributions is an important component of many models of cellular polarity establishment ., Given the stochastic nature of biochemical reactions involved in the polarity circuit , local heterogeneities are expected to emerge everywhere along the cell ., Work in both non-Turing type 21 , 23 , 24 , and Turing-type systems 6 , 7 , 32 , 33 has shown that stochasticity can aid pattern formation ., Here , we provide the first simulations of particle-based Turing-type yeast polarity establishment ., Both our 2D and quasi-3D particle-based simulations capture microscopic stochastic effects , which indeed facilitate polarization ., As anticipated , differences between the particle-based and reaction-diffusion approaches were most obvious around the bifurcation point ( Fig 6 , Fig 10 ) ., Stochastic fluctuations allowed for polarization outside of the Turing unstable regime and more rapid polarity establishment across all parameters tested ., Turing-type patterning mechanisms have been described as slow relative to other hypothesized patterning mechanisms , such as wave-pinning 5 , making it a less likely biological mechanism in some contexts ., Our simulations highlight that molecular fluctuations can alleviate such issues ., Given our simulations do not include other sources of fluctuations , such as endocytic and exocytic events 47 , 48 , our results represent the minimal level of variability expected to be observed in polarity establishment ., This minimal variability is sufficient to generate significant variations in competition times across multiple realizations of a single parameter set ( Fig 5 , Fig 9 ) , even at molecular abundances representative of whole yeast cells ., Therefore , particle-based simulations are an important computational tool for understanding the dynamics and control of biological pattern formation ., Polarity establishment is often modeled using reaction-diffusion equations that ignore the discrete nature of biomolecules , and treat concentrations of molecular species as continuous variables ., The chemical rate constants that appear in these equations represent macroscopic quantities that depend on microscopic properties , such as diffusion coefficients and molecular size ., In three-dimensional domains where particles diffuse with a single diffusion coefficient , theories for computing macroscopic rate constants from the underlying microscopic dynamics are well established 15 , 40 ., However , for two-dimensional systems , second-order rate constants in the diffusion limit are not well-defined 30 ., Additionally , in the polarity system , molecular species transition between the cytoplasm , where diffusion is relatively fast , to the plasma membrane , where diffusion is relatively slow ., Developing theories for computing appropriate rate constants under these conditions is an active area of research , and we did not attempt to provide a theoretical framework here ., Instead , we took an empirical approach , estimating effective second-order rate constants by fitting rate equations to the results of particle-based simulations of isolated reactions ( Fig 3; Figs I and M in S1 Text ) ., This approach allowed us to make fair comparisons between our particle-based and RDE simulation simulations , as evidenced by quantitative similarities in polarization ( Fig 9; Fig E in S1 Text ) and equivalent kinetics under non-polarizing conditions ( Figs B , K , and P in S1 Text ) ., Still , our empirical approach to estimating rate constants cannot capture the correct kinetics under all conditions: in general , a single rate constant is inappropriate for describing 2D diffusion-limited reactions 30 ., While this discrepancy presents challenges for comparing particle-based simulations to RDEs

Introduction, Results, Discussion, Models and methods

Polarity establishment , the spontaneous generation of asymmetric molecular distributions , is a crucial component of many cellular functions ., Saccharomyces cerevisiae ( yeast ) undergoes directed growth during budding and mating , and is an ideal model organism for studying polarization ., In yeast and many other cell types , the Rho GTPase Cdc42 is the key molecular player in polarity establishment ., During yeast polarization , multiple patches of Cdc42 initially form , then resolve into a single front ., Because polarization relies on strong positive feedback , it is likely that the amplification of molecular-level fluctuations underlies the generation of multiple nascent patches ., In the absence of spatial cues , these fluctuations may be key to driving polarization ., Here we used particle-based simulations to investigate the role of stochastic effects in a Turing-type model of yeast polarity establishment ., In the model , reactions take place either between two molecules on the membrane , or between a cytosolic and a membrane-bound molecule ., Thus , we developed a computational platform that explicitly simulates molecules at and near the cell membrane , and implicitly handles molecules away from the membrane ., To evaluate stochastic effects , we compared particle simulations to deterministic reaction-diffusion equation simulations ., Defining macroscopic rate constants that are consistent with the microscopic parameters for this system is challenging , because diffusion occurs in two dimensions and particles exchange between the membrane and cytoplasm ., We address this problem by empirically estimating macroscopic rate constants from appropriately designed particle-based simulations ., Ultimately , we find that stochastic fluctuations speed polarity establishment and permit polarization in parameter regions predicted to be Turing stable ., These effects can operate at Cdc42 abundances expected of yeast cells , and promote polarization on timescales consistent with experimental results ., To our knowledge , our work represents the first particle-based simulations of a model for yeast polarization that is based on a Turing mechanism .

Many cells need to generate and maintain biochemical signals in specific subcellular regions ., This phenomenon is broadly called polarity establishment , and is important in fundamental processes such as cell migration and differentiation ., A key polarity factor found in diverse organisms , including yeast and humans , is the protein Cdc42 ., In yeast , Cdc42-dependent polarization occurs through a self-reinforcing biochemical signaling loop ., Directional cues can guide polarity establishment , but interestingly , yeast can polarize in the absence of such a cue ., The mechanism thought to underlie this symmetry breaking involves the amplification of inhomogeneities in molecular distributions that arise from molecular-level fluctuations ., We investigated the effects of random fluctuations on polarization by performing particle-based simulations of the Cdc42 signaling network ., We found that fluctuations can facilitate polarization , allowing faster polarization , and polarization over a broader range of concentrations ., Our observations may help understand how polarity works in other systems .

cell physiology, reactants, simulation and modeling, cell polarity, fungi, model organisms, experimental organism systems, cellular structures and organelles, research and analysis methods, saccharomyces, chemistry, cell membranes, yeast, cytoplasm, biochemistry, biochemical simulations, eukaryota, cell biology, biology and life sciences, yeast and fungal models, saccharomyces cerevisiae, computational biology, physical sciences, chemical reactions, organisms

journal.pcbi.1002512

Mechanical Stress Inference for Two Dimensional Cell Arrays

Genetics and biochemistry are central to all aspects of biological function ., Physics is often less recognized yet also important at many levels , everywhere from intramolecular to organismal scales ., For example , many important aspects of cell behavior depend directly and indirectly on its mechanical state defined by its interaction with neighboring cells and adhesion to the extracellular matrix 1–3 ., Cytoskeletal mechanics and cell-cell adhesion determine geometric properties of cells 1 , 4–6 , as well as the dynamics of biological tissues 5 , 7–13 ., In plants , cells do not move , but the rigidity of cellulose membranes makes mechanical stress an obvious factor for cell division and proliferation 14 , 15 ., It is known that animal cell proliferation also depends on substrate adhesion and the degree of cell confinement 2 , 16–19 ., It has also been demonstrated that ( stem ) cell differentiation is affected by substrate rigidity 20 ., More speculatively , mechanical feedback interactions have been conjectured to have a role in coordination of growth during development 1 , 9 , 21 , 22 ., Mechanical transformation of epithelial tissue is of course itself central to many morphogenetic processes: gastrulation 7 and convergent extension 1 , to name a few ., Understanding how mechanical changes in cells orchestrate morphological reorganization of tissues is an open problem and a subject of much current work 1 , 7 , 8 Our present understanding of the role of mechanics as one of the regulatory inputs into the cell is strongly impaired by the difficulty of quantitatively characterizing the mechanical state ( i . e . stress and deformation ) of the cell ., Among the available techniques are laser tweezers 23 and “traction force microscopy” 18 , 24 performed on cultured cells ., UV laser ablation allows the mechanical perturbation of tissues 8 , 25 , 26 on the cellular scale with the time-lapse imaging of subsequent relaxation providing information on the mechanical state of the tissue ., The ablation approach is widely used on live preps , for example , in the study of Drosophila embryonic development ., Yet , this technique is definitely not a “non-destructive” one ., On the other hand one of the major recent technical advances in developmental biology is the improvement of live fluorescent imaging ., These provide high quality time lapse movies of developmental processes , including interesting morphological transformations such as gastrulation and convergent extension 7 , 26 , 27 ., The purpose of the present investigation is to explore what insight into the mechanical state of cells may be gleaned from a quantitative examination of high quality images of the type shown in Fig . 1A ., Our goal is to use image analysis as a non-destructive approach to obtaining quantitative measures of stress in these systems ., A similar strategy has been pursued by the recently proposed “Video Force Microscopy” ( VFM ) approach by Brodland et al 28 ., Our approach will differ from VFM in its assumptions about mechanical state of tissue , in the parameterization of forces and in the way imaging data is utilized ., Below we shall define a general model parameterizing the mechanical state of cells in two dimensional epithelial tissue and provide a computational method for inferring these parameters from the observed geometry of the cell array ., We shall study the sensitivity of the proposed Mechanical Inverse ( MI ) method to errors in measured cell geometry and identify conditions under which robust inference is possible ., We then illustrate the proposed MI method by applying it to the analysis of two different biological processes: cochlear neurogenesis 29 and ventral furrow formation27 ., Our approach is based on the assumption that epithelial monolayers are in an instantaneous mechanical equilibrium , characterized by a static balance of the forces acting at intercellular junctions ., The second important assumption is that epithelial mechanics is dominated by the actomyosin cortices and inter-cellular Adherens Junctions 1 both localized at cell boundaries , which form a visible two-dimensional web , as shown in Fig . 1A ., Thus we assume that the mechanical state of a cell can be described by an effectively two-dimensional model with tension at the interface and the hydrostatic pressure in the cell interior ., Yet , because cells can independently regulate their mechanical state , e . g . by modulating myosin activity or cell-cell adhesion , we allow for the possibility of each intercellular interface to have a different effective tension , , and for each cell to have a different internal pressure ( where labels cells and labels the interface between cells and ) , as shown in Fig . 1D ., Mechanical equilibrium then corresponds to the condition that the forces acting on each “vertex” ( defined as a junction of three cells and therefore of three interfaces ) add up to zero ., Let and be the vertices belonging to the interface and let be the vector from vertex to ., The force exerted by this interface on vertex is ( 1 ) where labels vector components in the plane and is the anti-symmetric tensor ( and ) ., As shown in Fig . 2 , this expression accurately represents the Young-Laplace balance between interfacial tension and the pressure differential across the interface , as long as the interfacial curvature is small ( see Supplementary Text S1 ) ., This fact enables us to formulate all mechanical balance conditions in terms of a polygonal approximation of the cell array , thus allowing us to reduce the problem to a generalized “vertex model” 8 , ., Remarkably , the forces given by ( 1 ) correspond to the mechanical energy in the form of the following simple Hamiltonian ( 2 ) where is the area of cell , is the length of the interface between cells and and denotes the set of interfaces belonging to cell ., Both and s are defined in the polygonal approximation ., This Hamiltonian is a generalization of the vertex models often used to describe epithelial sheet mechanics 8 , 9 , 26 ., Pressure and tension are defined by considering the differential form of : ( 3 ) where we define and ., The sum runs over all edges , i . e . pairs of neighboring cells , ., This tangent representation of mechanical energy expresses interfacial tension and intracellular pressure as conjugate variables to edge lengths and cell areas , respectively ., ( The reader will notice that strictly speaking our refers to a two-dimensional pressure which relates to the hydrostatic pressure only with the additional assumption that entails a change of cell volume . Alternatively may be thought of as the axial component of the three-dimensional stress tensor . ), Mechanical equilibrium means that is minimized with the respect to vertex positions ( 4 ) which defines the static force balance constraints ( the sum is over the vertices neighboring ) ., Our analysis will be based on the assumption that the cell layer is close to mechanical equilibrium in the sense of the magnitude of the net resultant force acting on vertices being much smaller than the average magnitude of the component forces that vectorially add to the resultant ., Stated in other words , we assume that the internal forces that balance each other in the ( approximate ) instantaneous mechanical equilibrium state are much larger than the unbalanced residual force that drives residual physical motion and ( through viscous effects ) defines its velocity ., More generally , the dynamics of passive relaxation towards this mechanical equilibrium would be described by , where is the “effective viscosity” and the sum is again over the vertices that neighbor ., In principle , given that vertex velocities can be directly measured by time-lapse microscopy in live tissues , can be obtained directly from the experiment , allowing a straightforward extension of the method described below toward the VFM method 28 ( where viscous forces were assumed to dominate ) ., We can now inquire to what extent the knowledge that a given cell array geometry is in a mechanical equilibrium constrains the parameters , describing the mechanical state of cells ., We proceed by a simple count of mechanical constraints and of the free parameters for two cases, i ) a closed cell array , shown in Fig . 3A and, ii ) an open cell array , shown in Fig . 3B ., Let us begin with the closed cell array where are the total number of vertices , edges , and cells , respectively ., For vertices in two dimensions , we have exactly mechanical constraints , where the extra three degrees of freedom are associated with global translation and rotation symmetries ( alternatively , three constraints are redundant because the total force and total torque in the closed system are equal to zero ) ., On the other hand , the number of unknown tension parameters is , and the number of unknown pressures is , so that the total number of parameters is ., Our closed system , if we count the exterior as an additional “cell” , is topologically equivalent to a sphere so that Eulers theorem reads ( 5 ) Combining this relation with the condition that vertices are points where three edges meet and each edge impinges on two vertices , that is , we obtain the result ( 6 ) This implies , which means that our unknown parameters can be determined up to four free constants ., One of the latter is the arbitrary overall scale of and which cannot be constrained by the force balance conditions ( note that since is only defined up to an additive constant , one can set the pressure in the exterior of the domain to zero ) ., Yet the good news is that the number of free constants is finite , while the number of nontrivial constraints scales with the number of cells !, This counting argument can be readily generalized ( see SI ) to the case where a fraction of vertices has more than three incoming edges: the so called “rosettes” that can be quite common in certain tissues 30 ., Repeating the counting procedure for the open system , one finds that , where is the number of cells at the boundary of the domain ., It follows that ., Thus mechanical parameters are determined up to free constants: we can still choose the overall scale while the additional degrees of freedom may be regarded as the boundary conditions such as s of the cells at the edge of the domain ., Again , for a large array , because while , the number of parameters and constraints is much larger than the number of free constants ., To actually determine the and parameters , we use the fact that they appear only linearly in the force balance equations ( 4 ) ., This results in a linear system for ( 7 ) in the form ( 8 ) with being an matrix where the 1st rows impose the force balance conditions from equation ( 4 ) , and the additional row imposes the scale ., This is performed by constraining the average tension to be equal to one ., Correspondingly the top entries of the column vector are zero , while the bottom row ., The rectangular system ( 8 ) is solved via a pseudo-inverse 31 with the general solution of the form ( 9 ) with ( 10 ) ( 11 ) where is the pseudo-inverse of the rectangular matrix and the free parameters , , are the amplitudes of the “zero modes” ., Additional details regarding the formulation and solution of the inverse problem are provided in the Supplementary Text S1 ., Fixing the remaining degrees of freedom requires introducing additional constraints: e . g . one may have reasons to seek a solution that minimizes variation of s or s ., In choosing such additional assumptions one may want to use all the information that one has for specific applications , as we shall do below ., However , before proceeding to the applications we must consider the issue of error sensitivity ., Our approach to mechanical parameter inference is based on the observed geometry of the cell array ., How sensitive are the results to the inaccuracy of vertex positions ?, Such inaccuracies will inevitably arise in the process of imaging , image segmentation , and more importantly from the fact that the cells themselves fluctuate ., ( These fluctuations are of course related to the fact that mechanical equilibrium is itself at best approximate . ), To quantify the stability of the inverse we consider the effect of an arbitrary small perturbation in vertex positions , ., Since the inhomogeneous term in ( 8 ) is independent of cell geometry , the first order response of the parameters to positional error is given by ( 12 ) ( 13 ) Ideally the error response matrix would have small eigenvalues providing a relatively robust inverse ., On the other hand , large eigenvalues of would indicate high error sensitivity ., These sensitive modes appear via the pseudoinverse matrix ., A histogram of singular values of the matrix is shown in blue in Fig . 4 ( for a closed system with cells ) ., One notes that a substantial fraction of modes have eigenvalues larger than one ., As a result , small errors in positions can result in large error in inferred parameters ., The simplest way to solve the sensitivity problem is to reduce the number of parameters ., For example , as we shall argue below , in some contexts it may be reasonable to neglect variation in cell pressure and set which eliminates parameters , reducing from to ., In that case the mechanical constraint system given by ( 8 ) becomes overdetermined and can be solved only in the sense of least square minimization: i . e . minimization of ( 14 ) The solution of the minimization problem is still given by the pseudo-inverse of the rectangular matrix which extends the force balance matrix by including additional ( linear ) equations that constrain ., Fig . 4 shows ( in red ) the distribution of singular values governing the sensitivity of the reduced or partial inverse problem ., We note a substantial reduction in sensitivity ., The partial inverse approach is then tested in silico ., To that end we consider a closed array of cells and define cell geometry by minimizing elastic energy given by ( 15 ) with uniformly distributed ., The absence of area terms imposes a constant pressure ., ( The closed cell array is relaxed under toroidal boundary conditions to prevent a collapse into the zero tension ground state . ), The vertex model parameters are computed via equation ( 2 ) ., These quantities are then compared to values obtained by applying the partial inverse algorithm to the vertex “data” corrupted by random noise with an r . m . s . variation of 5% of the average length of cell edge ( see Fig . 5 ) ., The correlation coefficient between inferred and computed parameters is 0 . 85 , which confirms the ability of our method to extract information from noisy data ., The Supplementary Figure S4 shows results under a 10% random corruption in vertex positions ., In that case the correlation coefficient is reduced to 0 . 65 , which remains serviceable ., We note that the”soft modes” that contribute to the sensitivity of the full inverse problem are quite interesting ., The formulation of the minimally constrained problem is analogous to the isostatic systems studied in jamming transitions of amorphous solids 32 ., These isostatic systems live on the boundary of Maxwells criterion for rigidity , and much like amorphous solids they must satisfy both the local and global rigidity conditions ., In our mechanical inverse formulation , “rigidity” corresponds to a fully constrained set of mechanical ( and ) parameters ., Amusingly , local soft modes for the MI problem correspond to special local geometries: specifically , polygons that can be inscribed into circles ( i . e . a generalization of regular polygons ) - a category which includes triangles of any shape ., These interesting mathematical aspects of the problem will be discussed in a separate publication ., In cochlear development , which takes place during the first two weeks of chick embryonic development , cells in an initially homogeneous two dimensional epithelial layer differentiate into pro-neural ( hair-cell ) and support cell fates 29 ., The process is driven by Delta/Notch-mediated cell-contact signaling 33 , which causes lateral inhibition: cells which express Delta ligand on their surface prevent their immediate neighbors from doing the same ., Expression of Delta is an early marker of the pro-neural fate of cells ., Fig . 1A , B presents a micrograph of the cochlea epithelium , obtained by Goodyear and Richardson 29 at the stage of development shortly after the onset of differentiation ., The images in Fig . 1A , B were obtained as described in 29 using a double fluorescent antibody labeling: antibody to the tight junction protein cingulin allowing visualization of cell boundaries and 275 kDa hair-cell antigen staining labeling pro-neural cells ., Note that the two cell types already have discernibly different morphology: pro-neural cells are somewhat smaller and have curved edges ., This dimorphism is supported by direct labeling of specific pro-neural markers , shown in Fig . 1B and demonstrated in 29 ., Our goal is to infer , based on the analysis of the image in Fig . 1A , the variation in the mechanical parameters between cells ., The visible positive curvature associated with pro-neural cells suggests that they are under higher internal pressure ., Can the Mechanical Inverse method determine pressure differentials between cells ?, Because our approach requires only positions of cellular vertices , it does not use the information provided by the interfacial curvatures which are readily measurable on the image ., This additional information will be used as an a posteriori validation of the inferred results ., To reduce the number of parameters , we assume that interfacial tensions can be expressed as in terms of constant cortical tensions , of adjacent cells which reduces the number of parameters by ., This is sufficient to render a robust partial inverse ( in the sense of least squares ) , yielding and for every cell ., Fig . 6 shows the distribution of inferred intracellular pressures and cortical tensions for the two cell types ., We see that pro-neural cells have on average higher tension and pressure ., While pressure shows some correlation with cell area , there is no correlation between interfacial tension and its length ., However , there is no reason to expect any specific correlation between these quantities ., On the other hand , Laplaces Law predicts which we are in a position to check directly , thanks to the fact that interfacial curvatures are directly measurable on the images such Fig . 1A ., Fig . 7 presents the “empirical” Laplaces Law obtained on the basis of the inferred and ., Because the Mechanical Inverse algorithm did not in any way use the interfacial curvature information , the fact that inferred parameters approximately obey the Laplaces Law provides a validation of the inverse method ., During the initial stage of development a Drosophila embryo is comprised of an ellipsoidal monolayer of cells ., The first step toward more complex morphology that is achieved after gastrulation , is the formation of a ventral furrow that begins with the contraction of the apical surfaces of cells along the ventral midline 27 , 34 , 35 ., Fig . 8A presents the ventral view of a Drosophila embryo at the beginning of this mechanical transformation ., The high quality of these images ( kindly provided by the Weischaus lab 27 ) makes it possible to attempt the Mechanical Inverse analysis ., Since the process begins even before cellularization is completed it is reasonable to assume that cells have the same internal pressure , allowing us to reduce the number of parameters enough to achieve a robust partial inverse and infer for every cell boundary ., Note that in contrast to the developing avian cochlea , cell-cell interfaces exhibit little curvature and ( apical surfaces of ) cells are well approximated by polygons ( see in Fig . 8A ) , which is consistent with pressure differentials being weak compared to interfacial tensions ., Interestingly , comparing images separated by merely two minutes ( Fig . 9 ) we found that the inferred at the later time-slice exhibited statistically significant anisotropy with estimated tensions of cell interfaces along the AP axis being on average about 15% higher than those along the DV axis ., The inferred increase in AP tension ( relative to DV ) is consistent with the laser ablation measurements made in the Wieschaus lab 7 , 27 ., Yet , mechanical inverse inference gives information not only on the global , tissue-wide level , but also on the scale of a single cell and interface ., The analysis also clearly demonstrates the ability to make specific predictions ( for interfacial tensions ) that can be directly tested by combining high quality live imaging with UV pulsed laser ablation ., The variation of tension from one interface to another implies the existence of traction forces acting between cells ., This traction , or shear stress , must be entirely borne by the cadherins and other cell adhesion molecules which bridge cellular membranes and connect actomyosin cortices of apposing cells 1 ., In Fig . 10 we zoom in on an interface decomposing interfacial tension into the cortical tensions on the opposite sides of the interface , now allowing for the possibility that the latter are not constant along the interface and vary as a function of position along the edge ., This transfer of tension from the cortical bundle in one cell to the other is possible because of cadherin mediated traction forces acting between cells ., The total shear stress on the interface is ., In the Supplementary Text S1 we show that because cortical tensions are constrained by the continuity conditions at cell “corners” they can be readily expressed in terms of interfacial tensions leading to the following simple expression for the traction force acting between cells and ., ( 16 ) Fig ., 8B shows inferred tractions calculated for the ventral furrow data taken two minutes prior to invagination ., We observe a significant variability in tractions at different interfaces ., Because traction forces stretch trans-cellular cadherin dimers , they may be physiologically important ., Since at present there is no way of measuring them directly the possibility of indirect inference is particularly interesting ., We have demonstrated that the readily visualized two dimensional network of cellular interfaces in an epithelial tissue holds , potentially , a wealth of information on the relative strength of mechanical stresses acting in the tissue ., The main precondition is that the tissue is close to the mechanical equilibrium in which internal cytoskeletal forces are balanced by intercellular interactions ., Any imbalance of forces corresponding to directed or fluctuating motion must be small in comparison to the magnitude of internal forces that balance each other in mechanical equilibrium ., Force balance is achieved by the suitable adjustment of cell geometries ( parameterized by the positions of vertices ) ., Conversely we envision changes in tissue geometry to be driven adiabatically - i . e . without disruption of the mechanical equilibrium - by changes in cytoskeletal forces within cells ., This picture is at once similar and dissimilar to the case of soap froths ., The geometry of a soap froth 36–38 is also defined by the instantaneous force balance and changes adiabatically ( when gas diffuses out of cells with higher internal pressure ) ., Yet epithelial cells , in contrast to soap bubbles , can control interfacial tension by regulating myosin activity within actomyosin cortices and therefore can generate variation in tension on sub-cellular scale , even between different interfaces of the same cell ., Our Mechanical Inverse method is fundamentally different from the Video Force Microscopy 28 ., In contrast to our assumption that cytoskeletal forces are in an approximate instantaneous balance , VFM is based on the assumption that bulk forces acting within the tissue are balanced by viscosity ., It is therefore based on the observed velocity of tissue motion and employs finite element methods to define forces on a computational grid rather than the underlying cells ., The two methods are complementary in the sense that VFM provides information about the distribution of unbalanced bulk force which drives motion on the scale of the embryo , while our Mechanical Inverse is focused on the internal balance of forces in relation to cell geometry and its local changes ., Our approach can be extended to include measured velocities which can be used to define net forces on the vertices , as explained below Eq ., ( 4 ) , leading to a modified inverse problem ., This generalization would bridge the static inference presented here with the VFM approach ., Yet , to the extent that the dynamics of normal epithelial cell rearrangement unfolds relatively slowly ( on the time scale of minutes ) compared to the rapid ( time scale of seconds ) viscosity limited retraction of laser ablated interfaces , it is reasonable to assume that the contribution of viscous forces during slow normal developmental dynamics is small compared to the balancing internal forces , which is the assumption underlying our Eqn ., ( 4 ) ., Recent experiments have demonstrated that actomyosin structures transiently assembling on the apical or basal surfaces of the cell , play an active role in defining its mechanical state 7 , 26 , 39 ., In particular , 7 and 26 argue that coalescing pulses of “medial myosin” on the apical surface drive a ratchet of apical surface contraction ., Presently , our mechanical model does not explicitly incorporate such effects , which in full generality would require introduction of many more parameters ( characterizing intracellular heterogeneity and anisotropy ) ., On the other hand , these effects are not observed in all epithelial tissues at all times , leaving the present approach with many possible applications ., Furthermore it may be possible to generalize our approach to model medial myosin as well , especially if additional information from cell imaging is used ., For example , during convergent extension investigated in 26 one often observes intracellular medial myosin filaments attaching to the lateral cortex and causing measurable deformation of cell-cell boundary ., It that case it may be possible to define an additional “vertex” corresponding to the attachment point , apply considerations of mechanical balance discussed above and obtain an estimate of the force applied by the medial myosin as compared to the cortical tension ., Alternatively , when medial actomyosin structures appear to be isotropic , their effect may be well approximated by a uniaxial stress which is already parameterized already by our existing model ., Studying the effect of medial myosin would be an interesting direction for future work ., The proposed Mechanical Inverse method converts clearly stated assumptions about the nature of cellular stresses into readily falsifiable predictions ., Using the example of avian cochlea , we were able to demonstrate that mechanical parameters inferred via the Mechanical Inverse satisfy non-trivial cross-checks provided by independent additional information ( interfacial curvature measurements ) read off the tissue images ., Thus our approach is capable , in realistic applications , of inferring mechanical parameters and to uncover interesting aspects of the internal state of the cell ., By combining high quality live imaging with UV pulsed laser ablation , one will be able to put predictions for local interfacial tensions obtained via the Mechanical Inverse , to a rigorous experimental test ., We note however , that the predictions do not have to be very accurate to be useful ., Even if inferred tensions each carry only a single bit of information - i . e . identify interfaces with high or low tension - correlating tension with the observed level of myosin , cadherin and/or other proteins involved in regulation of cell mechanics could be extremely informative ( in addition , since a large number of cells can be imaged and analyzed , the method is effectively “high throughput” ! ) ., Finally , our approach allows for inference of quantities such as inter-cellular traction forces ( or shear stress ) , which may be important for the stability of Adherens Junctions but cannot be directly measured by any means presently available ., Future development of FRET based molecular sensors of stress 40 may nevertheless make such measurements possible in the future ., Hence we expect that further development , validation , and application of the Mechanical Inverse method will lead to new insights into the molecular biology of epithelial cells and tissues .

Introduction, Materials and Methods, Results, Discussion

Many morphogenetic processes involve mechanical rearrangements of epithelial tissues that are driven by precisely regulated cytoskeletal forces and cell adhesion ., The mechanical state of the cell and intercellular adhesion are not only the targets of regulation , but are themselves the likely signals that coordinate developmental process ., Yet , because it is difficult to directly measure mechanical stress in vivo on sub-cellular scale , little is understood about the role of mechanics in development ., Here we present an alternative approach which takes advantage of the recent progress in live imaging of morphogenetic processes and uses computational analysis of high resolution images of epithelial tissues to infer relative magnitude of forces acting within and between cells ., We model intracellular stress in terms of bulk pressure and interfacial tension , allowing these parameters to vary from cell to cell and from interface to interface ., Assuming that epithelial cell layers are close to mechanical equilibrium , we use the observed geometry of the two dimensional cell array to infer interfacial tensions and intracellular pressures ., Here we present the mathematical formulation of the proposed Mechanical Inverse method and apply it to the analysis of epithelial cell layers observed at the onset of ventral furrow formation in the Drosophila embryo and in the process of hair-cell determination in the avian cochlea ., The analysis reveals mechanical anisotropy in the former process and mechanical heterogeneity , correlated with cell differentiation , in the latter process ., The proposed method opens a way for quantitative and detailed experimental tests of models of cell and tissue mechanics .

Mechanical forces play many important roles in cell biology and animal and plant development ., In contrast to inanimate matter , forces in living matter are generated by active and highly regulated processes within and between cells ., The ability to directly measure forces and mechanical stress on the cellular scale within living tissues is critically important for understanding many morphogenetic processes but is a serious experimental challenge ., The present work proposes an alternative approach based on the analysis of images that provide a visualization of cell boundaries in two dimensional epithelial tissues ., The method uses the assumption of force balance within the epithelial layer to infer , on the basis of image-derived geometric data , the mechanical state of each cell ., The proposed Mechanical Inverse method is illustrated on the analysis of two examples: the initial step of the gastrulation process in the Drosophila embryo , and the process of neurogenesis in the developing avian cochlea .

physics, biology, biophysics

journal.pntd.0001335

Imaginal Discs – A New Source of Chromosomes for Genome Mapping of the Yellow Fever Mosquito Aedes aegypti

Ae ., aegypti is a principal vector for yellow fever , dengue and chikungunya viruses 1 , 2 ., These diseases have a significant worldwide impact on human health ., Yellow fever affects up to 600 million lives and is responsible for about 30 , 000 deaths annually 3 ., Dengue fever is a threat to >2 . 5 billion people in tropical and subtropical regions , where between 50 to 100 million infections occur each year 2 , 4 , 5 ., The incidence of dengue fever is increasing globally 6 , for example in developed areas like Singapore where dengue was thought to be well-controlled 7 and is a growing threat to the United States 8 ., Despite all control campaigns , Ae ., aegypti has expanded its range to most subtropical and tropical regions during the last several decades ., This mosquito prefers to feed on humans and breeds in areas that humans inhabit 9 ., To facilitate the development of genome-based strategies for mosquito control , genomes for three major disease vectors--the African malaria mosquito Anopheles gambiae , the southern house mosquito Culex quinquefasciatus , and the yellow fever mosquito Ae . aegypti--have been sequenced 10 , 11 ., Among genomes of these three species , the genome of Ae ., aegypti is the largest 11 ., The draft genome sequence consists of 1 , 376 million base pairs , which is ∼5 times larger than the An ., gambiae genome 10 and ∼2 times larger than the Cx ., quinquefasciatus genome 12 ., About half of the genome consists of transposable elements ., The genome shows “short period interspersion” meaning that , in general , ∼1–2 kb fragments of unique sequences alternate with ∼0 . 2–4 kb fragments of repetitive DNA 13 ., Abundance of repetitive elements in the genome leads to low levels of replication and poor spreading of polytene chromosomes of Ae ., aegypti 14 , 15 ., The yield of chromosome preparations useful for cytogenetic studies was only 0 . 5% for salivary glands 15 ., At the same time , the large size of the genome makes mitotic chromosomes of this mosquito large and easily identifiable ., The average size of the biggest metaphase chromosome in Ae ., aegypti was estimated as 7 . 7 µm 16 , which is bigger than the average sizes of human metaphase chromosomes and comparable with the size of the human chromosomes at prometaphase 17 ., The average size of the biggest human chromosome at prometaphase was estimated as 9 . 24 µm ., Most of the classical cytogenetic studies on Ae ., aegypti undertaken in the past were performed on mitotic or meiotic chromosomes from larval brain or male testis 18 , 19 , 20 ., It has been demonstrated that Ae ., aegypti has a karyotype typical to that found in other mosquitoes and includes three pairs of chromosomes ., These chromosomes were originally designated as chromosomes I , II , and III in the order of increasing size 18 ., However , later chromosomes were renamed in accordance with Ae ., aegypti linkage groups as chromosomes 1 , 2 , and 3 21 ., Chromosome 1 was described as the shortest metacentric chromosome; chromosome 2 as the longest , also a metacentric chromosome; and chromosome 3 as a medium-length submetacentric chromosome with the secondary constriction on the longer arm ., However , precise measurement of the centromeric index made on spermatagonial metaphase chromosomes has indicated that all Ae ., aegypti chromosomes fall into the category of metacentric chromosomes according to the standard classification 22 , 23 ., Unlike the anophelines , the sex chromosomes are homomorphic in all culicine mosquitoes , including Ae ., aegypti 18 ., The sex determination alleles were linked to chromosome 1 and described as Mm in males and mm in females 24 ., M . Motara and K . Rai proposed to name sex chromosomes as “m” and “M” chromosomes for female and male determining chromosomes , respectively 20 ., However , it was also popular to refer to sex chromosomes in Aedes as “X” and “Y” 19 ., The precise measurement of the sex chromosomes in males and females has indicated that the female chromosome 1 is slightly bigger in size 22 ., The C-banding technique has also demonstrated differences between male and female sex chromosomes in Ae ., aegypti 20 ., Typically females have pericentromeric and additional distinct intercalary bands on both chromosomes 1 which are absent on the putative male determining sex chromosome ., C-banding has been found to be variable in different strains of Ae ., aegypti ., For example , an intercalary band can be present on the male chromosome in some strains , and intercalary bands may be differ in size in females 25 , 26 ., The silver staining technique 26 and in situ hybridization of 18S and 28S ribosomal genes 27 indicated the location of ribosomal locus on both sex chromosomes of Ae ., aegypti ., The genetic mapping of the Ae ., aegypti genome has been conducted in parallel with cytogenetic studies ., An early genetic map included about 70 morphological , insecticide-resistance and isozyme markers 28 ., Later , additional genetic maps were developed using restriction fragment-length polymorphism ( RFLP ) markers , random-amplified polymorphic DNA ( RAPD ) loci , single-strand conformation polymorphism ( SSCP ) , and single-nucleotide polymorphism ( SNP ) markers 29 , 30 , 31 ., A composite map for RFLP , SSCP , and SNP markers incorporated 146 loci and covered 205 cM 13 ., These maps provided the tools to localize a number of quantitative trait loci ( QTLs ) related to the mosquitos ability to transmit the filarioid nematode Brugia malayi 32 , the avian malaria parasite Plasmodium gallinaceum 33 , 34 , and dengue virus 35 , 36 ., Advent of the fluorescent in situ hybridization technique allowed mapping of BAC clones , cosmids , and cDNA probes on mitotic chromosomes from the ATC-10 cell line of Ae ., aegypti 16 ., The chromosome positions of these clones were measured by FLpter: a fractional length from the short arm telomeric end p-terminus ., The physical map was integrated with the genetic map by the direct placing cDNA genetic markers that contained the RFLP marker sequence to the chromosomes 37 ., Nevertheless , molecular cytogenetic studies on Ae ., aegypti mitotic chromosomes remain challenging ., The current physical map has relatively low resolution and includes ∼180 markers 11 ., Only ∼31% of the Ae ., aegypti draft genome assembly has been placed to chromosomes , but without order and orientation ., In contrast , a physical map of the malaria vector An ., gambiae includes more than 2000 markers and covers about 88% of the genome 10 , 38 ., Successful physical mapping for any organism relies on a robust source of high-quality , easily obtainable chromosome preparations ., Recently we discovered that imaginal discs ( IDs ) of 4th instar larva can be an excellent source for a high number of large , easily spreadable , banded chromosomes ., In this study , we optimized all cytogenetic procedures required for the successful in situ hybridization ., Idiograms for each individual chromosome at the metaphase stage have been developed ., Based on the banding pattern , 10 BAC clones and a 18S rDNA probe were mapped to their precise chromosomal positions ., We propose to use this new cytogenetic tool for the detailed physical mapping of the Ae ., aegypti genome ., In this study , we used the Liverpool strain , a parental strain for the Liverpool IB-12 strain , which was used for sequencing the Ae ., aegypti genome 11 ., Eggs were hatched at 28°C , and after several days , 2nd or 3rd instar larvae were transferred to16°C to obtain a high number of mitotic divisions in IDs ., For in situ hybridization and idiogram development , slides were prepared from 4th instar larvae of Ae ., aegypti ., Before dissection , larvae were placed on ice for several minutes , then transferred to a slide with a drop of cold hypotonic solution ( 0 . 5% sodium citrate ) , and after that dissected under a Olympus SZ microscope ( Olympus America , Inc . , Melville , NY , USA ) ., Larvae were decapitated , and cuticle from the ventral side of the larval thorax was slightly cut by dissecting scissors ( Fine Science Tools , Foster City , CA , USA ) ., The cuticle was opened to expose the IDs to treatment in hypotonic solution for 10 min ., Hypotonic solution was removed using filter paper , and larvae were treated with Carnoys solution ( ethanol/glacial acetic acid in 3∶1 ratio ) for 1 min ., After Carnoys application , IDs immediately turned white and became easily visible under the microscope ., Using dissecting needles ( Fine Science Tolls , Foster City , CA , USA ) , IDs were isolated from larvae , transferred to another slide in a drop of 50% propionic acid , and covered with a 22x22-mm cover slip ., After 10 min of propionic acid treatment , IDs were squashed and briefly analyzed using an Olympus CX41 microscope ( Olympus America , Inc . , Melville , NY , USA ) at ×200 magnification ., Slides suitable for in situ hybridization , which had >50 chromosome spreads , were then placed in liquid nitrogen , and cover slips were removed ., Slides were dehydrated in a series of ethanol ( 70% , 80% , 100% ) and air dried ., The percentage of the slides suitable for in situ hybridization was >90% ., For the analysis of mitosis dynamics in IDs and brain ganglia , larvae were fixed in Carnoys solution ( ethanol/glacial acetic acid in 3∶1 ratio ) ., After 24 hours , IDs and brain ganglia were dissected from larvae and squashed in 50% propionic acid ., Small drops of lactic acid were placed on each corner of the cover slip to prevent slides from drying ., Slides were analyzed under the Olympus CX41 microscope at x400 magnification ., BAC clone DNA was isolated using the Qiagen Large Construct kit ( Qiagen Science , Germantown , MD , USA ) ., BAC-DNA was labeled by nick translation ., Each reaction mix contained: 1 µg of DNA; 0 . 05 mM each of unlabeled dATP , dCTP , and dGTP , and 0 . 015 mM of dTTP ( Fermentas , Inc . , Glen Burnie , MD , USA ) ; 1 µl of Cy3 or Cy5 dUTP ( GE Healthcare UK Ltd , Buckinghamshire , UK ) ; 0 . 05 mg/ml of BSA ( Sigma , St . Louis , MO , USA ) ; 5 µl of 10x nick translation buffer; 20 u of DNA polymerase I ( Fermentas , Inc . , Glen Burnie , MD , USA ) ; and 0 . 0012 u of DNAse I ( Fermentas , Inc . , Glen Burnie , MD , USA ) ., DNA polymerase/DNAse ratio was selected empirically to obtain the probe with the size range from 300 to 500 bp ., To obtain a C0t1 DNA fraction , the genomic DNA was isolated from adult Ae ., aegypti mosquitoes using a blood and cell culture maxi kit ( Qiagen Science , Germantown , MD , USA ) ., DNA was digested by DNAse I with a concentration 0 . 0002 u/µl ( Fermentas , Inc . , Glen Burnie , MD , USA ) to obtain fragments <100 bp ., After that , DNA was denatured at 97°C for 10 min , and DNA fragments were allowed to reassociate in TE buffer for 1 hour at 37°C ., Then single-stranded DNA was digested using S1 nuclease ( Invitrogen Corporation , Carlsbad , CA , USA ) with a concentration of 2 . 58 u/µl for 15 min at 37°C ., Double-stranded C0t1 DNA fraction was collected by standard ethanol precipitation for further application ., Fluorescent in situ hybridization ( FISH ) was performed using a standard protocol 39 ., Slides were pretreated with 0 . 1 mg/ml of pepsin ( USB corp . , Cleveland , Ohio ) for 5 min at 37°C; denatured in deionized 70% formamide in 2xSSC at 72°C for 5 min; and dehydrated in an alcohol series ( 70% , 80% , and 100% ) for 5 min each ., Hybridization mix contained 50% formamide , 10% dextran sulfate ( Sigma , St . Louis , MO , USA ) , 200 ng of each probe per slide , and 4 µg of C0t1 DNA fraction ., To eliminate nonspecific hybridization to the chromosomes , the probe was prehybridized with C0t1 fraction in a tube at 37°C DNA for 30 min ., After that , the final 10 µl volume of hybridization mix per slide was overlaid with a 22x22 cover slip and glued by rubber cement ., Hybridization on the slide was performed at 37°C in a dark humid chamber over night ., Afterward , the slides were washed in a Coplin jar with 0 . 4x SSC , 0 . 3% Nonidet-P40 at 72°C for 2 min , and in 2x SSC , 0 . 1% Nonidet-P40 at RT for 5 min ., Slides were thereafter counterstained using 1 µM YOYO-1 iodide solution ( Invitrogen Corporation , Carlsbad , CA , USA ) in 1× PBS for 15 min and enclosed under antifade Prolong Gold reagent ( Invitrogen Corporation , Carlsbad , CA , USA ) by a cover slip ., Slides were analyzed using a Zeiss LSM 510 Laser Scanning Microscope ( Carl Zeiss Microimaging , Inc . , Thornwood , NY , USA ) at ×1000 magnification ., To develop idiograms , the best images of the chromosomes stained with YOYO-1 were selected ., The colored images were inverted in black and white images and contrasted in Adobe Photoshop as described before 40 ., The chromosomal images were straightened using ImageJ program 41 and were aligned for comparison ., In total , 150 chromosomes at various stages of condensation were analyzed ., The sizes of IDs were measured using an SZ dissecting microscope ( Olympus America Inc . , Melville , NY , USA ) ., The lengths of the chromosomes were measured using Zen 2009 Light Edition software 42 ., The statistic analysis was performed using the JPM8 software program at 95% confidence intervals Heiberger 43 ., To obtain polytene chromosomes for cytogenetic analysis of Ae ., aegypti , we have screened several tissues from different developmental stages including 4th instar larvae , pupae , and adults ., Polytene chromosomes were found in salivary glands , Malpighian tubules , and ovarian nurse cells ., However , polytene chromosomes had poor banding patterns and formed multiple ectopic contacts in all examined tissues ., To improve the quality of the polytene chromosomes , we maintained the larval stages at 16°C ., Reduced rearing temperature was effectively used to improve the quality of the polytene chromosome in salivary glands of Culex pipiens 44 ., In our study , we did not detect any such improvement in the polytenization level or chromosome structure in Ae ., aegypti ., Finally , we confirmed that polytene chromosomes in Ae ., aegypti are not suitable for the physical mapping of the genome ., In addition to polytene chromosomes , we analyzed mitotic chromosomes from IDs and brain ganglia ., Six IDs , which will develop into legs at the adult stage , are located right under the cuticle on the ventral side of the thorax in larva ( Fig . 1 ) ., Although IDs become visible under the dissecting microscope from the 2nd instar larval stage , the best stage for the chromosome preparation is 4th instar larvae when IDs start to develop into legs and accumulate large numbers of mitotic divisions ., IDs at different stages of their development are shown in Fig . 1 ., The size of IDs in 4th instar larvae ranged from 0 . 1 to 0 . 8 mm ., Fig . 1D represents overdeveloped IDs , which are not suitable for slide preparation because of the abundance of already differentiated tissues ., In this study , the number of mitotic divisions per slide was compared between:, 1 ) IDs of two sizes--0 . 1–0 . 25 mm and 0 . 3-0 . 45 mm ( Fig . 2A ) ;, 2 ) IDs from larvae reared at 28°C and 16°C ( Fig . 2B ) ; and, 3 ) one ID and two brain ganglia ( Fig . 2C ) ., The largest number of mitotic divisions ( ∼175 ) was detected in IDs with an oval shape and length of 0 . 3–0 . 4 mm ( Fig . 1A , B ) ., The 16°C temperature stimulated the accumulation of ∼1 . 5 times higher number of mitotic divisions per slide as compared to the normal temperature ( Fig . 2B ) ., Finally , our comparison indicated a ∼6 fold difference in number of mitotic divisions between one ID and two brain ganglia ( Fig . 2C ) ., This parameter is extremely important for utilizing chromosome preparations for successful in situ hybridization ., The major phases of mitosis in IDs of Ae ., aegypti are shown in Fig . 1: prophase ( A-C ) prometaphase ( D ) ; metaphase ( E ) and anaphase ( F ) ., The interesting feature , which characterizes mitosis in Ae ., aegypti , is that homologous chromosomes have strong somatic synapsis during interphase and stay paired up to early metaphase ( Fig . 3A-D ) ., As a result of chromosomal pairing , only three separate chromosomes can be detected in all cells at the early mitotic stages ., At metaphase , homologous chromosomes finally segregate from each other , and the visible number of chromosomes becomes equal to 6 ( Fig . 3E ) ., The synapsis of the homologous chromosomes in Aedes cells has been described before 45 ., Prometaphase and metaphase chromosomes ( Fig . 3D , E ) are the most abundant in IDs ( ∼42% ) and easily identifiable by their relative lengths and morphological characteristics ., Long prophase chromosomes ( Fig . 3A-C ) , which are present in IDs at the level of ∼35% , are convenient for the mapping and orientation of relatively short scaffolds with sizes ∼1 Mb ., Thus , ∼77% of all chromosome spreads on the preparations of squashed IDs can be utilized for the cytogenetic analysis and the physical mapping of Ae ., aegypti genome ., Another important feature of the mitotic chromosomes in IDs of Ae ., aegypti is a clearly visible and reproducible banding pattern that can be used for developing idiograms--the diagrammatical representation of the chromosomes ., In this study , idiograms for mid-metaphase chromosomes , the most convenient stage for chromosome recognition , have been developed ., To calculate the correct proportion of the idiograms , chromosomes were measured using Zen2009 Light Edition software 42 ., The results of these measurements are summarized and compared with previous data in Table 1 ., The average lengths of the chromosomes were 7 . 15 µm , 9 . 46 µm , and 8 . 36 µm for chromosomes 1 , 2 , and 3 , respectively ., The relative lengths of the chromosomes were 28 . 48% , 37 . 93% , and 33 . 39% ., Centromeric indexes ( the relative length of the p-arm ) were 46 . 92% , 48 . 61% , and 47 . 42% , respectively , for chromosomes 1 , 2 , and 3 ., Therefore , all three chromosomes should be considered as metacentric regarding current chromosomal nomenclature 23 ., The average lengths of the chromosomes from IDs at the metaphase stage were just slightly ∼0 . 8 µm bigger than that from ATC-10 line 16 ., The relative lengths of the chromosomes were found to be very similar to the chromosomes from brain 18 , spermatogonia 22 , and ATC-10 line 16 ., Interestingly , the centromeric indexes in our study were more similar to that from brain and spermatogonia than to the cell line ( Table 1 ) ., Fig . 4 shows the major steps of the idiogram development ., The images of the YOYO-1 stained chromosomes ( Fig . 4A ) were converted in black and white images ( Fig . 4B ) and further contrasted in Adobe Photoshop 40 to obtain clear banding patterns ., After that , chromosomes were straightened using Image J program plug-in 41 and aligned to each other for the pattern comparison ., In total , 150 chromosomes were analyzed ., Chromosomal arms were first determined by FISH of the BAC clones with known chromosomal positions ( Fig . 5 ) ., These BAC clones contained genetic markers previously genetically mapped to the chromosomes 46 ., Based on the human cytogenetic nomenclature , we determined bands with 4 different intensities – intense , medium intensity , low intensity , and negative 47 ., The total number of bands per three chromosomes at mid metaphase was equal to 78 ., The following regions can be used as cytogenetic landmarks for the chromosomal arm recognition: intense band in the middle of the 1q arm , intense double band in the 2q arm , and 2 low intense bands in the area next to the telomeric band on the 3q arm ( Fig . 4 ) ., These regions have consistent distinct morphology and can be easily utilized for the chromosomal arm recognition ., To test the reliability of chromosomal banding patterns for physical mapping , 10 BAC clones ( Table, 2 ) were placed to their precise chromosomal positions on idiograms ( Fig . 4C ) by FISH ., All BAC clones contained genetic markers ( Jimenez et al . , 2004 ) , and their positions on the chromosomes were predicted by previous genetic mapping 29 ., In our study , most of the BAC clones followed the order of the previous genetic mapping ., Only one BAC clone with genetic marker LF103 was found in slightly different order ., The expected position of this BAC clone was between genetic markers LF253 and LF106 on the 3p arm ., The actual position of this BAC clone was close to the centromere on 3q arm ., Thus , the idiograms for the mitotic chromosomes from the ID cells of Ae ., aegypti , which are presented here , can be successfully utilized for the physical mapping of the Ae ., aegypti genome ., The genome of Ae ., aegypti has several features that make physical mapping and genome assembly difficult ., First , Ae ., aegypti and other aediines have the largest genomes within the Culicidae family investigated thus far 11 ., Second , the Ae ., aegypti genome is extremely enriched with DNA repeats: about half of the genome consists of transposable elements ., Third , Ae ., aegypti lacks well-developed spreadable polytene chromosomes 14 , 15 ., Initial physical mapping of the Ae ., aegypti genome was performed on metaphase chromosomes from the ATC-10 cell line 16 ., Using FLpter , a fractional length from the p-terminus ( short arm telomeric end ) for measuring the location of the signal on each chromosome , provided a very approximate localization on the chromosomes ., In addition , using chromosomes from the permanent ( immortalized ) cell lines for the genome mapping can be misleading because these cells usually accumulate chromosomal rearrangements ., Two large chromosomal translocations were described in the ATC-10 line 16 ., It has been shown that in the cell culture of Ae ., albopictus ∼30% of the cells were tetraploid and 30% of the diploid cells had chromosomal aberrations 48 ., As a result of these difficulties and limitations , less than one third of the Ae ., aegypti draft genome assembly has been placed to chromosomes mostly based on results from genetic recombination mapping efforts , but without order and orientation 11 ., Using chromosomes from IDs of 4th instar larvae for the physical mapping of the Ae ., aegypti genome as proposed here will help to overcome the above problems ., Preparation of the chromosome spreads from IDs is a simple , robust procedure ., In this study more than 90% of the slides were suitable for in situ hybridization ., The number of the chromosome spreads per slide in IDs was also high ., We were able to find ∼150 chromosome spreads per individual ID at the stages appropriate for the mapping ., Finally , presence of these chromosomes in the IDs makes any individual mosquito at the larval stage available for cytogenetic analysis and allows avoiding having to use cell culture chromosomes for the physical mapping ., The chromosome spreads from ID cells have two features important for physical mapping ., First , chromosomes at all stages of mitosis have reproducible banding pattern which can be easily visualized by fluorescent staining with YOYO-1 ., Band-based physical mapping can be easily applied to these chromosomes instead of previously used distance–based mapping ( FL-pter , fractional length from the p-terminus ) 16 ., This approach will lead to the precise positioning of the BAC clones and genome assemblies on the chromosomes ., In addition to band-based mapping , the direct labeling of the DNA probe , which we used in our study , provides more precise location of the signal on the chromosome as compared to antibody-detected probes used before 16 ., Second , the significant number of chromosome spreads in IDs ( up to ∼30% ) might be found at early stages of mitosis ., Prometaphase and especially prophase chromosomes reflect significantly lower chromatin condensation and can be utilized for the orientation of relatively short scaffolds up to size ∼1 Mb ., The average size of the scaffolds in the current Ae ., aegypti genome assembly is 1 . 5 Mb 11 ., In order to map and orient scaffolds on the chromosomes , the probes for the BAC clones from the opposite sides of the scaffolds must be labeled with two different colors ., This approach was successfully used for the mapping of An ., gambiae heterochromatic scaffolds 38 ., Recently maps for mitotic chromosomes were created and successfully used for the physical mapping of Dr . melanogaster heterochromatin 49 , 50 , 51 , 52 , 53 ., Among other organisms , the most detailed cytogenetic analysis was performed for human and mammalian genomes 47 ., The highly populated FISH-based physical maps of mammalian genomes included 9528 and 851 markers for human and canine , respectively 54 , 55 ., The importance of chromosome-based physical mapping for comparative genomics was recently emphasized by H . Lewin and coauthors in the article titled “Every genome sequence needs a good map” 56 ., The authors suggested looking “back in the future” for developing high-resolution physical maps as an important framework for genome annotation and evolutionary analysis ., Finding an appropriate source of chromosomes and developing chromosomal idiograms , as conducted in this study , is the first important step toward the assembly and further utilization of the genomic information for the yellow fever mosquito Ae ., aegypti .

Introduction, Methods, Results, Discussion

The mosquito Aedes aegypti is the primary global vector for dengue and yellow fever viruses ., Sequencing of the Ae ., aegypti genome has stimulated research in vector biology and insect genomics ., However , the current genome assembly is highly fragmented with only ∼31% of the genome being assigned to chromosomes ., A lack of a reliable source of chromosomes for physical mapping has been a major impediment to improving the genome assembly of Ae ., aegypti ., In this study we demonstrate the utility of mitotic chromosomes from imaginal discs of 4th instar larva for cytogenetic studies of Ae ., aegypti ., High numbers of mitotic divisions on each slide preparation , large sizes , and reproducible banding patterns of the individual chromosomes simplify cytogenetic procedures ., Based on the banding structure of the chromosomes , we have developed idiograms for each of the three Ae ., aegypti chromosomes and placed 10 BAC clones and a 18S rDNA probe to precise chromosomal positions ., The study identified imaginal discs of 4th instar larva as a superior source of mitotic chromosomes for Ae ., aegypti ., The proposed approach allows precise mapping of DNA probes to the chromosomal positions and can be utilized for obtaining a high-quality genome assembly of the yellow fever mosquito .

Dengue fever is an emerging health threat to as much as half of the human population around the world ., No vaccines or drug treatments are currently available ., Thus , disease prevention is largely based on efforts to control its major mosquito vector Ae ., aegypti ., Novel vector control strategies , such as population replacement with pathogen-incompetent transgenic mosquitoes , rely on detailed knowledge of the genome organization for the mosquito ., However , the current genome assembly of Ae ., aegypti is highly fragmented and requires additional physical mapping onto chromosomes ., The absence of readable polytene chromosomes makes genome mapping for this mosquito extremely challenging ., In this study , we discovered and investigated a new source of chromosomes useful for the cytogenetic analysis in Ae ., aegypti – mitotic chromosomes from imaginal discs of 4th instar larvae ., Using natural banding patterns of these chromosomes , we developed a new band-based approach for physical mapping of DNA probes to the precise chromosomal positions ., Further application of this approach for genome mapping will greatly enhance the utility of the existing draft genome sequence assembly for Ae ., aegypti and thereby facilitate application of advanced genome technologies for investigating and developing novel genetic control strategies for dengue transmission .

biology

journal.pbio.0060206

An Enzymatic Atavist Revealed in Dual Pathways for Water Activation

Textbooks extol the extraordinary catalytic power and specificity of enzymes , yet the ability of many enzymes to promote several different chemical transformations is even more remarkable ., In examples such as the polyketide synthases , the substrate is tethered to a flexible linker and swings gymnastically between separate active sites 1 ., The evolutionary path to the assembly of such enzymes seems reasonably straightforward: gene duplication and recombination , followed by optimization of a promiscuous activity 2–6 ., In contrast , enzymes such as IMP dehydrogenase ( IMPDH ) move around a stationary substrate , restructuring the active site to accommodate different transition states 7 ., Such enzymes pose an evolutionary conundrum: it seems unlikely that Nature could simultaneously install multiple sets of catalytic machinery into the ancestral protein ., IMPDH controls the entry of purines into the guanine nucleotide pool , which suggests that the origins of IMPDH are primordial , so the ancestral IMPDH probably utilized a simpler catalytic strategy ., IMPDH catalyzes two very different chemical transformations: ( 1 ) a dehydrogenase reaction between IMP and NAD+ that produces a Cys319-linked intermediate E-XMP* and NADH , and ( 2 ) a hydrolysis reaction that releases XMP ( Figure 1A ) 7 , 8 ., A mobile flap is open during the hydride transfer reaction , permitting the association of NAD+ ., After NADH departs , this flap occupies the dinucleotide site , carrying Arg418 and Tyr419 into the active site and converting the enzyme into a hydrolase ( Figure 1B ) ., Thus , the dehydrogenase and hydrolase reactions utilize mutually exclusive conformations of the active site ., All enzymes that catalyze hydrolysis reactions have some strategy to activate water ., This strategy has been difficult to recognize in IMPDH because the hydrolytic water interacts with three residues that are usually protonated at physiological pH: Thr321 , Arg418 , and Tyr419 ( Figure 1C ) 9 ., The rate of the hydrolysis step decreases by a factor of 103 when Arg418 is substituted with Ala or Gln , whereas a decrease of approximately 20 is observed when Tyr419 is substituted with Phe 10 , 11 ., Neither Arg418 nor Tyr419 is involved in the dehydrogenase reaction , as expected , given their position on the mobile flap ., In contrast , Thr321 is found on the same loop as the catalytic Cys319 , and both the dehydrogenase and hydrolysis reactions are decreased by a factor of 20 when this residue is substituted 11 ., These observations suggest that Arg418 is the most likely candidate for the role of general base in the IMPDH reaction 11 , 12 ., We performed a series of hybrid quantum mechanical/molecular mechanical ( QM/MM ) simulations to further investigate the mechanism of the hydrolysis reaction of IMPDH ., Surprisingly , these simulations find that IMPDH possesses two mechanisms to activate water: the Arg418 pathway as previously proposed , and a second pathway utilizing Thr321 ., Phylogenetic analysis indicates that the Thr321 pathway was present in the ancestral enzyme ., These observations suggest that the primordial IMPDH used the Thr321 pathway exclusively , and elimination of the Arg418 pathway by mutation of modern IMPDH creates an enzymatic atavist ., When Arg418 is deprotonated in the starting condition , the lowest energy pathway for the hydrolysis reaction involves the transfer of a proton to the neutral Arg418 ( the Arg418 pathway , Figure 2A–2C ) ., Proton transfer is virtually complete at the transition state , and the developing hydroxide is stabilized by interactions with Tyr419 , Thr321 , and another water molecule ., Importantly , a stable hydroxide intermediate is not observed; the developing hydroxide instantaneously reacts with E-XMP* ., These results are in remarkable agreement with experimental observations: solvent isotope effects ( SIE ) demonstrate that proton transfer is rate limiting ( SIE = ∼2 11 ) , and Bronsted analysis indicates that proton transfer is virtually complete in the transition state ( β = ∼1 12 ) ., However , the calculated energy barrier is only 8 . 0 kcal/mol , much less than the experimentally observed barrier of 16 kcal/mol 10 ., This difference may reflect uncertainties in the calculation , but we believe this is unlikely ., A more intriguing source of discrepancy arises from the starting condition of neutral Arg418; if only a small fraction of the enzyme exists in this state , the energy barrier will be correspondingly increased ., Indeed , if the pKa of Arg418 is 12 . 5 , as for a typical Arg residue , the barrier would be increased by approximately 6 kcal/mol ., The pKa of a Tyr residue is usually two units lower than an Arg , which suggests that a deprotonated Tyr419 might activate water while Arg418 remains protonated ., Further simulations argue against such a mechanism; instead , the deprotonated , negatively charged Tyr419 interacts strongly with positively charged Arg418 and cannot interact with water ., Therefore , Tyr419 is unlikely to play the role of general base in the wild-type enzyme ., However , the situation changes when Arg418 is substituted with Gln: now the Tyr419 phenolate can accept a proton from water ., The barrier is approximately 17 kcal/mol ( Figure 3 ) ., Assuming a pKa of 10 , as is usual for a Tyr residue , then deprotonation of Tyr419 will further increase the barrier to 21–22 kcal/mol , which is very similar to the barrier observed in the reactions of the Arg418Gln and Arg418Ala variants ( ∼20 kcal/mol 10 , 11 ) ., As above , the simulations suggest that proton transfer is rate limiting and essentially complete in the transition state ., Whereas the landscape contains a shallow valley suggesting the presence of a hydroxide intermediate , the barriers are less than a kT , so the intermediate would not have a finite lifetime ., This simulation is generally consistent with experiments , where SIEs of 3–5 are observed when Arg418 is substituted 11 ., However , the magnitude of these SIEs is greater than expected if the transition state is indeed late as suggested by the simulations ., Interestingly , no activity is observed in the Arg418Ala/Tyr419Phe double mutant , though this fact may equally well be attributed to the inability to form the closed conformation required for the hydrolysis reaction as to the loss of the general base catalyst 12 ., Together , the simulations and experiments suggest Tyr419 may act as a surrogate general base in the absence of Arg418 ., Similar surrogate residues have been invoked to explain residual activity in other enzyme systems 26 ., In RNase T1 , His40 residue assumes the role the general base when Glu58 is substituted with Ala 27 ., Similarly , in ketosteroid isomerase , Asp99 may catalyze proton transfers in the Asp38Ala variant 28 ., Water or buffer molecules can also replace the function of missing catalytic residues 29 , 30 ., These examples illustrate the resilience and plasticity of enzyme catalysis ., Surprisingly , the simulations suggest a second pathway for water activation when the starting condition is protonated Arg418: Thr321 abstracts a proton from water while simultaneously transferring its own proton to Glu431 ( Figure 4 ) ., As in the Arg418 pathway , the developing hydroxide is stabilized by Tyr419 and another water molecule , and the hydroxide attack occurs instantaneously; the protonated Arg418 also stabilizes the developing hydroxide by 1–2 kcal/mol ., The calculated free energy barrier for the Thr321 pathway is approximately 20 kcal/mol ., The proton transfers are simultaneous and rate limiting ., When a simulation was performed with Glu431 treated as a molecular mechanical ( MM ) residue , which eliminates the possibility of proton transfer while maintaining electrostatic interactions , the energy barrier increases to at least 35 kcal/mol ( Figure S2 ) ., Likewise , when Glu431 is substituted with Gln , the barrier increases to at least 38 kcal/mol ., Therefore , the presence of Glu431 is essential for the operation of the Thr321 pathway ., The simulations suggest that the Thr321 pathway is favored at low pH , whereas the Arg418 pathway becomes dominant at high pH , which predicts that the pH-rate profile will shift to the right when the Thr321 pathway is disrupted by the Glu431Gln mutation ., This prediction was confirmed experimentally ( Figure 5 ) : the Glu431Gln mutation shifts the pKa from 7 . 2 ± 0 . 1 to 7 . 6 ± 0 . 1 , but has only a small effect on the pH-independent value of kcat ( kcat = 2 . 2 and 1 . 4 s−1 for wild type and Glu431Gln , respectively; these values are in good agreement with previous reports 31 , 32 ) ., Assuming that the pKa shift is entirely attributable to the loss of the Thr321 pathway , the barrier for the Thr321 pathway is approximately 19 kcal/mol , as predicted by the simulations ( see Figure S3 ) ., When Arg418 is substituted with Gln , the barrier for the Thr321 pathway is approximately 21 kcal/mol , which is similar to the barrier observed experimentally in the Arg418Ala and Arg418Gln variants 10 , 11 ., Therefore , both the Thr321 pathway and the Tyr419 pathway can account for the residual activity of the Arg418Ala and Arg418Gln variants ., However , since the Thr321 pathway involves the simultaneous transfer of two protons , this pathway can account for the large solvent isotope effects observed in the Arg418 variants ( SIE = 3–5 10 , 11 ) ., Therefore , we constructed the Arg418Gln/Glu431Gln variant , which should disrupt the Thr321 pathway but leave the Tyr419 pathway intact ., The simulations predict that the activity of this variant should be approximately the same as the Arg418Gln , but that the solvent isotope effect should be reduced ., These predictions were confirmed in subsequent experiments: ( 1 ) the value of kcat for Arg418Gln/Glu431Gln is decreased by 50% relative to that of Arg418Gln , as expected if the Thr321 pathway was lost ( 0 . 0020 ± 0 . 0002 s−1 and 0 . 0040 ± 0 . 0004 s−1 , respectively ) ; and ( 2 ) though the errors on the SIE are larger than ideal , nonetheless , a smaller SIE is observed in the reaction of Arg418Gln/Glu431Gln , consistent with the loss of the Thr321 pathway ( SIE = 2 . 1 ± 0 . 3 and 2 . 3 ± 0 . 4 for Arg418Gln/Glu431Gln in two independent determinations versus 2 . 9 ± 0 . 5 for Arg418Gln and 5 ± 2 for Arg418Ala 10 , 11 ) ., These experiments confirm the operation of the Thr321 pathway in IMPDH ., To the best of our knowledge , the presence of dual mechanisms for water activation in an enzyme active site is unprecedented ., Why would an enzyme have two pathways to accomplish the same task ?, We believe the Thr321 pathway may be vestige of evolution , and phylogenetic analysis is consistent with this hypothesis ( Figure 6; see Figure S4 for the complete phylogenetic tree ) ., The closest relative of IMPDH is GMP reductase ( GMPR ) , which catalyzes the conversion of GMP to IMP and ammonia with concomitant oxidation of NADPH ( Figure 6 ) 33 ., Cys319 , Thr321 , and Glu341 are also conserved in GMPR , which suggests that these residues were present in the IMPDH/GMPR ancestor ., X-ray crystal structures show that the conserved Cys , Thr , and Glu display similar interactions in both GMPR and IMPDH ( Figure 6 ) , suggesting that these residues may have similar functions in both enzymes ., To confirm that GMPR activity depends on the presence of Cys186 , Thr188 , and Glu289 , we tested the effect of mutations of these residues on the activity of Escherichia coli GMPR in a complementation assay ( Figure 7 ) ., E . coli H1173 requires both adenosine and guanosine for growth due to mutations in both purA and guaC 34 ( Figure 7 ) ., Growth on guanosine alone is restored with plasmid pGS682 , which carries the wild-type E . coli guaC gene 35 ., However , mutations in Cys186 , Thr188 , and Glu289 clearly compromise the ability of pGS682 to restore GMPR activity , demonstrating that selective pressure exists to conserve these residues ., Although the mechanism of the GMPR reaction has not been characterized , some clear parallels can be drawn with the IMPDH reaction , and E-XMP* may well be an intermediate ., Importantly , if E-XMP* forms as proposed , the active site must be constructed to prevent the hydrolysis reaction ., Kinetic and structural experiments clearly indicate that the reaction only proceeds when NADPH is bound in the active site and can block the access of water 33 , 36 , 37 ., Moreover , GMPR does not contain the Arg418-Tyr419 dyad , and the flap is truncated relative to the corresponding region of IMPDH , as expected , given that the hydrolysis of E-XMP* must be avoided during the GMPR reaction ., Therefore , the Arg418-Tyr419 dyad could have been installed as IMPDH optimized ., Alternatively , the dyad may have been present in the ancestral IMPDH/GMPR , but was subsequently remodeled in the GMPR lineage; since the flap binds in the same site as NAD+ , this scenario suggests that the ancestral IMPDH/GMPR was a hydrolase ., While we cannot rule out the latter scenario , we note that IMPDH is a member of the FMN oxidoreductase superfamily of ( β/α ) 8 barrel proteins ( unfortunately , none of these proteins is sufficiently similar to permit rooting of the tree ) 38–40 ., Therefore , it seems more likely that the ancestral enzyme was a promiscuous dehydrogenase , and the flap carrying the hydrolase activity was the later addition ., In contrast , the Thr321 pathway was likely present in the ancestral IMPDH/GMPR ., All IMPDHs and GMPRs contain Thr321 ( Figures 6 and S4 , and Text S1 ) ., As noted above , Thr321 also plays a role in the dehydrogenase reaction of IMPDH 11 , which suggests that Thr321 , like Cys319 , was inherited from the ancestral redox enzyme ., Glu431 is conserved among GMPRs , suggesting that the Thr321 pathway has a crucial function in this reaction , perhaps operating in the reverse to protonate the ammonia leaving group ., Curiously , although Glu431 is highly conserved among IMPDHs , it is substituted with Gln in the eukaryotic branch as well as in a few prokaryotic IMPDHs ., We suggest that the ancestral IMPDH/GMPR utilized the Thr321 pathway exclusively , but this pathway became expendable once the Arg418 pathway was established ., Phylogenetic analysis is consistent with this view: maximum likelihood analysis indicates that the ancestral enzyme almost certainly contained Glu at position 431 ( probability = 0 . 87 ) 41 ., Why then is Glu431 conserved in the majority of prokaryotic IMPDHs ?, The presence of the Thr pathway increases turnover , which may be important in maintaining the high concentration of guanine nucleotides required to support the rapid proliferation of prokaryotes ., More intriguingly , Glu431 provides 5–10-fold resistance to mycophenolic acid , a natural product that specifically inhibits IMPDH 32 ., Approximately 5% of microorganisms contain some mechanism to modify mycophenolic acid , which suggests that this compound is reasonably prevalent in the environment 42 ., Indeed , the extraordinary divergence of the adenosine subsite of IMPDH may be a response to the assault of natural product inhibitors such as mycophenolic acid and mizoribine 43 ., This divergence occurs despite the multiple functional constraints imposed by interactions with both NAD+/NADH and the flap ., The presence of the Thr pathway could facilitate this adaptation , making the evolutionary challenge of the IMPDH reaction much less formidable ., Plasmid pGS682 , a pUC plasmid carrying the 1 . 4-kb guaC insert from pGS89 35 , was a generous gift from Simon Andrews ( University of Sheffield ) ., E . coli strain H1173 was obtained from the E . coli Genetic Stock Center ( Yale University ) ., Atoms within a radius of 22 Å around the reaction center were treated as the dynamic region; this region was propagated with regular Newtonian dynamics by applying leapfrog integrator and 1-fs time step ., The atoms in the layer between the radii of 22 Å and 25 Å were treated as the buffer region; the heavy atoms in this region were harmonically restrained with the force constants scaled linearly with the distance from the sphere center ., The force constants around the boundary of the 25 Å sphere were set as implied by the B factors of the crystal structure ., In the buffer region , Langevin dynamics were applied with the friction coefficients also linearly scaled with the distance from the sphere center ., The friction coefficients around the boundary 25 Å sphere were set as 60 ., CHARMM 22 force fields 22 were utilized as the molecular mechanical potentials in these simulations ( colored in blue in Figures 2–4 ) and SCCDFTB ( self-consistent charge density-functional tight-binding ) method was applied as the quantum mechanical potential on the atoms involved in the chemical reactions ( colored in red in Figures 2–4 ) ., For the nonbonded interactions , an extended electrostatic treatment was applied with the electrostatic interactions within 12 Å described by group-based coulombic interactions ., IMP , acetylpyridine adenine dinucleotide ( APAD+ ) , Tris , and MES were purchased from Sigma ., DTT was purchased from Research Organics ., Wild-type and Glu431Gln IMPDH from T . foetus were expressed in E . coli and purified as described previously 10 , 32 ., All assays were performed at 25 °C ., The release of NADH is partially rate limiting 11 , 31 ., Therefore , to ensure that hydrolysis is completely rate limiting , these experiments used APAD+ 31 ., Pre-steady-state experiments were performed to demonstrate that hydride transfer and APADH are rapid over the entire pH range ( 11 and unpublished data ) ., Standard IMPDH assays contained saturating concentrations of IMP ( 2 mM ) and varying concentrations of APAD+ in 100 mM KCl , 1 mM DTT , and 50 mM of the appropriate buffer ( MES for pH 5 . 0–7 . 0 , and Tris-HCl for pH 7 . 3–9 . 3 ) ., Activity was measured by monitoring the absorbance of APADH at 363 nm on a Hitachi U-2000 UV-visible spectrophotometer ., Steady-state parameters with respect to APAD+ were derived at saturating IMP concentrations by plotting the initial velocity against APAD+ concentration and fitting to an equation describing uncompetitive substrate inhibition using SigmaPlot ( SPSS ) :, where ( kcat ) app are the apparent values at each pH , ( kcat ) indep are the pH-independent values , and Ka is the acid dissociation constant for the most acidic ionization ., IMPDH/GMPR amino acid sequences ( IMPDH IPR005990 , GMPR1 IPR005993 , and GMPR2 IPR005994 ) were retrieved from the InterPro database ( http://www . ebi . ac . uk/interpro/ ) ., Additionally , BLAST 44 searches with the T . foetus IMPDH ( P50097 ) and human GMPR1 ( P36959 ) amino acid sequences were performed ., Sequences from the BLAST search that were already part of the InterPro dataset were removed , and an initial multiple sequence alignment was performed with MUSCLE 45 ., A neighbor joining tree ( unpublished data ) was constructed in PAUP* 4 . 0b10 46 , and 95 sequences were selected for a Bayesian phylogenetic analysis ., The sequences of this subset were realigned with Espresso 47 , 48 ., A Bayesian phylogenetic analysis was performed with the parallel version of MrBayes 3 . 1 . 2 49 , 50 ., Amino acid substitution rates and state frequencies were fixed to the WAG parameters 51 ., A uniform ( 0 . 0 , 200 . 0 ) prior was assumed for the shape parameter of the gamma distribution of substitution rates 52 , an unconstrained exponential prior with rate 10 . 0 for branch lengths , and all labeled topologies were a priori equally probable ., Two independent MCMC analyses were run , each with one cold chain and three heated chains , with the incremental heating schema implemented in MrBayes ( λ=0 . 2 ) ., Convergence was assumed after the topology samples from the two cold chains had reached an average standard deviation of split frequencies of less than 0 . 01 ( after 1 , 610 , 000 generations ) ., Accession numbers , detailed results , and the full tree are found in Text S1 ., E . coli strain H1173 ( F- , guaC23 , tonA2 , proA35 , lacY1 , tsx-70 , supE44 ?, , gal-6 , l- , trp-45 , tyrA2 , rpsL125 , malA1 ( lR ) , xyl-7 , mtl-2 , thi-1 , purH57 ) contains mutations in purH and guaC , and therefore requires both adenosine and guanosine for growth ., Bacteria were transformed with pGS682 carrying either the wild-type guaC gene or variants containing C186A , T188A , and E289Q mutations ., Cultures were grown overnight in LB or LB/ampicillin and 5 μl of 1/20 serial dilutions were plated on M9 minimal media containing 0 . 5% casamino acids , 100 μg/ml l-tryptophan , 0 . 1% thiamin , 50 μg/ml guanosine , and/or 50 μg/ml adenosine .

Introduction, Results and Discussion, Materials and Methods

Inosine monophosphate dehydrogenase ( IMPDH ) catalyzes an essential step in the biosynthesis of guanine nucleotides ., This reaction involves two different chemical transformations , an NAD-linked redox reaction and a hydrolase reaction , that utilize mutually exclusive protein conformations with distinct catalytic residues ., How did Nature construct such a complicated catalyst ?, Here we employ a “Wang-Landau” metadynamics algorithm in hybrid quantum mechanical/molecular mechanical ( QM/MM ) simulations to investigate the mechanism of the hydrolase reaction ., These simulations show that the lowest energy pathway utilizes Arg418 as the base that activates water , in remarkable agreement with previous experiments ., Surprisingly , the simulations also reveal a second pathway for water activation involving a proton relay from Thr321 to Glu431 ., The energy barrier for the Thr321 pathway is similar to the barrier observed experimentally when Arg418 is removed by mutation ., The Thr321 pathway dominates at low pH when Arg418 is protonated , which predicts that the substitution of Glu431 with Gln will shift the pH-rate profile to the right ., This prediction is confirmed in subsequent experiments ., Phylogenetic analysis suggests that the Thr321 pathway was present in the ancestral enzyme , but was lost when the eukaryotic lineage diverged ., We propose that the primordial IMPDH utilized the Thr321 pathway exclusively , and that this mechanism became obsolete when the more sophisticated catalytic machinery of the Arg418 pathway was installed ., Thus , our simulations provide an unanticipated window into the evolution of a complex enzyme .

Many enzymes have the remarkable ability to catalyze several different chemical transformations ., For example , IMP dehydrogenase catalyzes both an NAD-linked redox reaction and a hydrolase reaction ., These reactions utilize distinct catalytic residues and protein conformations ., How did Nature construct such a complicated catalyst ?, While using computational methods to investigate the mechanism of the hydrolase reaction , we have discovered that IMP dehydrogenase contains two sets of catalytic residues to activate water ., Importantly , the simulations are in good agreement with previous experimental observations and are further validated by subsequent experiments ., Phylogenetic analysis suggests that the simpler , less efficient catalytic machinery was present in the ancestral enzyme , but was lost when the eukaryotic lineage diverged ., We propose that the primordial IMP dehydrogenase utilized the less efficient machinery exclusively , and that this mechanism became obsolete when the more sophisticated catalytic machinery evolved ., The presence of the less efficient machinery could facilitate adaptation , making the evolutionary challenge of the IMPDH reaction much less formidable ., Thus our simulations provide an unanticipated window into the evolution of a complex enzyme .

biochemistry, computational biology

How does nature construct complex catalysts? Molecular simulations revealed two sets of catalytic residues in the enzyme IMPDH, one of which seems to represent a primitive catalytic machinery that may be a vestige of evolution.

journal.pcbi.1005759

Modeling the adenosine system as a modulator of cognitive performance and sleep patterns during sleep restriction and recovery

When sleep is restricted , cognitive performance declines , recovering again when adequate sleep is obtained ., The dynamics of performance decline and recovery depend on the timescales over which sleep loss occurs ., During 1–2 nights of sleep deprivation ( continuous wakefulness ) , cognitive performance declines rapidly , and then returns to baseline after 1–2 nights of recovery sleep 1 , 2 ., However , when sleep restriction is chronic ( i . e . , multiple nights of insufficient sleep ) , such as 1–2 weeks of 3–6 hours of sleep per night , performance declines steadily on a timescale of weeks or longer 3 , 4 , and performance remains significantly impaired even after 2–3 nights of recovery sleep 5 ., Mathematical models have been developed to describe the effects of different sleep schedules on physiology and cognitive performance ., In 1984 , Daan et al . 6 developed a mathematical model , called the “two-process model” , to describe the effects of regular sleep schedules and sleep deprivation on EEG slow-wave activity , which is one marker of “sleep debt” ., The two-process model assumes that sleep is regulated by two independent processes: a circadian process , which describes the approximately 24-hour rhythm in sleepiness , and the sleep homeostatic process , which describes the tendency to accrue sleep debt and become sleepier the longer one is awake ., The dynamics of the sleep homeostatic process consist of exponential saturation towards an upper threshold during wake , with a time constant of ~20 h , and exponential decay towards a lower threshold during sleep , with a time constant of ~4 h in young adults ., Variants of the two-process model have also been used to describe changes in cognitive performance with sleep loss 7–12 ., This whole family of models , however , fails to describe the long-timescale changes in cognitive performance that occur under chronic sleep restriction 3 , 5 , 13 , 14 ., This is because the models lack any time constants longer than ~20 h , so the effects of any particular sleep regime ( restriction or recovery ) rapidly saturate ., To address this problem , extensions of the two-process model were developed 15–18 ., These models include an additional long-timescale process that modulates the upper and/or lower saturation thresholds for the sleep homeostatic process ., Adding this additional degree of freedom allows the models to capture changes in cognitive performance under both acute sleep deprivation and chronic sleep restriction ., These long-timescale model processes are , however , ad hoc , and not based on physiology ., Additional insights and opportunities for intervention design could be gained if these models were based on the physiological processes that underlie the sleep homeostatic process , including the adenosine system in particular ., Sleep-promoting substances , including adenosine , accumulate in multiple regions of the brain during wakefulness 19 ., In addition , adenosine A1 receptors in the brain are up-regulated by sleep loss 20 , 21 ., McCauley et al . 17 noted that the dynamics of their model “could be a mathematical representation of the interaction between a neurotransmitter or neuromodulator and its receptor , with the density of both changing dynamically across time awake and time asleep”; they identified the adenosine system as a probable candidate 22 ., However , no explicit link was made between their model and the underlying physiology; we show below that their model’s dynamical structure differs from our model of the adenosine system ., Understanding the physiological basis for cognitive impairments associated with sleep restriction is important , given that approximately 30% of the adult US population sleeps less than 7 hours per night , which is below the 7–9 hour range recommended by the National Sleep Foundation 23 ., Moreover , impaired cognition due to sleep loss is associated with errors and accidents 24 ., Here , we develop an explicit mathematical model of the adenosine system , with the goal of testing the hypothesis that dynamic changes in the concentrations of both adenosine molecules and receptors can account for changes in cognitive performance and sleep patterns under acute sleep deprivation , chronic sleep restriction , and recovery from chronic sleep restriction ., Our model is tested against two previously published experimental data sets:, ( i ) psychomotor vigilance test ( PVT ) data during acute and chronic sleep restriction , and recovery from chronic sleep restriction; and, ( ii ) sleep durations during long sleep opportunities in individuals recovering from low-level chronic sleep restriction ., Extracellular adenosine concentration increases during wakefulness and decreases during sleep in multiple brain regions , including the basal forebrain where its action is important to sleep regulation 25 ., Time-courses for increase and decrease of adenosine concentration have not been precisely characterized , but we can make reasonable physiological assumptions; see e . g . , 26 ., Specifically , we assume adenosine is produced at a constant rate in wakefulness and at a constant ( but lower ) rate in sleep , due to the lower ( on average ) brain metabolism during sleep 27 ., We also assume that adenosine follows first-order pharmacokinetics , i . e . , it is cleared at a rate proportional to its concentration in both wake and sleep , with clearance being faster during sleep due to active removal of metabolites 28 ., Adenosine concentrations can vary between different brain regions 19; here we calibrate the model against data collected from the basal forebrain , given its important role in sleep regulation 25 ., We do not model regional differences in concentrations within the brain ., These assumptions yield the following ordinary differential equation for total adenosine concentration ,, χdAtotdt=μ−Atot ., ( 1 ), The solution of this is exponential towards the saturation value μ with a time constant χ ., The values of χ and μ depend on sleep/wake state , which is a binary input to the model ( i . e . , the model can be either awake or asleep at any given time ) ., Across sleep/wake transitions , we demand continuity of Atot ., Total adenosine concentration includes: concentration of unbound molecules , denoted Au; concentration of molecules bound to A1 receptors , denoted A1 , b; and concentration of molecules bound to A2A receptors , denoted A2 , b ., We do not consider A2B or A3 receptors here , due to their much lower affinity for adenosine 29 ., Thus ,, Atot=Au+A1 , b+A2 , b ., ( 2 ), Concentrations of the different pools of adenosine depend on the availability of A1 and A2A receptors ., For receptor type n ( where n is 1 or 2A , abbreviated by 2 ) , we denote total receptor concentration by Rn , tot ., Total receptor concentration includes: concentration of unbound receptors , denoted Rn , u; and concentration of bound ( occupied ) receptors , denoted Rn , b , which by definition equals An , b ., Thus ,, R1 , tot=R1 , u+R1 , b ,, ( 3 ), R2 , tot=R2 , u+R2 , b ., ( 4 ), We use mass-action kinetics to describe the rates of binding and unbinding at each receptor type ,, dAudt=−k1 , bAuR1 , u−k2 , bAuR2 , u+k1 , uR1 , b+k2 , uR2 , b ,, ( 5 ), dR1 , bdt=k1 , bAuR1 , u−k1 , uR1 , b ,, ( 6 ), dR2 , bdt=k2 , bAuR2 , u−k2 , uR2 , b ., ( 7 ), Parameters kn , b and kn , u are rate constants for binding and unbinding at receptor type n , respectively ., The equilibrium conditions for Eqs ( 5 ) – ( 7 ) can be written in terms of ratios of rate constants ,, Kd1=k1 , uk1 , b=AuR1 , uR1 , b ,, ( 8 ), Kd2=k2 , uk2 , b=AuR2 , uR2 , b ., ( 9 ), These ratios are the dissociation constants for each reaction and have units of concentration ., The value of Kdn can be interpreted as the concentration of unbound adenosine , Au , for which there are equal numbers of bound and unbound receptors: Rn , b = Rn , u ., It has been experimentally observed that the concentration of A1 receptors increases when sleep is restricted and adenosine concentrations are elevated , and decreases to normal following recovery 30 ., We hypothesize here that this phenomenon reflects the dynamics of a ( homeostatic ) cellular response to maintain stable levels of A1 receptor occupancy ., When adenosine levels are elevated , a greater fraction of A1 receptors will be occupied ., To return to a homeostatic level of occupancy , more receptors must be synthesized ., This is a physiologically reasonable hypothesis , as receptor occupancy rates could be sensed and integrated on a per-cell basis ., We model these dynamics as:, λdR1 , totdt=R1 , b−γR1 , tot ,, ( 10 ), where 0 < γ < 1 is the target occupancy fraction , with equilibrium at R1 , b/R1 , tot = γ , and λ is a time constant that determines how quickly receptors are up-regulated or down-regulated in response to a change in occupancy ., We assume that R2 , tot is fixed , because only A1 receptors have been convincingly demonstrated to up-regulate in response to chronic sleep restriction 30 ., The evolution of Atot and R1 , tot occurs on the timescale of hours to weeks , which is much slower than the chemical rate constants ., The system can therefore be timescale separated by assuming Eqs ( 5 ) - ( 7 ) are at equilibrium ( i . e . , they are quasistatic ) ., We can then solve Eqs ( 8 ) and ( 9 ) for A1 , b and A2 , b in terms of Atot , R1 , tot , and R2 , u ( the concentration of unbound A2A receptors , which can be assumed to be approximately constant and is estimated below ) ,, A1 , b=12Atot+R1 , tot+Kd11−β− ( Atot+R1 , tot+Kd11−β ) 2−4AtotR1 , tot, ( 11 ), A2 , b=β ( Atot−A1 , b ) ,, ( 12 ), where, β=R2 , uR2 , u+Kd2 ,, ( 13 ), Substituting Eq ( 11 ) into Eq ( 10 ) gives a closed two-dimensional system for Atot and R1 , tot that depends on the values of Kd1 and Kd2 , which are known from experiment , and does not depend on the values of the individual kn , b and kn , u parameters ., Human cognitive performance and sleep depend on both the circadian process and the sleep homeostatic process 6 ., Here , we model the sleep homeostatic process as the concentration of bound A1 adenosine receptors , R1 , b ., In this respect , our model differs from previous models , which have usually treated the sleep homeostatic process per se as a predictor for EEG slow wave activity or cognitive performance ., We choose bound receptor concentration as the relevant sleep homeostatic variable over total receptor concentration or total adenosine concentration , because it is the downstream consequence of binding that mediates physiological effects ., The sleep homeostatic process could also depend on R2 , b , as well as other sleep-promoting molecules and receptors , such as cytokines 31 , but we choose the adenosine model here as the most parsimonious basis for a computational model , and first probe its explanatory power ., We model the circadian process by a sinusoid with a period of 24 hours ., This is a reasonable assumption for individuals who are entrained to the 24-hour day during chronic sleep restriction or recovery , or for individuals who are undergoing a brief acute sleep deprivation under constant conditions 32 , which are the experimental conditions modeled here ., Under more complicated scenarios , such as shifting time-zones , a dynamic circadian model should be used to account for changes in amplitude and phase 33 ., We assume that the overall influence of the circadian and homeostatic processes on sleep and performance is represented by a linear combination of the two processes , which we call the overall sleep drive ,, D=R1 , b+acos\u2061ω ( t−ϕ ) ,, ( 14 ), where a > 0 is the circadian amplitude , ω = ( 2π/24 ) h-1 , and ϕ is the clock-time ( modulo 24 , where zero is midnight ) at which the circadian process maximally promotes sleep ., This is typically near the core body temperature minimum in the early hours of the morning 33 ., Using D , we can predict when the model is sleepy ( high values of D ) or alert ( low values of D ) ., There is evidence of nonlinear interactions between the circadian and homeostatic processes 34 , 35 , and these have been introduced in some models 8 , 36 ., Here , we choose to start with the linear form , with the possibility of extending the model in future ., As a metric for cognitive performance , we use the number of lapses ( defined as responses slower than 500ms ) on the 10-minute PVT task ., This metric is bounded ., It is not possible to have fewer than 0 lapses , and the length of the test imposes a maximum possible number of lapses ., Thus , we choose a sigmoid function for converting D into an estimate of PVT lapses ,, P=pmax1+e ( Dmid−DDs ) ,, ( 15 ), where pmax is the maximum possible number of lapses , Dmid is the value of D for which the half-maximum value of lapses is achieved , and Ds determines the width of the sigmoid ., The variables D and P are both continuous functions of time ., For comparison with experiments , we sample P at times when a PVT test was performed ., In this section , we describe how the model parameters are fit and how the model outputs are compared to experimental data ., We first estimate physiological ranges for a subset of the model parameters and the mathematical relationships between others ., Due to the nature of the model and data , we then fit all the parameter values using an iterative approach ., For Experiment 1 , the dependent variable is PVT lapses , and sleep/wake timing is given as an input to the model via Eq ( 1 ) ., Values of all model parameters are fit at this stage , with the exception of λ , which takes a nominal value ., This is valid due to the fact that the model predictions for Experiment 1 are only weakly dependent on λ ., The model is then applied to simulate Experiment 2 . For this experiment , the dependent variable is daily sleep duration ., Two new parameters , Dsleep , and Dwake , are introduced to allow the model to make automatic transitions between sleep and wake ( i . e . , the model no longer requires sleep/wake timing as an input ) ., The values of the three parameters λ , Dsleep , and Dwake are thus fit to Experiment 2 , with all other parameters taking the values previously obtained by fitting to Experiment 1 . Finally , the values of Dmid and Ds are recalibrated to ensure the model still optimally fits data from Experiment 1 . Details of this fitting procedure are provided below ., In the adenosine system model , some parameters can be estimated from existing data ., The dissociation constants for Kd1 and Kd2 for A1 and A2A receptors are 1-10nM and 100–10 , 000nM , respectively 37 ., The fact that Kd1 ≈ Kd2/100 means that A1 receptors have much higher affinity for binding adenosine molecules ( i . e . , equivalent binding at 1/100 the concentration ) ., The time constants in Eqs ( 1 ) and ( 10 ) can be estimated based on the time-course of sleep homeostasis ., The hypothesis we wish to test with our model is that variations in Atot on timescales of hours to days can account for short timescale variations in sleep homeostatic pressure such as acute sleep deprivation , whereas variations in R1 , tot on timescales of weeks to months can account for long timescale effects of chronic sleep restriction ., The value of λ must therefore be large ., The longest inpatient chronic sleep restriction experiment to date lasted 3 weeks , with large decreases in PVT performance between weeks 1 and 2 , followed by non-significant decreases in PVT performance between weeks 2 and 3 3 ., This suggests λ is on the order of 1–2 weeks ., The dynamics of the model in Experiment 1 are found to be relatively insensitive to the value of λ , so we use a nominal value of 300 h ( the fit value ends up being close to this initial guess ) ., The value of λ is then refined by fitting the model to Experiment 2 . Since R1 , tot is slowly varying , it can be considered approximately constant on a timescale of hours or shorter ., On this timescale , only Atot significantly varies , so the dynamics of sleep homeostatic pressure described by the two-process model should approximately correspond to the dynamics of Atot described by Eq ( 1 ) ., Thus , we use the time constants of the original two-process model , χwake = 18 . 18 h and χsleep = 4 . 20 h 6 ., Biochemical data also give us typical values for the concentrations R1 , tot , R2 , tot , and Au , which can be used to establish approximate quantitative relationships between some of the model’s parameters ., In the human cerebral cortex , R1 , tot , is approximately 600nM , and R2 , tot , is approximately 300nM , with some variation in both between brain regions 38 ., In mammals , microdialysis measurements of extracellular unbound adenosine concentration , Au , report concentrations around 30nM 19 ., Since the dissociation constant for A2A receptors is large compared to physiological concentrations of Au , it is reasonable to assume R2 , u ≈ R2 , tot in Eq ( 13 ) ., In an individual who is well rested ( i . e . , not sleep restricted ) and keeping a regular daily sleep/wake cycle , Au will make daily oscillations about a stable level in response to sleep/wake cycles , causing daily oscillations in R1 , b = Ab , 1 , following the relationship described in Eq ( 11 ) ., In steady state , Eq ( 10 ) can be written <R1 , b> = γ ( <R1 , b> + <R1 , u> ) , where <∙> denotes expected value , and Eq ( 8 ) ( valid only at equilibrium ) can be rewritten in term of its time average as <R1 , b>=Kd1−1<Au><R1 , u>1+h . o . , where, h . o ., denotes the time average of higher order terms ( multiplicative cross products ) ., Combining these yields:, γ≈〈Au〉〈Au〉+Kd1 ., ( 16 ), Given Au is typically around 30nM and Kd1 is 1-10nM , this gives an estimated value of 0 . 50–0 . 91 for γ ., Using Eqs ( 2 ) and ( 11 ) – ( 13 ) , the parameters Atot , Au , Kd1 , and β are related by, 2Au= ( 1−β ) ( Atot−R1 , tot−Kd11−β+ ( Atot+R1 , tot+Kd11−β ) 2−4AtotR1 , tot ), ( 17 ), Rearranging for Atot gives, Atot=Au ( Au+Kd1+R1 , tot ( 1−β ) ) ( Au+Kd1 ) ( 1−β ) ., ( 18 ), Given values of Kd1 and β , we use the typical value of Au ≈ 30nM to estimate a typical value of Atot ., Finally , we relate the value of Atot to the parameters μwake and μsleep in Eq ( 1 ) ., During wake , the solution of Eq ( 1 ) as a function of time into the wake episode is, Atot , wake ( t ) =μwake+ ( Atot , wake ( 0 ) −μwake ) e−t/χwake ., ( 19 ), Similarly , during sleep , the solution of Eq ( 1 ) as a function of time into the sleep episode is, Atot , sleep ( t ) =μsleep+ ( Atot , sleep ( 0 ) −μsleep ) e−t/χsleep ., ( 20 ), For an individual who keeps a regular 24-hour sleep/wake cycle with a block of T hours of sleep per day , continuity of Eqs ( 19 ) and ( 20 ) require that, Atot , wake ( 24−T ) =Atot , sleep ( 0 ) ,, ( 21 ), Atot , sleep ( T ) =Atot , wake ( 0 ) ., ( 22 ), Combining these conditions gives the initial values of Atot at the beginning of each sleep and wake episode , respectively ,, Atot , sleep ( 0 ) =μwake ( 1−eT−24χwake ) +μsleep ( 1−e−Tχsleep ) eT−24χwake1−eT−24χwake−Tχsleep ,, ( 23 ), Atot , wake ( 0 ) =μsleep ( 1−e−Tχsleep ) +μwake ( 1−eT−24χwake ) e−Tχsleep1−eT−24χwake−Tχsleep ., ( 24 ), For a human with a typical schedule ( T = 8 h ) , substituting the numerical values of χwake and χsleep in Eqs ( 21 ) and ( 22 ) , and averaging these , we obtain an approximate estimate of a typical total adenosine concentration in the model ,, Atot=0 . 36μwake+0 . 65μsleep ., ( 25 ), Given values of Kd1 and Kd2 within their respective physiological ranges , we use Eq ( 18 ) to estimate Atot ., For each value of Atot we then obtain a range of values for μwake and μsleep using Eq ( 25 ) ., We require μwake > μsleep > 0 ., In the performance model described in Eq ( 15 ) , the parameter pmax can also be estimated ., In the standard 10-minute PVT , trials occur every 2–10 seconds ., Inter-trial intervals are drawn uniformly randomly from this time interval , with an average inter-trial interval of 6 seconds ., For a typical response time of δ , the number of trials per PVT is 600/ ( 6 + δ ) ., The theoretical maximum number of lapses would occur in an individual who responded in exactly 500ms on each trial , giving 92 lapses ., In reality , lapses are often much longer than 500ms 39 ., Individuals subject to a combination of acute sleep deprivation and severe chronic sleep restriction approach 4 seconds as a median response time 3 ., This suggests a theoretical ceiling of pmax ≈ 60 lapses ., The first test of the model is whether it can account for PVT data collected in humans undergoing acute sleep deprivation or different levels of chronic sleep restriction ., In Experiment 1 four groups of healthy young adults were exposed to different conditions of sleep restriction 4 ., One group ( n = 13 ) underwent acute sleep deprivation for 88, h . The other groups underwent chronic sleep restriction for 13 nights ( 4 h time in bed per night for n = 13 , 6 h time in bed per night for n = 13 , and 8 h time in bed per night for n = 9 ) , followed by 2 recovery nights ( 8 h time in bed per night ) ., Average sleep times per night during the chronic sleep restriction were approximately 3 . 7 h , 5 . 5 h , and 6 . 8 h , for the 4 h , 6 h , and 8 h time in bed conditions , respectively ., In each condition , participants awoke at the same time of 7:30am , which we plotted as 8am for convenience ., Prior to beginning the experiment , all four groups had three baseline nights with 8 h time in bed ., In the 5 nights prior to entering the laboratory , participants reported getting an average of 7 . 8 h sleep per night ., During wakefulness , participants completed 10-minute PVT tests every 2 hours ., Group-average PVT lapses were reported for each experimental condition in McCauley et al . 17 , beginning 4 hours after awakening each day to avoid effects of sleep inertia on performance ., We used these data ( recorded manually from the previous paper ) as our performance metric , P . The same data set was used previously to develop a model of the effects of chronic sleep restriction on human performance 17 and a similar data set 5 was used to develop another model 18 ., It is therefore an important first test of our physiological model ., Model parameter values were chosen within the estimated ranges given in Table 1 to achieve a least-squares fit to the experimental data ., This optimization was performed numerically using the Levenberg-Marquardt algorithm for global convergence ., The implementation used was the nlinfit function in Matlab ( version R2014A , Natick MA , USA ) ., The optimization was initialized using parameter values that fell within physiological ranges ., The model was initialized by simulating a schedule with 7 . 8 h sleep , matching the sleep duration participants reported getting prior to the inpatient schedule ., This schedule was repeated until convergence to a limit cycle was achieved ., Three baseline nights were then simulated with 7 . 0 h sleep per night , matching the average sleep duration participants achieved during baseline inpatient conditions ., Actual sleep durations were then simulated for each condition ., For consistency between all conditions , we chose to simulate recovery nights in the same manner as baseline nights , with 7 . 0 h sleep per night ., Morning awakenings are all plotted as occurring at 8am ., For reference , we also plotted the predictions of the McCauley et al . model ., For these , we used the published equations and initial conditions ., In Experiment 2 , 16 healthy young adults lived under “long night” conditions for 28 days 40 ., During these days , they were required to spend 14 hours per night ( 6pm to 8am ) in bed in a completely dark room , with no activities allowed , besides using the bathroom ., Participants were free to sleep for as much of this time as they liked , and sleep was recorded with polysomnography ., In the week prior to this , the participants were given 8 h time in bed per night , beginning around midnight ., During this week , they averaged approximately 7 h total sleep per night , and thus likely had some residual sleep debt ., While it was not primarily designed for this purpose , the experiment can be viewed as a long-term recovery from chronic low-level sleep restriction ., This is extremely valuable , since most chronic sleep restriction experiments have involved a week or less of recovery , making it difficult to determine the timescale of recovery ., The experimental data are strongly suggestive of a slow recovery process from an accrued sleep dept; individuals slept an average of 10 . 3 h across nights 1–3 , 9 . 1 h across nights 4–7 , 8 . 7 h across nights 8–14 , 8 . 7 h across nights 15–21 , and 8 . 2 h across nights 22–28 ., Interestingly , some individuals developed “split” sleep patterns , in which they had two main nighttime sleep bouts with a period of awakening in the middle ., This finding has been used as empirical support for the historical claim that humans in pre-industrial times had split sleep patterns 41 ., The estimation and fitting methods described above for Experiment 1 provide values for all parameters of the adenosine model , except λ ., The length of Experiment 2 allows us to accurately fit the value of this parameter ., In addition , we introduce two new parameters , Dsleep , and Dwake that allow the model to generate its own sleep/wake patterns ., During times when sleep is allowed by the schedule , the following rules are used to determine sleep/wake transitions ,, {D>Dsleep:TransitiontosleepifcurrentlyawakeD<Dwake:Transitiontowakeifcurrentlyasleep, This is motivated by the two thresholds used for sleep/wake transitions in the two-process model 6 ., The values of λ , Dsleep , and Dwake were estimated by least-squares fitting the model’s daily total sleep durations to the experimental group-average daily total sleep durations for days 1–28 of Experiment 2 . The Levenberg-Marquardt algorithm was found to perform poorly in this application , due to many points in parameter space achieving similarly good fits to the data ., We therefore finely gridded parameter space to find the optimal values of λ , Dsleep , and Dwake , each to at least 3 significant figures ., The model was initialized by simulating a schedule with 7 h sleep per day , beginning at midnight ., This schedule was repeated until convergence to a limit cycle was achieved ., The experimental protocol was then simulated by allowing sleep between 6pm and 8am each night ., More specifically , the model was forced to be awake from 8am to 6pm each day , and then freely selected sleep and wake times in the interval between 6pm and 8am each night using the thresholds described above ., These conditions were maintained for 50 days , allowing the model’s behavior in the first 28 days to be compared to data and allowing us to observe the model’s predicted longer-term behavior ., Finally , the parameters Dmid and Ds were recalibrated against Experiment 1 data to yield their final values , resulting in modest changes in both parameters ., The Levenberg-Marquardt algorithm was again used ., This recalibration was necessary , because the values of λ , Dsleep and Dwake determine the model’s natural sleep duration during baseline and the level of initial sleep homeostatic pressure ., The PVT function parameters must therefore be adjusted to allow the model to still optimally fit Experiment 1 , while maintaining the same outputs for Experiment 2 ( because the PVT function parameters do not affect sleep/wake outputs ) ., All other parameters therefore remained fixed at their previously fit values ., Within the physiological constraints discussed in Materials and Methods , the model achieves a good fit to the PVT data from Experiment 1 , with an adjusted R2 value of 0 . 66 ., The same adjusted R2 value was obtained both with the initial fit to Experiment 1 and with the final ( recalibrated ) parameter values ., Values for fit parameters are in Table 1 ., Notably , values for Kd1 and Kd2 are both at the lower end of the allowed ( physiological ) range ., This suggested that a better fit may exist outside of the range ., Relaxing the lower bounds on Kd1 and Kd2 , a best fit was achieved at Kd1 = 0 . 011 nM and Kd2 = 5 . 0 nM , with a slightly better adjusted R2 value of 0 . 73 , but this solution was discarded on the grounds of physiological constraints ., The system’s ability to capture PVT performance to similar accuracy both inside and outside the empirically-observed ranges for Kd1 and Kd2 suggests that these ranges exist due to other biological constraints ., Fits to each of the experimental conditions are shown in Fig 3 ., The model performs especially well in fitting the 4-h time in bed and 8-h time in bed conditions ., Some minor discrepancies between model and data are also observed ., Under acute sleep deprivation , there is a slight mismatch in circadian phase; this is likely due to, ( i ) the model fitting an average circadian phase to all conditions ,, ( ii ) drift away from a period of 24 hours under constant conditions , and, ( iii ) data being restricted to certain circadian phases in the other three conditions ., There is also a tendency for the model to underestimate PVT lapses in the 6-h time in bed condition ., This same issue was faced by McCauley et al . when they fit their model to the same dataset; they attributed this to one outlier participant in the 6-h group who was unusually sensitive to the effects of sleep restriction 17 ., The predictions of the McCauley et al . model are shown in Fig 3 for reference , since our model’s parameters were fit to the exact same dataset ., In general , the models closely agree under these simulated conditions ., Both models predict a characteristic within-day variation in performance , with a sudden decline in performance in the final hours of awakening ., This corresponds to the onset of the circadian night , as the circadian phase of maximal alertness is passed and homeostatic sleep pressure continues to build ., This is consistent with dependence of PVT performance on circadian phase 3 ., For PVT data , we find adjusted R2 = 0 . 66 ., Using the McCauley et al . model , which has a similar number of total parameters and no explicit physiological constraints , we find adjusted R2 = 0 . 70 on the same data set ., The same model used to simulate PVT lapses in Experiment 1 can also account for changes in sleep duration and timing during recovery from chronic sleep restriction in Experiment 2 ., Depending on the value of λ and the separation between the thresholds Dsleep and Dwake , the model was found during the fitting procedure to generate a variety of different sleep patterns , from one sleep bout per night to multiple sleep bouts per night ., Smaller values of λ and smaller separations between the thresholds favored more sleep bouts per night , due to the shorter time required to transit between thresholds ., This finding is consistent with previous results found in the two-process model 6 and physiological models of mammalian sleep 26 , 42 ., Fig 4 shows the model’s optimal fit to Experiment 2 ., In general , the model and data closely agree , with a root mean square error of 0 . 36 h ., The model underestimates sleep duration on the first night by 1 . 3 h , then is within 0 . 7 h of the experimental data on all subsequent nights ., During the recovery process , the model exhibits both monophasic sleep ( one sleep bout per night ) , and biphasic sleep ( two sleep bouts per night ) ., This is interesting , since both sleep patterns were experimentally observed in different participants in Experiment 2 ., Some participants consistently had one sleep bout throughout the experiment , others consistently had two sleep bouts , while others alternated between one and two sleep bouts on different nights 40 ., This suggests that the human population may span the region of parameter space that encompasses these two different modes of s

Introduction, Materials and methods, Results, Discussion

Sleep loss causes profound cognitive impairments and increases the concentrations of adenosine and adenosine A1 receptors in specific regions of the brain ., Time courses for performance impairment and recovery differ between acute and chronic sleep loss , but the physiological basis for these time courses is unknown ., Adenosine has been implicated in pathways that generate sleepiness and cognitive impairments , but existing mathematical models of sleep and cognitive performance do not explicitly include adenosine ., Here , we developed a novel receptor-ligand model of the adenosine system to test the hypothesis that changes in both adenosine and A1 receptor concentrations can capture changes in cognitive performance during acute sleep deprivation ( one prolonged wake episode ) , chronic sleep restriction ( multiple nights with insufficient sleep ) , and subsequent recovery ., Parameter values were estimated using biochemical data and reaction time performance on the psychomotor vigilance test ( PVT ) ., The model closely fit group-average PVT data during acute sleep deprivation , chronic sleep restriction , and recovery ., We tested the model’s ability to reproduce timing and duration of sleep in a separate experiment where individuals were permitted to sleep for up to 14 hours per day for 28 days ., The model accurately reproduced these data , and also correctly predicted the possible emergence of a split sleep pattern ( two distinct sleep episodes ) under these experimental conditions ., Our findings provide a physiologically plausible explanation for observed changes in cognitive performance and sleep during sleep loss and recovery , as well as a new approach for predicting sleep and cognitive performance under planned schedules .

Sleep loss is known to cause significant decrements in cognitive performance , but the physiological mechanisms responsible for this response are not well understood ., Computational models have been developed to predict how individuals will cognitively perform under acute or chronic sleep loss , but they currently lack an explicit physiological foundation , and do not specifically predict sleep timing ., Adenosine is hypothesized to be an important mediator in the effects of sleep loss , as it is a sleep-promoting substance that accumulates in the brain during wakefulness ., We developed a mathematical model of the adenosine system in the brain and showed that it can parsimoniously account for not only changes in cognitive performance during acute sleep deprivation , chronic sleep restriction , and recovery , but also changes in sleep patterns during long-term recovery ., The model thus provides a quantitative link between complex whole-organism behaviors and underlying molecular and physiologic mechanisms .

cell physiology, glycosylamines, medicine and health sciences, sleep deprivation, cognitive neurology, sleep, mathematical models, neuroscience, homeostatic mechanisms, physiological processes, cognitive neuroscience, receptor physiology, homeostasis, adenosine, research and analysis methods, mathematical and statistical techniques, cognitive impairment, biochemistry, cell biology, physiology, nucleosides, neurology, biology and life sciences, cognitive science, glycobiology

journal.pcbi.1003968

Bayesian History Matching of Complex Infectious Disease Models Using Emulation: A Tutorial and a Case Study on HIV in Uganda

Complex computer models ( hereafter called simulators ) are now being used in many scientific disciplines and are becoming increasingly common in basic science , climate modelling , communicable and non-communicable disease epidemiology and public health 1–6 ., The simulators utility for prediction and planning relies on how well they are calibrated to empirical data and how well they can be analysed to assess the validity of their predictions 7 , 8 ., Simulators can be calibrated using a multitude of approaches ., Simple ‘goodness of fit’ methodologies , such as least squares , are often used , however these approaches are difficult to apply to high-dimensional and computationally expensive individual-level simulators ., More rigorous statistical techniques have been developed , usually based around the concept of a likelihood function ., These techniques are very flexible , and can be used to fit a wide variety of simulators , ranging in complexity ., Nonetheless , implementing likelihood-based inference techniques for complex simulators is challenging , particularly when considering large-scale , missing or partially observed data ., Recent advances that have been usefully applied in the field of dynamic epidemic modelling include maximum likelihood via iterated filtering 9; data augmented and/or reversible-jump Markov chain Monte Carlo ( MCMC; 10–13 ) and stochastic differential equations 14 ., However , these systems can sometimes become mathematically or computationally intractable , leading to the development of various approximation techniques 15 , 16 ., A common theme in approximation methods for dynamic simulators is to replace dependence on the likelihood with outputs from simulator runs , since although a models likelihood may be intractable , running the simulator is straightforward ., These approaches can be implemented in various ways , for example: embedded in a particle filter 17; using Approximate Bayesian Computation 18–20; or using pseudo-marginal methods 21 ., Another technique that could be applied in this field is particle MCMC 22 ., Despite the variety of calibration methods , their application to the analysis of complex simulators is lacking ., A systematic review of cancer simulators found that of 131 studies only two thirds ( 87 ) provided any information on what methods were employed ., Of these only about a third ( 27 ) used a formal goodness of fit measure ( two used likelihood-based methods and 25 distance-based metrics , such as least squares or chi-square ) 23 ., The remainder did not state how the simulators were calibrated or used visual inspection to assess how well the simulator described the data ., Similarly , a systematic review of simulators of HIV transmission in men who have sex with men found only 18% of the 115 simulators had been formally calibrated to data and remarkably that calibration had become less common over time 24 ., One of the key reasons that complex simulator calibration is uncommon is that most formal methods ( including distance-based and likelihood-based measures ) require that simulators are run many times 25 ., This poses a considerable problem for complex simulators that require several minutes or even hours for the evaluation of a single scenario , making most of the above calibration methods utterly impractical ., The problem is compounded for stochastic simulators because hundreds or thousands of realisations are required for each scenario ., Current standard methods for formal sensitivity and uncertainty analysis 26 are also impractical for complex simulators because of the heavy computational burden 27 ., Simulator simplification , although desirable , is not appropriate if a complex simulator is required to satisfactorily address the research question and it increases the probability of simulator inadequacy 25 ., As the number of simulator parameters increases , the number of runs required for an adequate exploration of the parameter space increases rapidly ., Robust fitting and uncertainty analysis of complex simulators with dozens of parameters is often impossible , even with increasing computer power and advances in parallelisation ., Another important aspect of the calibration of complex models that remains unaddressed in the epidemiology literature is that of model discrepancy 25 , 28–30 ., This represents an upfront acknowledgement of the limitations of the complex model and helps tailor the search for acceptable input parameters by providing a more rigorous and realistic definition of match quality between the model outputs and observed data ( see section ‘History matching’ ) ., In this work we present a novel method based on Bayesian history matching , emulation and model discrepancy , that is designed to address all of the above issues while simultaneously avoiding unnecessary complexity ., This method has the potential to greatly improve the calibration of complex infectious disease simulators ., We present this in the form of a tutorial ( section ‘Methods’ ) and a case study where we history match a dynamic , event-driven , individual-based stochastic HIV simulator , using extensive demographic , behavioural and epidemiological data from Uganda ( section ‘Results’ ) ., The online supplementary material includes the details required for building an emulator and a simulation study that demonstrates the performance of history matching on synthetic data ., A major issue that affects the calibration algorithms discussed so far arises from simulators that are slow to evaluate ., Although computers are becoming increasingly powerful , running times of hours or days are not uncommon ( as modellers tend to develop more complex simulators to exploit increased computing power ) ., This can render any calibration algorithm that relies on a large number of simulator evaluations utterly impractical ., Another issue is that modern simulators tend to have a large number of inputs and outputs and the task of matching several outputs while varying a large number of inputs simultaneously can be very intensive computationally ., Both of these conditions can be addressed with history matching and emulation ., History matching 28 is designed to identify the set of inputs that would give rise to acceptable matches between the model outputs and the observed data ., It has three characteristics that distinguish it from most calibration methods ., Firstly , many calibration algorithms ( for example Bayesian MCMC ) attempt to make full probabilistic statements about the input values that are most likely to match the simulators output to the empirical data ., This represents a challenging and computationally intensive task , involving complex and frequently intractable calculations ., Critically , often such detailed calculations are unwarranted as the complex model is not thought to be an accurate enough representation of reality to justify them ., History matching instead provides a more tractable calculation involving expectations and variances , that is often of primary interest to modellers ., Secondly , history matching works by excluding parts of the input space that are unlikely to provide a good match ., These parts of the space are known as implausible ., The third characteristic is that the implausible space is not excluded all at once , but in iterations of the process , known as waves ., As a result , the non-implausible space ( i . e . the complement of the implausible space ) , shrinks at each iteration of the process ., The above characteristics give some desirable properties to history matching ., First , the calculations involved are far more efficient and straightforward to implement ., Second , the exclusion of implausible space is possible without considering the full set of inputs and outputs simultaneously , thus reducing the burden of high dimensionality ., For example , if the simulator fails to match one output for a particular input value , then this value is implausible regardless of the other outputs behaviour ., This should be compared to fully probabilistic approaches ( for example full Bayesian MCMC or maximum likelihood methods ) which attempt to model how likely an input is , usually using a likelihood function , thus representing a far more complex calculation that must use all outputs and all observed data and information simultaneously ., Third , as the volume of non-implausible space shrinks with consecutive waves , often to a tiny fraction of the original , the simulators behaviour typically becomes more predictable and smooth , as the range of the inputs is significantly smaller ., Once this point is reached , handling the full set of inputs and outputs is normally more manageable: at this point more detailed probabilistic calibration methods can be employed if necessary ( see section ‘Posterior sampling’ ) ., Finally , it is possible that a simulator is incapable of matching the observation data , due to either incorrect modelling assumptions , poor error specification , or coding errors ., History matching can identify this condition by characterising all the input parameter space as implausible , whereas alternative methods will always attempt to return a posterior distribution , regardless of how well , if at all , the simulator fits the data ., A long established method for handling computationally expensive simulators is to first construct an emulator: a statistical model of the simulator that can be used as a surrogate 31 ., The simulator is first run at a manageable number of input values , to provide training data to build the emulator ., The emulator will give a joint probability distribution of the simulator outputs for any set of input values , and the distribution can be used both to provide estimates of the outputs , and quantify uncertainty in the estimates ., Building an emulator will involve some computational effort in obtaining the training data ( the simulator runs ) and fitting the emulator to the data ., However , once built , the emulator can provide estimates ( with a quantification of uncertainty ) of the simulator output near instantaneously , even for very large numbers of inputs ., Emulators enable rapid exploration of high dimensional input spaces , and have been used within fully probabilistic calibration 25 , 32 , 33 , including simulators with high dimensional output 34 ., Emulators can be used within history matching if the simulator is computationally expensive , as is the case in this work ., History matching together with emulation has been successfully applied across a range of scientific disciplines including galaxy formation simulations ( 35 , 36 or for an overview see 37 ) , oil reservoir models 28 , 38 , systems biology models 39 , 40 , climate models 41 and rainfall runoff models 30 ., History matching is a method designed for reducing the simulators input space but is not designed to make probabilistic statements about the inputs , such as producing posterior distributions ., Thus , it can be seen as a pre-calibration method or as a calibration method but in the broader sense ., We would assert that for many situations involving model development and assessment , the results of a history match are all that are required by the modeller ., When specifying the initial input ranges , we may have substantial uncertainty about what the acceptable input values are , so that the acceptable region of the input space ( that would contain say the posterior distribution ) is a tiny proportion of the initially specified input space , and thus hard to discover ., The iterative nature of history matching and the fact that it discards the implausible space instead of looking for input values that are close to the empirical data simplify significantly this task ., It is also important to bear in mind that alternative ‘probabilistic’ calibration methods would most likely struggle with a model of the complexity and input-output dimensionality such as the one studied here ., Therefore , should one wish to probabilistically calibrate a well tested and accurate simulator , it is still advantageous to greatly reduce the input space under consideration first , using history matching as a precursor ., We continue this tutorial by describing how history matching is set up ( section ‘History matching’ ) , and we then present the procedure of history matching ( section ‘Procedure’ ) along with a toy example that illustrates the fundamental concepts ., The tutorial then proceeds with two more technical sections , one containing details on how an emulator is built ( section ‘Emulation’ ) , and another describing an essential component of history matching , the implausibility measure ( section ‘Implausibility measure’ ) ., Finally , we present an approximate method for drawing samples from the simulators posterior distribution ( section ‘Posterior sampling’ ) ., History matching assumes the existence of a physical process that is measured through observations ( Fig . 1 ) ., The acquisition of observations takes place with finite accuracy and introduces some uncertainty , which we term observation uncertainty ( OU ) ., History matching also assumes the existence of a simulator ( computer model ) that attempts to describe the process ., The simulator has inputs ( parameters ) , assumed to be continuous ., We consider a stochastic simulator: a simulator which when run twice at the same value of can produce different outputs ., We suppose that the simulator output consists of a vector of quantities , which we denote with the vector ., To represent the stochastic nature of the simulator , if we keep the input vector fixed and run the simulator times , we would observe , for the run , with : ( 1 ) where is the mean value of the output ( if the simulator were to be run repeatedly at the same input value ) , and is a random variable with expectation 0 ., We suppose that the physical process corresponds , to some level of accuracy or tolerance , to a realisation of the simulator output , at some particular input , rather than the mean output ., In our search for non-implausible inputs , we need to take into account the variability of around ., We refer to this term as Ensemble Variability ( EV ) ., As mentioned earlier , the calibration of complex simulators can be infeasible if the calibration method depends on a large number of simulator evaluations that take considerable time to complete ., For this reason , we rely on a statistical model of the simulator , known as an emulator , which is trained using a relatively small number of simulator runs and which we use to provide an estimate of in a fraction of the time required for a simulator run ., The emulator represents our beliefs about the at all , yet to be evaluated inputs , and our uncertainty about such values ., The fact that the simulator ( code ) is not evaluated for every possible value of , creates another source of uncertainty , which we term Code Uncertainty ( CU ) and is quantified via the emulator ., There is one final source of uncertainty , which is important though perhaps the most difficult to consider ., Due to our incomplete understanding of the process and our inability to model all of its aspects , we do not believe the simulator to be a perfect representation of reality 28 ., This has three implications for calibration ., Firstly , an input that gives a good match to historical data will not necessarily give a good prediction of future data; the simulator may be overfitted ., Secondly , an input that does not give a good match to one physical output quantity may still give a good prediction of another physical output quantity , if the simulator models some quantities more accurately than others ., Thirdly , if the inputs are physically observable quantities ( that could , in principle , be learnt independently of the simulator ) , failing to account for an imperfect simulator can lead to overconfident posterior distributions that are centred on the wrong values 42 ., We refer to this final source of uncertainty as Model Discrepancy ( MD ) 25 ., Incorporating model discrepancy protects against overfitting , ensures that we do not exclude possible values of future observations , when the exclusion would be unwarranted , is necessary for inferring the true values of simulator inputs , when the notion of a true input value is clearly understood , and is required for making realistic forecasts ( 35 , 43 ) ., In summary , we link the observation of the physical process to the best simulator input , which we denote by via ( 2 ) where is a vector of errors representing observation uncertainty , is a vector of errors representing ensemble variability , is a vector of errors representing model discrepancy , and , , and are judged to be independent 25 ., Fig . 2 shows a typical history matching workflow ., The first step is the selection of a number of input values ( design points ) at which the simulator is run ., The initial inputs are chosen using a maximin Latin hypercube design 44 , which generates uniformly distributed points , but also aims to fill the entire input space , by maximising the minimum distance between the points generated ., The number of points in this initial design depends on the available computational resources ., A very approximate rule of thumb is to use at least for training the emulator and points for validation 45 ., Once the initial design space , , is specified , the simulator is run at the selected points ., Following the notation set out in equation 1 , we construct separate emulators: one for the mean of each output , with ., For the j-th output , the training data takes the following form ., We choose the training inputs and for each input value , we run the simulator times , to generate observations ., We then calculate the sample mean and variance of the simulator runs at input : ( 3 ) ( 4 ) The training data point for input is then , where is an estimate of ., The number of runs ( ) per input point are determined by the simulators complexity and the available computational power ., A relatively large number of repetitions ( e . g . ) will ensure that the error in the estimate is approximately normally distributed with expectation 0 and variance even if the individual terms are not normally distributed ., Once we have built the th emulator , we can efficiently obtain an expected value of and variance for any , in particular for input values where we have not run the simulator ., We denote this expectation and variance by and , where the superscript indicates that the expectation and variance refer to code uncertainty: the fact that is an uncertain quantity ., Other approaches of emulating separately the simulators mean and variance are also possible 33 , 39 , 40 ., It should be noted that often some of the outputs are difficult to emulate in the first few waves , in which case we would emulate a subset of the outputs initially , emulating the remaining outputs in later waves , when the process becomes easier due to the reduced size of the input space ., More detail on emulation is given in section ‘Emulation’ ., Fig . 3, ( a ) shows a simple example of a one dimensional emulator ., The ( toy ) simulator used is the deterministic function ., Because the simulator is deterministic , it holds that ., The value of is considered unknown apart from the six points where the simulator is run and are represented by the black dots in the figure ., The blue line is the emulators posterior mean , and the red lines represent its posterior uncertainty ( 95% CI ) ., The 3 horizontal lines represent the empirical data ( ) and the 95% CI ( ) that we use to history match the simulator ., The next step involves choosing an implausibility measure and defining its various components ., The implausibility is an essential element of history matching and is a measure that estimates whether the input is likely to result in an output that will match the observations ., It essentially weighs the difference between and with all the uncertainties that are present in the system ., The implausibility is large when the emulators posterior mean is far from the empirical data , relative to the uncertainties present in the system ( observation and code uncertainty in this case ) ., An analytical description of how the implausibility can be formulated is provided in section ‘Implausibility measure’ ., Fig . 3, ( b ) shows the implausibility for the emulator and empirical data from Fig . 3, ( a ) ., The horizontal green line is an implausibility cut-off , which determines whether an input is implausible or not ., The implausibility plot shows that a match between the simulators output and the empirical data is unlikely to be found for values of smaller than 30 and larger than 45 ., With the emulators and the implausibility measure at our disposal we can then carry out two key functions of history matching: the first is to sample the non-implausible space and study its distribution ., This can reveal input combinations that can lead to acceptable matches , correlations between inputs and detailed insight into the models structure ., The second function is the creation of a design that is space filling over the current non-implausible space , which will be used to run the simulator in the next wave ( iteration ) of history matching ., The simplest method for sampling the non-implausible space , is to draw samples uniformly from the entire input space and reject those that fail the implausibility criteria ., This method is computationally straightforward , but it can become inefficient when the non-implausible space is a tiny fraction of the original space , which is often true , especially in later waves ., A method for solving this problem using an evolutionary Monte Carlo algorithm was proposed in 46 ., In this paper , we propose a simpler but also effective method ., Suppose that in wave we have a number of non-implausible points ., For each of these , we draw samples from a variate normal distribution that is centered on the value of the generating point ., The wave implausibility is then evaluated on the new samples and the variance of the normal distribution is selected so that a small percentage of them ( ) are non-implausible ., The low acceptance rates should ensure that the new samples are sufficiently different from the old ones ., This method can efficiently generate an adequate number of data points that can be used in subsequent waves ., A subset of the non-implausible samples drawn are then used to run the simulator and repeat another wave of history matching ., The code or emulator uncertainty decreases with each iteration for the following reasons ., At each wave , the emulators are only constructed over a smaller region of input space compared to the previous wave , and therefore the mean of the simulator outputs are usually smoother functions of the input parameters and hence easier to emulate accurately ., Also there is a higher density of simulator runs in the new reduced input space , which again leads to improvements due to the Gaussian process part of the emulator as described in section ‘What is an emulator ? ’ ., There may be additional benefits due to active variable selection as discussed in 35 , 37 , and new outputs that were previously difficult to emulate may now become available ., A major reason for the power of the history matching approach described here , is due to the above improvements to the emulation process at each wave , allowing the iterative exploration of complex input spaces ., Fig . 4 shows the second wave of history matching for the running example of this section ., The simulator was run for the non-implausible value of and this point was included in the training data ., Note how the emulators posterior variance has decreased in the region of interest ., Consequently , the non-implausible region has shrunk dramatically , indicating that a match can only be found for and , where indeed the function takes values between −0 . 8 and −0 . 63 ., The procedure can continue with more waves until one or more stopping criteria are met ., One such criterion is when all the input space is deemed non-implausible , meaning that the simulator cannot match the observations given the current error specifications ., In this case one would then vary the size of the model discrepancy to determine how large it would have to be to obtain a match: a very large model discrepancy would suggest that the simulator is inadequate as a model for the physical process in question , and that further model development is required ., Another stopping criterion occurs when the emulators have a posterior variance smaller than the remaining uncertainties in the system ( the observation uncertainty , model discrepancy and the ensemble variability ) , as this condition implies that the non-implausible space contains acceptable matches and is unlikely to decrease in size in the next iteration , unless the remaining uncertainties in the system can be revised and decreased as well ., Here we would check the acceptable matches against any other outputs that were not used in the emulation process ., A final condition for stopping could be the fact that the simulator runs obtained in the current wave are close enough to the empirical data and we do not wish to continue any further ., In these two cases , we would investigate the sensitivity and robustness of the non-implausible region obtained from the history match , to alterations in the observation uncertainties and model discrepancy 47 ., By design , history matching excludes parts of the input space that produce poor fits to the empirical data and leads to , where possible , the generation of a large number of runs that give acceptable fits to the observed data ., Often this is sufficient for model analysis , e . g . to help understand the biological or epidemiological mechanisms underlying the physical system , and model development to determine the next improvement to the mathematical model and hence the next extension to the computer code ., However , history matching does not provide full Bayesian posterior distributions for the uncertain quantities of interest ( for example , the input parameters ) ., Were it the case that full posterior distributions are required , for say a highly accurate , well tested and well understood epidemiology model , we now show how the results of history matching , specifically the identification of the final non-implausible region of input space ( which should enclose the posterior distribution ) , can be used to obtain samples from the posterior ., Such information from the posterior can be used for many tasks such as comparing competing models , and also when making forecasts into the future ( a typical use of epidemic models ) ., Let be the non-implausible samples from the last wave of history matching ., We formulate the following proposal distribution ( 11 ) which is a multivariate normal distribution , with mean chosen to be the sample mean of the non-implausible samples and with variance chosen to be multiplied by the sample variance-covariance matrix of the non-implausible samples ., The constant is used to inflate the variance of the non-implausible samples , so that the method explores a larger part of the input space , as this was found to improve the sampling process ., We define the approximate likelihood of the input parameters with observed data as ( 12 ) with ., This follows from equation 2 combined with the additional assumptions that the observation error , model discrepancy , ensemble variability and the emulator representation of are all normally distributed ., Such normality assumptions for and are very common ( see for example 25 ) , and for the emulator this is simply equivalent to having sufficient runs to ensure the emulators t-distribution can be viewed as approximately normal , as is described in S1 Text ., The assumption of normality of the ensemble variability ( which is directly analogous to the assumptions made in standard regression ) may not be justified in some cases , however the above approximations can still be accurate when is smaller that the other variance components so that their total can still be considered approximately normal ., More detailed modelling of is of course possible 39 , 40 , and more complex distributions can be specified , however this would only be warranted if the epidemiology model was judged to be a sufficiently accurate mimic of reality that the details of the distribution of its ensemble variability were physically meaningful ., In the case study described in section ‘Results’ , this was definitely not thought to be the case ., We assume that the prior is constant over the final non-implausible region found from the history match ., This is in many cases reasonable as the volume of this region is often many orders of magnitude smaller that the original input space , and hence any prior that was not strongly informative would likely be approximately constant over such a small space ., Note however , that the algorithm below can be easily adapted for any prior , by a suitable scaling of the weights ., The algorithm proceeds as follows: we first use the proposal distribution to generate a number of samples ., We then calculate a weight for each sample as ., Finally , we draw the desired number of posterior samples from the set of , with a probability defined by the weights ., Provided the weights are reasonably well behaved , this will generate direct draws from the approximate posterior ., This case study was based on a research project that explored the effects of partnership concurrency ( overlapping sexual partnerships ) on HIV transmission in Uganda 56 ., The simulator used in the research study , named Mukwano , was a dynamic , stochastic , individual based computer model that simulates heterosexual sexual partnerships and HIV transmission ., In an individual based micro simulation model , the life histories of hypothetical individuals are simulated over time in a computer program ., Each individual is represented by a number of characteristics , of which some remain constant during simulated life ( e . g . gender and date of birth ) , whereas others change ( e . g . HIV status ) ., Changes in personal characteristics result from events such as the start and the end of sexual relationships ., These events are stochastic: if and when an event occurs is determined by Monte-Carlo sampling from probability distributions ., To generate model outcomes for a simulated population , the characteristics of the simulated individuals are aggregated ., The simulator had been fitted to empirical data in a number of scenarios by eye and by changing the values of inputs , which control various demographic , behavioural and epidemiologic characteristics of the simulated population 56 ., Births , deaths , partnership formation and dissolution and HIV transmission were modelled using time-dependent rates ., At birth , simulated individuals were assigned to one of two sexual activity groups ( ‘high activity’ and ‘low activity’ ) and to one of two concurrency groups ( ‘high concurrency’ and ‘low concurrency’ ) ., Each sexual activity group had associated male and female sexual contact rates , which determined the rate at which individuals formed new partnerships , which were of two types ( ‘short duration’ and ‘long duration’ ) ., For the present case study , we apply our history matching and emulation methodology to the primary baseline scenario from 56 , rather than fitting ‘by-eye’ ., Twenty behavioural and two epidemiologic inputs were varied , including a mixing parameter , which determines the tendency for individuals to preferentially form partnerships with people in their own activity group , and an input which determines the duration of the long and short duration partnerships ., The behavioural inputs are permitted to take different values in each of three calendar time periods ., This enables sexual behaviour to vary over time ., The full list of the 22 simulator inputs and their original plausible ran

Introduction, Methods, Results, Discussion

Advances in scientific computing have allowed the development of complex models that are being routinely applied to problems in disease epidemiology , public health and decision making ., The utility of these models depends in part on how well they can reproduce empirical data ., However , fitting such models to real world data is greatly hindered both by large numbers of input and output parameters , and by long run times , such that many modelling studies lack a formal calibration methodology ., We present a novel method that has the potential to improve the calibration of complex infectious disease models ( hereafter called simulators ) ., We present this in the form of a tutorial and a case study where we history match a dynamic , event-driven , individual-based stochastic HIV simulator , using extensive demographic , behavioural and epidemiological data available from Uganda ., The tutorial describes history matching and emulation ., History matching is an iterative procedure that reduces the simulators input space by identifying and discarding areas that are unlikely to provide a good match to the empirical data ., History matching relies on the computational efficiency of a Bayesian representation of the simulator , known as an emulator ., Emulators mimic the simulators behaviour , but are often several orders of magnitude faster to evaluate ., In the case study , we use a 22 input simulator , fitting its 18 outputs simultaneously ., After 9 iterations of history matching , a non-implausible region of the simulator input space was identified that was times smaller than the original input space ., Simulator evaluations made within this region were found to have a 65% probability of fitting all 18 outputs ., History matching and emulation are useful additions to the toolbox of infectious disease modellers ., Further research is required to explicitly address the stochastic nature of the simulator as well as to account for correlations between outputs .

An increasing number of scientific disciplines , and biology in particular , rely on complex computational models ., The utility of these models depends on how well they are fitted to empirical data ., Fitting is achieved by searching for suitable values for the models input parameters , in a process known as calibration ., Modern computer models typically have a large number of input and output parameters , and long running times , a consequence of their increasing computational complexity ., The above two things hinder the calibration process ., In this work , we propose a method that can help the calibration of models with long running times and several inputs and outputs ., We apply this method on an individual based , dynamic and stochastic HIV model , using HIV data from Uganda ., The final system has a 65% probability of selecting an input parameter set that fits all 18 model outputs .

mathematics, statistics (mathematics), physical sciences, statistical methods

journal.pcbi.1000034

Chromophore Protonation State Controls Photoswitching of the Fluoroprotein asFP595

Fluorescent proteins have been widely used as genetically encodable fusion tags to monitor protein localizations and dynamics in live cells 1–3 ., Recently , a new class of green fluorescent protein ( GFP ) -like proteins has been discovered , which can be reversibly photoswitched between a fluorescent ( on ) and a non-fluorescent ( off ) state 4–10 ., As the reversible photoswitching of photochromic organic molecules such as fulgides or diarylethenes is usually not accompanied by fluorescence 11 , this switching reversibility is a very remarkable and unique feature that may allow fundamentally new applications ., For example , the reversible photoswitching , also known as kindling , may provide nanoscale resolution in far field fluorescence optical microscopy much below the diffraction limit 12–15 ., Likewise , reversibly switchable fluorescent proteins will enable the repeated tracking of protein location and movement in single cells 16 ., Since fluorescence can be sensitively read out from a bulky crystal , the prospect of erasable three-dimensional data storage is equally intriguing 17 ., The GFP-like protein asFP595 , isolated from the sea anemone Anemonia sulcata , is a prototype for a reversibly switchable fluorescent protein ., The protein can be switched from its non-fluorescent off state to the fluorescent on state by green light of 568 nm wavelength 5 , 6 , 18 , 19 ., From this so-called kindled on state , the same green light elicits a red fluorescence emission at 595 nm ., Upon kindling , the intensity of the absorption maximum at 568 nm diminishes , and an absorption peak at 445 nm appears ., The kindled on state can be promptly switched back to the initial off state by this blue light of 445 nm ., Alternatively , the off state is repopulated through thermal relaxation within seconds ., In addition , if irradiated with intense green light over a long period of time , asFP595 can also be irreversibly converted into a fluorescent state that cannot be quenched by light any more 5 ., The nature of this state is hitherto unknown ., The switching cycle of asFP595 is reversible and can be repeated many times without significant photobleaching ., These properties render asFP595 a promising fluorescence marker for high-resolution optical far-field microscopy , as recently demonstrated by Hofmann and coworkers 20 ., Currently , however , with its low fluorescence quantum yield ( <0 . 1% and 7% before and after activation , respectively 6 , 16 ) and rather slow switching kinetics , the photochromic properties of asFP595 need to be improved ., To systematically exploit the potential of such switchable proteins and to enable rational improvements to the properties of asFP595 , a detailed molecular understanding of the photoswitching mechanism is mandatory ., The aim of this study is to obtain a detailed mechanistic picture of the photoswitching mechanism of asFP595 at the atomic level , i . e . , to understand the dynamics of both the activation process ( off-to-on switching ) and the de-activation process ( on-to-off switching ) ., High-resolution crystal structures of the wild-type ( wt ) asFP595 in its off state 19 , 21 , 22 , of the Ser158Val mutant in its on state 19 , and of the Ala143Ser mutant in its on and off states 19 were recently determined ., Similar to GFP , asFP595 adopts a β-barrel fold enclosing the chromophore , a 2-acetyl-5- ( p-hydroxybenzylidene ) imidazolinone ( Figure 1 ) ., The chromophore is post-translationally formed in an autocatalytic cyclization-oxidation reaction of the Met63-Tyr64-Gly65 ( MYG ) triad ., As compared to the GFP chromophore , the π-system of MYG is elongated by an additional carbonyl group 23 ., Reversible photoswitching of asFP595 was possible even within protein crystals , and x-ray analysis showed that the off-on switching of the fluorescence is accompanied by a conformational trans-cis isomerization of the chromophore 19 ., In a recent study 24 , we have shown that the isomerization induces changes of the protonation pattern of the chromophore and some of the surrounding amino acids , and that these changes account for the observed shifts in the absorption spectrum upon kindling ., Based on the comparison between measured and calculated absorption spectra , the major protonation states in the ground state have been assigned to the zwitterion ( Z ) and the anion ( A ) for the trans conformer , whereas the neutral ( N ) chromophore is dominant for the cis conformation ( Figure 1B ) ., Here , we study the photochemical behavior of each of the previously identified protonation states ., We have addressed the following questions: How does light absorption induce the isomerization of the chromophore within the protein matrix , and how do the different protonation states affect the internal conversion mechanism ?, Which is the fluorescent species , and how can the fluorescence quantum yield be increased ?, To address these questions , we have carried out nonadiabatic molecular dynamics ( MD ) simulations using a hybrid quantum-classical QM/MM approach ., This approach includes diabatic surface hopping between the excited state and the ground state ., The forces acting on the chromophore were calculated using the CASSCF 25 , 26 multi-reference method , which , although not always yielding highly accurate excitation and fluorescence energies , has shown to be a reliable method for mechanistic studies of photochemical reactions involving conical intersections 27 ., A number of approaches for modeling nonadiabatic dynamics have been described in the literature , such as Tullys fewest switches surface hopping 28 , and multiple spawning 29 ., For recent reviews , see 30 , 31 ., In the context of QM/MM simulations , the surface hopping approach to photobiological problems has been pioneered by Warshel and coworkers 32 , 33 ., The diabatic surface hopping approach used in this work differs from the other approaches in two main respects ., First , in our approach a binary decision ( hop or no hop ) is made at each integration time step of the trajectory , based only on the current wavefunctions of the ground and excited states ., Second , hopping is only allowed at the conical intersection ( CI ) seam , where hopping probability approaches unity ., This could in principle underestimate the crossing probabilty , because we do not allow for transitions in regions of strong coupling but no real crossing ., However , for ultra-fast photochemical reactions in large polyatomic systems , decay predominantly takes place at the CI seam , as also shown by others 31 ., Thus , most surface hops are essentially diabatic , justifying our approach ., In addition , both energy and momentum are conserved upon a transition , as the trajectory never leaves the diabatic energy surface ., The theoretical background and algorithmic implementation of the diabatic surface hopping are detailed in the Supporting Information ( Text S4 ) ., Several theoretical studies on the photochemistry of the GFP chromophore have been conducted , applying both static ab initio 34–36 and DFT calculations 37 , and dynamics simulations based on a semi-empirical Hamiltonian 38 ., In addition , vertical excitation energies of asFP595 model chromophores in the gas phase and in a continuum dielectric were calculated by DFT and ab initio methods 39 , 40 , as well as in a minimal protein environment by means of DFT and CASSCF calculations within a QM/MM approach 41 ., By identifying key residues in the cavity of the asFP595 chromophore , our nonadiabatic QM/MM molecular dynamics simulations elucidate how the protein surrounding governs the photoreactivity of this photoswitchable protein ., Based on the simulations , we provide a new mechanism that qualitatively explains measured decay times and quantum yields , and that predicts the structures and protonation states of the photochemical intermediates and of the irreversibly fluorescent state ., We also suggest excited state proton transfer ( ESPT ) to play an important mechanistic role ., However , the detailed study of such ESPT processes is beyond the scope of this paper ., Our predictions can be probed by , e . g . , time-resolved Fourier transform infrared ( FTIR ) spectroscopy and x-ray crystallography ., The five excited state simulations that were initiated from the ground state trajectory of the trans neutral chromophore Ntrans are listed in Table 1 ., Trans-to-cis photoisomerization of the chromophore was observed in one of these simulations ( run b , Table 1; Video S2 in Supporting Information ) ., Figure 2 shows a schematic representation of the S0 ( green ) and S1 ( red ) potential energy surfaces of the neutral chromophore , along with a photoisomerization MD trajectory ( yellow dashed line ) ., Two coordinates are shown , the isomerization coordinate and a skeletal deformation coordinate of the chromophore ( see below ) ., The dynamics can be separated into three distinct phases:, ( i ) evolution on the electronic ground state S0 ,, ( ii ) excitation and evolution on the excited state S1 , and, ( iii ) decay back to S0 at the surface crossing seam followed by subsequent relaxation on the ground state surface ., The position of the surface crossing seam controls the passage of the trajectory from S1 to S0 ., The seam is accessed from a global twisted minimum on S1 , which is separated by a small S1 barrier from a local planar minimum near the Franck-Condon ( FC ) region ., In our simulations , individual excited state ( S1 ) lifetimes between 0 . 224 ps and 0 . 718 ps were observed ( Table 1 ) ., A simple exponential fit to the observed lifetimes yielded a decay time of τ\u200a=\u200a0 . 34 ps ( σ+\u200a=\u200a0 . 21 ps , σ−\u200a=\u200a0 . 13 ps , with σ being the statistical error ) ., Given the low number of trajectories , the statistical error of our estimated lifetime may seem unexpectedly low , but results from a rigorous analysis assuming an underlying single exponential decay 42 ., Recent femtosecond time-resolved pump/probe experiments by Schüttrigkeit and coworkers have yielded excited state decay time constants of 0 . 32 ps ( 78% ) , 2 . 6 ps ( 19% ) , and 12 . 1 ps ( 3% ) as well as a fluorescence lifetime of 2 . 2 ns for asFP595 43 ., However , although the simulated decay times seem to be in good agreement to the experimental results , we believe that the results should not be directly compared ., In previous work 24 , we demonstrated that the Ntrans protonation state is hardly populated in asFP595 and therefore is unlikely to contribute to the observed excited state decay ., Instead , the species that is predominantly responsible for the ultra-fast radiationless decay observed in the experiments is the anionic trans chromophore Atrans , as shown in detail below ., Figure 3A shows the snapshot from the isomerization trajectory ( run b , Table 1 ) shortly before the surface crossing seam was encountered ., After photon absorption ( blue arrow in Figures 2 and 3B ) , the chromophore spontaneously rotated around torsion A ( imidazolinone-twist ) , and the ring-bridging CH group pointed downwards ( away from His197 ) by almost 90° ., The time-evolution of the S0 and S1 potential energies and of the two ring-bridging torsion angles during trans-cis photoisomerization are shown in Figures 3B and 3C ., After excitation to S1 , the chromophore rapidly relaxed from the FC region into a nearby planar S1 minimum , as is evident from the decreasing S1 energy in panel b ( red curve ) ., The system stayed in this planar minimum for about 0 . 2 ps; subsequently the global twisted S1 minimum was reached through rotation around torsion A ( Figures 2 , 3B , and 3C ) ., The system then oscillated around this minimum until the conical intersection seam was encountered ., After the surface hop to S0 , torsion B ( hydroxyphenyl-twist ) rotated after a short delay of about 0 . 5 ps ., Previously , an ideal “hula-twist” isomerization mechanism of the zwitterionic chromophore was proposed based on a force field model 19 ., This hula-twist involves a simultaneous rotation around both torsion angles A and B . In the QM/MM simulations of the neutral chromophore presented here , rotation around both torsions was also observed ., However , twisting around torsion B was slower than around torsion A ( Figure 3C ) , and the isomerization hence proceeded via the twisted conformer shown in Figure 3A , with perpendicular imidazolinone and hydroxyphenyl moieties ., The hula-twist isomerization mechanism of the zwitterion at the QM/MM level will be discussed below ., During the initial equilibration of Ntrans , the hydrogen bonding network of the x-ray crystal structures of the anionic and zwitterionic chromophores ( see above ) changed to accommodate the non-native neutral chromophore ., First , a stable hydrogen bond formed between the hydroxyphenyl OH group of MYG and Glu145 ( Figure 3A ) ., Second , the hydrogen bonds between the imidazolinone nitrogen and Glu215 as well as between His197 and Glu215 ruptured ., Interestingly , these two hydrogen bonds were transiently re-established during the end of the second isomerization phase ( blue and orange curves in Figures 3D and 3E , respectively ) in which torsion B followed torsion A . In the twisted intermediate structure , the imidazolinone nitrogen atom was sterically more exposed as compared to the planar conformation , which facilitated the formation of the hydrogen bonds ., During an extended 10 ns force field simulation , the His197-Glu215 and MYG-Glu215 hydrogen bonds repeatedly broke and re-formed at a timescale of several hundred picoseconds ( data not shown ) , further underlining the flexibility of these two hydrogen bonds ., Figures 3D and 3E furthermore show that during the isomerization , the hydrogen bonding network in the chromophore cavity was stable , as none of the hydrogen bonds that were established at the instant of photoexcitation ruptured ., A similar stability was found in all simulations , irrespective of the chromophore conformation or protonation state ., Thus , the hydrogen bonding network around the chromophore is flexible enough to allow for photoexcitation and even photoisomerization without being ruptured ., For the neutral cis chromophore Ncis , five excited state simulations were initiated from the ground state trajectory ( Table 2 ) ., Only two of these trajectories returned to the ground state within 10 ps ( Table 2 , runs a and b ) , which was the maximum affordable trajectory length in terms of computation time ., In one of these two simulations , a spontaneous cis-to-trans photoisomerization was observed ( run a; Video S1 in Supporting Information ) ., As expected , the isomerization pathway was similar to the reverse trans-to-cis pathway in that the conical intersection seam was accessed via rotation around torsion A , followed by a slightly delayed rotation around torsion B in S0 ( see Figure S4 in Supporting Information ) ., However , in contrast to the activation pathway , the ring-bridging CH group rotated upwards ( i . e . , towards His197 ) ., Thus , despite the anisotropic protein surrounding , both rotational orientations of the chromophore CH bridge are feasible ., In the second simulation , the CI seam was also encountered after rotation around torsion A , but the chromophore returned to the initial cis conformation ., For the other three trajectories , the chromophore remained trapped in a planar S1 minimum conformation near the FC region throughout the simulation ( data not shown in Table 2 ) ., The starting structures of these trajectories were used for three additional simulations in which the escape from the planar S1 minimum was accelerated by means of conformational flooding 42 , 44 ( Table 2 , runs c–e ) ., In these accelerated simulations , the flooding potential successfully induced the escape from the local S1 minimum , and the surface crossing seam was encountered in all cases ., Isomerization was observed in two of these simulations ., In total , cis-to-trans photoisomerization was seen in three out of five simulations initiated in the Ncis state ., Although the number of trajectories is small , our simulations suggest that the probability for cis-to-trans photoisomerization is larger than for the reverse trans-to-cis process discussed before ., To further characterize the potential energy surfaces underlying the photochemical conversion processes of the neutral chromophore ( Figure 2 ) , we have optimized three S1 minima and a minimum energy S1/S0 conical intersection ( MECI ) in the gas phase ( see Supporting Information , Figure S1 , Tables S1 and S5 ) ., We found a local planar S1 minimum for the trans isomer 76 kJ/mol below the FC region ., The structure corresponding to the global S1 minimum is twisted around torsion A by 85° and lies −131 kJ/mol relative to the FC region ., The structure of the nearby MECI is twisted around torsion A by 81° ., The MECI is energetically lower than the FC region by 62 kJ/mol , and the CI seam is therefore readily accessible ., Twisting around torsion B instead of torsion A also leads to a local minimum on S1 , whose energy is 28 kJ/mol below the FC energy ., The MD simulations reflect this surface topology ., Immediately after excitation , the system relaxed from the FC region to the global S1 minimum by rotation around torsion A ( Figure 3C ) ., The system oscillated around this minimum until the conical intersection seam was encountered , with a subsequent surface hop back to S0 ., The gradients on S0 and S1 are almost parallel at the CI , which indicates that the CI is sloped ., The gradient difference vector and the derivative coupling vector that span the branching space largely correspond to skeletal deformations of the imidazolinone moiety ( see Supporting Information , Figure S1 ) ., Thus , as shown in Figure 2 , the rotation coordinate around torsion A is parallel to the seam and does not lift the S1/S0 degeneracy ., The seam is accessible anywhere along this torsional rotation coordinate , and therefore such torsional rotation is in principle not essential for the radiationless decay ., The extended surface crossing seam parallel to the isomerization coordinate accounts for the low isomerization quantum yield seen in our simulations ., In most of our MD simulations , the seam was encountered rather “early” along the torsional rotation coordinate ( Figure 2 ) , and the system thus returned to the ground state before overcoming the S0 barrier maximum ., In these cases , relaxation on S0 after the surface hop led back to the starting conformation ., To elucidate the influence of the protein environment on the photoisomerization process of the chromophore , we have re-calculated the S1 and S0 energies along two excited state trajectories ( run b , Table 1 and run a , Table 2 ) in the gas phase ., In these simulations , the chromophore followed the same trajectory as before , but did not interact with the rest of the system ( protein and solvent surrounding ) ., We have not attempted to further characterize the electrostatic influence of the surrounding by , e . g . , pKa calculations ., Figure 4A and 4C show the obtained energy traces ., In the protein , both S1 and S0 are stabilized with respect to the gas phase ., For the trans-to-cis isomerization process , the protein stabilized the energies of the S1 and S0 states on average by −339 kJ/mol and −307 kJ/mol , respectively ., For the cis-to-trans process , the average stabilization energies were −173 kJ/mol and −126 kJ/mol , respectively ., Thus , the protein ( and solvent ) environment favors S1 over S0 by about 30–50 kJ/mol ., Figure 4B and 4D show the energy differences between the protein and the gas phase , ΔE\u200a=\u200aE ( protein ) −E ( gas phase ) ., The S1 stabilization was rather strong at the surface crossing seam ( Figure 4 ) ., We found S1 to be stabilized stronger than S0 by 78 kJ/mol and 93 kJ/mol at the conical intersection in both MD simulations ., In summary , the protein environment energetically stabilizes S1 more than S0 , thereby enhancing fast radiationless decay ., In total , 20 simulations of the anionic chromophore protonation state were carried out , 10 of which were initiated in the trans conformation and the other 10 were initiated in the cis conformation ., Ultra-fast radiationless deactivation was observed in all 20 trajectories ( Table S6 in Supporting Information ) ., However , trans-cis photoisomerization never occurred ., A simple exponential fit to the S1 lifetimes of the trans anion yielded a decay time of τ\u200a=\u200a0 . 45 ps ( σ+\u200a=\u200a0 . 19 ps , σ−\u200a=\u200a0 . 12 ps ) ., Since Atrans is one of the two dominant protonation states in the off state besides Ztrans 24 , we expect Atrans to significantly contribute to the experimentally observed decay ., The measured decay time of 0 . 32 ps 43 agrees well with the decay time derived from the simulations ., For Acis , an excited state decay time of τ\u200a=\u200a1 . 81 ps ( σ+\u200a=\u200a0 . 77 ps , σ−\u200a=\u200a0 . 48 ps ) was obtained , which is about four times longer as compared to the decay time of Atrans ., Figure 5A shows the conical intersection geometry adopted during a typical trajectory ., In contrast to the neutral chromophore , the CI seam was accessed through a phenoxy-twist ( rotation around torsion B , see Figure 5C ) , and the CH bridge remained in the imidazolinone plane ., Shortly after excitation , rotation around torsion B drove the system towards the surface crossing seam ( Figure 5B and 5C ) ., Back on S0 , the system returned to the initial configuration ., The hydrogen bonding network in the chromophore cavity was very similar to the network observed in the x-ray crystal structures and remained stable during the excited state MD simulations ., Since rotation around torsion B does not lead to trans-cis isomerization and rotation around torsion A did not occur , the quantum yield for the isomerization of the anion was zero in our simulations ., However , due to the limited number of trajectories ( 20 ) , we cannot rule out the trans-cis photoisomerization of the anion ., Our results agree with recent MRPT2 computations of Olsen and coworkers on an anionic DsRed-like model chromophore in the gas phase 45 , who have shown that the imidazolinone-twisted S1/S0 CI ( i . e . , twisted around torsion A ) , which leads to cis-trans isomerization , is disfavored by more than 150 kJ/mol as compared to the phenoxy-twisted CI ., The latter CI is lower in energy than the planar S1 minimum by 9 . 2 and 48 . 1 kJ/mol at the MRPT2 and CASSCF levels , respectively 45 ., In our calculations , this energy difference is about 30 kJ/mol ( see Supporting Information , Table S2 ) , in qualitative agreement with Olsen and coworkers ., Due to the slightly different chromophores in DsRed and asFP595 , a quantitative agreement cannot be expected ., The deviation from planarity of the trans chromophore observed in the crystal structures was speculated to enhance ultra-fast deactivation 24 , 43 ., Indeed , the difference between the S1 lifetimes of the cis and trans conformers seen in our simulations can be attributed to steric constraints imposed by the protein matrix ., In the trans conformation the phenoxy-ring deviates from planarity by about 20° , whereas the cis chromophore is essentially planar 19 , 24 ., We observed in our simulations that only a slight additional twisting was required for the trans conformer to reach the surface crossing seam ., Thus , the pre-twisting of the phenoxy-moiety due to the protein matrix facilitated fast internal conversion of the trans conformer ., As shown in Supporting Information ( Figure S2 , Table S2 ) , we have optimized the S1/S0 MECI , a planar and two twisted S1 minima ( imidazolinone-twist and phenoxy-twist ) for an isolated anionic chromophore ., The planar minimum lies 31 kJ/mol below the FC point ., The structure of the global S1 minimum is twisted around torsion B by 269° and its energy lies −82 kJ/mol relative to the FC energy ., The MECI structure is also twisted about torsion angle B by 269° and is energetically lower than the FC point by 61 kJ/mol , thus explaining the ultra-fast decay seen in our MD simulations ., Twisting around torsion A leads to a local S1 minimum that is 46 kJ/mol below the FC point ., The CI of the anion is sloped , and the gradient difference vector corresponds to a skeletal deformation of the imidazolinone ring , analogous to the neutral chromophore ( see above ) ., In contrast to the neutral chromophore , the derivative coupling vector involves rotation around torsion B . However , the amplitude of this vector is small ., Thus , the two electronic states may remain close in energy along torsion B , allowing the system to decay at various phenoxy-twist angles ., To study the influence of the protein matrix on the deactivation process of the anionic chromophore , we have re-evaluated the S0 and S1 energies along two representative excited state trajectories ( trans and cis ) with all interactions between the QM atoms of the chromophore and the MM surrounding switched off , as was done for the neutral chromophores ( see above ) ., As Figure 5D and 5E show , the protein ( and solvent ) environment stabilizes the chromophore with respect to the gas phase ., The S0 and S1 states of the trans chromophore are strongly stabilized by −840 kJ/mol and −832 kJ/mol , respectively ., The S0 and S1 states of the cis chromophore were stabilized relative to the gas phase by −407 kJ/mol and −433 kJ/mol , respectively , during a representative Acis trajectory ( Figure S5 in Supporting Information ) ., Similar to the neutral chromophore , the protein environment favors the trans conformation over cis ., Before reaching the CI seam , the S0 and S1 states of the chromophore were stabilized to the same extent ., At the CI , however , the protein environment lowered the energy of the S1 state more strongly than the energy of the S0 state by 26 kJ/mol and 20 kJ/mol for Atrans and Acis , respectively ., This preferential stabilization of S1 enhanced the ultra-fast radiationless deactivation seen in our MD simulations ., Ten simulations were carried out for the zwitterion ., Five simulations were started in the Ztrans conformation , and the other five simulations were initiated in the Zcis conformation ., No decay back to the ground state was observed within a maximum trajectory length of 10 ps , neither for Ztrans nor for Zcis ., The chromophore did not escape from a planar S1 minimum in the vicinity of the FC region in any of the excited state simulations ., This suggests that Ztrans and Zcis could be the fluorescent species in asFP595 , although the measured fluorescence lifetime of 2 . 2 ns 43 is still much longer than our maximal trajectory length ( 10 ps ) ., For Ztrans and for Zcis , we have carried out three additional simulations each , in which we applied the conformational flooding technique 42 , 44 , 46 to accelerate the escape from the S1 minimum in an unbiased manner ( see Materials and Methods ) ., The results of these flooding simulations are shown in the Supporting Information ( Figure S6 , Text S3 ) ., The flooding potential induced isomerization of the chromophore , which followed a hula-twist pathway , in agreement with our previous work 19 ., From the flooding simulations , we obtained a lower bound for the excited state lifetime of the order of 1 ns ., The qualitative agreement with the measured fluorescence decay time of 2 . 2 ns provides further support for the assignment of the zwitterion as the fluorescent species ., The results thus obtained for the zwitterionic chromophore suggest that a hula-twist CI may be spontaneously accessed if the trajectories were extended to ( significantly ) longer times , i . e . , nanoseconds ., However , at the nanosecond timescale , fluorescence ( and not isomerization ) will be the predominant decay process ., Note that in Ref ., 19 , the energy barrier for hula-twist isomerization of the zwitterion was underestimated , thus favoring this isomerization over fluorescence ., In the next paragraph , we characterize the CI and show that the minimum energy crossing point for the zwitterionic chromophore has a high energy , thus hampering radiationless decay through hula-twist isomerization ., To characterize the topology of the S1 and S0 potential energy surfaces and of the conical intersection that occurs between them , multiconfigurational calculations were carried out for the isolated zwitterionic chromophore , as was also done for the other protonation states ( see above ) ., We optimized a planar S1 minimum and a hula-twist S1/S0 MECI ( see Supporting Information , Figure S3 , Tables S3 and S4 ) ., In contrast to the anion and the neutral chromophore , no twisted S1 minima were found ., The gradient difference vector and the derivative coupling vector at the MECI do not involve torsional rotation of either torsion A or torsion B , indicating that the CI seam lies parallel to the isomerization coordinate ., The MECI lies 70 kJ/mol above the planar S1 minimum and 23 . 4 kJ/mol above the FC energy ., Hence , in contrast to the anion and the neutral chromophore , no low-lying CI is present for the zwitterion , demonstrating that radiationless decay in the gas phase cannot occur in an unactivated manner ., For the CI seam to become accessible , a significant stabilization of S1 relative to S0 by the protein environment would be required ., However , as shown in Figure S6 in Supporting Information , the protein surrounding does not reduce the S1/S0 energy gap anywhere along the isomerization coordinate ., Our results suggest that the zwitterionic chromophore is potentially fluorescent , irrespective of the conformation ., However , the x-ray analysis of the emitting species has shown that only the cis chromophore fluoresces , whereas the trans chromophore is dark 19 ., A possible explanation for this discrepancy is the presence of an alternative deactivation channel that does not involve isomerization ., This deactivation pathway would have to be more easily accessible for Ztrans than for Zcis ., Only the latter would therefore be trapped in S1 and fluoresce ., The hydrogen bond between the NH group of the imidazolinone ring and Glu215 strongly suggests that the alternative decay involves an excited state proton transfer ( ESPT ) ., Such ESPT would quench the fluorescence , because the resulting anion rapidly deactivates , as shown above ., However , by including only the chromophore into the QM subsystem , we have excluded the possibility of observing such ESPT in our QM/MM simulations ., To identify possible ESPT pathways , we have carried out extended force field MD simulations of both Ztrans and Zcis and analyzed the relevant hydrogen bonds ., Figure 6A shows that , during the simulation of Ztrans , two stable hydrogen bonds were formed between the protonated OH group of Glu215 and His197 as well as between the NH proton of MYG and Glu215 ., These two hydrogen bonds allow for a proton transfer from Ztrans to the rapidly deactivating Atrans ., The OH proton of Glu215 could transfer to the Nδ atom of His197 , with a simultaneous or subsequent transfer of the NH proton of the imidazolinone moiety to Glu215 ., In contrast , during the force field simulation of Zcis , the MYG-Glu215 hydrogen bond remained intact , whereas the Glu215-His197 hydrogen bond broke after about 1 ns ( Figure 6B ) ., This differential behavior of Ztrans and Zcis was confirmed by two additional independent MD simulations ( data not shown ) ., Based on these results , we assume that only the trans zwitterion can be converted to the anion through a short proton wire ., Therefore , an ultra-fast deactivation channel is available only for the trans zwitterion , and not for the fluorescent cis zwitterion ., From the presence of the hydrogen bonding network in our force field trajectories , we do not obtain insights into the energetics of proton transfer ., Studying these transfers along the

Introduction, Results/Discussion, Materials and Methods

Fluorescent proteins have been widely used as genetically encodable fusion tags for biological imaging ., Recently , a new class of fluorescent proteins was discovered that can be reversibly light-switched between a fluorescent and a non-fluorescent state ., Such proteins can not only provide nanoscale resolution in far-field fluorescence optical microscopy much below the diffraction limit , but also hold promise for other nanotechnological applications , such as optical data storage ., To systematically exploit the potential of such photoswitchable proteins and to enable rational improvements to their properties requires a detailed understanding of the molecular switching mechanism , which is currently unknown ., Here , we have studied the photoswitching mechanism of the reversibly switchable fluoroprotein asFP595 at the atomic level by multiconfigurational ab initio ( CASSCF ) calculations and QM/MM excited state molecular dynamics simulations with explicit surface hopping ., Our simulations explain measured quantum yields and excited state lifetimes , and also predict the structures of the hitherto unknown intermediates and of the irreversibly fluorescent state ., Further , we find that the proton distribution in the active site of the asFP595 controls the photochemical conversion pathways of the chromophore in the protein matrix ., Accordingly , changes in the protonation state of the chromophore and some proximal amino acids lead to different photochemical states , which all turn out to be essential for the photoswitching mechanism ., These photochemical states are, ( i ) a neutral chromophore , which can trans-cis photoisomerize ,, ( ii ) an anionic chromophore , which rapidly undergoes radiationless decay after excitation , and, ( iii ) a putative fluorescent zwitterionic chromophore ., The overall stability of the different protonation states is controlled by the isomeric state of the chromophore ., We finally propose that radiation-induced decarboxylation of the glutamic acid Glu215 blocks the proton transfer pathways that enable the deactivation of the zwitterionic chromophore and thus leads to irreversible fluorescence ., We have identified the tight coupling of trans-cis isomerization and proton transfers in photoswitchable proteins to be essential for their function and propose a detailed underlying mechanism , which provides a comprehensive picture that explains the available experimental data ., The structural similarity between asFP595 and other fluoroproteins of interest for imaging suggests that this coupling is a quite general mechanism for photoswitchable proteins ., These insights can guide the rational design and optimization of photoswitchable proteins .

Proteins whose fluorescence can be reversibly switched on and off hold great promise for applications in high-resolution optical microscopy and nanotechnology ., To systematically exploit the potential of such photoswitchable proteins and to enable rational improvements of their properties requires a detailed understanding of the molecular switching mechanism ., Here , we have studied the photoswitching mechanism of the reversibly switchable fluoroprotein asFP595 by atomistic molecular dynamics simulations ., Our simulations explain measured quantum yields and excited state lifetimes , and also predict the structures of the hitherto unknown intermediates and of the irreversibly fluorescent state ., Further , we find that the proton distribution in the active site of the asFP595 controls the photochemical conversion pathways of the chromophore in the protein matrix ., Our results show that a tight coupling between trans-cis isomerization of the chromophore and proton transfer is essential for the function of asFP595 ., The structural similarity between asFP595 and other fluoroproteins suggests that this coupling is a quite general mechanism for photoswitchable proteins ., These insights can guide the rational design and optimization of photoswitchable proteins .

biophysics/theory and simulation, biochemistry/theory and simulation, computational biology, biophysics, computational biology/molecular dynamics

journal.pntd.0002940

Factors Affecting Perceived Stigma in Leprosy Affected Persons in Western Nepal

Leprosy is a chronic granulomatous disease caused by Mycobacterium leprae ., Besides clinical sequel followed usually after infection , the consequences of stigma associated with leprosy outweigh the burden of physical afflictions 1 ., Three kinds of stigma associated with leprosy affected persons have been described ., Experienced or enacted stigma refers to the real discrimination or acts experienced by leprosy affected persons while perceived stigma refers to the development of fear within an affected person where the fear may arise out of potential discrimination from family members , friends or society ., As a consequence of both enacted and perceived stigma , a person over a long period of time may believe what others think and say about him , resulting to the loss of self-esteem and dignity which is referred to be a self-stigma or internalized stigma 2 ., Stigma affects the psychosocial well-being of the affected person ., A person may feel fear or shame which can lead to anxiety and depression ., The resultant anxiety and depression may lead to decreased social participation and social exclusion 3 ., Anticipation of stigma may cause affected person to conceal their condition 4 ., The burden of keeping this secret , of being ever watchful and careful takes an emotional toll and adversely affects health seeking behavior 3 ., Concealing the disease , avoiding the questions regarding the disease and at times even telling lie for the fear of disclosure was found to be a major concern for leprosy affected persons attending Green Pastures Hospital , Nepal 5 ., Stigma has been found to be associated with misconceptions about the disease , visible deformities and the development of ulcers 4 ., Disability is a broad term covering any impairment , activity limitation or participation restriction affecting a person ., According to WHO , grade 0 means no disability is found ., Grade I means that loss of sensation has been noted in the hand or foot while grade II means the visible damage or disability is noted 6 ., Visible deformities and disabilities have been found to be the prominent contributor of stigma development in leprosy affected persons 7 while it also triggers the development of negative attitudes towards leprosy among unaffected people 8 ., In a systematic review of risk factors contributing to stigma , the basis of stigma development was found to be the visibility of the disfigurements and disability augmented by the stereotypes of the society , knowledge and the status of the person in terms of economy , education and ability to participate in society 9 ., In Nepal , leprosy is still a stigmatizing disease ., Misconceptions about the disease have contributed to the development of negative attitudes to leprosy affected persons ., In a study conducted in eastern Nepal , fear of infection and gods curse were found to be the most prevalent causes of negative behavior towards leprosy affected persons 8 ., In the other study 10 conducted in eastern part of Nepal , the causes of stigma perception in leprosy affected persons were consistent with the causes of negative attitudes in unaffected community members 8 ., The beliefs and perceptions about leprosy were found to be the prominent causes of stigma 10 ., Fear of infection , was the most important cause of stigma different countries including China 11 and India 12 ., In India , in addition to the fear of infection , false beliefs about leprosy , ignorance about the disease and lower socio-economic status were associated with stigma in leprosy 12 ., Therefore , we hypothesized that there is association between the levels of perceived stigma in leprosy affected persons and the factors characterizing them ( demographic characteristics , knowledge about leprosy , natural history of disease , clinical presentation , disability grades and reaction ) While few studies are done in eastern part of Nepal , most of them are focused on the impact of the stigma , participation restriction and income generation ., There has been no research so far in leprosy stigma in a view to explore the factors associated with it ., The specific objective of this study was to determine the prevalence of perceived stigma and its association with factors such as socio-demographic , knowledge about leprosy and clinical presentation characterizing leprosy affected persons attending Green Pastures Hospital and Rehabilitation Centre ., Green Pastures Hospital and Rehabilitation Centre , the only known leprosy referral center in western region of Nepal provides the services for leprosy patients with disability management , treatment and vocational training ., Therefore , exploring the risk factors of stigma in leprosy affected persons attending GPH&RC can help to understand the leprosy stigma and therefore can direct the stigma reduction strategies and intervention programs ., Ethical permission for this research was obtained from Nepal Health Research Council and International Nepal Fellowship Research Committee ., People were eligible if they were affected by leprosy , age above 18 years and willing to participate ., Interviews were only conducted after the written consent was received and was conducted by principal investigator ., Interviews were conducted with all leprosy affected people attending GPH&RC from February 2013 to March 2013 ., Attempt was done to include equal number of participants from the ward and OPD , 5 from the ward and 3 from the OPD denied the written consent , however , there were no drop outs ., The interviewer taking into the consideration the sensitivity of the subject established a friendly rapport before the interview and encouraged participants to express their views ., The anonymity of the participants was secured by coding the participants name ., No incentives were offered or paid for their time ., All participants who met the eligibility criteria were recruited into the study after taking written consent ., Total 135 participants were asked with the questionnaire form ., Among the questions representing different aspects of perceived stigma in EMIC questionnaire , most affected areas of perceived stigma were concealment of the disease , self-esteem , disclosure concern and the shame and embarrassment due to leprosy ( Table 1 ) ., Among the total participants 65 . 9% affirmed that they would conceal the disease condition as long as it is possible while 57 . 8% anticipated decreased self-esteem due to the disease condition and 40 . 7% only disclosed the disease condition to the close ones ., Of the 135 leprosy affected participants , 58 . 5% of them were those who attended OPD at the hospital ., Total median score of EMIC scale was higher among those leprosy patients who were in the ward compared to those who attended OPD ( p\u200a=\u200a0 . 006 ) ., There was no significant difference in mean EMIC score between different age groups ( p\u200a=\u200a0 . 199 ) , sex ( p\u200a=\u200a0 . 344 ) , ethnicity ( p\u200a=\u200a0 . 934 ) , location ( p\u200a=\u200a0 . 072 ) , marital status ( p\u200a=\u200a0 . 477 ) and family type ( p\u200a=\u200a0 . 356 ) ., Similarly , participants were asked if they had any other member of their family affected by leprosy in past or present including if they had relatives or neighbors affected by leprosy ., Neither of them had significant difference in median score of stigma ( Table 2 ) ., There was a significant difference in median EMIC score ( p\u200a=\u200a0 . 008 ) between different level of education in participants classified as illiterate ( those who could not read and write ) , those who attended primary level ( <5 years of education ) and those who attended secondary and higher education ( >5 years ) ., On post hoc analysis , the illiterate and those who attended more than 5years of education had significant difference in median score ( p\u200a=\u200a0 . 03 ) ., Similarly , when EMIC scores among subjects with less than 5 years education were compared with those with more than 5 years there was a significant difference ( p\u200a=\u200a0 . 016 ) while EMIC scores of the illiterate and those who attended <5 years of education were not significantly different ( p\u200a=\u200a0 . 673 ) ., There was no significant difference in median score between religious groups Hindu and other ( p\u200a=\u200a0 . 309 ) , Occupation ( p\u200a=\u200a0 . 321 ) , and amount of income ( p\u200a=\u200a0 . 068 ) ., However , on post hoc analysis two different income groups ( the highest and lowest income group ) showed significant difference ( p\u200a=\u200a0 . 011 ) There was a significant difference in EMIC score between those who felt economic inadequacy and who did not ( p\u200a=\u200a0 . 014 ) ., Similarly , there was also significant difference in stigma score between those who had to change their occupation after being affected by leprosy and those who did not ( p\u200a=\u200a0 . 018 ) ., Knowledge and perceptions about leprosy and perceived stigma scores were analyzed in all participants ., The overall stigma score for those who had knowledge about leprosy was lower than those who lacked knowledge of leprosy ( Table 3 ) ., There was a significant difference in EMIC stigma score between those who had information on leprosy ( p\u200a=\u200a0 . 025 ) , knowledge on leprosy cause ( p\u200a=\u200a0 . 02 ) and knowledge on transmission ( p\u200a=\u200a0 . 046 ) ., Similarly , participants who did not have knowledge of leprosy signs and symptoms had lower stigma scores compared to those who knew one or more signs and symptoms of leprosy although this was statistically insignificant ( p\u200a=\u200a0 . 344 ) ., There was a difference in EMIC stigma score who perceived leprosy as a very infectious disease ( p\u200a=\u200a0 . 127 ) ., Similarly , there was a significant difference in perceived stigma score between groups who felt that leprosy is difficult to treat ( p<0 . 001 ) and a severe disease ( p<0 . 001 ) ., Brief history of disease and clinical presentations were asked and assessed respectively with all the participants ( Table 4 ) ., Participants age at diagnosis ( p\u200a=\u200a0 . 213 ) and years after diagnosis ( p\u200a=\u200a0 . 967 ) did not show any difference in EMIC score ., First sign and symptoms were categorized into skin involvement , nerve involvement and deformity development ., Neither of them showed significant difference in perceived stigma score ( p\u200a=\u200a0 . 792 ) ., Similarly , there was no significant difference in EMIC between participants who sought hospital or doctor soon after development of signs and symptoms and who did not ( p\u200a=\u200a0 . 079 ) ., The majority ( 55 . 6% ) of participants received first treatment from non-medical providers such as witch doctors and traditional healers ., There was no significant difference in EMIC score between groups of participants who received treatment from medical providers , non-medical providers and friends/family and others ( p\u200a=\u200a0 . 255 ) ., Similarly , there was no significant difference in EMIC score between those who completed treatment and who did not ( p\u200a=\u200a0 . 156 ) ., There was a significant difference in EMIC score in participants who had disfigurement or deformities ( p\u200a=\u200a0 . 014 ) , ulcer ( 0 . 022 ) and odorous ulcer ( 0 . 043 ) compared to those who did not ., However , there was no significant difference in EMIC between those who had reaction and who did not ( p\u200a=\u200a0 . 331 ) ., More than half ( 51 . 1% ) of the participants had grade II disabilities and higher EMIC stigma score compared to grade 0 and grade I disabilities ( p\u200a=\u200a0 . 161 ) ( Table 5 ) ., However , the difference in EMIC stigma score showed marginal significance between grade II and grade 0 combined with grade I ( p\u200a=\u200a0 . 056 ) , not shown in table ., In majorities of the leprosy affected persons as evident from EMIC profile , concealment of the disease , lowered self-esteem and the disclosure to the close ones were major aspects of the EMIC questionnaire which contributed to higher EMIC score compared to the marital problems , social exclusion acts and impacts to their family members ., Focus group discussion with leprosy affected persons concluded that the discrimination and stigma attached to the disease was felt to be decreasing over the time ., However , the reasons for most of the participants intention not to disclose their disease condition were the fear of discrimination , isolation and rejection ., The most often reported cause of fear was the strongly rooted stereotype attached to the disease ., The most common belief leprosy affected person presumed was the fear of transmission of the disease among others ., In addition to the prevalent false beliefs about the transmission , severity and myths attached with the disease , the deformities and ulcers were also reported to be the triggering factor for the disease disclosure ., While most of the participants realized that ulcers and disabilities due to leprosy were affecting them physically , its psychosocial burden was the greater problem ., Some patients never reported to their close ones about their causes of disabilities and ulcers ., Instead they often told the causes of disabilities and ulcers to be due to some other disease ., However , participants realized that keeping this secrecy was a huge burden for them ., This study was conducted in western region of Nepal , where only those people who visited hospital for treatment , rehabilitation and wound care were recruited while many other people affected by leprosy who did not have any symptoms were not included in the study which limits our finding to generalize over all leprosy affected persons ., Only perceived stigma was assessed in this study while two other types of stigma were not assessed therefore , stigma in this study cannot be the whole picture of stigma ., While clinical presentations of the participants were obtained from the hospital treatment card , many other questions might have encountered recall biases ., The full evaluation of the data using multiple regressions was not done in this study which could have strengthened our findings ., This study concludes that lower education level , perceived economic inadequacy , obligation to change the occupation due to leprosy , lack of knowledge and the wrong perceptions about leprosy were the significant factors contributing to higher levels of perceived stigma in leprosy affected persons ., In addition to these socio-demographic factors , the presence of visible deformities , ulcers and disabilities also contributed to higher perceived stigma in leprosy affected persons ., The major aspects of EMC stigma scale affected were the attitude to conceal the disease , and lowered self-esteem ., The major causes for these have been explained by focus group discussion as the perceived fear of discrimination , rejection and the societys fear of transmission ., The factors contributing to the development of stigma in leprosy affected persons from this study can direct the need of intervention programs focusing on health education ., Health education which might correct the wrong perceptions and might increase understanding of leprosy and the people affected can have a significant impact in both leprosy affected persons and leprosy unaffected persons ., In addition to the education and health awareness programs , empowerment of the leprosy affected persons by technical education , vocational training and social participation might be helpful to increase self-esteem and reduce perceived stigma ., Ulcers and visible deformities have been found as contributing factors for the higher level of perceived stigma ., Early case detection through training of health professionals and health education to the general public might prevent the delays in presentation , ulcers , and deformities which ultimately can reduce the stigma .

Introduction, Materials and Methods, Results, Discussion

There are various factors which construct the perception of stigma in both leprosy affected persons and unaffected persons ., The main purpose of this study was to determine the level of perceived stigma and the risk factors contributing to it among leprosy affected person attending the Green Pastures Hospital , Pokhara municipality of western Nepal ., A cross-sectional study was conducted among 135 people affected by leprosy at Green Pastures Hospital and Rehabilitation Centre ., Persons above the age of 18 were interviewed using a set of questionnaire form and Explanatory Model Interview Catalogue ( EMIC ) ., In addition , two sets of focused group discussions each containing 10 participants from the ward were conducted with the objectives of answering the frequently affected EMIC items ., Among 135 leprosy affected persons , the median score of perceived stigma was 10 while it ranged from 0–34 ., Higher perceived stigma score was found in illiterate persons ( p\u200a=\u200a0 . 008 ) , participants whose incomes were self-described as inadequate ( p\u200a=\u200a0 . 014 ) and who had changed their occupation due to leprosy ( p\u200a=\u200a0 . 018 ) ., Patients who lacked information on leprosy ( p\u200a=\u200a0 . 025 ) , knowledge about the causes ( p\u200a=\u200a0 . 02 ) and transmission of leprosy ( p\u200a=\u200a0 . 046 ) and those who had perception that leprosy is a severe disease ( p<0 . 001 ) and is difficult to treat ( p<0 . 001 ) had higher perceived stigma score ., Participants with disfigurement or deformities ( p\u200a=\u200a0 . 014 ) , ulcers ( p\u200a=\u200a0 . 022 ) and odorous ulcers ( p\u200a=\u200a0 . 043 ) had higher perceived stigma score ., The factors associated with higher stigma were illiteracy , perceived economical inadequacy , change of occupation due to leprosy , lack of knowledge about leprosy , perception of leprosy as a severe disease and difficult to treat ., Similarly , visible deformities and ulcers were associated with higher stigma ., There is an urgent need of stigma reduction strategies focused on health education and health awareness programs in addition to the necessary rehabilitation support .

A total of 135 leprosy affected persons were interviewed with a questionnaire containing EMIC questions designed to assess the level of perceived stigma and the questionnaire containing variables for socio-demographic characteristics , knowledge about leprosy and the clinical presentations of the participants ., Clinical presentation as disability was graded according to WHO guidelines , where grade 0 means no disability found , grade I means loss of sensation has been noted in the hand or foot while grade II means visible damage or disability ., Total EMIC score was analyzed between sub-variables to see the factors associated with the higher level of perceived stigma score ., Additionally , among the total participants , we included 20 of them who were admitted at hospital for various reasons ., Two sets of focus group discussions were conducted with additional questions to derive the reasons behind frequently affected EMIC stigma domains ., The factors associated with higher perceived stigma score were illiteracy ( those who could not read and write ) , perceived economical inadequacy , lack of knowledge on leprosy , the perceptions as difficult to treat and severe disease and presence of visible deformities and ulcers ., Considering our findings pertaining to higher perceived stigma , there is an urgent need of stigma reduction strategies which should focus on health education about leprosy that can change the perceived stigma in leprosy .

medicine and health sciences, social sciences

journal.pbio.2006145

JMJD5 links CRY1 function and proteasomal degradation

Circadian rhythms are endogenous , approximately 24-hour oscillations in behavior and physiology that evolved as an adaptation to the day–night cycle ., These rhythms are generated by a cell-autonomous timekeeping mechanism known as the molecular circadian oscillator ., At its most basic , the oscillator is a transcription–translation circuit formed by two interlocked delayed negative feedback loops 1 ., In one loop , the transcription factors circadian locomotor output cycles protein kaput ( CLOCK ) and brain and muscle ARNT-like protein 1 ( BMAL1 ) drive expression of the genes coding for their own repressors , the CRYPTOCHROME ( CRY ) and PERIOD ( PER ) proteins , leading to alternative cycles of transcription activation and repression—the molecular basis of the clock ., In a second loop , the opposing actions of REV-ERB and ROR nuclear hormone receptors ( NHRs ) generate strong oscillations in Bmal1 gene transcription , which contributes to robust amplitude in circadian rhythms ., However , the function of the circadian oscillator involves a much larger repertoire of factors that include other transcription regulators , kinases , phosphatases , ubiquitin ligases and peptidases , and chromatin regulators ., Together , this large cohort of molecules acts in concert to generate circadian rhythms , coordinate the clock with other physiological processes , and enable environmental information to be integrated into its function ., A key mode by which circadian rhythms are generated and fine-tuned is by the regulation of the protein levels of the core oscillator components 2 ., For instance , phosphorylation of PER proteins by Casein kinase I ( CKI ) decreases their stability by stimulating their interaction with and ubiquitylation by the Skip-Cullin-F box ( SCF ) β-TRCP1/2 ubiquitin ligase complex 3 , 4 ., Similarly , degradation of REV-ERBα by Homologous to the E6-AP Carboxyl Terminus ( HECT ) - ( ARF-BP1 ) and Really Interesting New Gene ( RING ) -class E3 ligases ( PAM ) is induced by phosphorylation by glycogen synthase kinase 3β ( GSK3β ) 5 , 6 ., BMAL1 is also phosphorylated by GSK3β , leading to its destabilization 7 ., Although BMAL1 ubiquitylation has been found to be catalyzed by the HECT-class E3 ligase UBE3A 8 , a link between this process and GSK3β-mediated phosphorylation has not been found ., Taken together , these observations show that , although the mechanisms that control the stability of the clock proteins are similar , involving coordinated phosphorylation and ubiquitylation , there is divergence in the machinery that targets different components ., In mammals , CRY degradation is mediated by SCF ubiquitin ligase complexes that contain one of two closely related F-box/leucine-rich repeat proteins ( FBXLs ) , FBXL3 and FBXL21 9–13 ., Although the two ligases can both ubiquitylate CRYs , their actions are antagonistic ., In the nucleus , FBXL3 promotes K48-linked polyubiquitylation of CRYs , leading to its degradation , whereas FBXL21 binds with greater affinity to CRY yet catalyzes K48 polyubiquitylation less efficiently than FBXL3 13 ., Thus , presence of FBXL21 diminishes the overall CRY degradation ., Despite its presence in the nucleus , FBXL21 localizes primarily to the cytoplasm , where it promotes CRY degradation , highlighting the complexity of CRY1 regulation ., As with other clock proteins , CRY1 degradation is controlled by phosphorylation , most notably by the AMP-regulated protein kinase ( AMPK ) 14 ., AMPK-mediated phosphorylation of CRY1 strengthens interactions with FBXL3 , thereby leading to CRY destabilization ., Yet , despite the fact that both mammalian CRY paralogs largely share the same degradation machinery , differences in how their levels are controlled appear to exist 15–17 ., Members of the JmjC domain–containing family of proteins are characterized by a cupin-type domain of about 150 amino acids , known as the JmjC domain , which is able to confer lysine demethylase activity to some but not all proteins that harbor it 18 ., In recent times , members of the JmjC family have emerged as important regulators involved in a variety of physiological processes , including control of circadian rhythms in plant , mammalian , and insect systems 19–22 ., Previously , a genetic study identified Arabidopsis thaliana Jmjd5 ( AtJmjd5 ) as a regulator of the circadian system in plants that exhibits sufficient functional conservation with its mammalian ortholog JmjC domain–containing protein 5 ( JMJD5 ) as to exhibit reciprocal rescue of circadian phenotypes arising from genetic ablation in plants or small interfering RNA ( siRNA ) knockdown in U2OS cells 19 ., Similarly , in Drosophila , genetic deletion of JMJD5 leads to reduced period length in locomotor activity and decreased sleep 23 ., However , though these studies firmly establish JMJD5 as an evolutionarily conserved participant of the clock , its mechanism of action in the clock has remained undefined ., Although JMJD5 has been suggested to be a lysine demethylase , such a function remains highly debated and is not yet firmly established 24–26 ., Nonetheless , JMJD5 has been reported to influence gene transcription through several mechanisms , including modulation of protein levels , nuclear entry of transcription factors , and proteolytic processing of histone subunits 25 , 27–30 ., We found that JMJD5 is recruited to CRY1–FBXL3 complexes , in which it facilitates CRY1 interaction with the proteasome ., Furthermore , we report that JMJD5-dependent CRY1 destabilization is intertwined with the repressive function of CRY1 ., To determine whether JMJD5 plays a role in the mammalian oscillator , we first analyzed the impact of its deletion on the circadian clock of mouse embryo fibroblasts ( MEFs ) ., We measured gene expression levels of core circadian oscillator components in a circadian timeline from Jmjd5+/+ and Jmjd5−/− MEFs harvested at 4-hour intervals from 12 to 56 hours post synchronization with dexamethasone ( Fig 1A and S1 Fig ) ., Cells that lack JMJD5 exhibit marked down-regulation of Clock and Bmal1 mRNAs , the two central circadian transactivators ., Consistent with a decrease in CLOCK–BMAL1 activity , mRNAs of their regulatory targets Dbp , Cry1 , Nr1d1 , and Rora showed similar decreases ., In contrast , the impact of JMJD5 deficiency on Cry2 , Per1 , and Per2 gene expression was divergent , with decreased Per2 , unaffected Cry2 , and increased Per1 mRNA abundances ., Real-time bioluminescence measurements from a Per2 promoter-luciferase reporter also showed circadian dysfunction ( S2 Fig ) , as Jmjd5-null cells exhibited shortened period length and decreased amplitudes in their oscillation , which not only confirmed our gene expression observations but also were consistent with previous reports 19 , 23 ., Next , we assessed the impact of JMJD5 on circadian clock gene expression in vivo ., Full-body Jmjd5-null mutant mice are embryonic lethal 26 , but hepatocyte-specific Jmjd5-ablated animals ( Jmjd5 liver knockouts Jmjd5LKO ) are viable and exhibit no overt phenotype ., We generated a circadian liver timeline from wild-type and Jmjd5LKO animals at a 4-hour resolution ., As observed in fibroblasts , JMJD5 deficiency disrupts clock gene expression in a similar albeit nonidentical manner ., Specifically , in both cells and liver that lack JMJD5 , Per1 mRNA was increased , and those of Clock , Per2 , and Rora were decreased ( Fig 1A and 1B ) ., In contrast to fibroblasts , JMJD5-null livers showed no defect in the expression patterns of Dbp , Bmal1 , Cry1 , and Nr1d1 mRNA ( Fig 1B and S1 Fig ) ., Next , we assessed the impact of JMJD5 on CLOCK–BMAL1-mediated transcription in a series of real-time luciferase-reporter assays performed in the non-oscillating human embryonic kidney 293T ( HEK293T ) cell line ., When coexpressed , JMJD5 decreased CLOCK–BMAL1 activation from Per1- and Per2-driven promoter-driven luciferase reporters in a dose-response manner ( Fig 2A and 2B ) ., Repression of CLOCK–BMAL1 by JMJD5 is dependent on the presence of a functional E-box ( Fig 2C ) ., We observed that in the E-box mutant promoter , CLOCK–BMAL1 had a suppressive effect; it is possible that this effect is due to sequestration of limiting factors by CLOCK–BMAL1 away from the mutant promoter construct ., Further , inclusion of JMJD5 with CLOCK–BMAL1 did not change this effect ., In the absence of CLOCK–BMAL1 , JMJD5 had only a minor repressive effect on the wild-type Per1 promoter and no such effect in the E-box mutant construct , demonstrating that CLOCK–BMAL1-mediated activation is required for JMJD5 repression ( Fig 2D ) ., Although disputed , JMJD5 has been suggested to possess catalytic activity by virtue of its JmjC domain ., To define whether any such activity is necessary for its effect on CLOCK–BMAL1 activity , we assessed CLOCK–BMAL1 repression by JMJD5H321A , a mutant construct that harbors a mutation in a conserved residue required for cofactor binding by the JmjC domain , thus precluding any enzymatic function 30–32 ., JMJD5H321A repressed CLOCK–BMAL1 activation of both Per1- and Per2-luciferase reporter constructs , indicating that the circadian function of JMJD5 does not require catalytic activity ( Fig 2E and 2F ) ., The only other two JmjC proteins that have been shown to participate in the mammalian clock—JARID1A and FBXL11/lysine-specific demethylase 2A ( KDM2A ) —also do so in a catalytically independent manner 21 , 22 ., JMJD5 has previously been reported to influence other transcription factors via regulation of their stability 25 ., Thus , we performed cycloheximide chase assays to assess whether JMJD5-mediated repression of CLOCK–BMAL1 was due to induction of their degradation ., JMJD5 did not influence CLOCK or BMAL1 degradation but instead markedly destabilized CRY1 in a catalytically independent manner ( Fig 3A and 3B ) ., In contrast , JMJD5 had no effect on other clock proteins , including CRY2 ( Fig 3C and S5 Fig ) ., Consistent with our cycloheximide assays , we found elevated CRY1 levels in both liver nuclear and whole extracts of Jmjd5-null livers compared to wild-type controls ( Fig 3D and 3E and S6 Fig ) ., In nuclear extracts from JMJD5-null fibroblasts , CRY1 was slightly increased , even though Cry1 mRNA levels were much decreased ( Fig 3F ) , which is the exact same situation observed in livers of FBXL3 mutant animals 9 ., To determine whether JMJD5 participates directly in regulation of CRY1 degradation , we interrogated its ability to associate with CRY1–FBXL3 complexes ., In coimmunoprecipitation studies , we found that JMJD5 interacts with CRY1 , but not CRY2 , and that this association was enhanced when FBXL3 was coexpressed ( Fig 4A and S9 Fig ) ., CRY1 degradation by FBXL3 is induced via AMPK-mediated phosphorylation of CRY1 residues S71 and S280 14 ., Interaction analyses between FBXL3 and CRY1 constructs harboring phospho-null or phospho-mimetic mutations at these sites showed that although phosphorylation of CRY1 by AMPK increases its binding to FBXL3 , it is not required for basal interaction between these two proteins 14 ., In a series of coimmunoprecipitation experiments , we found that the binding pattern of JMJD5 to the different phosphosite mutants tested by Lamia and colleagues 14 and to a non-ubiquitylatable CRY1 mutant 13 paralleled that of FBXL3 , which further confirmed the existence of CRY1–FBXL3–JMJD5 complexes ( Fig 4B and 4C ) ., Association of JMJD5 with CRY1 is dependent on the presence of FBXL3 , as RNA interference ( RNAi ) -mediated knockdown of the latter led to a decrease of CRY1–JMJD5 association ( Fig 4D and 4E ) ., In contrast , knockdown of CRY1 did not impact FBXL3 association with JMDJ5 , nor did knockdown of JMJD5 abrogate CRY1–FBXL3 interactions ( Fig 4F and 4G ) ., Together , these data suggest that FBXL3 bridges the interaction between CRY1 and JMJD5 ., As FBXL3 mediates CRY1 ubiquitylation , we assessed whether a defect in this process was responsible for the increased CRY1 levels we observed in a JMJD5-null genetic background ., To achieve this , we treated control and JMJD5-null MEFs that expressed FLAG-tagged CRY1 with MG132 to block proteasomal degradation ( Fig 4H ) ., Ubiquitylation was not affected ., At any given timepoint , the intensity of ubiquitylated CRY1 signal was greater in JMJD5-null than in control cells ., However , non-ubiquitylated CRY1 was also increased so that the ratio of these two remains unaffected ( S10C Fig ) , indicating that ubiquitylation of CRY1 was normal ., We noted that total CRY1 levels in Jmjd5+/+ MEFs increased to 400% of baseline levels after 8 hours of MG132 treatment , yet no similar increase over baseline occurred in Jmjd5−/− cells ( Fig 4I and S10 Fig ) ., Quantification of the non-ubiquitylated band alone yielded similar results ( S10 Fig ) ., These results suggest a reduction in CRY1 degradation by the proteasome in JMJD5-null cells , even while the normal process of ubiquitylation is unaffected ., Coincidentally , JMJD5 has been reported to copurify with 19S proteasome regulatory particle non-ATPase 1 ( RPN1 ) , the largest 19S proteasome cap subunit 33 ., Of note , RPN1 constitutes a docking site for shuttling proteins that help target ubiquitylated substrates to the proteasome 34 ., Based on these observations , we hypothesized that JMJD5 was required for normal CRY1 interaction with the proteasome ., To test this , we transfected Jmjd5+/+ and Jmjd5−/− cells with an HA-RPN1 expression construct and assessed its ability to associate with endogenous CRY1 ., We found that CRY1 association with RPN1 was significantly diminished in JMJD5-null cells ( Fig 4J and 4K ) , which argues that JMJD5 facilitates CRY1 targeting to the proteasome ., The seeming paradox posed by the repressive effect of JMJD5 on CLOCK–BMAL1 while simultaneously promoting CRY1 degradation could be resolved if the repressive function of CRY1 was coupled to its degradation ., To test this possibility , we performed real-time luciferase assays in non-oscillating HEK293T cells to compare the repressive potential of wild-type to the stable CRY171A/280A mutant ., Consistent with the idea that the repressive function of CRY1 is linked with its degradation , CRY171A/280A repression of CLOCK–BMAL1 activation of a Per1-luciferase reporter was markedly impaired , a defect that was most pronounced when comparing conditions with similar protein levels of wild-type and mutant CRY1 ( Fig 5A–5C and 5G ) ., Repression of CLOCK–BMAL1 by CRY171A/280A was much less impacted on a Per2-luciferase reporter construct ( Fig 5D–5F ) ., As JMJD5 inhibition of CLOCK–BMAL1 was lower on a Per2 than on a Per1-luciferase construct ( Fig 2A and 2B , S4B and S4C Fig ) , it is possible that these observations reflect differences in the regulation of these promoters , consistent with previous reports of differential regulation of PER genes 35–37 ., Next , we determined whether JMJD5 could act in concert with CRY1 to repress CLOCK–BMAL1 ., To this end , we measured the repressive activity of wild-type CRY1 in the presence or absence of JMJD5 coexpression ( Fig 5H and 5I ) ., To be able to determine the existence of either cooperation or synergism between CRY1 and JMJD5 , we transfected suboptimal amounts of CRY1 and JMJD5 plasmids and found that they co-repressed CLOCK–BMAL1 ., We also observed that the stable CRY171A/280A mutant cooperated with JMJD5 to an extent similar to the wild type , indicating that JMJD5 could rescue CRY1 activity ( Fig 5H and 5I , S11 Fig ) ., Consistently , JMJD5 coexpression destabilized CRY171A/280A to the same extent as wild-type CRY1 ( Fig 5J and 5K ) ., A possible explanation for the ability of JMJD5 to rescue CRY171A/280A is that increased JMJD5 availability in the context of basal FBXL3–CRY1 interaction leads to increased proteasome degradation ., In all , these results strongly support the idea that the repressive ability of CRY1 , at least in some contexts , is linked to its degradation ., A key feature of the circadian oscillator is its ability to integrate environmental and cellular information with its machinery ., This occurs via modulation of its different molecular components by different signaling pathways ., AMPK is a master regulator of energy homeostasis that relays information to the circadian clock via CRY1 14 ., As JMJD5 is required for normal CRY1 degradation , we next explored whether it played a role in AMPK-induced CRY1 degradation ., To do this , we assessed the effect of AMPK activation on CRY1 levels in wild-type and Jmjd5−/− MEFs 28 ., In the absence of JMJD5 , the basal stability of CRY1 was much greater than in wild-type cells , and AMPK activation by 5-Aminoimidazole-4-carboxamide 1-β-D-ribofuranoside ( AICAR ) treatment failed to induce the rapid destabilization of CRY1 seen previously by Lamia and colleagues ( Fig 6A ) 14 ., In contrast , reconstitution of JMJD5 sensitized CRY1 levels to the effects of AMPK activation ( Fig 6B ) ., Together , these data indicate that JMJD5 has a critical role in control of CRY1 stability by AMPK–FBXL3 axis ., We next asked whether JMJD5 impinges on other biological functions of FBXL3 or CRY1 besides circadian transcription ., We first looked at c-MYC levels because its stability is regulated by FBXL3 in conjunction with CRY2 38 ., We interrogated nuclear extracts of control and Jmjd5−/− MEFs ., Across all time points assessed , we found elevation of c-MYC levels in the absence of JMJD5 ( Fig 6C ) ., Next , since CRY1 interacts with and modulates the activity of several NHRs 39 , 40 , we assessed whether hepatic JMJD5 ablation impacted the expression profile of genes regulated by NHR partners of CRY1 ( Fig 6D ) ., In JMJD5-null livers , we observed increased levels of genes regulated by liver X receptor α and β ( LXRα/β ) and liver receptor homolog 1 ( LRH1 ) ( Abcg5 and Abcg8 ) 41–43 , peroxisome proliferator–activated receptor δ ( PPARδ ) ( Lpl ) 44 , and pregnane X receptor ( PXR ) ( Gstm3 ) 45 , 46 ., On the other hand , in the absence of JMJD5 , we saw decreases in expression of genes regulated by hepatocyte nuclear factor 4α ( HNF4α ) and PPARγ ( Cyp27a1 ) 47 , 48 and by glucocorticoid receptor ( GR; Angptl4 , Pck1 , Lipg ) 39 , 49 , 50 ., In contrast to increased Abcg5 and Abcg8 levels , the expression of another LXR target , Cyp51 51 , was decreased in Jmjd5LKO livers ., Finally , we also found decreased expression of HNF4α , an NHR that regulates a broad range of hepatic processes and whose promoter region is bound by CRY1 52 ., In this study , we shed light on the mechanisms by which JMJD5 participates in the circadian oscillator ., Specifically , we show that JMJD5 plays a role in CRY1 function and in the regulation of stability ., First , JMJD5 expression destabilizes CRY1 but not other circadian proteins ., Our data indicate that FBXL3–JMJD5 complexes promote CRY1 degradation by the proteasome ., In agreement with this , CRY1–proteasome association is greatly diminished in the absence of JMJD5 ., We found that although JMJD5 is required for normal CRY1 degradation , it nonetheless cooperates with CRY1 to repress CLOCK–BMAL1 , which indicates that CRY1 destabilization and function are , in some cases , positively linked ., Indeed , repression of CLOCK–BMAL1 by a degradation-resistant CRY1 mutant is drastically impaired , and JMJD5 simultaneously rescues its functional and stability defects ., Though the phenomenon of activation-coupled degradation has been observed in other tightly controlled transcription factors—including the Aryl hydrocarbon Receptor ( AhR ) , Estrogen Receptor alpha ( ERα ) , Sma and Mad homolog 2 ( SMAD2 ) , signal transducer and activator of transcription ( STATS ) , and even CLOCK–-BMAL1 53 , 54—this is the first time it has been described for a circadian repressor ., Until now , the view regarding the relationship between the repressive function of CRY1 and its protein levels has been that these have direct correlation ., As this perspective has been largely shaped by studies involving mechanisms that regulate levels of both CRY1 and CRY2 , it is likely that , in such context , differences in the function and regulation between the two CRY paralogs may be obscured ., Since the mechanism we describe here is specific to CRY1 , we are now able to better define how regulation of CRY1 levels relates to its activity ., Cells deficient in JMJD5 exhibit dysregulation in circadian gene expression , albeit with a pattern diverging from simple E-box regulation , which is consistent with previous studies ., A Gene Dosage Network Analysis ( GDNA ) by Baggs and colleagues , for instance , showed that clock gene expression responses to circadian network perturbations are complex , depend on the specific oscillator component that is being disrupted , and do not always follow predicted changes based on transcriptional relationships 55 ., In that study , siRNA-mediated depletion of Clock reduces Nr1d1 and Nr1d2 levels , has a marginal impact on Per1 , has no effect on Cry2 and Per2 , and results in slight increases in Cry1 mRNA , all of which are canonical target genes of CLOCK–BMAL1-mediated activation ., In addition , Cry1 depletion in U2OS cells clearly increases levels of Per2 and Cry2 but has no apparent impact on Per1 , Nr1d1 , or Nr1d2 ., However , our observations in both Jmjd5-deficient cells and liver do have overlap with findings by Baggs and colleagues ., Specifically , the opposite changes in Per1 and Per2 mRNA expression we observed in JMJD5-null cells ( increase and decrease , respectively ) are consistent with the unidirectional Period paralog compensation in gene expression observed by Baggs and colleagues , in which Per1 depletion increases Per2 levels but not in the reverse 55 ., Our functional assays suggest that , at least in certain contexts , JMJD5 may have a more prominent role in control of Per1 transcription than in that of Per2 ( Fig 2B and S4 Fig ) ., Consistently , increased CRY1 stability had a much greater impact to repress a Per1-luciferase construct than one driven by the Per2 promoter ( Fig 5A–5F ) ., These observations may reflect reports that control of Per1 and Per2 transcription is not identical and raises the intriguing possibility that JMJD5 has a role in the mechanisms behind paralog compensation ., Similarly , complex patterns of clock expression occur in tissues of clock component knockout mice ., Single knockout of CRY1 or single knockout of CRY2 affects transcription not only differently across genotypes but also across tissues within a genotype 56 ., For example , Per2 mRNA levels in the liver of CRY1 knockout mice are elevated and rhythmic but are arrhythmic and mostly reduced in the cerebellum ., CRY2 ablation , on the other hand , does not result in derepression of Per2 in liver , a canonical E-box driven target , but quite the opposite ., In CLOCK knockout mice , Per1 mRNA levels drop in the hypothalamic suprachiasmatic nucleus ( SCN ) but rise in the liver , whereas the phase of Per2 mRNA rhythm is shifted without any impact on its levels 57 ., In contrast , Per2 mRNA in CLOCK-null mouse liver is elevated only during the nadir of expression , whereas Dbp and Nr1d1 levels are decreased despite drastically elevated Bmal1 gene expression 57 ., Finally , a recent study by Ramanathan and colleagues found that knockdown of canonical clock genes ( e . g . , Cry1 , Per1 , Per2 , Nr1d1 ) do not always result in the same circadian effect in different cell lines 58 ., Altogether , these observations indicate that deletion of a single clock regulator—even of canonical clock components—can lead to nonintuitive effects , which help explain our observations here ., Hepatic ablation of JMJD5 also resulted in abnormalities in circadian gene expression ., As with Jmjd5−/− cells , we observed elevation of Per1 levels with a simultaneous decrease in Per2 levels , again suggestive of paralog compensation ., As with Per genes , paralog compensation in Cryptochromes occurs unidirectionally , so that knockdown of Cry1 gene expression in U2OS cells leads to an increase in Cry2 transcript but not vice versa ., In JMJD5-deficient liver , we noted a slight increase in Cry2 mRNA levels , which could reflect a decrease in CRY1 function even if Cry1 transcripts remain unaltered ., In contrast to cells , we observed no changes in the levels of Bmal1 , Dbp , Cry1 , or Nr1d1 transcripts in JMJD5-null liver tissue ., A possible explanation for the differences in the impact that JMJD5 deletion has on cells and liver is that circadian clock control is not identical in all cell types and/or due to divergence in how the clock is regulated ( e . g . , differences in paralog levels ) in cultured cells versus in vivo ., Knockdown of canonical clock genes ( e . g . , Cry1 , Per1 , Per2 , Nr1d1 ) does not always result in the same circadian effect in different cell lines 58 , which lends further support to this idea ., In a previous study , Huber and colleagues demonstrated that CRY2–FBXL3 specifically regulates c-MYC protein stability ., Likewise , we find that c-MYC levels are affected , suggesting that JMJD5 may impact FBXL3 function beyond CRY1 degradation , yet this effect could be indirect , given that JMJD5 did not impact or interact with CRY2 ., Huber and colleagues also noted a moderate yet noticeable increase in overexpressed c-MYC upon Cry1 ablation 38 ., Thus , elevation of c-MYC levels in the absence of JMJD5 is consistent with a deficit in CRY1 function ., A possibility is that ablation of JMJD5 disturbs the balance between CRY1–FBXL3 and CRY2–FBXL3 complexes ., Nonetheless , the precise mechanisms by which JMJD5 influences c-MYC function remain to be discovered ., Second , CRY1 interacts with and participates in NHR-mediated transcription control ., In JMJD5-deficient livers , we found abnormal expression of several genes regulated by one or more NHRs that are known to interact strongly with CRY1 ., The genes we assessed are known to code for important components of metabolic processes , including cholesterol metabolism ( Abcg5 , Abcg8 , Cyp27a1 , Cyp51 ) , lipid metabolism ( Angptl4 , Lpl , Lipg ) , glucose metabolism ( Pck1 ) , and xenobiotic detoxification ( Gstm3 ) ; their dysregulation suggests that JMJD5 function may have an important role in regulation of liver physiology by CRY1 ., We found both up-regulation and down-regulation in the expression of these genes , which is reminiscent of what we observed in oscillator components ., We found decreased expression of genes regulated by NHRs that are repressed to a similar extent by both CRY1 and CRY2 ( e . g . , the GR target Angptl4 ) , which is consistent with increased CRY1 levels ., In contrast , genes controlled by PPARδ and PXR , which are more strongly repressed by CRY2 than by CRY1 , were moderately derepressed 40; a possible explanation is that this effect is due to a decreased association of CRY2 with PPARδ and PXR as a consequence of increased CRY1 availability ., Finally , our observation that JMJD5 seems to impinge on the core clock through CRY1 but not through other core components is intriguing ., Several observations suggest that CRY1 serves a unique repressive function ., First , CRY1 is able to bind and repress CLOCK–BMAL1 independently of PER 59 ., In liver , CRY1 has a markedly different temporal genomic occupancy pattern than that of other circadian oscillators 52 ., Furthermore , a recent study found that under certain conditions , most circadian proteins are only detectable as part of a large multiprotein complex , with the exception of CRY1 and CKIδ; in that study , both were detected as uncomplexed from other clock components 60 ., When considered together , these observations support the existence of CRY1-specific regulatory mechanisms and thereby suggest that CRY1 constitutes a unique node through which the molecular oscillator machinery is fine-tuned ., This work involved the killing of animals by cervical dislocation , as approved by the University of Kansas Medical Center Institutional Animal Care and Use Committee ( IACUC ) ( protocol # 2015–2292 ) ., HEK293T cells were purchased from the American Type Culture Collection ( ATCC ) ., Cells were cultured in Dulbecco’s Modified Eagle Medium ( DMEM ) ( Corning Cat# 10-013-CV ) supplemented with 10% FBS ( Atlanta Biologicals Cat# S11595H ) and 1% antibiotics and antimycotics ( Thermo Fisher Cat# 15240062 ) in a 37 °C incubator maintained at 5% CO2 ., Transfections were performed using Trans-IT LT1 ( TLT-1 ) ( Mirus Bio Cat# MIR 2304 ) according to the manufacturer’s instructions ( specific conditions described below ) ., HEK293T cells were seeded out in 24-well plates at 80 , 000 cells per well ., Twenty-four hours later , they were transfected with 200 ng of FLAG-CRY1 , 40 ng of FLAG-JMJD5 , 260 ng of pCDNA3 . 1 vector ( 500 ng total ) , and 1 . 5 μl of transfection reagent ., Forty-eight hours post transfection , cycloheximide ( Sigma Cat# C7698-1G ) was added to each well to a final concentration of 100 μg/ml , and the cells were harvested every 2 hours ., Jmjd5+/+ and Jmjd5−/− immortalized MEFs were obtained from Dr . Ralf Janknecht’s laboratory and have been previously described 28 ., Jmjd5+/+ and Jmjd5−/− MEFs were seeded out in 6-well plates at a concentration of 350 , 000 cells per well ., After 20 hours , the cells were transfected with 1 μg of FLAG-CRY1 and 1 . 5 μl of TLT-1 ., Forty-eight hours post transfection , the cells were treated with a mixture of cycloheximide ( 100 μg/ml ) ± 3 mM AICAR ., For rescue experiments , Jmjd5−/− were seeded out in 6-well plates at a concentration of 350 , 000 cells per well ., After 20 hours , the cells were transfected with 1 μg of FLAG-CRY1 and 100 ng of FLAG-JMJD5 ( or pCDNA 3 . 1+ vector ) ., Forty-eight hours post transfection , the cells were treated with a mixture of cycloheximide ( 100 μg/ml ) ± 3 mM AICAR and harvested at the indicated time points ., Jmjd5+/+ and Jmjd5−/− MEFs were seeded out in 6-well plates at 350 , 000 cells per well and 24 hours later were transfected with 1 μg FLAG-CRY1 ., After 48 hours of transfection , the cells were treated with 10 μM MG132 , and cells were harvested every 4 hours for 12 hours ., FLAG-CRY1 , FLAG-CRY1S71A , FLAG-CRY1S280A , and FLAG-CRY1AA in pcDNA3 . 1+ expression backbone were a gift from Katja Lamia ., FLAG-JMJD5 and V5-JMJD5 in the pEV3S backbone and HA-JMJD5 in the pQCXIH backbone were generated by Ralf Janknecht ., The CRY1 ( K:R ) -HA construct was a kind gift of Dr . Joe Takahashi ., Real-time luciferase assays were performed in a 96-well plate format by reverse transfecting HEK293T ( 40 , 000 cells per well ) with a total of 250 ng of DNA ( 10 ng of pGL3 Per1: Luc reporter , 30 ng CMV-CLOCK , 10 ng of CMV-Bmal1 , and up to 200 ng of test plasmids ) and 7 . 5 μl of TLT-1 ., The cells were seeded out in phenol red–free DMEM/Ham’s F-12 50/50 mix ( Corning Cat# 16-405-CV ) supplemented with 10% FBS , 1% Antibiotic-Antimycotic ( Life Technologies ) , 25 mM HEPES , and 125 μM of D-Luciferin ., The plate was sealed tight with TempPlate Optical film ( USA Scientific ) ., The plate was immediately transferred to the Tecan Infinite M200 maintained at 37 °C , and luminescence was measured in kinetic mode ( every 20 minutes ) for at least 72 hours ., To determine the relative expression of flag components , lysates were prepared at the time corresponding to the peak of CLOCK–Bmal1 activity and analyzed by western blots ., Jmjd5+/+ and Jmjd5−/− MEFs were seeded out in a 35-mm dish with 350 , 000 cells per dish ., Sixteen hours later , they were transfected with 2 μg of a Per2-Luc reporter construct ., Forty-eight hours post transfection , they were shocked with 0 . 1 μM dexamethasone for 2 hours ., The media were replaced with DMEM:F12 media without phenol red containing 1% antibiotic-antimycotic , 10% FBS , and 25 mM HEPES ., The plates were sealed tight and placed in an incubating luminometer ( Atto Kronos ) , and the luminescence was measured for 5 days ., Whole-cell lysates from cells and livers were prepared using lysis buffer containing 150

Introduction, Results, Discussion, Materials and methods

The circadian oscillator is a molecular feedback circuit whose orchestration involves posttranslational control of the activity and protein levels of its components ., Although controlled proteolysis of circadian proteins is critical for oscillator function , our understanding of the underlying mechanisms remains incomplete ., Here , we report that JmjC domain–containing protein 5 ( JMJD5 ) interacts with CRYPTOCHROME 1 ( CRY1 ) in an F-box/leucine-rich repeat protein 3 ( FBXL3 ) -dependent manner and facilitates targeting of CRY1 to the proteasome ., Genetic deletion of JMJD5 results in greater CRY1 stability , reduced CRY1 association with the proteasome , and disruption of circadian gene expression ., We also report that in the absence of JMJD5 , AMP-regulated protein kinase ( AMPK ) -induced CRY1 degradation is impaired , establishing JMJD5 as a key player in this mechanism ., JMJD5 cooperates with CRY1 to repress circadian locomotor output cycles protein kaput ( CLOCK ) –brain and muscle ARNT-like protein 1 ( BMAL1 ) , thus linking CRY1 destabilization to repressive function ., Finally , we find that ablation of JMJD5 impacts FBXL3- and CRY1-related functions beyond the oscillator .

In mammals , circadian rhythms are generated by a molecular oscillator in which the circadian locomotor output cycles protein kaput ( CLOCK ) –brain and muscle ARNT-like protein 1 ( BMAL1 ) transcription factors drive expression of the genes coding for their own repressors , the CRYPTOCHROME ( CRY ) and PERIOD ( PER ) proteins ., A key feature of the oscillator is that the protein stability of its components is highly regulated ., Previous studies had implicated the JmjC domain–containing protein 5 ( JMJD5 ) in regulation of the circadian clock in plants and flies ., Here , we show that cells and livers that lack JMJD5 exhibit dysregulation of circadian gene expression ., Mechanistically , JMJD5 is required for CRY1 degradation , including its destabilization by AMP-regulated protein kinase ( AMPK ) , by facilitating its interaction with the proteasome ., We found that JMJD5 is needed for normal CRY1-mediated transcriptional repression , thereby uncovering an inverse relationship between CRY1 stability and circadian repression ., Finally , we showed that JMJD5 impinges on non-clock roles of F-box/leucine-rich repeat protein 3 ( FBXL3 ) and CRY1 ., Altogether , our studies demonstrate that JMJD5 is a novel link between the oscillator and other physiological processes .

gene regulation, regulatory proteins, dna-binding proteins, circadian oscillators, muscle proteins, transcription factors, chronobiology, small interfering rnas, proteins, gene expression, biochemistry, circadian rhythms, rna, genetic oscillators, nucleic acids, genetics, protein domains, biology and life sciences, non-coding rna

journal.pntd.0005283

Modeling the Potential for Vaccination to Diminish the Burden of Invasive Non-typhoidal Salmonella Disease in Young Children in Mali, West Africa

In industrialized countries , non-typhoidal Salmonella ( NTS ) predominately causes gastroenteritis 1 , 2 ., However , in sub-Saharan Africa the NTS serovars S . Typhimurium and S . Enteritidis have become recognized as important causes of severe invasive bacterial disease ( e . g . , septicemia , meningitis , bacteremia ) with high case fatality rates 2 , 3 , 4 ., Infants age 6–11 months and toddlers age 12–23 months exhibit the highest incidence of severe invasive NTS ( iNTS ) disease 5 ., Whereas host factors such as malnutrition and co-infection with malaria and HIV may contribute to the higher burden of iNTS disease ( i . e . , high case fatality rate and prevalent cause of bacteremia ) in this region compared to industrialized countries 6–7 , fundamental differences in the circulating NTS strains from sub-Saharan Africa are also evident ., Available evidence suggests that the vast majority of the S . Typhimurium strains from cases of iNTS disease in sub-Saharan Africa are multi-locus sequence type 313 ( ST313 ) , a genotype unique to Africa that has undergone extensive genomic degradation 8–9 ., As the burdens of invasive disease due to Haemophilus influenzae type b ( Hib ) and Streptococcus pneumoniae have plummeted in recent years in sub-Saharan Africa following the introduction of Hib conjugate and multivalent pneumococcal conjugate vaccines 10–11 , recognition of the need to address iNTS disease has increased 12 ., Lack of information on the reservoirs and vehicles of transmission of iNTS in sub-Saharan Africa limits opportunities to utilize classic epidemiologic interventions to control iNTS disease ., However , successful vaccination programs implemented to control other invasive diseases prevalent among pediatric populations in Mali and other countries of sub-Saharan Africa have stimulated interest in the development of vaccines to control iNTS disease ., Several candidate vaccines under development have shown promise in protecting against invasive S . Typhimurium and S . Enteritidis disease in animal models 13 , 14 ., These include a bivalent conjugate vaccine based on covalently linking the core and O-antigen polysaccharides of S . Typhimurium ( a Group B O:4 serovar ) and S . Enteritidis ( a Group D O:9 serovar ) to the respective Phase 1 flagellin subunits ( FliC ) of each of these serovars 14–16 , a live attenuated oral vaccine 17–18 , and a bivalent Generalized Modules for Membrane Antigens ( GMMA ) vaccine consisting of outer membrane protein blebs from S . Typhimurium and S . Enteritidis that include lipopolysaccharide 12 , 14 ., These vaccines also have the potential to provide cross protection against other NTS serovars within Salmonella O Group B ( e . g . , S . Stanleyville ) and O Group D ( e . g . , S . Dublin ) 13 , 16 , 18 ., Our research modeled the decrease in the number of cases and deaths attributable to iNTS in children < 5 years of age following the programmatic introduction of a NTS vaccine utilizing the same Expanded Program on Immunization ( EPI ) infrastructure that successfully delivered Hib and pneumococcal conjugate vaccines and that drastically reduced the number of cases of invasive disease caused by those pathogens ., Census data from the National Institute of Statistics ( INSTAT ) of Mali from 2009 19 provided the pediatric population of different age groups of interest in Bamako as denominators for the model ., Crude birth rate and age-specific all-cause mortality data for specific pediatric age groups in Bamako came from Demographic and Health Surveys ( DHS ) of 2001 , 2006 and 2012 20–22 ( Table 1 ) ., The highest and lowest rates reported across the years of DHS reports were used to establish a range of probable values with the intermediate of the three values used as the initial parameter value ., The burden of invasive bacterial disease caused by NTS in Mali was first identified during systematic surveillance of the incidence of bacterial pathogens begun in 2002 at l’Hôpital Gabriel Touré ( HGT ) , Bamako , Mali ., The surveillance program established by the Center for Vaccine Development , Mali ( CVD-Mali ) and the Center for Vaccine Development ( CVD ) , University of Maryland School of Medicine , was designed to identify bacterial pathogens associated with invasive disease among consented enrolled patients <15 years of age admitted to HGT with fever or clinical signs of invasive bacterial disease 5 ., This hospital-based surveillance was conducted under a protocol approved by the Ethics Committee of the Faculté de Médecine , Pharmacie et Odontostomatologie in Bamako , Mali and the University of Maryland Institutional Review Board ., Consent was documented on a written form ., If the participants parent or guardian was illiterate , they listened to an audiotaped version of the consent form in their local language and questions were so answered in the presence of a witness ., We used anonymized data on 515 pediatric patients under five years of age who were admitted to HGT with laboratory-confirmed iNTS disease between July 1 , 2002 and June 30 , 2014 to develop and validate the model parameters ., Numbers of cases within specific age groups , pooled across even years of the HGT surveillance ( i . e . , 2002 , 2004 , 2006 , 2008 , 2010 , 2012 ) , were used with denominators from INSTAT 19 to generate the age group-specific hospitalization rates of severe iNTS disease ( Table 2 ) ., Case fatality rates for the model were fatal cases divided by total cases per age group ( Table 3 ) ., The model was then validated by comparing the number of cases and deaths due to iNTS per year estimated by the model , without accounting for vaccination effects , against the data from the odd years of HGT surveillance ., Data were pooled as means across the years of surveillance because the number of cases per age-group per year was small ( Fig 1 ) ., Multiple years of data were included to provide more robust estimates for the Malian pediatric population and to encapsulate some of the variability over time ., The proportion of hospitalizations with S . Typhimurium and S . Enteritidis serovars has been seen to change over time ., In particular , from 2008 to the present , the incidence of invasive S . Typhimurium infections has decreased , while the incidence of invasive S . Enteritidis infections has increased 5 ., Moreover , these serovars exhibit different case fatality rates ., Therefore , serovar-specific case fatality rates for hospitalized children were also calculated based on the HGT surveillance data ., The expected coverage for an iNTS vaccine was estimated based on data from Hib vaccine implementation in Bamako ., Vaccination coverage estimates came from an immunization coverage survey undertaken in 2015 among a sample of infants 6–8 months of age in the population as part of prospective demographic surveillance in the Djikoroni-para quartier of Bamako ., The demographic surveillance system allowed population-based estimates to be derived as was done for the Global Enteric Multicenter Study ( GEMS ) 23–24 ., Sixty-one mothers or other caretakers of infants 6–8 months of age were asked if they had an immunization card and 60 were able to show the card ., The narrow infant age range was selected to document not only evidence of receipt of Hib vaccine but to provide information on the timeliness of immunization which is important to the success of Hib and NTS vaccination as a public health tool ., Among these 61 Djikoroni-para infants , 60 had received at least one dose of Hib vaccine 60/61 ( 98 . 4% ) and 55 had received all three doses of Hib vaccine ( 55/61 , 90 . 2% full coverage ) ., This Hib vaccine coverage information from Bamako was used as the starting point for our simulations , since it was drawn directly from our modeled population ., Hib coverage data from Kenya 26 was utilized to generate wider intervals of coverage values from a larger study sample and broader population data ., Coverage data for alternative vaccination programs and booster vaccinations were based on the measles vaccine program implemented in Mali , which targeted children of the same scheduled ages as proposed in the model 33 ., Assumptions on the expected efficacy of the vaccines under development to prevent iNTS disease in Mali were based on assessments of the efficacy of Hib conjugate in a randomized clinical trial in The Gambia 24 and from post-licensure impact evaluations on Hib disease in Mali 10 , Kenya 26 and Uganda 29 and a 9-valent pneumococcal conjugate vaccine efficacy trial in The Gambia 30 ., Hib conjugate was highly effective in diminishing the disease burden when administered routinely through the Expanded Program on Immunization in Mali 10 , Kenya 26 and other African countries 29 ., Efficacy for booster vaccination doses and a catch-up campaign program were based on anticipated results similar to those exhibited by the Hib catch up campaign and booster vaccine interventions performed in the United Kingdom 34–35 ., In certain populations , such as those with a high prevalence of HIV cases , immune suppression decreases the amount of protection granted by vaccination against Hib 25 ., While the pediatric population of Bamako does not exhibit high levels of HIV , a scenario with low vaccine efficacy due to immune suppression such as seen by Madhi et al . , in South Africa ( a 20% decrease in each vaccine efficacy parameter ) was modeled ., The invasive disease such as meningitis , septicemia , bacteremia and septic arthritis caused by invasive non-typhoidal Salmonella is clinically indistinguishable from those types of clinical infections caused by Hib ., Each of these pathogens traverses a mucosal barrier leading to a bacteremia during which the bacterial pathogens are cleared by fixed macrophages residing in organs of the reticuloendothelial system ., In the case of Hib , it is upper respiratory mucosa that is traversed , while for iNTS it is believed to be intestinal mucosa ., Bacteremic organisms that reach the meninges , synovia and pleura can cause meningitis , septic arthritis and empyema , respectively ., The NTS conjugate vaccines under development elicit serum antibodies that exhibit both bactericidal and opsonophagocytic functional properties 31–32 , 36 , like the antibodies stimulated by Hib conjugate vaccines 25 , 37–40 ., Thus , there exist striking pathogenetic , clinical and epidemiologic similarities between iNTS and Hib pathogens and similar functional activities are exhibited by the antibodies stimulated by the parenteral NTS conjugate vaccines ( in animals ) and by Hib conjugate vaccine in human infants ., Therefore , we assumed a similar efficacy and coverage for the NTS vaccine as was observed with Hib conjugate vaccine in the infant and toddler population in Bamako ( and elsewhere in sub-Saharan Africa ) ., Relying on Hib vaccination efficacy data allowed us to validate the iNTS vaccine implementation within the model and allowed for reliable comparisons of the protection potentially granted by the iNTS vaccine and various immunization schedules ., The incidence and epidemiologic features of iNTS infections among young children sufficiently severe to result in hospitalization were captured using an age-structured Markov chain infectious disease model including Susceptible , Infected , and Recovered status groups ., A diagram of the model ( Fig 2A ) with vaccine administered at 6 , 10 , and 14 weeks of life , as was used in the Hib vaccination initiative , was used as the baseline for developing a generalized model capturing all modeled vaccination schedules as illustrated in Fig 2B ., Each age group ( a ) included an age-specific number of children susceptible to ( Sa , 10 groups , where a = 1–10 ) , hospitalized with ( Ia , 4 groups , where a = 1–3 , 4–8 , 9 , 10 ) , or vaccinated by dose ( d ) against ( Va , d , where a varied for different scenarios and d = 1–5 ) severe iNTS disease ., Children who recovered ( R ) from severe hospitalization were considered as a single group ., Age categories were established by examining statistically significant variation in incidence and case fatality rates of iNTS ( p<0 . 05 ) within the HGT surveillance data , and ages at which the EPI vaccinations were scheduled ., Neonates entered the model based on the population birth rate reported for Bamako ( ν , Table, 1 ) and were considered to be hospitalized with iNTS disease at the same incidence as other infants < 2 months of age ( Table 2 ) ., Children at any age were subject to the age-specific all-cause mortality rates ( μa , Table, 1 ) reported in the DHS ., Susceptible children ( Sa ) were moved to a hospitalized status ( Ia ) at age-specific iNTS hospitalization rates ( βa , Table 2 ) ., Children hospitalized with iNTS disease experienced mortality at age-specific case fatality rates ( ηa , Table, 2 ) or moved to a recovered status ( R ) ., The length of iNTS infection was constrained to a single two-week time step corresponding to the observed duration of iNTS clinical disease in hospitalized Bamako children ., Any susceptible children who did not suffer hospitalized iNTS disease , iNTS fatality , or all-cause mortality graduated into the next susceptible age group at a rate appropriate to the two-week time step of the model ., All surviving children exited the model upon reaching five years of age , when the rate of hospitalizations due to iNTS rapidly declines 5 ., Vaccination against iNTS was initially modeled as a program which occurred at ~6 , ~10 , and ~14 ( a = 2 , 4 , 6 , respectively ) weeks of life to match the same three-dose infant immunization schedule as Hib conjugate ., Children potentially received one , two , or three ( dose d = 1–3 , respectively ) doses of the vaccine with age-specific and dose-specific coverage rates ( ϑa , d ) ., For vaccinated children , the rate of hospitalization due to iNTS disease was applied only to the proportion of children without an effective vaccination ( 1-εd ) ., Vaccine protection was assumed to persist through early childhood , so successfully immunized children remained protected until they aged out of the model ., Our major outcome measures were the number of cases and deaths due to iNTS disease ., To generate an overall distribution of potential values describing the natural epidemiologic behavior of the disease without implementation of a vaccination program , 1000 simulations were run with key model variables ( i . e . , birth rate , incidence of hospitalization of iNTS cases , case fatality , and background all-cause mortality rates ) randomly drawn each time from a uniform distribution based on the limits of the probable range around each parameter ., After being used to simulate the ‘average’ dynamics in absence of vaccine use , a second set of 1000 simulations was used to assess the effect of different assumptions about vaccine efficacy and coverage in the model ., Key variables were sampled as before and values for vaccine efficacy and coverage drawn from triangular distributions based on each parameter value and its associated probable range under different scenarios ., The influence of each parameter on the number of iNTS cases and fatal iNTS cases generated by the model was assessed by sampling each parameter individually , while holding all other parameter values constant ., The ranges of iNTS cases and fatal iNTS cases generated by these simulations were summarized in tornado plots ., Next , specific effects of vaccination were examined in a birth cohort of 75 , 978 children , corresponding to the annual births for Bamako ., The expected number of cases and number of fatal cases in children < 36 months of age within this cohort was generated under unvaccinated conditions , vaccinated conditions with 100% coverage and efficacy implemented at 6 weeks of life , and vaccinated conditions with coverage and efficacy matching the model parameters described in Table 4 at 6 , 10 , and 14 weeks of life ., Additionally , the effects of high , mid , and low vaccine efficacy levels on the overall number of cases and case fatalities were examined by dose , based on the ranges around these parameters ., To assess the effects of serovar- specific severity , the vaccination program was modeled using case fatality rates representing only S . Typhimurium or only S . Enteritidis as causal agents ( Table 3 ) ., Since the overall incidence of iNTS disease has not changed significantly with the documented serovar shift , we maintained the same overall incidence rates regardless of underlying causal agent ., Another 1000 simulations were performed to assess effects of varying the parameter values across the probable ranges for case fatality rates of these two serovars on the number of severe iNTS cases and fatal iNTS cases ., The same approach was used to simulate effects of alternative three-dose EPI vaccination regimens ., For example , the first two doses were administered at 6 ( a = 2 ) and 10 ( a = 4 ) weeks or at 10 ( a = 4 ) and 14 ( a = 6 ) weeks of life ( concomitant with two doses of pentavalent and pneumococcal conjugate vaccine ) and the third NTS vaccine dose was administered as a booster ( dose d = 4 ) at either age 9 months ( a = 7 ) ( with measles containing vaccine dose 1 MCV1 ) or at 12 or 15 ( a = 9 ) months of age concomitant with MCV2 ., We also modeled a rapid mass immunization catch-up campaign ( dose d = 5 ) targeting all children of age 6–23 ( a = 6–9 ) , 9–23 ( a = 7–9 ) , or 12–23 ( a = 9 ) months of age concomitant with the onset of adding iNTS vaccination to the routine young infant EPI ., In each of these schedules , the precise ages at time of vaccination , although ideally targeted at specific weeks of life , in fact vary and are often delayed by several weeks in Bamako ., To allow for this , the number of children admitted to the HGT who received one , two , or three doses of vaccine any time within the relevant month of life was used to calculate the sample mean coverage levels among hospital admissions , and not by specific week of implementation ., These coverage levels fell within the confidence intervals of the Kenyan coverage data reported by Cowgill et al . 26 that were used to parameterize the model ., Model development and analyses were performed using R version 3 . 0 . 1 and utilizing the Markovchain package , version 0 . 5 27 for the development of the model and the Triangle package , version 0 . 1 28 for vaccine efficacy and vaccine coverage variable distribution analysis ., The code developed for the model and the case data used to develop the model parameters are available on a public GitHub repository 41 ., Based on the parameter values determined from even years , our model predicted a similar number of hospitalized iNTS disease cases for most age groups as was observed during the odd years of HGT surveillance data ( Fig 1 ) , with 37 cases per year , including 7 fatal cases , occurring in a non-vaccinated population ., The 1000 model runs generated a range of 14–64 cases per year ( with interquartiles of 29 and 39 ) and a range of 2–14 fatal cases per year ( with interquartiles of 5 and 8 ) , sampling from the probable range of model parameters ., Hospital surveillance records did not include neonatal cases who died of iNTS infection before leaving the hospital after birth , so our model assumed a similar level of incidence among this youngest age group as was observed for other children less than one month old ., This assumption generated a simulated overestimation of cases in the youngest age group compared to the observed , but attempted to include neonatal cases and iNTS related deaths in our outcome measurements ., Varying the parameter estimates used for hospitalized infection rate , all-cause mortality rate , and case fatality rate across the range of each parameter led to a mean of 34 cases per year ( ranging from 14–64 , with interquartiles of 29 and 39 ) and 7 fatal cases per year ( ranging from 2–14 , with interquartiles of 5 and 8 ) , based on 1000 runs of the model ., One thousand model runs with varying parameter estimates for birth rate , incidence of hospitalization of iNTS cases , case fatality , background all-cause mortality rates , vaccine efficacy and coverage with three doses of vaccine administered at 6 , 10 , and 14 weeks of life generated a mean of 9 cases per year ( ranging from 5–16 , with interquartiles of 11 and 6 ) and 3 fatal cases per year ( ranging from 2–12 , with interquartiles of 2 and 5 ) ., The greatest change in the number of iNTS cases occurred as the incidence rate was sampled across its probable range , while the least amount of change was generated by sampling across vaccine coverage ( Fig 3 ) ., The greatest change in the number of fatal iNTS cases was driven by the case fatality rate ( Fig 4 ) ., In observing the total number of cases and the number of deaths due to iNTS among a birth cohort , as presented in Fig 5 , the vaccination parameters of the model functioned as expected ., If a NTS vaccine was implemented with 100% coverage and 100% efficacy , all cases following an initial dose of vaccine were prevented and all fatal cases were averted ., Furthermore , when the model was run with parameters based on the Hib vaccination field trials and post-introduction impact assessments in Africa , the results were very similar to a vaccine with perfect coverage and efficacy ., Almost all cases among the birth cohort were prevented even after a single dose , and all cases in the cohort were prevented after two doses ., Modeling vaccine coverage and efficacy for an iNTS vaccine equivalent to levels observed with Hib ( with routine young infant immunization only and with no catch-up campaign ) , the number of hospitalized iNTS cases per year decreased by 73% ( from 37 to 10 cases ) and the number of deaths decreased by 43% ( from 7 to 4 deaths ) ., These estimates , based on the distribution of S . Enteritidis and S . Typhimurium cases that was observed during the 2002–2014 surveillance , reflect the effect of direct protection alone ( i . e . , with no adjustment for indirect protection from “herd immunity” ) and without a catch-up campaign ., Effects of varying the parameter values for vaccine efficacy , number of doses , serovar distributions , and vaccination schedules are shown in Table, 5 . Even at the lowest ranges of vaccine efficacy , our model predicted a range of only 4–23 cases per year , based on 1000 simulation runs ., This equates to prevention of more than half of the pediatric cases each year if the vaccine exhibits similar efficacy as might be seen with the Hib vaccine in immunodeficient populations ., At higher levels of efficacy , as much as 78% of severe iNTS cases ( 29 cases/year ) were averted ., The highest level of protection was granted by a 3-dose vaccination schedule targeting infants at 6 , 10 , and 14 weeks of life , which resulted in a 73% decrease in the annual number of severe iNTS cases with moderate vaccine efficacy levels ., This varied from 21–29 cases prevented and 2–3 child lives saved per year among the pediatric population of Bamako , based on the vaccine efficacy levels observed in the Hib conjugate field trial in The Gambia 24 and the post-implementation effectiveness assessment in Mali 10 ., The increased immunity over time among the population was captured through the number of hospitalized iNTS cases and deaths per 6-month intervals after vaccine implementation , compared to a 6-month pre-vaccination baseline interval , as illustrated in Fig, 6 . If a catch-up vaccination campaign was implemented simultaneously along with the introduction of a three-dose young infant EPI vaccination strategy , protection occured sooner among older age children who remained at risk ., Fig 7 illustrates the effects of such a catch-up campaign targeting various age groups ., The addition of the catch-up campaign prevented at least four and as many as 12 additional cases of severe iNTS during the first three years following the catch-up campaign and start of the vaccine initiative compared to the three-dose schedule without the catch-up campaign ., Investigating the effects of different serovar distributions that have been observed over time , our model predicted that if S . Enteritidis , with its higher case fatality rate , was the only iNTS serovar causing disease , 44% of deaths per year would be averted through vaccination , with a range of 5–37 cases per year including 2–18 fatal cases , across 1000 simulation runs ., If S . Typhimurium returned as the dominant serovar , up to 80% of iNTS deaths per year would be averted , with a range of 4–31 cases per year , including 1–8 fatal cases ., Implementation of programmatic use of the Hib conjugate vaccine among the pediatric population in Bamako , Mali was extremely successful and reduced the burden of invasive Hib disease hospitalizations by 83% among infants within three years of vaccine introduction 10 ., Our results show that if we utilize the same EPI infrastructure to deliver a bivalent conjugate vaccine to prevent invasive disease due to S . Enteritidis and S . Typhimurium , it is reasonable to expect a comparable level of protection and reduction of iNTS cases as was seen with Hib vaccine ., Our iNTS model captured the current burden of iNTS in the population and assessed potential effects of various vaccine implementation scenarios ., Even alternative vaccination schedules with fewer doses of iNTS vaccine predicted a notable reduction in the burden of iNTS disease ., Additionally , if a one-time catch-up campaign is implemented concomitantly with routine vaccination of young infants and if it targets the highest risk age group ( children in the second half of the first year of life ) , our model predicted additional cases of severe iNTS disease deaths would be averted in the first years after introducing the vaccine ., If field trials confirm the efficacy of the NTS vaccines currently under development , their future implementation within the EPI could markedly diminish the morbidity and mortality from one of the predominant remaining burdens of invasive bacterial disease among pediatric populations in sub-Saharan Africa ., Since neither the reservoir of infection nor the modes of transmission of NTS to young children have heretofore been identified , vaccination currently represents the most plausible interventional strategy for reducing the burden of iNTS disease ., Furthermore , the model we have developed could be applied to estimate the effects of implementing an iNTS vaccine in other regions of sub-Saharan Africa , providing the same integrity of information on age-specific case and fatality rates ., Two candidate NTS vaccines are progressing towards clinical trials ., One candidate developed by the GSK Vaccines Institute of Global Health consists of a bivalent parenteral Salmonella Enteritidis/S ., Typhimurium vaccine based on Generalized Modules for Membrane Antigens ( GMMA ) technology 14 , 42–43 ., The second NTS candidate , developed by the Center for Vaccine Development of the University of Maryland School of Medicine ( CVD ) and its industrial partner , Bharat Biotech , International ( BBI ) of Hyderabad , India , contains S . Enteritidis and S . Typhimurium conjugate vaccines consisting of the core plus O polysaccharide of those serovars covalently linked to Phase 1 flagellin subunits of the homologous serovar 13–16 , 44 ., As each vaccine moves towards clinical trials , a Target Product Profile ( TPP ) must be created that by necessity incorporates multiple assumptions and predictions that guide the development of the project for multiple years before clinical data become available to corroborate or refute the TPP assumptions ., The TPP , which must be crafted early in the development of the candidate vaccine , provides a roadmap as it defines the type of vaccine , the route of administration , the target populations and sub-populations to be vaccinated , the number of doses to be administered and the intended immunization schedule for the target populations ., The TPP also proposes limits for the expected reactogenicity ( local and systemic ) , the level of efficacy to be achieved , the duration of protection , when a booster dose might be needed , the storage conditions , the vaccine formulation ( s ) , the presentation of the vaccine , whether an adjuvant will be included and what preservative will be present in multi-dose vials ., The design of the preclinical toxicology test , the formulations of vaccine to be tested , the design of the Phase 1 and 2 clinical trials , and of the ultimate pivotal Phase 3 efficacy trial all follow guidance provided by the TPP ., The mathematical model described herein includes assumptions contained within one TPP ., Even at early stages in development of the candidate vaccines to prevent iNTS disease , a mathematical model of what the vaccine might achieve at the future public health level becomes a useful , hopefully predictive , tool ., Modifying the parameters of the model offers insights on what the vaccine can achieve ., Our model has been used to assess the impact of introducing a bivalent NTS vaccine on decreasing the number of hospitalized iNTS cases and fatalities caused by the two most prevalent iNTS serovars currently found in the Malian pediatric population , S . Enteritidis and S . Typhimurium ., However , other serovars have been identified among a minority of hospitalized cases of iNTS that theoretically could also be prevented ., Indeed the bivalent vaccines currently in development offer the prospect of cross protection against other serovars ., The bivalent conjugate vaccine described by Simon et al . 16 , for example , may offer such cross protection by targeting shared O-polysaccharides ., If the effectiveness of the vaccine is reliant on these targeted polysaccharides , which are shared among all serovars within the same serogroup , this vaccine would offer cross protection against all Group B and Group D serovars , including the serovars with the next highest prevalence among hospitalized cases in Bamako , Mali ( S . Stanleyville and S . Dublin , respectively ) 13 , 15–16 ., Our findings have some limitations because of the lack of data on segments of the pediatric population where iNTS disease may be occurring but not detected with our surveillance ., By focusing only on iNTS cases admitted to hospital , we did not model the overall burden of NTS in the Bamako pediatric population that would include children in the community with iNTS infections who were not ill enough for their caretakers to seek health care or who were brought to traditional healers ., It also did not include children with severe iNTS disease who may not have had easy access to the hospital and thus may have died at home ., Moreover , blood cultures and cultures of ordinarily sterile body fluids are not 100% sensitive in detecting invasive bacterial infections , particularly if antibiotics were administered prior to reaching the hospital ., Thus , some iNTS cases may have been missed by our surveillance techniques , leading to an underestimation of the number of cases and the burden of NTS in the study population ., However , cases hospitalized at HGT likely capture a substantial proportion of the more severe clinical forms of iNTS disease which likely have a higher case fatality than milder forms of iNTS disease ., The impact of a NTS vaccine in preventing culture-negative severe iNTS cases could be estimated by noting the difference between the decrease in hospitalized iNTS cases following vaccine implementation and a decrease in all-cause hospitalizations 10 ., Despite some knowledge gaps about the epidemiologic behavior of iNTS disease in the community , our model provides an informative view that should adequately assess the impact of introducing an effective vaccine ., The model presented herein is one of the first attempts to capture mathematically the epidemiologic dynamics of endemic pediatric iNTS disease in Africa and to predict the effects of future implementation of NTS vaccines currently in development on the disease burden ., Several key features of the epidemiology of iNTS disease in B

Introduction, Methods, Results, Discussion

In sub-Saharan Africa , systematic surveillance of young children with suspected invasive bacterial disease ( e . g . , septicemia , meningitis ) has revealed non-typhoidal Salmonella ( NTS ) to be a major pathogen exhibiting high case fatality ( ~20% ) ., Where infant vaccination against Haemophilus influenzae type b ( Hib ) and Streptococcus pneumoniae has been introduced to prevent invasive disease caused by these pathogens , as in Bamako , Mali , their burden has decreased markedly ., In parallel , NTS has become the predominant invasive bacterial pathogen in children aged <5 years ., While NTS is believed to be acquired orally via contaminated food/water , epidemiologic studies have failed to identify the reservoir of infection or vehicles of transmission ., This has precluded targeting food chain interventions to diminish disease transmission but conversely has fostered the development of vaccines to prevent invasive NTS ( iNTS ) disease ., We developed a mathematical model to estimate the potential impact of NTS vaccination programs in Bamako ., A Markov chain transmission model was developed utilizing age-specific Bamako demographic data and hospital surveillance data for iNTS disease in children aged <5 years and assuming vaccine coverage and efficacy similar to the existing , successfully implemented , Hib vaccine ., Annual iNTS hospitalizations and deaths in children <5 years , with and without a Salmonella Enteritidis/Salmonella Typhimurium vaccine , were the model’s outcomes of interest ., Per the model , high coverage/high efficacy iNTS vaccination programs would drastically diminish iNTS disease except among infants age <8 weeks ., The public health impact of NTS vaccination shifts as disease burden , vaccine coverage , and serovar distribution vary ., Our model shows that implementing an iNTS vaccine through an analogous strategy to the Hib vaccination program in Bamako would markedly reduce cases and deaths due to iNTS among the pediatric population ., The model can be adjusted for use elsewhere in Africa where NTS epidemiologic patterns , serovar prevalence , and immunization schedules differ from Bamako .

A surveillance program at Gabriel Touré Hospital in Mali observed a high burden of invasive disease caused by non-typhoidal Salmonella ( iNTS ) ., This surveillance program was originally instituted to measure the amount of invasive disease ( e . g . , septicemia , meningitis ) caused by two bacteria that invade the respiratory tract: Haemophilus influenzae type b ( Hib ) and Streptococcus pneumoniae ( pneumococcus ) ., While documenting the burden of these pathogens , the surveillance program also found that serotypes of iNTS , mainly Salmonella Typhimurium and Salmonella Enteritidis , were common causes of severe invasive disease ., As the number of cases of Hib and pneumococcus markedly decreased following the introduction of relevant vaccines , the relative threat of iNTS increased ., Little is known about the reservoir of iNTS , whether it resides in humans , animals , or the environment , or how it is spread to susceptible children ., Without this knowledge , it is not possible to employ certain disease control methods useful in interrupting the transmission of other pathogens ., Therefore , vaccination remains the one promising control strategy for this disease ., Our research modeled the potential effects of introducing an iNTS vaccine ., The findings are of great importance to Mali and other developing countries where young children are at a high risk of developing iNTS disease .

medicine and health sciences, mali, pathology and laboratory medicine, pathogens, immunology, geographical locations, microbiology, pediatrics, vaccines, preventive medicine, bacterial diseases, enterobacteriaceae, vaccination and immunization, bacteria, bacterial pathogens, africa, salmonella typhimurium, public and occupational health, infectious diseases, booster doses, medical microbiology, microbial pathogens, salmonella, people and places, biology and life sciences, conjugate vaccines, vaccine development, organisms

journal.pcbi.1005684

Confirmation bias in human reinforcement learning: Evidence from counterfactual feedback processing

Goal-directed behaviour is composed of two core components 1: one component is the decision-making process that starts with representing the available options and terminates in selecting the option with the highest expected value; the second component is reinforcement learning ( RL ) , through which outcomes are used to refine value expectations , in order to improve decision-making ., Human decision-making is subject to biases ( i . e . deviations from the normative prescriptions ) , such as the framing effect 2 ., While the investigation of decision-making biases has a long history in economics and psychology , learning biases have been much less systematically investigated 3 ., This is surprising as most of the decisions we deal with in everyday life are experience-based and choice contexts are recurrent , thus allowing learning to occur and therefore influencing future decision-making ., In addition , it is important to investigate learning biases as there is evidence that RL processes play a role in psychiatric conditions and maladaptive economic behaviour 4 , 5 ., Standard RL algorithms learn action-outcome associations directly from obtained outcomes on a trial and error basis 6 ., We refer to this direct form of learning as “factual learning” ., Despite the fact that standard models , built around the notion of computational and statistical optimality , prescribe that an agent should learn equally well from positive and negative obtained outcomes 7–9 , previous studies have consistently shown that humans display a significant valence-induced bias ., This bias generally goes in the direction of preferential learning from positive , compared to negative , outcome prediction errors 10–14 ., This learning asymmetry could represent a RL counterpart of the “good news/bad news” effect that is observed for probabilistic reasoning 15 ., However , human RL cannot be reduced simply to learning from obtained outcomes ., Other sources of information can be successfully integrated in order to improve performance and RL has a multi-modular structure 16 ., Amongst the more sophisticated learning processes that have already been demonstrated in humans is counterfactual learning ., Counterfactual learning refers to the ability to learn from forgone outcomes ( i . e . the outcomes of the option ( s ) that were not chosen ) 17 , 18 ., Whether or not a valence-induced bias also affects counterfactual learning remains unknown ., To address this question , we ran two experiments involving instrumental learning and computational model-based analyses ., Two groups of healthy adults performed variants of a repeated two-armed bandit task involving probabilistic outcomes 19 , 20 ( Fig 1A ) ., We analysed the data using a modified Rescorla-Wagner model that assumes different learning rates for positive and negative , and factual and counterfactual , prediction errors ( Fig 1B ) 21 , 22 ., The first experiment aimed to replicate previous findings of a “positivity bias” at the level of factual learning ., In this first experiment , participants were presented only with the obtained outcome ( chosen outcome: RC; Fig 1A ) 10 ., In the second experiment , in order to investigate whether or not counterfactual learning rates are also affected by the valence of prediction errors , we used a variant of the same instrumental learning task , in which participants were also presented with the forgone outcome ( unchosen outcome: RU; Fig 1B ) ., Our design allowed us to test three competing hypotheses concerning the effect of valence on counterfactual learning ( Fig 2A ) ., The first hypothesis–“no bias”—was that unlike factual learning , counterfactual learning would be unbiased ., The second hypothesis , —“positivity bias”—was that factual and counterfactual learning would present the same valence-induced bias , such that positive counterfactual prediction errors would be more likely to be taken into account than negative counterfactual prediction errors ., In this scenario , factual and counterfactual learning biases would be consequences of a more general positivity bias , in which positive prediction errors have a greater impact on learning , regardless of whether the option was chosen or not ., Finally , the third hypothesis–“confirmation bias”—was that valence would affect factual and counterfactual learning in opposing directions , such that negative unchosen prediction errors would be more likely to be taken into account than positive unchosen prediction errors ., In this scenario , factual and counterfactual learning biases would be consequences of a more general confirmation bias , in which outcomes that support the current choice are preferentially taken into account ., To investigate both factual and counterfactual reinforcement learning biases , we designed an instrumental task based on a previous paradigm , in which we showed a significant positivity bias in factual learning 10 ., Here , we used two variants of the task , which differed in that the task used in Experiment 1 involved participants ( N = 20 ) being shown only the outcome of their chosen option , whereas in Experiment 2 ( N = 20 ) the outcome of the unchosen option was also displayed ( Fig 1A ) ., To test our hypotheses concerning valence-induced learning biases ( Fig 2A ) we fitted the data with a Rescorla-Wagner model assuming different learning rates for positive and negative outcomes , which respectively elicit positive and negative prediction errors ( Fig 1B ) ., The algorithm used to explain Experiment 1 data involved two learning rates for obtained outcomes ( αc+ and αc− for positive and negative prediction errors of the obtained outcomes , respectively ) ., In addition to the obtained outcome learning rates , the algorithm used to explain Experiment 2 data also involved two learning rates for forgone outcomes ( αu+ and αu− for positive and negative prediction errors of the forgone outcomes , respectively ) ., Replicating previous findings , in Experiment 1 we found that the positive factual learning rate ( αc+ ) was significantly higher than the negative one ( αc−; T ( 19 ) = 2 . 4; P = 0 . 03 ) ( Fig 2B , left ) ., In Experiment 2 , we analysed learning rates using a repeated-measure ANOVA with prediction error valence ( positive or negative ) and prediction error type ( factual or counterfactual ) as within-subjects factors ., Falsifying the “positivity bias” hypothesis , the ANOVA revealed no main effect of prediction error valence ( F ( 1 , 19 ) = 0 . 2; P>0 . 6 ) ., We also did not find any effect of prediction error type , indicating that , on average , factual and counterfactual learning were similar ( F ( 1 , 19 ) = 0 . 5; P>0 . 4 ) ., Consistent with the “confirmation bias” hypothesis , we found a significant interaction between valence and type ( F ( 1 , 19 ) = 119 . 2; P = 1 . 3e-9 ) ., Post-hoc tests indicated that the interaction was driven by effects of valence on both factual ( αc+>αc−; T ( 19 ) = 3 . 6; P = 0 . 0017 ) and counterfactual learning rates ( αu−>αu+; T ( 19 ) = 6 . 2; P = 5 . 8e-06 ) ( Fig 2B , right ) ., To verify the robustness of this result in the context of different reward contingencies , we analysed learning rates in each task condition separately ., In both experiments , our task included three different conditions ( S1 Fig ) : a “Symmetric” condition , in which both options were associated with a 50% chance of getting a reward; an “Asymmetric” condition , in which one option was associated with a 75% chance of getting a reward , whereas the other option was associated with only a 25% chance; and a “Reversal” condition , in which one option was initially associated with a 83% chance of getting a reward and the other option was associated with a 17% chance of getting a reward , but after 12 trials the reward contingencies were reversed ., For Experiment 1 , we analysed factual learning rates using a repeated-measure ANOVA with prediction error valence ( positive and negative ) and task condition ( Symmetric , Asymmetric and Reversal ) as within-subjects factors ( S1B Fig ) ., Confirming the aggregate result , the ANOVA showed a significant main effect of valence ( F ( 1 , 19 ) = 26 . 4 , P = 5 . 8e-5 ) , but no effect of condition ( F ( 2 , 38 ) = 0 . 7 , P>0 . 5 ) , and , crucially , no valence by condition interaction ( F ( 2 , 38 ) = 0 . 8 , P>0 . 4 ) ., For Experiment 2 , we analysed factual and counterfactual learning rates using a repeated-measure ANOVA with prediction error valence ( positive and negative ) , prediction error type ( factual or counterfactual ) and condition ( Symmetric , Asymmetric and Reversal ) as within-subjects factors ( S1C Fig ) ., Confirming the aggregate result , the ANOVA showed no effect of prediction error type ( F ( 1 , 19 ) = 0 . 0 , P>0 . 9 ) , no effect of valence ( F ( 1 , 19 ) = 0 . 3 , P>0 . 5 ) , but a significant valence by type interaction ( F ( 1 , 19 ) = 162 . 9 , P = 9 . 1e-11 ) ., We also found an effect of condition ( F ( 2 , 38 ) = 5 . 1 , P = 0 . 01 ) , reflecting lower average learning rates in the Reversal compared to the Asymmetric condition ( T ( 19 ) = 2 . 99; P = 0 . 007 ) , which was not modulated by valence ( F ( 2 , 38 ) = 0 . 2 , P>0 . 7 ) , or type ( F ( 2 , 38 ) = 1 . 2 , P>0 . 3 ) ., The three-way interaction was not significant ( F ( 2 , 38 ) = 1 . 8 , P> . 1 ) , indicating that learning biases were robust across different task contingencies ., To further test our hypotheses and verify theparsimony of our findings , we ran a model comparison analysis including the ‘Full’ model ( i . e . , the model with four learning rates; Fig 1C , right ) and reduced , alternative versions of it ( Fig 3A ) ., The first alternative model was obtained by reducing the number of learning rates along the dimension of the outcome type ( factual or counterfactual ) ., This ‘Information’ model has only two learning rates: one for the obtained outcomes ( αC ) and another for the forgone outcomes ( αU ) ., The second alternative model was obtained by reducing the number of learning rates along the dimension of the outcome valence ( positive or negative ) ., This ‘Valence’ model has only two learning rates ( one for the positive outcomes ( α+ ) and another for the negative outcomes ( α- ) ) and should win according to the “positivity bias” hypothesis ., Finally , the third alternative model was obtained by reducing the learning rate as a function of the outcome event being confirmatory ( positive obtained or negative forgone ) or disconfirmatory ( negative obtained or positive forgone ) ., This ‘Confirmation’ model has only two learning rates ( one for confirmatory outcomes ( αCON ) and another for the disconfirmatory outcomes ( αDIS ) ) and should win according to the “confirmation bias” hypothesis ., Bayesian Information Criterion ( BIC ) analysis indicated that the ‘Full’ model better accounted for the data compared to both the ‘Information’ and the ‘Valence’ models ( both comparisons: T ( 19 ) >4 . 2; P<0 . 0005; Table 1 ) ., However the ‘Confirmation’ model better accounted for the data compared to the ‘Full’ model ( T ( 19 ) = 9 . 9; P = 6 . 4e-9 ) ., The posterior probability ( PP ) of belonging to each model , calculated for each subject , ( i . e . , the averaged individual model attributions ) of the ‘Confirmation’ model was higher than chance ( . 0 . 25 for a model space including 4 models; T ( 19 ) = 13 . 5; P = 3 . 3e-11 ) and higher than the posterior probability all the other models ( all comparisons: T ( 19 ) >9 . 0; P<2 . 1e-8 ) ( Fig 3B ) ., The learning rate for confirmatory outcomes was significantly higher than that for disconfirmatory outcomes ( αCON>αDIS; T ( 19 ) = 11 . 7; P = 3 . 9e-10 ) ( Fig 3C ) ., These results support the “confirmation bias” hypothesis and further indicate that , at least at the behavioural level , chosen and unchosen outcomes may be processed by the same learning systems ., To evaluate the capacity of our models to reproduce the learning curves , we plotted and analysed the trial-by-trial model estimates of choice probabilities ( Fig 4 ) 23 ., The model estimates were generated using the best fitting set of parameters for each individual and model ., In the Symmetric condition ( where there is no correct response ) , we considered the preferred option choice rate ( i . e . , the option/symbol that was chosen more than >50% ) ., In the Asymmetric condition we considered the correct choice rate ., In the Reversal condition ( where the correct response is reversed after the first half of the trials ) we considered the choice rate of the initially more advantageous option ( i . e . , the correct option during the first half ) ., Qualitative observation of the learning curves indicated that the biased models ( Experiment 1: αc+≠αc−; Experiment 2: αCON≠αDIS ) tended to reproduce the learning curves more closely ., To quantify this , we compared the average square distance between the biased and the unbiased models ( Experiment 1: αc+=αc−; Experiment 2: αCON = αDIS ) ., We found that the square distance was shorter for the biased models compared to the unbiased models in both experiments ( Experiment 1: 0 . 074 vs . 0 . 085 , T ( 19 ) = 3 . 5 P = 0 . 0022; Experiment 2: 0 . 056 vs . 0 . 064 , T ( 19 ) = 3 . 5 P = 0 . 0016 ) ., We calculated the Pearson correlation between the parameters ( Fig 5A ) and found no significant correlation when correcting for multiple comparisons ( corrected P value = 0 . 05÷6 = 0 . 008; lowest uncorrected P value = 0 . 01 , highest P2 = 0 . 30 ) ., The correlation between αCON and αDIS was weak , but positive , which rules out the possibility that the significant difference between these two learning rates was driven by an anti-correlation induced by the model fitting procedure ., We then applied the same model fitting procedure to the synthetic datasets and calculated the correlation between the true and the retrieved parameters ( Fig 5B ) ., We found that , on average , all parameters in both experiments were well recovered ( 0 . 70 ≤ R ≤ 0 . 89 ) and that our model fitting procedure introduced no spurious correlations between the other parameters ( |R| ≤ 0 . 5 ) ., We also checked the parameter recovery for discrete sets of parameter values ( S2 & S3 Figs ) ., For Experiment 1 , we simulated unbiased ( αc+=αc− ) and biased ( αc+>αc− ) participants ., For Experiment 2 , we simulated unbiased ( αc+=αc− and αu+=αu− ) , semi-biased ( αc+>αc− and αu+=αu− ) and biased ( αc+>αc− and αu+>αu− ) participants ., We simulated N = 100 virtual participants per set of parameters ., The results of these analyses are presented in the supplementary materials and confirm the capacity of our parameter optimisation procedure to correctly recover the true parameters , regardless of the presence ( or absence ) of learning rate biases ., To investigate the behavioural consequences of the learning biases , we median-split the participants from each experiment into two groups according to their normalised learning rate differences ., We reasoned that the effects of learning biases on behavioural performance could be highlighted by comparing participants who differed in the extent they expressed the bias itself ., Experiment 1 participants were split according to their normalised factual learning rate bias: ( αc+−αc− ) / ( αc++αc− ) , from which we obtained a high ( M = 0 . 76±0 . 05 ) and a low bias ( M = 0 . 11±0 . 14 ) group ., Experiment 2 participants were split according their normalised confirmation bias: ( αc+−αc− ) − ( αu++αu− ) / ( αc++αc−+αu++αu− ) , from which we also obtained a high bias group ( M = 0 . 72±0 . 04 ) and a low bias group ( M = 0 . 36±0 . 04 ) ., From the Symmetric condition we extracted preferred choice rate as a dependent variable , which was the choice rate of the most frequently chosen option ( i . e . the option that was chosen on >50% of trials ) ( Fig 6A ) ., We hypothesised that higher biases were associated with an increased tendency to develop a preferred choice , even in the absence of a “correct” option , which naturally emerges from overweighting positive factual ( and/or negative counterfactual ) outcomes , as observed in our previous study 10 ., We submitted the preferred choice rate to an ANOVA with experiment ( 1 vs . 2 ) and bias level ( high vs . low ) as between-subjects factors ., The ANOVA showed a significant main effect of bias level ( F ( 1 , 36 ) = 8 . 8 , P = 0 . 006 ) ., There was no significant main effect of experiment ( F ( 1 , 36 ) = 0 . 6 , P>0 . 6 ) and no significant interaction between experiment and bias level ( F ( 1 , 36 ) = 0 . 3 , P>0 . 5 ) ., Replicating previous findings , the main effect of bias level was driven by higher preferred choice rate in the high , compared to the low bias group in both Experiment 1 ( T ( 18 ) = 1 . 8 P = 0 . 08 ) and Experiment 2 ( T ( 18 ) = 2 . 3 P = 0 . 03 ) ( Fig 6B & 6C ) ., From the remaining conditions we extracted the correct choice rate , which was the choice rate of the most frequently rewarded option ., In the Reversal condition , correct choice rate was split across the first half of the trial ( i . e . , before the reversal of the contingencies ) and second half ( i . e . , after the reversal of the contingencies ) ( Fig 6A ) ., We hypothesised that in the second half of the Reversal condition , where correct choice rate depends on un-learning previous associations based on negative factual prediction errors ( and positive counterfactual prediction errors , in Experiment 2 ) , high bias subjects will display reduced performance ., We submitted the correct choice rate to a mixed ANOVA with experiment ( 1 vs . 2 ) and bias group ( high vs . low ) as between-subjects factors , and condition ( Asymmetric , Reversal: first half , and Reversal: second half ) as a within-subjects factor ., There was a main effect of experiment ( F ( 1 , 36 ) = 4 . 1 , P = 0 . 05 ) , indicating that correct choice rate was higher in Experiment 2 than Experiment 1 , which is consistent with previous studies showing that counterfactual feedback enhances learning20 , 24 ., We also found a significant effect of bias level ( F ( 1 , 36 ) = 10 . 8 , P = 0 . 002 ) , a significant effect of condition ( F ( 2 , 72 ) = 99 . 5 , P = 2 . 0e-16 ) , and a significant bias level by condition interaction ( F ( 2 , 72 ) = 9 . 6 , P = 0 . 0002 ) ., Indeed , in both experiments , the correct choice rate in the second half of the Reversal condition was lower in the high bias compared to the low bias group ( Experiment 1: T ( 18 ) = 3 . 9 P = 0 . 0003; Experiment 2: T ( 18 ) = 2 . 5 P = 0 . 02 ) ( Fig 6B & 6C ) ., Importantly , we found that the temperature did not differ between low and high bias subjects in Experiment 1 ( low vs . high: 3 . 38±0 . 82 vs . 3 . 78±0 . 67; T ( 18 ) = 0 . 4 , P = 0 . 7078 ) or in Experiment 2 ( low vs . high ., 3 . 29±0 . 56 vs . 2 . 13±0 . 36; T ( 18 ) = 1 . 7 , P = 0 . 0973 ) ., Of note , the difference in temperature goes in two different directions in the two experiments , whereas the behavioural effects ( i . e . , increased preferred response rate in the Symmetric condition and decreased performance in the second half of the Reversal condition ) go in the same direction ., Finally , we used Pearson’s correlations to verify that the relevant results remained significant when assessed as continuous variables ., As predicted , the normalised learning biases were significantly and positively correlated with the preferred choice rate in the Symmetric condition in both experiments ( Experiment 1: R = 0 . 54 , P = 0 . 013; Experiment 2: R = 0 . 46 , P = 0 . 040 ) ., Similarly , the normalised learning biases were significantly and negatively correlated with the correct choice rate in the second half of the Reversal condition ( Experiment 1: R = -0 . 66 , P = 0 . 0015; Experiment 2: R = -0 . 47 , P = 0 . 033 ) ., Two groups of healthy adult participants performed two variants of an instrumental learning task , involving factual ( Experiment, 1 ) and counterfactual ( Experiments 1 & 2 ) reinforcement learning ., We found that prediction error valence biased factual and counterfactual learning in opposite directions ., Replicating previous findings , we found that , when learning from obtained outcomes ( factual learning ) , the learning rate for positive prediction errors was higher than the learning rate for negative prediction errors ., In contrast , when learning from forgone outcomes ( counterfactual learning ) , the learning rate for positive prediction errors was lower than that of negative prediction errors ., This result proved stable across different reward contingency conditions and was further supported by model comparison analyses , which indicated that the most parsimonious model was a model with different learning rates for confirmatory and disconfirmatory events , regardless of outcome type ( factual or counterfactual ) and valence ( positive or negative ) ., Finally , behavioural analyses showed that participants with a higher valence-induced learning bias displayed poorer learning performance , specifically when it was necessary to adjust their behaviour in response to a reversal of reward contingencies ., These learning biases were therefore significantly associated with reduced learning performance and can be considered maladaptive in the context or our experimental tasks ., Our results demonstrated a factual learning bias , which replicates previous findings by showing that , in simple instrumental learning tasks , participants preferentially learn from positive compared to negative prediction errors 11–13 ., However , in contrast to previous studies , in which this learning bias had no negative impact on behavioural performance ( i . e . , correct choice rate and therefore final payoff ) , here we demonstrated that this learning bias is still present in situations in which it has a negative impact on performance ., In fact , whereas low and high bias participants performed equally well in conditions with stable reward contingencies , in conditions with unstable reward contingencies we found that high bias participants showed a relatively reduced correct choice rate ., When reward contingencies were changed , learning to successfully reverse the response in the second half of the trials was mainly driven by negative factual ( and positive counterfactual ) prediction errors ., Thus in this case , participants displaying higher biases exhibited a lower correct choice rate ., In other words , these learning biases significantly undermined participants’ capacity to flexibly adapt their behaviour in changing , uncertain environments ., In addition to reduced reversal learning , and in accordance with a previous study 10 , another behavioural feature that distinguished higher and lower bias participants was the preferred response rate in the Symmetric condition ., In the Symmetric condition , both cues had the same reward probabilities ( 50% ) , such that there was no intrinsic “correct” response ., This allowed us to calculate the preferred response rate for each participant ( defined as the choice rate of the option most frequently selected by a given participant , i . e . the option selected in > 50% of trials ) ., The preferred response rate can therefore be taken as a measure of the tendency to overestimate the value of one cue compared to the other , in the absence of actual outcome-based evidence ., In both experiments , higher bias participants showed higher preferred response rates , a behavioural pattern that is consistent with an increased tendency to discount negative factual ( and positive counterfactual ) prediction errors ., This can result in one considering a previously rewarded chosen option as better than it really is and an increased preference for this choice ., Thus , these results illustrate that the higher the learning bias for a given participant , the higher his/her behavioural perseveration ( the tendency to repeat a previous choice ) , despite the possible acquisition of new evidence in the form of negative feedback ., Previous studies have been unable to distinguish whether this valence-induced factual learning bias is a “positivity bias” or a “confirmation bias” ., In other words , do participants preferentially learn from positive prediction errors because they are positively valenced or because the outcome confirms the choice they have just made ?, To address this question we designed Experiment 2 in which , by including counterfactual feedback , we were able to separate the influence of prediction error valence ( positive vs . negative ) from the influence of prediction error type ( chosen vs . unchosen outcome ) ., Crucially , whereas the two competing hypotheses ( “positivity bias” vs . “confirmation bias” ) predicted the same result concerning factual learning rates , they predicted opposite effects of valence on counterfactual learning rates ., The results from Experiment 2 support the confirmation bias hypothesis: participants preferentially took into account the outcomes that confirmed their current behavioural policy ( positive chosen and negative unchosen outcomes ) and discounted the outcomes that contradicted it ( negative chosen and positive unchosen outcomes ) ., Our results therefore support the idea that confirmation biases are pervasive in human cognition 25 ., It should be noted that , from an orthodox Bayesian perspective , a confirmation bias would involve reinforcing ones own initial beliefs or preferences ., Previous studies have investigated how prior information—in the form of explicit task instructions or advice—influences the learning of reinforcement statistics and have provided evidence of a confirmation bias 26–28 ., However , consistent with our study , their computational and neural results suggest that this instruction-induced confirmation bias operates at the level of outcome processing and not at the level of initial preferences or at the level of the decision-making process 29 , 30 ., Here , we take a slightly different perspective by extending the notion of confirmation bias to the implicit reinforcement of ones own current choice , by preferentially learning from desirable outcomes , independently from explicit prior information ., We performed a learning rate analysis separately for each task condition and the results proved robust and were not driven by any particular reward contingency condition ., Our results contrast with previous studies that have found learning rates adapt as a function of task contingencies , showing increases when task contingencies were unstable 31 , 32 ., Several differences between these tasks and ours may explain this discrepancy ., First , in previous studies , the stable and unstable phases were clearly separated , whereas in our design , participants were simultaneously tested in the three reward contingency conditions ., Second , we did not explicitly tell participants to monitor the stability of the reward contingency ., Finally , since in our task the Reversal condition represented only one quarter of the trials , participants may not have explicitly realised that changing learning rates were adaptive in some cases ., To date , two different views of counterfactual learning have been proposed ., According to one view , factual and counterfactual learning are underpinned by different systems that could be computationally and anatomically mapped onto subcortical , model-free modules , and prefrontal , model-based modules 17 , 18 , 33 ., In contrast , according to another view , factual and counterfactual outcomes are processed by the same learning system , involving the dopaminergic nuclei and their projections 34–36 ., Our dimensionality reduction model comparison result sheds new light on this debate ., If the first view was correct , and factual and counterfactual learning are based on different systems , different learning rates for positive and negative prediction errors would have better accounted for the data ( the ‘Information’ model ) ., In contrast , our results showed that the winning model was one in which the learning process was assumed to be different across desirable and undesirable outcomes , but shared across obtained and forgone outcomes ( as in the “Confirmation” model ) , This supports the second view that factual and counterfactual learning are different facets of the same system ., Overall , we found that correct choice rate was higher in Experiment 2 than in Experiment 1 , indicating that the presence of complete feedback information improved performance ., Previous literature in psychology and economics suggest that this beneficial effect of counterfactual information is conditional on the payoff structure of the task ., Specifically , studies have shown that the presence of rare positive outcomes could impair performance in the presence of complete feedback 37–40 ., Further research is needed to assess whether or not the learning biases we identified extend to these payoff schemes and how they relate to the observed performance impairment ., Another series of studies in psychology and economics have used paradigms that dissociate information sampling ( i . e . , choosing an option to discover its value without getting the outcome ) from actual choice ( i . e . , choosing an option in order to obtain the associated outcome ) 3 ., Other paradigms have been used to investigate learning from outcomes derived from choices performed by either a computer or another player ( i . e . , observational learning ) 41 , 42 ., Future research should assess whether or not information sampling and observational learning present similar valence-induced learning biases ., Why do these learning biases exist ?, One possibility is that these learning biases arise from neurobiological constraints , which limit human learning capacity ., However , we believe this interpretation is unlikely because we see no clear reason why such limits would differentially affect learning from positive and negative prediction errors ., In other words , we would predict that neurobiological constraints on learning rate would limit all learning rates in a similar way and therefore not produce valence-induced learning asymmetries ., A second possibility is that these learning biases are not maladaptive ., For instance , it has been shown that in certain reward conditions agents displaying valence-induced learning biases may outperform unbiased agents 9 ., Thus , a possible explanation for these learning biases is that they have been positively selected because they can be adaptive in the context of the natural environment in which the learning system evolved 43 ., A third , intermediate possibility is that these learning biases can be maladaptive in the context of learning performance , but due to their adaptive effects in other domains of cognition , overall they have a net adaptive value ., For example , these biases may also manifest as “self-serving” , choice-supportive biases , which result in individuals tending to ascribe success to their own abilities and efforts , but relatively tending to neglect their own failures 44 ., Accordingly , we could speculate that these learning biases may help promote self-esteem and confidence , both of which have been associated with overall favourable real life outcomes 45 ., In summary , by investigating both factual and counterfactual learning , the current experiments demonstrate that , when presented with new evidence , people tend to discard information that suggests they have made a mistake ., This selective neglect of useful information may have adaptive value , by increasing self-confidence and self-esteem ., However , this low level reinforcement-learning bias may represent a computational building block for higher level cognitive biases such as belief perseverance , that is , the phenomenon that beliefs are remarkably resilient in the face of empirical challenges that logically contradict them 46 , 47 ., The study included two experiments ., Each experiment involved N = 20 participants ( Experiment 1: 7 males , mean age 23 . 9 ± 0 . 7; Experiment 2: 4 males , mean age 22 . 8 ± 0 . 7 ) ., The local ethics committee approved

Introduction, Results, Discussion, Methods

Previous studies suggest that factual learning , that is , learning from obtained outcomes , is biased , such that participants preferentially take into account positive , as compared to negative , prediction errors ., However , whether or not the prediction error valence also affects counterfactual learning , that is , learning from forgone outcomes , is unknown ., To address this question , we analysed the performance of two groups of participants on reinforcement learning tasks using a computational model that was adapted to test if prediction error valence influences learning ., We carried out two experiments: in the factual learning experiment , participants learned from partial feedback ( i . e . , the outcome of the chosen option only ) ; in the counterfactual learning experiment , participants learned from complete feedback information ( i . e . , the outcomes of both the chosen and unchosen option were displayed ) ., In the factual learning experiment , we replicated previous findings of a valence-induced bias , whereby participants learned preferentially from positive , relative to negative , prediction errors ., In contrast , for counterfactual learning , we found the opposite valence-induced bias: negative prediction errors were preferentially taken into account , relative to positive ones ., When considering valence-induced bias in the context of both factual and counterfactual learning , it appears that people tend to preferentially take into account information that confirms their current choice .

While the investigation of decision-making biases has a long history in economics and psychology , learning biases have been much less systematically investigated ., This is surprising as most of the choices we deal with in everyday life are recurrent , thus allowing learning to occur and therefore influencing future decision-making ., Combining behavioural testing and computational modeling , here we show that the valence of an outcome biases both factual and counterfactual learning ., When considering factual and counterfactual learning together , it appears that people tend to preferentially take into account information that confirms their current choice ., Increasing our understanding of learning biases will enable the refinement of existing models of value-based decision-making .

learning, decision making, social sciences, neuroscience, learning and memory, optimization, analysis of variance, cognitive psychology, mathematics, statistics (mathematics), cognition, research and analysis methods, learning curves, human learning, behavior, mathematical and statistical techniques, economics, economic history, psychology, biology and life sciences, physical sciences, cognitive science, statistical methods

journal.pntd.0003369

Risk Factors Associated with Malnutrition in One-Year-Old Children Living in the Peruvian Amazon

Malnutrition is the leading cause of mortality in preschool-age children ( i . e . children under five years of age ) in low- and middle-income countries ( LMICs ) ., Over 150 million children suffer from one or more forms of malnutrition , including stunting , underweight and wasting 1 , 2 ., Malnutrition also predisposes to infection , creating a vicious infection-malnutrition cycle that contributes to over 35% of the disease burden of early childhood 1 , 3 ., Infection and micronutrient and other deficiencies from an inadequate diet are the primary causes of malnutrition in childhood 4 ., Early childhood before the age of two years is a particularly critical time for growth faltering 5 ., This window of time corresponds to weaning and the introduction of complementary foods ., As mobility increases , the risk of early acquisition of certain infectious pathogens also increases during this time ., The soil-transmitted helminths ( STHs ) , or worm infections , are one such pathogen cluster that is transmitted through contaminated food , water and/or the environment in warm , tropical and subtropical climates ., The STH disease cluster includes ascariasis ( caused by the roundworm Ascaris lumbricoides ) , trichuriasis ( caused by the whipworm Trichuris trichiura ) and ancylostomiasis or hookworm disease ( caused either by Ancylostoma duodenale or Necator americanus ) ., The geographical distribution of these three diseases is overlapping , mainly in areas of poverty with poor sanitation and limited access to potable water ., STHs are one of the most important Neglected Tropical Diseases ( NTDs ) and one of the most common infections worldwide ., Recent estimates indicate that 1 . 45 billion people are infected with STHs in over 100 endemic countries 6 ., It is estimated that they contribute 4 . 98 million years lived with disability ( YLD ) and 5 . 18 million disability-adjusted life years ( DALYs ) 6 ., STHs are a significant contributor to poor health and nutritional status in all age groups , and especially in childhood ., Traditionally , the occurrence of STH infection had been perceived to be low in children under two years of age ., However , there has been increasing empirical evidence which shows that the opposite is true 7 ., In Belén , a community of extreme poverty in the Peruvian Amazon , while the prevalence of Ascaris or Trichuris was only 4% in children at seven to nine months of age , it rose to almost 30% at 12 to 14 months of age 8 ., In a cohort of preschool-age children in Ecuador , over 20% suffered from Ascaris or Trichuris infection at least once in the first two years of life , with infection first appearing around seven months of age 9 ., There is also evidence to suggest that hookworm infection may be high in early preschool-age children ., A study in Zanzibar by Stoltzfus et al ( 2004 ) , demonstrated that 31 . 3% of children under 30 months of age were infected with hookworm 10 ., It is becoming increasingly recognized that STH infection in early childhood may have important adverse effects on health and nutrition 11 , 12 ., One such reason for this is that the parasites take up a greater proportion of the body in younger children 13 ., However , the importance of STH infection and its link with malnutrition in preschool-age children has been inadequately studied ., Few studies have included preschool-age children in their study population ., Even fewer studies have provided age-disaggregated data to examine differing effects and sequelae in the critical growth window before two years of age ., Evidence from the World Health Organization ( WHO ) Child Growth Standards demonstrates that , with appropriate nutrition and health interventions provided early in life , all children have a similar potential for healthy growth and development 14–16; however , children living in areas of greatest poverty suffer the most from health and social inequities due to increased disease burden and lack of access to necessary health interventions and services 17 ., Improving the health of the youngest children has been a focus of many international efforts , including Canadas Muskoka Initiative , and the Millennium Development Goals ( MDGs ) which aim to reduce poverty worldwide by 2015 ., With focus now shifting to the post-2015 MDG agenda , it is imperative to fill in knowledge gaps on the burden of disease and risk factors in early childhood to improve health in the short and the long term 18 ., The principal objective of this study was to determine the association between malnutrition ( i . e . stunting and underweight ) and soil-transmitted helminth infection and other child , maternal and household characteristics in 12 and 13-month old children , living in an area of extreme poverty in the Peruvian Amazon ., This study received ethics approval in Peru from the Comité Institucional de Ética of the Universidad Peruana Cayetano Heredia and the Instituto Nacional de Salud , in Lima , and the local Ministry of Health office ( Dirección Regional de Salud Loreto ) in Iquitos ., Ethics approval was obtained in Canada from the Research Ethics Board of the Research Institute of the McGill University Health Centre in Montréal , Québec ., Written informed consent was obtained by the parents or guardian of each child that participated in the study ., This study was conducted in neighbouring districts in and around the city of Iquitos , the capital of the Loreto region in the Peruvian Amazon ( Fig . 1 ) ., The study area included four districts ( Belén , Iquitos , Punchana and San Juan ) where poverty is widespread , STH infections are highly endemic and malnutrition prevalence is high ., Both malnutrition and STH prevalence have been identified as priority concerns by stakeholders in the community 19 ., The study population included children attending their routine 12-month growth and development ( “Crecimiento y Desarrollo” or CRED ) clinic visit in the study area , and whose parents had agreed to their participation in a randomized controlled trial ( RCT ) to determine the benefit of deworming ( mebendazole ) on growth and development ( ClinicalTrials . gov #NCT01314937 ) ., The current cross-sectional survey was nested within the deworming RCT , and describes information obtained at the baseline 12-month CRED trial visit ., Preschool-age children are scheduled to attend routine government-sponsored CRED visits ( similar to well baby clinics ) at health clinics in Peru once-monthly from birth to 11 months of age ( with two visits before one month of age ) , and every two months from 12 to 24 months of age ( with less frequent visits thereafter to school age ) ., During routine CRED visits , anthropometric measurements ( e . g . length and weight ) are taken , developmental milestones are recorded , and children receive routine age-appropriate vaccinations and micronutrient supplements ., Parents also receive nutrition and other health counselling for their child 20 ., Using information provided by the Peruvian Ministry of Health on health centre location and attendance , 12 study health centres ( “Centros de Salud” ( C . S . ) and “Puestos de Salud” ( P . S . ) ) were identified in the study area ., These included:, 1 ) P . S . America;, 2 ) C . S . Belen;, 3 ) C . S . Bellavista Nanay;, 4 ) C . S . Cardozo;, 5 ) P . S . 1 de Enero;, 6 ) C . S . 6 de Octubre;, 7 ) C . S . 9 de Octubre;, 8 ) P . S . Masusa;, 9 ) P . S . Porvenir;, 10 ) C . S . Progreso;, 11 ) C . S . San Juan; and, 12 ) P . S . Tupac Amaru ., Inclusion criteria for participating in the study were:, 1 ) children attending any one of the study health centres for their 12-month CRED visit; and, 2 ) children living in Belén , Iquitos , Punchana or San Juan districts ., Exclusion criteria preventing participation in the study were:, 1 ) children attending the health centre for suspected STH infection;, 2 ) children who had received deworming treatment in the six months prior to the study;, 3 ) children whose families planned to move outside of the study area within the next 12 months;, 4 ) children under 12 months of age or 14 months of age or older; and, 5 ) children with any serious congenital or chronic medical condition ( e . g . chronic severe malnutrition , extremely preterm birth ( i . e . < 28 weeks gestation ) , newborn hypoxia and neural tube defects ) ., All inclusion and exclusion criteria were based on considerations related to participation in the deworming trial ., All children who were enrolled in the deworming RCT were included in the current study ., The sample size of the RCT was estimated to be 1760 children , or 440 children per intervention group ( MC4G Software© , GP Brooks , Ohio University , 2008 ) ., This was based on detecting a minimum difference of 0 . 20 kg in mean weight gain among different deworming interventions ( 3 intervention groups , and 1 control group ) ., Canvassing of the local population was undertaken between April 2011 and August 2011 prior to recruitment to assist in identifying potentially eligible children for the study ., In households where any child under 12 months of age was present , information was recorded on the childs date of birth and address ., Lists of CRED attendance from each health centre were also provided to identify children who would be potentially eligible to participate in the study based on place of residence and age of the child ., Nine trained research assistants ( RAs ) , primarily nurses and nurse-midwives , were assigned to one or two health centres each to recruit study participants in the respective communities and health centres and to obtain all study outcomes ., Additional nurse-technicians were hired and trained to assist in participant recruitment ., For parents of eligible children , an informed consent form was administered and signed ., A questionnaire , which included questions on socio-demographic and health information about the child and family , was then administered by the RA during a household interview with that parent who was the primary caregiver ., The questionnaire was adapted from previous studies 8 , 21–23 , but included additional questions related to child nutrition ., These included history and duration of breastfeeding , and first introduction of liquids and solid foods ., The latter was confirmed by redundancy among the questions and a 24-hour dietary recall ., At the end of the home visit , parents were also provided with the information and materials needed to collect a stool specimen from the child ., Parents were then given an appointment at the health centre , at which time they would deposit the stool specimen and the childs anthropometric measures and development would be ascertained ., All forms and questionnaires were returned to the study offices at the end of each work day , and reviewed by the Project Director , the local Study Coordinator , and , when needed , by the local Principal Investigator , to confirm the eligibility of each child ., During the visit at the health centre , the quality of the stool specimen was first verified ., If no specimen or an inadequate specimen ( i . e . liquid specimen and/or insufficient quantity ) was provided , then anthropometry was ascertained and a subsequent visit was scheduled to arrange for another stool specimen ., If any child was discovered to be ill on the day of his or her health centre visit , the visit was postponed until the child had recovered ., After verification of the quality and quantity of the stool specimen , the child was undressed and weighed ( in duplicate ) using a portable electronic scale ( Seca 334 , Seca Corp . , Baltimore , MD , USA ) ., Length ( i . e . the recommended measurement for height in children less than two years of age ) was measured ( in duplicate ) as recumbent crown-heel length on a flat surface using a stadiometer ( Seca 210 , Seca Corp . , Baltimore , MD , USA ) ., Cognition , receptive and expressive communication ( i . e . language ) and fine motor development were assessed using the Bayley Scales of Infant and Toddler Development , Third Edition ( Bayley-III ) ( Pearson Education Inc , Texas , 2006 ) ., The latter instrument was translated into Spanish and adapted for local cultural appropriateness and validity by the Project Director ( SAJ ) and an experienced psychologist from the Instituto de Investigación Nutricional ( IIN ) in Lima , Peru ., All RAs were trained on administration of the Bayley-III by SAJ and the IIN psychologist ., As little variability was anticipated in gross motor skills at 12 months of age , and to reduce the length of time of assessment , the WHO gross motor milestones ( i . e . walking alone , standing alone , walking with assistance , hands and knees crawling , and standing with assistance ) was used instead of the Bayley-III Gross Motor subtest ., Childrens gross motor skill achievement was assessed by observation by RAs 24 ., Upon completion of all baseline outcome measurements and the provision of an adequate stool specimen , participants were enrolled into the deworming trial and randomly assigned to one of three intervention groups or the control group ., The stool specimen was labeled with a unique number between 1 and 1760 , corresponding to the randomly assigned treatment code for the deworming trial ., Stool specimens were transferred to the laboratory at the local research facility ( Asociación Civil Selva Amazónica ) to be read by one of two experienced laboratory technologists ., Two different techniques were required for reading stool specimens in the nested study based on the childs treatment allocation in the larger deworming RCT ., Stool specimens from participants who were randomized to receive active deworming treatment were analyzed immediately by the Kato-Katz method , as recommended by WHO ( within 24 hours of initial collection , as a fresh specimen is required for this technique ) to determine both prevalence and intensity of STH infection 25 , 26 ., This procedure of immediately analyzing stool specimens only of those randomly allocated to the intervention groups receiving active deworming treatment takes into account the ethical imperative of treating those who would be found to have positive results ., Stool specimens of those receiving inactive placebo tablets were stored in 10% formalin and examined by the direct method upon completion of the trial , at which time all participants received deworming treatment ., To maintain blinding , each specimen code was replaced with a laboratory code by the local study coordinator for use by the laboratory technologists ., Laboratory technologists were provided with a list of those laboratory codes which would be analyzed and those which were to be stored ., Each list was kept on a password-protected computer , one in the coordinators office and one in the lab accessible only to the laboratory supervisor ., A master list linking all information was stored at the research office in Canada ( Research Institute of the McGill University Health Centre ) ., Quality control was conducted on 10% of all Kato-Katz slides to ensure agreement in species identification and egg counts between laboratory technologists ., To classify child anthropometric measurements ( i . e . length and weight ) into categories of stunting , underweight and wasting , WHO Anthro software ( Version 3 , 2011 ) was used to calculate length-for-age z scores ( LAZ ) , weight-for-age z scores ( WAZ ) , and weight-for-length z scores ( WLZ ) , respectively ., Z scores are calculated taking into account a childs sex and age and are based on a comparison to a WHO international standard population ., Moderate-to-severe categories of stunting , underweight and wasting are based on LAZ , WAZ and WLZ of <−2SD ., Severe stunting , underweight and wasting are defined as LAZ , WAZ and WLZ of <−3SD , respectively 27 ., Categories of STH infection intensity were determined from established WHO guidelines 28 ., For Ascaris infection , light , moderate and heavy intensity are based on egg counts per gram of feces ( epg ) of 1–4999 , 5000–49999 and 50000 and greater , respectively ., For Trichuris infection , the categories for light , moderate and heavy intensity infection are an epg of 1–999 , 1000–9999 and 10000 and greater , respectively ., Light , moderate and heavy intensity hookworm infection are based on epgs of 1–1999 , 2000–3999 and 4000 and greater , respectively ., Both arithmetic and geometric mean epg were calculated and reported ., The development score was calculated as the mean crude score for each subtest of the Bayley-III , as well as a composite score of all four subtests combined ., The range of possible scores was 0 to 91 for cognition , 0 to 49 for receptive communication , 0 to 48 for expressive communication and 0 to 66 for fine motor skills ., Scaled scores were derived from scaling the total raw score in each individual subtest to a metric between 1 ( i . e . the lowest possible score ) and 19 ( i . e . the highest possible score ) according to the subtest and age of the child in months and days 29 , 30 ., As scaled scores are based on a study population that may not be representative of the general population , these scores were used for descriptive purposes only and not to quantify the level of developmental deficit ., The WHO gross motor milestones were categorized into a dichotomous variable indicating whether the child had achieved the most advanced milestone of walking alone ., The variable was coded as one , if the child was able to walk without any assistance or support , regardless of the other milestones achieved , and zero , if the child could not walk without assistance , but achieved at least one of the other gross motor milestones ., Principal Component Analysis was used to create an asset-based index for socioeconomic status ( SES ) to be included in multivariable analyses ( StataCorp . 2013 . Stata Statistical Software: Release 13 . College Station , TX: StataCorp LP ) ., Variables included in the index were house material , type of cooking fuel , television ownership , radio ownership and electricity in the home 31 , 32 ., The socioeconomic status index explained 40 . 1% of the variance and was divided into quartiles for subsequent analyses ., All associations with the outcomes of stunting and underweight were examined initially in univariable analyses ., Variables with a p value<0 . 20 , or that were deemed to be important from previous published research , were included in multivariable modelling to determine the most parsimonious model ., If variables were highly correlated , the most informative variable ( i . e . with more variation , more accurate measurements and/or important factors in previous literature ) was chosen to be included in multivariable model building ., Multivariable associations with stunting and underweight were examined using a generalized linear model with a log link , a Poisson distribution , and a robust variance estimator to estimate the risk ratio for the dichotomous outcomes of moderate-to-severe stunting and moderate-to-severe underweight , where no and mild categories of stunting , and no and mild categories of underweight , respectively , comprised the reference groups 33 , 34 ., Analyses were first restricted to children whose stool specimens were examined by the Kato-Katz method 26 ., Analyses were then performed including all children in the study population ., A complete case approach was used to analyze the data ., All statistical analyses were performed using the Statistical Analysis Systems statistical software package version 9 . 3 ( SAS Institute , Cary , NC , USA ) ., Between September 2011 and June 2012 , parents of 2297 children 12 to 13 months of age were approached to participate in the study in order to meet the sample size requirements of 1760 eligible children ., Three-hundred and eighty-five children did not meet the inclusion criteria , parents of 126 children declined to participate , and 26 children were recruited but the sample size was reached before they were enrolled in the study ., Anthropometric and development measurements and stool specimens were obtained from all 1760 enrolled children ., Baseline characteristics of the study population are described in Table 1 ., The average number of CRED visits before enrolment in the study ( i . e . from birth to 11 months , inclusive ) was 7 . 6 ( ±3 . 5 ) ., Less than 4% of children had no previous CRED attendance ( n\u200a=\u200a62 ) ., Only 25 . 5% ( n\u200a=\u200a447 ) had all vaccinations up-to-date according to Peruvian Ministry of Health guidelines ( i . e . one dose of Bacille Calmette-Guérin ( BCG ) , one dose of hepatitis B , three doses of polio , three doses of pentavalent , two doses of rotavirus , three doses of pneumococcal and one dose of measles , mumps and rubella ( MMR ) vaccines ) 20; however , as MMR vaccine and the third dose of pneumococcal vaccine are scheduled at the 12-month CRED visit , many children had not yet received these latter vaccinations ., Including only vaccinations scheduled prior to 12 months , coverage of up-to-date vaccinations reached 80 . 3% ., In terms of family and household characteristics , the average maternal age was 26 . 5 ( ±7 . 1 ) years ., The average number of people living in the household was 6 . 6 ( ±2 . 7 ) ., Sixty-nine percent of children had one or more siblings ., Roughly half of the children ( 50 . 1% ) had received liquids ( other than water and water-based drinks ) or food before the age of six months ., Baseline socio-demographic and epidemiological characteristics were similar in the 880 children whose stool specimens were examined by the Kato-Katz method compared to the entire study population of children ( n\u200a=\u200a1760 ) ( results not shown ) ., Twenty-five percent of the study population suffered from one or more forms of malnutrition ., Prevalence of moderate-to-severe underweight , stunting and wasting were 8 . 6% , 24 . 2% and 2 . 3% , respectively ( Table 2 ) ., Co-morbidity with two or three concurrent forms of malnutrition was present in 8 . 3% ( n\u200a=\u200a146 ) of participants ., Mean z scores for the study population were below average ( i . e . below 0 ) for all three indices ( i . e . LAZ , WAZ and WHZ ) ., Severe malnutrition ( i . e . a z score of <−3 SD for LAZ , WAZ or WLZ ) affected 5 . 5% ( n\u200a=\u200a96 ) of the population ., The overall prevalence of any STH infection in children whose stool specimens were analyzed by the Kato-Katz method was 14 . 5% ( Table 3 ) ., The prevalence of infection was 11 . 5% for Ascaris , 4 . 5% for Trichuris and 0 . 6% for hookworm ., Eighteen children ( 2 . 1% ) were infected with two STH species , but none with all three ., For those who had their stool specimens stored and analyzed by the direct method , the prevalence was lower for all three STH species ( i . e . 9 . 5% , 0 . 9% and 0 . 1% for Ascaris , Trichuris , and hookworm , respectively ) ., Using the Kato-Katz method as the gold standard , and assuming equal STH prevalence due to randomization , the direct method , therefore , underestimated Ascaris infection by 17 . 4% , Trichuris infection by 80 . 0% , hookworm infection by 83 . 3% , and any STH prevalence by 29 . 0% ., For the 880 children whose stool specimens were examined using the Kato-Katz method and who were found to be STH positive , most were found to have low intensity infection , with 86 . 1% , 92 . 5% and 100% harbouring light infections of Ascaris , Trichuris and hookworm , respectively ( Table 4 ) ., There were no cases of heavy intensity infection of any STH species ., In terms of developmental functioning in all 1760 children , the mean composite development score on the Bayley-III was 98 . 1 ( ± SD 6 . 0 ) with a range between 73 and 123 points ., On individual subtests , the mean score was 42 . 5 ( ±3 . 0 ) for cognition , 12 . 9 ( ±1 . 6 ) for receptive communication , 13 . 5 ( ±2 . 1 ) for expressive communication and 29 . 2 ( ±1 . 5 ) for fine motor skills ., This translated to a mean scaled score of 9 . 9 ( ±1 . 84 ) , 7 . 2 ( ±1 . 9 ) , 8 . 1 ( ±1 . 7 ) and 9 . 2 ( ±1 . 5 ) for the cognitive , receptive language , expressive language and fine motor subtests , respectively ., The mean scores were slightly higher for 13-month old children compared to 12-month old children ( i . e . 43 . 2 vs . 42 . 5 for cognition , 13 . 3 vs . 12 . 9 for receptive communication , 13 . 8 vs . 13 . 4 for expressive communication , and 29 . 4 vs . 29 . 2 for fine motor skills , respectively ) ., Twenty-three percent and 35 . 6% of 12 and 13-month old children , respectively , were able to walk without support ., In determining the risk factors for malnutrition in the group of children whose specimens were analyzed by the Kato-Katz method , stunting was found to be statistically significantly associated with the presence of any STH infection , male sex , older age ( i . e . 13 months old ) , one or more hospitalizations since birth , lower SES , and lower birth weight in both unadjusted and adjusted analysis ( Table 5 ) ., The crude score of each individual Bayley-III subtest was significantly associated with stunting in univariable analyses ., The overall composite development score was included in the multivariable model , with a lower score associated with an increased risk of stunting in the adjusted model ( aRR 0 . 97; 95% CI: 0 . 95 , 0 . 99 ) ., Risk factors for underweight in unadjusted and adjusted analyses included lower birth weight , lower development score , and lower SES ( Table 5 ) ., Continued breastfeeding at one year of age was associated with a decreased risk of underweight in unadjusted and adjusted analyses ., No statistically significant association was found between underweight and any STH infection in either unadjusted or adjusted analyses ., No independent associations were found between malnutrition and up-to-date vaccinations , vitamin A supplementation , walking alone , maternal employment outside of the home , place of residence , place of delivery or antenatal care attendance ( Table 5 ) ., The timing of introduction of liquids and foods was not associated with stunting or underweight in either unadjusted or adjusted analyses ., STH infection was not associated with wasting in either unadjusted or adjusted analyses ( results not shown ) ., Multivariable results for stunting , underweight and wasting were similar when analyses were extended to include participants with specimens analyzed by both the Kato-Katz and the direct method ( results not shown ) ., The scientific literature to date has provided insufficient evidence of an association between malnutrition and STH infection in early preschool-age children ., This nested cross-sectional study in 1760 preschool-age children aged 12 and 13 months in a community of extreme poverty in the Peruvian Amazon contributes to filling this research gap ., We demonstrate an important association between malnutrition and STH infection and developmental deficits ., Previous studies in the area of Belen have found similar associations between malnutrition and STH infection in a wider age range of preschool-age children 8 , 23 ., In contrast to previous studies , however , this association was apparent even with low intensity STH infection 8 ., The current study updates previous estimates and provides in-depth data for that critical time period around one year of age when interventions are likely to be considered to be integrated into vaccination programs or well baby clinics ., Consistent with previous studies , lower socioeconomic status and older child age were associated with a higher risk of malnutrition 8 , 23 , 35 ., Nonetheless , the latter result is somewhat unexpected , as the age range was quite restricted in the present study ., This finding , along with a greater number of children who were walking alone at 13 months of age , support the concept of a critical window in which children are rapidly developing and growing before two years of age 5 ., This has the potential to translate to an even greater impact of parasite infection and nutritional deficits on child health in this time period ., An interesting finding in this study was that STH prevalence ( from either detection method ) and malnutrition prevalence were lower compared to previous work in the area 8 ., The current study was embedded within the existing health infrastructure of routine growth and development clinic visits ., Although previous attendance was not an inclusion criterion , there may have been higher-risk populations with low CRED attendance that would not have been easily reached , but who may have been included in the previous community-based surveys ., We attempted to solve this problem by conducting community canvassing prior to enrolment to identify all children in the eligible jurisdictions , not only those who had had the opportunity to access health services previously ., An increase in research attention and community-based health and nutrition campaigns may also explain some of the improvements ., In particular , deworming campaigns directed towards school-age children , may have contributed to a reduction in overall environmental contamination in the area ., This could have resulted in lower infection rates in younger children not directly targeted by campaigns , as has been shown in other settings 36 ., A recent study also demonstrated a decrease in the prevalence of stunting in preschool-age children in Peru from 1991 to 2011 , possibly due to economic growth and an increased emphasis on pro-poor social programs 37 ., However , the overall prevalence of stunting has remained unacceptably high , with children between 12 and 23 months , those living in the Amazon or Andean region , and those of lower SES , suffering disproportionately from malnutrition 37 ., Prevalence of stunting was also higher in males compared to females under the age of 36 months , which is consistent with our findings ., Despite the positive trends in a reduction in stunting and STH infection in this and other studies , the current results demonstrate that even low STH prevalence and intensity of infection can be associated with poor growth in children in this vulnerable age group ., This study benefits from a large sample size of children , representative of the wider population of children living in the STH high-risk flooding areas of Iquitos ., This representativity was helped in part by the community-wide canvassing and by the inclusion of health centres from a wide catchment area ., Nevertheless , hard-to-reach and hidden populations of children suffering from severe malnutrition or other chronic illnesses may be under-represented in the study ., An additional strength of the study is the focus on children of a narrow age range in the critical growth window ., Other studies have included populations of children at heterogeneous growth and development stages and have been unable to disaggregate outcome results by age ., In-depth information on potential risk factors was also collected to ensure that the impact of other child , maternal and household characteristics were taken into account in the analysis ., The ascertainment of nutritional information , such as when liquids and foods were first introduced and the age of weaning may have been limited by recall bias; however , the collection of information on the age of introduction of specific local foods and a 24-hour recall were used to increase validity of the responses ., This study also incorporated comprehensive developmental testing ., To our knowledge , this is the first study that has incorporated the Bayley-III , one of the most rigorous development tests available for preschool-age children , in conjunction with STH infection ., We were also able to take into account the potential effects of SES by using an asset-based proxy index ., The study was limited by the fact that , for ethical reasons , the Kato-Katz method could only be used to analyze half of the specimens from randomly-allocated participants ( i . e . those who were randomly a

Introduction, Methods, Results, Discussion

Children under two years of age are in the most critical window for growth and development ., As mobility increases , this time period also coincides with first exposure to soil-transmitted helminth ( STH ) infections in tropical and sub-tropical environments ., The association between malnutrition and STH infection , however , has been understudied in this vulnerable age group ., A nested cross-sectional survey was conducted in 12 and 13-month old children participating in a deworming trial in Iquitos , an STH-endemic area of the Peruvian Amazon ., An extensive socio-demo-epi questionnaire was administered to the childs parent ., Length and weight were measured , and the Bayley Scales of Infant and Toddler Development were administered to measure cognition , language , and fine motor development ., Stool specimens were collected to determine the presence of STH ., The association between malnutrition ( i . e . stunting and underweight ) and STH infection , and other child , maternal , and household characteristics , was analyzed using multivariable Poisson regression ., A total of 1760 children were recruited between September 2011 and June 2012 ., Baseline data showed a prevalence of stunting and underweight of 24 . 2% and 8 . 6% , respectively ., In a subgroup of 880 randomly-allocated children whose specimens were analyzed by the Kato-Katz method , the prevalence of any STH infection was 14 . 5% ., Risk factors for stunting in these 880 children included infection with at least one STH species ( aRR\u200a=\u200a1 . 37; 95% CI 1 . 01 , 1 . 86 ) and a lower development score ( aRR\u200a=\u200a0 . 97; 95% CI: 0 . 95 , 0 . 99 ) ., A lower development score was also a significant risk factor for underweight ( aRR\u200a=\u200a0 . 92; 95% CI: 0 . 89 , 0 . 95 ) ., The high prevalence of malnutrition , particularly stunting , and its association with STH infection and lower developmental attainment in early preschool-age children is of concern ., Emphasis should be placed on determining the most cost-effective , integrated interventions to reduce disease and malnutrition burdens in this vulnerable age group .

Malnutrition , including stunting and underweight , is one of the leading causes of morbidity and mortality in preschool-age children ., Children under two years of age are at a particularly critical period for growth and development , and for first exposure to worm infections in tropical and subtropical environments ., The association between malnutrition and worm infection , however , is not well understood in this age group ., A nested cross-sectional survey was therefore conducted between September 2011 and June 2012 in 1760 children 12 and 13 months of age living in a worm-endemic area of the Peruvian Amazon ., Length , weight , development ( i . e . cognitive , language and motor development ) , worm infection , and socio-demographic information were obtained ., Results showed a high prevalence of stunting , and a significant association with worm infection and lower development ., Overall , these adverse effects have the potential to negatively impact short-term and long-term health and nutrition , and educational and social achievement , into school-age and adulthood ., Emphasis is needed on determining the most appropriate and effective interventions to reduce poor health and nutrition outcomes in this age group .

helminth infections, medicine and health sciences, nutrition, epidemiology, biology and life sciences, soil-transmitted helminthiases, parasitic diseases, malnutrition

journal.ppat.1004765

Exome and Transcriptome Sequencing of Aedes aegypti Identifies a Locus That Confers Resistance to Brugia malayi and Alters the Immune Response

The rate at which parasites are transmitted by mosquitoes is an important determinant of the prevalence of vector-borne diseases in human populations ., Alongside factors like the number of mosquitoes and their biting preferences , the rate of transmission depends on the ability of mosquitoes to acquire the parasite when feeding on an infected person and subsequently transmit it ., This is referred to as their vector competence , and is affected by both environmental and genetic factors 1 , 2 ., Even within a population of a single mosquito species there can be tremendous genetic variation in vector competence , often as a result of differences in the immune response of the mosquito to the parasites they are vectoring 2 ., For example , variation exists in susceptibility of Anopheles gambiae to the malaria parasite Plasmodium falciparum 3 , 4 and in Aedes aegypti to dengue and filarial nematodes 5 , 6 ., This has attracted much attention as it could one day lead to be better disease control by manipulating mosquito populations to reduce vector competence ., For example , field trials are underway that are releasing Ae ., aegypti mosquitoes carrying the bacterial symbiont Wolbachia , which reduces the mosquitoes’ ability to transmit dengue virus 7 ., The tropical disease lymphatic filariasis , or elephantiasis , is a leading cause of morbidity and disability worldwide , with especially high parasite burdens in Africa and south and south-east Asia 8 ., It is estimated to affect 120 million people worldwide , and symptoms include lymphedema and swelling of the extremities 9 ., In humans , the disease is caused by the filarial nematodes Wuchereria bancrofti , Brugia malayi and Brugia timori , and is vectored by a range of mosquitoes , including species of Culex , Mansonia , Anopheles and Aedes 9 ., W . bancrofti is the major cause of filariasis worldwide , leading to 90% of the cases of lymphatic filariasis , and Brugia species , which are only found in Asia , cause the remaining 10% 8 ., B . malayi is the main laboratory model for studying lymphatic filariasis , and it grows readily in some strains of the mosquito Ae ., aegypti ., Despite having overlapping ranges , Ae ., aegypti does not naturally vector any of the nematodes that cause lymphatic filariasis in humans ., It is however a natural vector of Dirofilaria , which causes filariasis in dogs 10 ., B . malayi , along with other filarial nematodes , are heteroxenous , requiring both a vertebrate host and a mosquito vector for their life cycle 8 , 9 ., Humans , cats , and monkeys can all serve as vertebrate hosts for B . malayi 10 ., Male and female worms reproduce sexually in the vertebrate , producing microfilariae which circulate in the bloodstream and are ingested by mosquitoes during blood feeding ., After penetrating the mosquito midgut , the filarial nematodes develop inside various tissues within the mosquito ., In the case of B . malayi , the microfilariae migrate to the thoracic muscles of the mosquito , where they undergo successive molts until they become L3 larvae 11 ., They then migrate to the mosquito proboscis , where they are transferred to the vertebrate host during blood feeding ., Beginning in the 1960’s , mosquito strains and species have been identified that are naturally refractory ( resistant ) to infection by filarial nematodes 12 ., Proposed mechanisms of resistance include reduced ingestion of parasites , physical killing of parasites in the foregut , barriers to penetration of the midgut , and hemolymph factors that kill the parasite in the thoracic cavity and lead to melanotic encapsulation 13 ., Some species such as Armigeres subalbatus , a natural vector of Brugia pahangi , are completely refractory to infection by B . malayi while being highly susceptible to B . pahangi 14 ., Others , such as Ae ., aegypti , are polymorphic within species for resistance 15 ., In laboratory lines of Ae ., aegypti , genetic variation in resistance to B . malayi has a simple genetic basis , and is primarily determined by a single dominant locus on the first chromosome 16 ., This genetic resistance extends to some other species of nematodes , such as B . pahangi and W . bancrofti , but not to Dirofilaria , for which Ae ., aegypti is a natural vector 16 ., In this mosquito , sex is also determined by a region on the first chromosome , and the resistance locus is tightly linked to the sex-determining region 17–19 ., The immune responses of Ae ., aegypti have been extensively studied , but it remains unknown which factors are important in killing filarial nematodes and whether genetic differences in susceptibility are caused by differences in immune responses ., Despite the mechanisms being unclear , the mosquito immune response does appear to control filarial nematodes ., Fewer parasites reached the L3 stage when the immune system was upregulated by inoculating mosquitoes with bacteria before they fed on blood carrying microfilariae 20 ., Similarly , parasite numbers were reduced when the mosquito was infected with the bacterium Wolbachia , which also upregulated the immune response 21 ., Anti-microbial peptide ( AMP ) production may be responsible for these effects as cecropin negatively affects worm motility 22 ., However , activation of the two main immune signaling pathways , Toll and IMD , by RNAi depletion of their negative regulators , Cactus and Caspar , produced no measurable effect on resistance to B . malayi 11 ., Genetic mapping of parasite resistance in mosquitoes has so far been done by individually testing markers 3 , 6 , 18 or with high-density genotyping using SNP arrays or RAD-sequencing 19 , 23 ., These approaches often utilize randomly selected markers sparsely interspersed in the genome and rely on markers being in linkage disequilibrium with the causative polymorphism , which itself is unlikely to be sampled ., In species like An ., gambiae , linkage disequilibrium extends very short distances in wild populations 24 , and it is preferable to concentrate efforts on regions that are likely to be involved in the trait of interest ., In humans , the solution has been to use exome capture to sequence only protein-coding regions of the genome , which has been met with much success in identifying the mutations that cause Mendelian diseases 25 ., This is especially desirable in species like humans and Ae ., aegypti , where the large and repetitive genomes mean that whole genome sequencing is prohibitively costly and that much of the non-coding sequence cannot be investigated because relatively short sequence reads cannot be uniquely mapped to the genome ., We have investigated the genetic and mechanistic basis of resistance to B . malayi in Ae ., aegypti using a combination of genomic and transcriptomic approaches ., First , we resequenced the exome using probes we designed for Ae ., aegypti and performed an association study to map the locus causing resistance with unprecedented precision ., Using RNA-seq , we then measured gene expression in resistant and susceptible genotypes of the mosquito to understand how this locus alters the transcriptional response to filarial nematode infection ., To minimize the contribution of random genetic differences between the resistant and susceptible lines , we performed genetic crosses to isolate the resistance locus in a common genetic background ., This allowed us to identify differences in immune and non-immune response gene expression that will facilitate our understanding of mechanisms of resistance ., A wild outcrossed population was established for association mapping ., Mosquito eggs were collected in July 2010 from a 120 km stretch between Kilifi , Malindi , and Mombasa in coastal Kenya using oviposition traps 26 ., Each trap consisted of a black plastic cup , hay infused water ( 4 g dried grass in 1 L of water for 4 days ) and a strip of creped cardboard paper ., Eggs from each collection site ( median of 42 eggs/trap with 1–16 traps used per collection site ) were hatched in the laboratory and reared separately ., Strains were established from two collection sites near Kilifi ( St . Thomas and Mabarikani ) and one site each near Malindi ( Muthangani ) and Mombasa ( Mtwapa ) ., At the F2 generation all strains were reciprocally crossed to each other and to themselves , with similar numbers of males and females in each group ., Fifteen males and fifteen females from each cross ( 480 individuals total ) were used to start an outcrossed population , where they were allowed to mate randomly for six generations ., Each generation was maintained at a minimum population size of 900 adults and was not allowed to overlap with the previous generation ., We measured the effect of the resistance locus on gene expression by taking advantage of sex linkage to generate susceptible and resistant mosquitoes that are genetically equivalent across most of their genome ., Resistance has previously been mapped to approximately 4-21cM from the sex-determination locus 17 , 19 and is dominant in action ., The Liverpool IB12 ( LVP-IB12R ) strain of Ae ., aegypti is a highly inbred line that was used for the genome sequencing project 27 and was previously found to be resistant to infection19 ., It is derived from the Liverpool strain which has been maintained in culture since 1936 and was originally collected from West Africa 12 ., A strain of Liverpool susceptible to infection by B . malayi ( LVP-FR3S ) 19 was obtained from the NIAID/NIH Filariasis Research Reagent Resource Center ( FR3 , Atlanta , Georgia , USA ) ., We refer to the strains as LVPR or LVPS from this point on ., To obtain resistant progeny , we crossed LVPR virgin females to LVPS males and backcrossed F1 males to LVPS virgin females ., To obtain susceptible progeny , we crossed LVPS virgin females to LVPR males and backcrossed F1 males to LVPS virgin females ., All mosquitoes were reared at a larval density of 200 individuals in 1 . 8 L of water ., They were fed liver powder as larvae and 10% w/v fructose with 0 . 1% para-aminobenzoic acid ( PABA ) as adults and kept at 28°C ( ± 1°C ) with 75% ( ±5% ) humidity and a 12 hour light:dark cycle ., Females were blood fed using an artificial membrane feeder ( Hemotek Limited , UK ) with donated human blood obtained from Blood Transfusion Services at Addenbrooke’s Hospital , Cambridge , UK ., The temperature of the blood was maintained at 37°C in the feeders ., To infect mosquitoes for association mapping , B . malayi was obtained from Darren Cook and Mark Taylor at the Liverpool School of Tropical Medicine ( LSTM ) , where they were reared in gerbils ., Microfilariae were harvested into RPMI medium , which was then centrifuged at 700 rpm for 5 minutes and 0 . 5 mL of the pellet was transferred to 40 mL of blood ., Microfilariae were incubated in the blood at 37°C for at least one hour prior to feeding ., Outcrossed and control LVPS mosquitoes were fed on blood containing parasites at a concentration of 457 microfilariae per 20 μl of blood ., Female mosquitoes were 6 to 9 days old on the day of infection ., Unfed mosquitoes were discarded , and infected mosquitoes were maintained on a 10% fructose solution with 0 . 1% PABA for 10–11 days post-infection ., To check for infection , individual mosquitoes were separated at the head and thorax at 10 or 11 days after infection and incubated in 100 μl of 1X phosphate buffered saline ( PBS ) for one hour at 37°C ., We found this caused L3 larvae to migrate into the PBS and gave similar estimates of infection as individually dissecting mosquitoes ., The supernatant was transferred to a microscope slide , the number of L3 parasites was counted , and the mosquito carcasses were stored at -80°C until DNA extraction could be performed ., Mosquitoes were classified as susceptible to infection if they had one or more L3 parasites and were classified as resistant if they had none ., For measuring gene expression , resistant and susceptible progeny from the crosses described in the previous section were collected from the following treatments: immediately prior to blood feeding and 12 and 48 hours post-feeding with either a control blood meal or a blood meal containing microfilariae ., Microfilariae were harvested into 50 mL RPMI medium and incubated overnight with 0 . 5 mL gentamicin ( 10 mg/ml in water ) at 28°C , and 0 . 5 mL of the pellet formed overnight was transferred to 16 mL of blood ., The infective blood meal contained 160 microfilariae per 20 ul of blood ., A non-infective control of 50 mL RPMI with 0 . 5 mL gentamicin was also incubated in the same manner , and 0 . 5 mL of solution was transferred to 16 mL of blood ., Both blood vials were then incubated at 37°C for at least one hour prior to feeding ., Female mosquitoes were 4 to 8 days old on the day of blood feeding ., Three to four replicate cages were maintained for each treatment and all time points were collected from the same cages ., After blood feeding , mosquitoes were maintained in paper cups in groups of 8 individuals and were given 10% fructose with 0 . 1% PABA after collection of the 12 hour time point ., We dissected five individual mosquitoes of each genotype at 24 , 48 , and 72 hours after infection to follow the progression of B . malayi development in resistant versus susceptible mosquitoes ., Pools of 8 individuals for each treatment were snap frozen at each time point and stored at -80°C prior to RNA extraction ., DNA was extracted from single mosquitoes using QiaAmp MicroDNA kit ( Qiagen ) with the following modifications ., Tissues were incubated with RNAse post-homogenization and no carrier RNA was used ., DNA was eluted in 50 μl AE buffer and 1 μl of eluate was quantified with a Qubit 2 . 0 fluorimeter ( Invitrogen ) ., Total RNA was extracted using Trizol ( Invitrogen ) and was treated with Turbo DNAse ( Ambion ) prior to library preparation ., RNA integrity was assessed using a Bioanalyzer ( Agilent ) ., We sequenced the exomes of individual mosquitoes ., DNA sequencing libraries were made using TruSeq DNA Sample Preparation kits ( Illumina ) ., Genomic DNA ( 600ng to 1ug of starting material ) was sheared to 500bp fragment sizes via sonication , and libraries were prepared following the instructions from the manufacturer ., Exome capture was then performed to enrich for coding sequences using custom SeqCap EZ Developer probes ( Nimblegen ) ., Overlapping probes covering the protein coding sequence ( not including UTRs ) in the AaegL1 . 3 gene annotations 27 were produced by Nimblegen based on exonic coordinates specified by us ., In total , 26 . 7Mb of the genome ( 2% ) was targeted for enrichment ., Exome capture coordinates are available at https://www . jiggins . gen . cam . ac . uk/data/Aaegypti_exome . bed ., Captures were performed on pools of 24 uniquely barcoded individuals , and the target enriched libraries were sequenced with either 100bp paired-end HiSeq2000 or 150bp paired-end MiSeq ( see S1 Table ) ., Library preparation , exome capture , and sequencing were performed by the High-Throughput Genomics Group at the Wellcome Trust Centre for Human Genetics ( Oxford , UK ) ., In addition to the exome sequencing , we also produced low coverage whole genome sequences from some mosquitoes ( these were largely different individuals but were from the same experiment , see S1 Table ) ., For production of these libraries , DNA was sheared and PCR adapters were added in a single transposase mediated ligation step using the Nextera Library kit ( Illumina ) ., Fifty ng of genomic DNA was used per individual and libraries were prepared following the instructions from the manufacturer ., Libraries were pooled in groups of 21–25 uniquely barcoded individuals and sequenced with 100bp paired-end HiSeq2000 by the Biosciences Core Laboratory at King Abdullah University of Science of Technology ( KAUST ) ( Thuwal , Saudi Arabia ) ., RNA sequencing libraries were made using the TruSeq RNA Sample Preparation kit version 1 ( Illumina ) starting with 3 ug of total RNA per library ., Libraries from different treatments were balanced between lanes ( see S2 Table ) , pooled in groups of 8–10 libraries per lane , and sequenced with four lanes of 100 bp paired-end HiSeq2000 by the Eastern Sequence and Informatics Hub ( EASIH ) ( Addenbrooke’s , Cambridge , UK ) ., Sequences from DNA sequencing libraries were quality trimmed from the 3’ end using Trimmomatic version 0 . 30 28 when average quality scores in sliding windows of 4 base pairs dropped below 20 or when the quality score at the end of the read dropped below 20 ., Sequences less than 50 base pairs in length and unpaired reads were discarded ., Sequences were aligned to the reference genome ( AaegL1 , Oct 2005 ) 27 with BWA version 0 . 6 . 1-r104 29 using the default parameters ., Alignments for individuals sequenced across different lanes were merged into single BAM files using Picard version 1 . 93 ., Alignments were sorted , indexed , and assigned read groups using SAMtools version 0 . 1 . 18 30 and Picard ., Indels were realigned using GATK version 2 . 3 31 , and PCR and optical duplicates were removed using Picard ., We have deposited the raw sequencing reads to the Short Read Archive with Accession Number SRP044393 ., We performed association mapping using a combination of high and low coverage sequences ., Average exome coverage from whole genome sequenced libraries was 0 . 73X per sample while average exome coverage from exome captured libraries was 32X for HiSeq sequencing and 2 . 3X for MiSeq sequencing ., For this reason , we tested for associations with infection status using genotype posterior probabilities , which incorporate uncertainty in genotype calls , rather than calling individual genotypes prior to mapping 32 using the doAsso function in ANGSD version 0 . 539 33 ., BAM files were used as input for ANGSD ., All SNPs called with a LRT statistic greater than 24 ( P<10–6 ) were tested for association with susceptibility to Brugia ., Only bases with a minimum base quality greater than 20 and only reads that were uniquely mapped and with a mapping quality greater than 20 were included ., Major and minor alleles were inferred from genotype likelihoods using the genotype likelihood model implemented in SAMtools 34 , and allele frequencies were estimated assuming a known minor allele using an EM algorithm 35 ., Associations were tested under an additive model with logistic regression , a dominance model or a recessive model ., The dominance and recessive models test for associations with infection status assuming the minor allele is dominant or recessive respectively ., In addition , the additive model was reimplemented setting the most significant marker from the original test as a covariate ( supercont1 . 398 , position 175496 ) to test for the presence of a second locus ., Only individuals with full genotypic information at this SNP with a posterior probability of 0 . 7 were included ( 73 of 140 individuals ) , and the covariate was coded under the dominant model ., At least 15 individuals were required to have each genotypic class for the additive , dominance , and covariate models , and at least 10 individuals were required to have each genotypic class for the recessive model ., To obtain a genome-wide significance threshold for each model that is corrected for multiple tests we permuted the phenotypes and repeated the analysis 200 times , each time retaining the lowest P-value across all variants to generate a null distribution ., This was used to set a genome-wide significance cutoff of P<0 . 01 and P<0 . 05 ., We also tested whether any indels were associated with resistance ., ANGSD can only test SNPs for associations directly from BAM files , so we provided indel genotype probabilities , which are used in an intermediate step in ANGSD , to test for associations ., Genotype probabilities for indels were produced using GATK’s UnifiedGenotyper and ProduceBeagleInput ., Only the additive model was tested using this method , and significance was assessed by permutation as described for SNPs ., The variant effect predictor 36 was used to assign variants to genes and classify their effects ( non-synonymous , synonymous , etc ) using gene annotation set AaegL1 . 3 ., We found that multiple SNPs in linkage disequilibrium were associated with resistance , so we excluded variants that explained the infection data significantly less well than our top hit ( supercont1 . 398 , position 175496 ) ., Using only the HiSeq exome sequenced individuals , we fitted a generalized linear model with a logit link function , where the response was the probability of a mosquito being infected with an L3 worm , and the predictor variables were the ‘top hit’ SNP and the SNP in question ., A SNP was rejected if it was not significant but the ‘top hit’ SNP was ., Sequences from RNA sequencing libraries were quality trimmed using the same method as used for DNA sequencing libraries , except that sequences less than 25 base pairs in length were discarded ., An average of 42 million paired-end reads were obtained from each of the 36 libraries ( S2 Table ) ., Reads were aligned to predicted transcripts in the Ae ., aegypti transcriptome ( gene annotation set AaegL2 . 0 ) with Bowtie2 version 2 . 1 . 0 37 using TopHat2 version 2 . 0 . 9 38 with 10 mismatches allowed , read gap length and read edit distance set to 5 , and no novel junctions allowed ., Reads were mapped to the B . malayi genome ( Ensembl version 3 . 0 . 19 ) 39 using TopHat2 as described above , but gene expression was not analyzed further due to low coverage ., We have deposited the raw sequencing reads to the Short Read Archive with Accession Number SRP044393 ., Differential expression analysis was performed using edgeR 40 after enumerating the number of reads per transcript with HTSeq 41 ., We made the following comparisons:, 1 ) Differential expression in response to infection , performed separately for each genotype and time point;, 2 ) Differential expression between genotypes prior to infection to measure constitutive expression differences;, 3 ) Difference in response to infection between genotypes ( interaction model ) , performed separately for each time point ., In all cases , we filtered out lowly expressed genes by requiring that each gene included in our comparison have at least 0 . 1 count per million ( 0 . 1 cpm ) in enough samples to equal our smallest replicate size for that comparison ( n = 2–4 ) ., All pairwise comparisons were made using exact tests , and the interaction models were fit using general linear models that accounted for genotype and infection status ., Significance was assessed either as having an experiment-wide FDR<0 . 20 ( pairwise comparisons ) or an individual gene significance of P<0 . 01 ( interaction model ) ., The biological coefficient of variation ( BCV ) , a measure of biological variability between replicates , ranged from 0 . 179 to 0 . 455 ( S3 Table ) ., After excluding the four libraries with the lowest library amplifications ( S2 Table ) , the BCV ranged from 0 . 179 to 0 . 353 ., The number of genes meeting our filtering criteria ranged from 12 , 549 to 13 , 594 ( of 17 , 165 ) after excluding poor libraries ( S3 Table ) ., We used the RNA-seq data and Popoolation2 42 to measure differentiation ( FST ) between resistant and susceptible progeny on a per SNP and per gene basis ., BAM alignments from all treatments from the same cross ( yielding either resistant or susceptible progeny ) were merged prior to analysis ., We compared our data on gene expression patterns with previously published microarray data for the Toll and IMD pathways 43 and JAK-STAT pathway 44 ., To determine which pathways were activated by infection , we examined gene expression patterns in response to B . malayi in those genes that were previously shown to be differentially expressed as a result of perturbation of each pathway ., We also compared expression patterns with the response to infection by Wolbachia strain wMelPop-CLA 45 ., Data for this comparison was downloaded from VectorBase 46 , and only genes that were significant at P<0 . 01 were used for comparison ., We classified immunity genes using ImmunoDB ( http://cegg . unige . ch/Insecta/immunodb/ ) and manual curation by ourselves of more recently identified immune genes ., To create a population that varied in susceptibility to B . malayi , we collected Ae ., aegypti eggs from the coastal region of Kenya where there is known to be a mixture of genetically resistant and susceptible individuals 15 ., These eggs were used to create a large outcrossed population that was maintained in the laboratory for 6 generations ., The mosquitoes were then fed on human blood containing B . malayi microfilariae , which resulted in 23% ( 88 of 388 ) becoming infected with L3 larvae , with an average of 2 . 4 L3’s in each infected mosquito ., This is a considerably lower infection rate than in the susceptible control line ( 86% of mosquitoes infected , 19 of 22 , with an average of 2 . 5 L3’s per infected mosquito ) , suggesting that there is genetic variation in susceptibility within our population ., To identify genes associated with resistance , we used a combination of exome sequencing or low coverage whole-genome sequencing ., The Ae ., aegypti genome is large and repetitive , so exome sequencing provided us with far higher coverage of the exonic regions than was possible with the whole genome sequencing ., The exome capture was highly efficient , resulting in 100 times greater coverage of the exome regions ( 26 . 7 MB , 2% of the genome ) compared with the non-exome regions ., So that we could combine the high and low coverage data , we performed our association mapping using an approach based on genotype probabilities at each site ( as opposed to calling genotypes and then testing each site for an association with infection ) ., In total we sequenced 67 L3-infected and 73 uninfected mosquitoes , which we classified as susceptible and resistant respectively ., We found that susceptibility to Brugia has a simple genetic basis in our population , with a small number of sites highly significantly associated with infection ( Fig . 1A ) ., Of the sites that have a known position in the genome , all of those with a genome-wide significance of P<0 . 01 were clustered together at 0 cM on chromosome 1 ( Fig . 1A ) ., To test whether there were multiple genes affecting susceptibility to Brugia , we repeated the association study including the most significant variant from the first analysis as a covariate ., This resulted in no significant associations ( Fig . 1B ) ., Furthermore , quantile-quantile ( qq ) plots comparing expected and observed P-values in the analysis with the top SNP as a covariate confirm that there are no additional associations ( S1 Fig ) ., Therefore we can conclude that there is a single variant causing the differences in susceptibility , and all the significant associations are caused by sites in linkage disequilibrium with this variant ., To test whether resistance is dominant , we repeated the association study using a model that assumed the minor allele to be either dominant or recessive ., The dominant model resulted in a cluster of significant associations on chromosome 1 , and the top associations were more significant than the previous analysis that assumed additive effects ( Fig . 1C ) ., In contrast , the recessive model generated only two marginally significant associations ( Fig . 1D ) , and inspection of these showed that they were caused by linkage disequilibrium with the highly significant dominant variants ., The minor allele of the most significant site was associated with increased resistance , so we can conclude that resistance is caused by a single dominant locus at 0 cM on chromosome 1 ( positions 1p23 and 1p25 on the physical map 47 ) ., This is within the same region that we previously genetically mapped in crosses between laboratory lines from West Africa ( 0 to 12 cM ) 19 , and in the same physical region where a sex-linked , dominant resistance locus was approximately mapped ( resistance: 1p31 , sex determining region: 1q21 ) 18 , 48 , so the associations we detected are likely to be caused by the same resistance region ., Therefore , the previously identified laboratory QTL is present at an appreciable frequency in East Africa in the wild ., We next examined whether the variant causing resistance could be identified in our dataset ., We found 26 SNPs associated with infection below a genome-wide significance threshold of 1% and an additional 27 below 5% ( Dominance Model; Fig . 1C and S1 Dataset ) ., No indels were significantly associated with infection ., Because of our much higher coverage of the exome , all of the significant associations were in or near to genes ., We used two criteria to exclude SNPs from the list of candidate loci ., First , our data only provides evidence for a single causative variant , so we are able to reduce this list from 53 SNPs to 19 SNPs by excluding variants that explain the phenotypic data significantly less well than our top hit ( see methods for details; based on HiSeq exome sequences alone , S1 Dataset ) ., Second , in the experiments described below we find that worms never develop in mosquitoes carrying the resistance allele , and only 7 of the 19 SNPs follow this pattern ( allowing a 10% phenotyping error and using only HiSeq exome data , S1 Dataset ) ., None of these SNPs alter the protein sequence ( 6 synonymous and 1 intronic , S1 Dataset ) ., Furthermore , these SNPs occur in five different genes , none of which have known immune functions ( gamma-tubulin complex component 3 AAEL008465 , cysteine synthase AAEL008467 , putative latent nuclear antigen AAEL002082 , conserved hypothetical protein AAEL008350 , and chaoptin AAEL008940 ) ., Therefore , we cannot identify a single variant that causes resistance ., A combination of two factors has likely prevented us from identifying the gene that is causing mosquitoes to be resistant ., First , linkage disequilibrium in our mapping population means that significant associations are found across multiple scaffolds , sometimes with identical segregation patterns ., This is visible in quantile-quantile plots where there is a large excess of sites with differing frequencies in resistant and susceptible chromosomes ( S1 Fig ) ., Second , low sequencing coverage means we have limited power to detect associations in non-coding sequence ., Therefore , if the variant causing resistance is in a low coverage region we may not detect it , but it will generate significant associations in nearby genes in linkage disequilibrium ., While we did not identify a causal variant , the list of sites and genes associated with resistance are strong candidates ., Even if none of these genes prove to be causing resistance , we have narrowed the region down to the 14 scaffolds that either map to the 0 cM genetic position on chromosome 1 ( scaffolds 1 . 48b , 1 . 97b , 1 . 166 , 1 . 222a , 1 . 267 , 1 . 272 , 1 . 296 , 1 . 327 , 1 . 398 , 1 . 461 , 1 . 487 , 1 . 512 , 1 . 696 , and 1 . 970 , containing 175 genes and 15 . 0 Mb ) or the 6 scaffolds with SNPs in the top 1% of hits whose location in the genome is uncertain ( scaffolds 1 . 226 , 1 . 360 , 1 . 676 , 1 . 1219 , 1 . 257 , 1 . 389 , containing at most 94 genes and 6 . 6 Mb ) ., FST data obtained from our RNA-seq experiment ( described in the following section ) supports four of these six scaffolds as being located on chromosome 1 ( 1 . 360 , 1 . 676 , 1 . 257 , and 1 . 389 ) ., This conflicts with a previous analysis showing that two of these scaffolds ( 1 . 360 and 1 . 257 ) are on chromosome 2 19 , but this is not unexpected as the genome has a high misassembly rate , and many scaffolds are chimeras of sequences found in different places in the genome 19 ., To examine the effects of the B . malayi resistance locus on mosquito gene expression and worm development we used two laboratory mosquito lines , one resistant ( LVPR ) and one su

Introduction, Materials and Methods, Results, Discussion

Many mosquito species are naturally polymorphic for their abilities to transmit parasites , a feature which is of great interest for controlling vector-borne disease ., Aedes aegypti , the primary vector of dengue and yellow fever and a laboratory model for studying lymphatic filariasis , is genetically variable for its capacity to harbor the filarial nematode Brugia malayi ., The genome of Ae ., aegypti is large and repetitive , making genome resequencing difficult and expensive ., We designed exome captures to target protein-coding regions of the genome , and used association mapping in a wild Kenyan population to identify a single , dominant , sex-linked locus underlying resistance ., This falls in a region of the genome where a resistance locus was previously mapped in a line established in 1936 , suggesting that this polymorphism has been maintained in the wild for the at least 80 years ., We then crossed resistant and susceptible mosquitoes to place both alleles of the gene into a common genetic background , and used RNA-seq to measure the effect of this locus on gene expression ., We found evidence for Toll , IMD , and JAK-STAT pathway activity in response to early stages of B . malayi infection when the parasites are beginning to die in the resistant genotype ., We also found that resistant mosquitoes express anti-microbial peptides at the time of parasite-killing , and that this expression is suppressed in susceptible mosquitoes ., Together , we have found that a single resistance locus leads to a higher immune response in resistant mosquitoes , and we identify genes in this region that may be responsible for this trait .

Within mosquito populations , genetic differences between individuals affect their ability to transmit human diseases such as malaria , dengue fever , and lymphatic filariasis ., In the mosquito Aedes aegypti , some individuals are genetically resistant to Brugia malayi , a mosquito-vectored parasite that causes a debilitating tropical disease called lymphatic filariasis ., To characterize the genetic basis of resistance , we identified resistant and susceptible mosquitoes from a wild Kenyan population , and sequenced the protein-coding region of their genomes ( the exome ) ., This allowed us to locate a single region of the mosquito genome that is causing resistance and to identify genes that may be controlling the trait ., To understand the mechanisms of resistance , we measured gene expression ., The susceptible mosquitoes have reduced expression of immunity genes after they are infected with B . malayi , including genes known to kill this group of parasites ., This is possibly because their immune response is being suppressed by the parasites ., We conclude that resistance is controlled by a single locus and show that resistance results in an increased immune response .

journal.pgen.0030121

Plasticity of Fission Yeast CENP-A Chromatin Driven by Relative Levels of Histone H3 and H4

In most eukaryotes , chromosomes contain a centromere that occupies a single locus ., The centromere acts as the site for assembly of the kinetochore that mediates the attachment of chromosomes to spindle microtubules and orchestrates their equational segregation to daughter nuclei at mitosis ., In many organisms , long tandem arrays of repetitive satellite DNA , such as alpha-satellite DNA in humans , are found at each centromere 1 , 2 ., Chromosomal DNA is packaged in chromatin composed of nucleosomes containing the four core histones H2A , H2B , H3 , and H4 ., Histone variants can play specific roles in the regulation of gene expression ., For example , H3 . 3 replaces H3 in regions of active transcription 3–5 ., Intriguingly , kinetochores contain a specific form of chromatin in which canonical histone H3 is replaced by the centromere-specific histone H3 variant known generally as CENP-A 1 , 2 , 6 , 7 ., CENP-A is essential for the assembly of a functional kinetochore and as such must represent a key component in establishing and/or maintaining the site of kinetochore assembly at the centromere 8–12 ., Although CENP-A proteins are found at centromeres in all organisms , there appears to be no specific conserved sequence that ensures the assembly of CENP-A chromatin 1 , 2 , 7 , 13 ., Indeed , the deposition of CENP-A appears to be malleable since inactivated human centromeres lack CENP-A , even though they retain alpha-satellite repeats 14 ., In addition , neocentromeres occasionally arise on chromosomal DNA that lacks any similarity to alpha-satellite repeats , and CENP-A can associate with noncentromeric sequences included in human artificial chromosomes 14–19 ., Similarly , in Drosophila , noncentromeric DNA can acquire the ability to assemble and propagate kinetochore proteins 20–22 ., These and other observations suggest that the site of CENP-A chromatin assembly is epigenetically regulated and propagated 1 , 2 , 13 ., Fission yeast centromeres are reminiscent of those of metazoa in that they contain repetitive elements ( outer repeats , otr ) that flank the central domain ( see Figure 1 ) ., The central domain is composed of inner repeats ( imr ) surrounding the central core ( cnt ) 23–25 ., Noncoding transcripts arising from the outer repeat provide a substrate for RNA interference ( RNAi ) that directs methylation of histone H3 on lysine 9 and the assembly of silent chromatin ( reviewed in 26 ) ., Within the central domain most histone H3 is replaced by the centromere-specific H3 variant CENP-ACnp1 to form the unusual chromatin that occupies most of the 10–12 kb comprising imr and cnt 11 , 23 , 27–29 ., Consistent with the notion that CENP-ACnp1 chromatin is a signature of kinetochore activity , kinetochore-specific proteins are confined to the central domain 11 , 29–35 ., At each centromere a cluster of tRNA genes demarcates the two distinct chromatin domains: outer repeat silent heterochromatin and kinetochore chromatin 23 , 36 ., Deletion of one of the two tRNA genes from one side of a centromere allows heterochromatin to infiltrate the central domain 37 ., The fact that all three fission yeast centromeres have a common organization of DNA elements suggests that the assembly of heterochromatin and CENP-ACnp1 chromatin domains could be strictly governed by sequence ., Marker genes placed within outer repeat chromatin are silenced , and this requires RNAi components , Clr4 histone H3K9 methyltransferase and Swi6 ( homologue of HP1 ) 36 , 38 ., Genes are also silenced in the central domain , but this silencing is strongly dependent on kinetochore integrity rather than RNAi-mediated heterochromatin 31 , 35 , 36 , 39 ., Mutations in several kinetochore proteins , including CENP-ACnp1 , alleviate silencing , specifically in the central domain , and affect the association of CENP-ACnp1 with the central domain and/or the unusual chromatin found within the central domain 11 , 29 , 35 , 36 ., Dissection of fission yeast centromeric DNA showed that outer repeat and central domain sequences are required for the assembly of a kinetochore that supports mitotic segregation 23 , 24 , 40 ., In Drosophila and humans , centromeres can arise at sites apparently lacking particular features at the primary DNA sequence level ., It is not known whether the similarity between fission yeast and metazoan centromeres extends to the ability of CENP-A chromatin to assemble on noncentromeric sequences or if CENP-A assembly in this organism more closely resembles the situation in budding yeast in which CENP-A associates at a specific sequence ., Here , we investigate the ability of fission yeast CENP-ACnp1 chromatin to assemble on noncentromeric sequences ., We determine whether the amount of CENP-ACnp1 incorporation directly correlates with silencing , kinetochore assembly , and kinetochore function ., Altering the relative ratios between H3 , H4 , and CENP-ACnp1 influences the assembly of CENP-ACnp1 chromatin , the recruitment of other kinetochore proteins , and the fidelity of chromosome segregation ., Surprisingly , there is no impediment to depositing H3 in the central domain ., Thus , our observations indicate that the relative levels of histones are crucial for the correct formation of CENP-ACnp1 chromatin ., The central domain of centromere 1 ( cen1 ) is composed of the cnt1 element , a portion of which is shared with cen3 , and the cen1-specific imr1 repeats that are virtually identical in sequence on the left and right sides 23 ., The outer repeats flanking the central domain are also found at cen2 and cen3 , totaling 17–18 copies at centromeres 23 , 25 ., The distribution of endogenous CENP-ACnp1 across cen1 was assessed using anti-CENP-ACnp1 antiserum for chromatin immunoprecipitation ( ChIP ) , confirming that endogenous CENP-ACnp1 associates with the central domain ( cnt1 and imr1 ) , but not with flanking outer repeats ( Figure S1 ) ., To determine whether CENP-ACnp1 chromatin forms on noncentromeric sequences in fission yeast , its association with a ura4+ gene inserted at different sites in cen1 was assessed using ChIP ( Figure 1A ) ., The use of ura4+ insertions also provides specificity in ChIP for cen1 , especially for duplicated cnt1 and imr1 regions and the more repetitive outer repeats ., The ura4-DS/E allele ( 279-bp deletion ) at its euchromatic locus provides an internal control for quantification ., CENP-ACnp1 shows no association with ura4+ in the outer repeats of cen1 ( sites 1 and 2 ) ., No association of CENP-ACnp1 is seen at a euchromatic control site R . int-cnt1:ura4+ ( R: between open reading frames SPBC342 . 01 and . 02 ) , even though in this case the ura4+ gene is also flanked by 1 . 7 and 1 . 6 kb of DNA from the central domain of cen1 ( Figure S9 ) ., However , just outside the tRNAala and tRNAglu genes that demarcate the edge of the central domain , a 4-fold enrichment of CENP-ACnp1 is detected ( site 3 ) ., A similar level of CENP-ACnp1 is detected inside the first tRNAala gene ( site 4 ) , but much higher levels ( 20- to 50-fold enrichment ) of CENP-ACnp1 are associated with ura4+ in the core of the central domain of cen1 ( site 5 and 6 ) ., This confirms and extends previous analyses utilizing C-terminally 3 × HA tagged CENP-A 11 and is similar to the pattern of association seen for Mis6-HA across cen1 36 ., These observations suggest that fission yeast CENP-ACnp1 can associate with noncentromeric DNA inserted within the central domain ., However , it is possible that the strong enrichment of ura4+ in ChIPs is simply due to the association of CENP-ACnp1 with adjacent centromeric DNA sequences in the IPs ., To test this rigorously we used two strains in which 1 . 7 kb or 4 . 7 kb of DNA bearing the ura4+ gene was inserted at site 6 in the middle of the central domain of cen1 , cnt1:ura4+ , and cnt1:bigura4+ , respectively ( Figure 1B ) ., The sonicated chromatin used for anti-CENP-ACnp1 ChIP was less than 800 bp ( Figure 1C ) ., The region within ura4+ monitored by PCR is 0 . 9 kb and 0 . 5 kb ( cnt1:ura4+ ) or 2 . 2 kb and 2 . 2 kb ( cnt1:bigura4+ ) away from centromeric sequences , and thus , any enrichment of ura4+ indicates association of CENP-ACnp1 with ura4+ DNA ., Using this stringent assay , we observe that CENP-ACnp1 assembles on ura4+ DNA in cnt1:ura4+ ( 19–35× enrichment ) and even associates with the middle of cnt1:bigura4+ which is at least 2 kb from any endogenous centromeric sequences ( 6–12× enrichment ) ., Thus , we conclude that CENP-ACnp1 can assemble on noncentromeric sequences ., The R . int-cnt1:ura4+ control ( R ) consists of the ura4+ gene flanked by 1 . 7 and 1 . 6 kb of central domain DNA 39 , yet no association of CENP-ACnp1 is seen at this location ., Thus , central domain sequences alone are not sufficient to induce CENP-ACnp1 assembly when placed on a chromosome arm ., This is consistent with previous analyses showing that plasmids containing just central domain DNA do not exhibit centromere function or features associated with functional centromeres 23 , 24 , 27 ., In addition to ura4+ , we have also detected CENP-ACnp1 on ade6+ when inserted in the central domain of centromeres ( Figure S1 ) ., Thus , the deposition of CENP-ACnp1 in Schizosaccharomyces pombe does not require specific underlying DNA sequences and can probably assemble on any noncentromeric DNA sequence provided that it is placed in the context of a functional centromere ., The ChIP analyses above indicate that there is more CENP-ACnp1 on cnt1:ura4+ than on cnt1:bigura4+ ., Both ura4+ insertions at site 6 are silenced , as indicated by growth on counter-selective 5-fluoro-orotic-acid ( FOA ) plates ., However , the cnt1:ura4+ insertion exhibits more growth on FOA plates indicating that it is more strongly silenced than the larger cnt1:bigura4+ insertion ( Figure 1D ) ., In support of this , more ura4+ mRNA is detected in cnt1:bigura4+ cells compared to cnt1:ura4+ cells by reverse transcriptase ( RT ) -PCR ( Figure 1E ) ., Consistent with the idea that silencing is due to spreading of CENP-ACnp1 chromatin onto ura4+ gene insertions , cnp1 mutants alleviated silencing of ura4+ within the central domain ( less growth on FOA; Figure 2A ) and less CENP-ACnp1 was detectable by ChIP on the central domain ( Figure 2B ) ; this is supported by previous analyses 11 , 35 , 41 ., The most severe cnp1 alleles showed the greatest reductions in CENP-ACnp1 levels: cnp1–1 ( L87Q ) 11 > cnp1–76 ( T74M ) > cnp1–87 ( E92K ) > cnp1–169 ( V52A ) ( previously referred to as sim2 mutants; 35 ) ., This suggests that the production of defective CENP-ACnp1 results in less CENP-ACnp1 chromatin in the central domain , and that this chromatin state is more compatible with ura4+ gene expression ., ChIP with antibodies recognizing the C-terminus of histone H3 indicates that this is accompanied by increased incorporation of histone H3 into the central domain chromatin in all cnp1 mutants ( Figure 2B ) , and this concurs with analyses of other mutants affecting CENP-ACnp1 deposition 41 , 42 ., Silencing also correlates with chromatin composition in the centromere insertions: more H3 is associated with cnt1:bigura4+ than with the more silent cnt1:ura4+ ( Figure 2C ) ., Interestingly , in Drosophila , H3 can take the place of CENP-ACID when CENP-ACID levels are reduced 43 ., In addition , cnp1 mutants are sensitive to excess H3 , displaying a correlation between allele severity and sensitivity to different H3 levels ( Figures 2D and S2 ) ., Moreover , overexpression of histone H4 suppressed the temperature sensitivity of cnp1 mutants , with higher levels of H4 being required to suppress more severe alleles ( Figures 2E and S2 ) , as reported for cnp1–1 42 ., These genetic interactions suggest that additional H4 may assist in incorporation of mutant CENP-ACnp1 into the central core by facilitating the formation of more H4/CENP-ACnp1 heteromers ., On the other hand , excess H3 would exacerbate the phenotype by competing with CENP-ACnp1 for H4 from the available pool , further favoring H3/H4 deposition into central domain chromatin ., Together , these data imply that assembly of CENP-ACnp1 chromatin on DNA inserted in the central domain is incompatible with gene expression and that when CENP-ACnp1 is defective , histone H3 takes its place , allowing greater expression ., Since defective CENP-ACnp1 alleviates silencing , we determined whether , conversely , overexpression of CENP-ACnp1 causes an increase in central domain silencing ., Unlike budding yeast , fission yeast are able to tolerate overexpression of CENP-ACnp1 , to an extent that the normally undetectable CENP-ACnp1 is detectable by western analysis 44 ( Figure S3 ) ., Overexpression of CENP-ACnp1 ( from the attenuated nmt1 promoter in prep81x ) in a wild-type strain bearing cnt1:ura4+ or cnt1:bigura4+ increased the level of silencing , as indicated by increased growth on FOA ( Figure 3A ) , and more CENP-ACnp1 was detected on cnt1:ura4+ and cnt1:bigura4+ ( Figure 3B ) ., This suggests that the ura4+ sequence inserted in the central domain is not saturated for CENP-ACnp1 ., There was also a detectable increase in the levels of CENP-ACnp1 on endogenous central domain sequences ( Figure 3B ) ., Most CENP-ACnp1 is normally confined to the central domain , which is flanked on both sides by tRNAala and tRNAglu genes , separated by 349 bp 37 ., When CENP-ACnp1 was overexpressed , increased silencing ( more growth on FOA ) occurred at site 4 ( ura4+ between the tRNAala and tRNAglu genes; Figure 3C ) , and a subtle increase in CENP-ACnp1 levels was reproducibly detected on ura4+ at this site ., However , in strains overexpressing CENP-ACnp1 , additional CENP-ACnp1 could not be detected in the heterochromatin domain beyond the tRNAala and tRNAglu genes ( unpublished data ) ., Moreover , there was no apparent effect on the extent of the CENP-ACnp1 domain in a strain lacking heterochromatin on the outer repeats ( clr4Δ; unpublished data ) ., These data suggest that CENP-ACnp1 chromatin is confined to the central domain but that overexpression of CENP-ACnp1 allows some expansion in the extent of the CENP-ACnp1 domain beyond the proximal tRNAala gene into an inserted ura4+ gene ., As the degree of central domain silencing correlates with increased H3 levels in both cnp1 mutants and cnt1:bigura4+ versus cnt1:ura4+ , we investigated the effects of simply overexpressing H3 on central domain silencing and composition ., Additional histone H3 was expressed from the nmt1 promoter ( prep3X-H3 ) in strains harboring cnt1:ura4+ or cnt1:bigura4+ ( Figure 4A ) ., Quantitative western analyses indicate that 2- to 3-fold more H3 is present and that the levels of CENP-ACnp1 are unaffected ( Figures S4 and S5 ) ., H3 overexpression resulted in increased ura4+ expression from cnt1:ura4+ as shown by loss of growth on FOA ., This was accompanied by increased H3 on cnt1:ura4+ and cnt1:bigura4+ ( Figure 4B ) and a concomitant decrease in CENP-ACnp1 levels ( Figure 4C ) ., The nmt1 promoter used in the above experiments is active throughout the cell cycle whereas histone gene expression is normally induced from endogenous promoters in early S phase 11 , 45 ., To rule out the possibility that the observed effects were due to constitutive expression of H3 from the nmt1 promoter , we altered the ratio of H3:H4 genes while retaining the normal endogenous promoter elements ., In fission yeast , three pairs of H3 and H4 genes are transcribed divergently from a putative common regulator at three distinct loci ( H3 . 1/H4 . 1 , H3 . 2/H4 . 2 , and H3 . 3/H4 . 3 ) 46 ., We completely deleted the H3 . 1/H4 . 1 pair and either the H3 . 2 gene , resulting in a H3:H4 gene ratio of 1:2 ( H4 > H3 ) , or the H4 . 2 gene , resulting in a H3:H4 gene ratio of 2:1 ( H3 > H4 ) 47 ( Figure S6 ) ., Increased H3 relative to H4 ( H3 > H4 ) alleviated silencing of both cnt1:ura4+ and cnt1:bigura4+ , resulting in failure to grow on counter-selective FOA plates ( Figure 5A ) , indicating a complete inability to silence ura4+ expression ., Conversely , excess H4 ( H4 > H3 ) enhanced silencing of cnt1:ura4+ as indicated by reduced growth on plates lacking uracil ., These phenotypic effects were confirmed by RT-PCR analysis ( Figure 5B ) ; more ura4 mRNA is produced from cnt1:ura4+ in the H3 > H4 strain ( although we were unable to detect a further increase in the comparatively high levels of ura4 mRNA in cells containing cnt1:bigura4+ ) ., These analyses also indicate that central domain silencing is fully intact in the control strain ( H3 = H4; H3:H4 ratio 2:2 ) used in these experiments ( Figure 5A and 5B; compare to wild type; H3:H4 ratio 3:3 ) ., The strain also displays robust ChIP of CENP-ACnp1 on endogenous centromeric sequences and ura4+ sequences inserted therein ( e . g . , Figure 5D ) ., Thus ( as with cnp1 mutants ) , disturbing the H3:H4 balance so that more H3 is expressed alleviates silencing in the central domain , while additional H4 enhances silencing ., To determine the impact of additional H3 on centromere integrity , we assessed CENP-ACnp1 association with centromeres by ChIP and immunolocalization ., Cells with a histone imbalance in favor of H3 ( H3 > H4 ) were confirmed to have increased H3 levels on cnt1:ura4+ and cnt1:bigura4+ by ChIP ( Figure 5C ) ., CENP-ACnp1 levels were greatly reduced on cnt1:ura4+ and cnt1:bigura4+ in cells with excess H3 ( Figure 5D ) , yet the cellular levels of CENP-ACnp1 are unaffected ( Figure S5 ) ., Lower levels of CENP-ACnp1 are normally detected on the ura4+ region of cnt1:bigura4+ relative to cnt1:ura4+ , and excess histone H3 caused a further decrease in the level of CENP-ACnp1 associated with the middle of the cnt1:ura4+ and a consistent but less dramatic effect on cnt1:bigura4+ ( Figure 5D ) ., We also observed a reduction of CENP-ACnp1 levels on endogenous centromeric sequences in strains with excess H3 ( Figure 5D ) ., In strains with an excess of H4 relative to H3 ( H4 > H3 ) , more CENP-ACnp1 was detected on cnt1:bigura4+ , although the effects on endogenous centromeric sequences and cnt1:ura4+ were more variable ( Figure 5D and unpublished data ) ., Chromatin immunoprecipitation of a protein reports the population average for its association with particular DNA sequences ., To assess CENP-ACnp1 levels in individual cells , immunolocalization was performed ., In fission yeast , the three centromeres cluster adjacent to the spindle pole body ( SPB ) in interphase and this was used as a marker for centromere position ., In both H3 = H4 and H4 > H3 strains a clear CENP-ACnp1 signal ( red ) was detected next to the SPB ( green ) in all cells ( Figure 6A ) ., However , in H3 > H4 cells , the CENP-ACnp1 signal was weaker or undetectable and displayed variation between cells ., The majority of cells had much weaker staining than H3 = H4 ( 54% versus 7% ) , and very few displayed bright/very bright staining ( 9% versus 93% in H3 = H4 ) ., In addition , the CENP-ACnp1 signal in H4 > H3 cells was quantified and found to be of greater intensity than in H3 = H4 cells ( 2 . 5-fold brighter on average in the H4 > H3 cells compared to the H3 = H4 cells; details of quantification methods described in Figure S7 and Materials and Methods ) ., Thus , the immunolocalization data confirm the ChIP analyses and indicate that the level of CENP-ACnp1 at centromeres in H3 > H4 cells exhibits variation within the population ., These observations are consistent with a scenario in which excess CENP-ACnp1 allows more CENP-ACnp1/H4 chromatin assembly in the central domain , while elevated H3 levels permit more H3/H4 nucleosomes to occupy the central domain at the expense of CENP-ACnp1 ., It also appears that an increase in the available H4 pool allows CENP-ACnp1 to compete more effectively with H3 for incorporation in the centromere ., Altering the ratios of H3:H4:CENP-ACnp1 have clear effects on central domain chromatin , but what are the consequences for chromosome segregation ?, It is possible that a functional kinetochore can be assembled despite alterations in the proportion of CENP-ACnp1/H3 in the central domain ., To determine whether chromosome segregation is aberrant in strains with altered histone ratios , fixed cells were stained for chromosomal DNA ( DAPI: red ) and anti-α-tubulin to identify cells with a mitotic spindle ( green ) ( Figure 6B ) ., The H3 > H4 strain exhibited significantly higher rates of chromosome segregation defects in anaphase ( lagging chromosomes and uneven segregation ) compared to the H4 > H3 and control ( H3 = H4 ) strains ( 16 . 6% , 0% , and 1 . 4% , respectively , see Figure S8 ) ., Together these data suggest that excess H3 interferes with the localization of CENP-ACnp1 at centromeres resulting in defective kinetochore function during mitosis ., It could be argued that disturbing histone ratios in the cell is likely to have pleiotropic effects that lead to chromosome missegregation ., For instance , mutants that affect silencing and function of centromeric outer repeat chromatin display high frequencies of lagging chromosomes ., However , alleviation of outer repeat silencing was not observed when H3 was overexpressed or in H3 > H4 strains ( unpublished data ) ., In addition , prep81x-Cnp1 rescued defective growth of the H3 > H4 strain , allowing growth at 32 °C ( Figure 5E ) , suggesting that the primary defect in H3 > H4 cells is in central domain function ., As expected , the defective growth of the H3 > H4 strain was also rescued by H4 overexpression ( Figure 5E ) ., Mutations in cnp1 or genes encoding kinetochore proteins alleviate silencing in the central domain 31 , 35 , 36 ., Previous analyses suggest that the kinetochore proteins Mis6 and Sim4 can associate with cnt1:ura4+ 35 , 36 ., To examine this further , ChIP was used to determine if three kinetochore proteins , CENP-CCnp3-GFP , Mal2-GFP , and Sim4-GFP , are also associated with the middle of cnt1:ura4+ and cnt1:bigura4+ ., Under conditions where chromatin was extensively sheared , all three kinetochore proteins were found to associate with the middle of the ura4+ gene when it was inserted within cnt1 at site 6 ( both cnt1:ura4+ and to a lesser degree for cnt1:bigura4+ ) , but not when it was flanked by 1 . 7 and 1 . 6 kb of central domain DNA at a euchromatic site ( R: R . int-cnt1:ura4+ ) ( Figure 7A ) ., Thus , not only CENP-ACnp1 , but three other kinetochore proteins tested associate with noncentromeric DNA when inserted in the context of a functional fission yeast centromere ., Elevated levels of CENP-ACnp1 or histone H3 have opposite effects , respectively leading to more and less deposition of CENP-ACnp1 on cnt1:ura4+ and cnt1:bigura4+ ., Since CENP-ACnp1 is required for kinetochore assembly , we tested whether the levels of Mal2-GFP , Sim4-GFP , and CENP-CCnp3-GFP on cnt1:ura4+ and cnt1:bigura4+ are also affected in strains expressing additional CENP-ACnp1 or histone H3 ( Figure 7B ) ., Cells expressing additional CENP-ACnp1 reproducibly showed more CENP-CCnp3-GFP , Mal2-GFP , and Sim4-GFP associated with ura4+ in the central core , especially for cnt1:bigura4+ ., Conversely , in cells expressing more histone H3 , less of these three kinetochore proteins were detected on cnt1:ura4+ ., This indicates that adjusting the levels of CENP-ACnp1 and H3 not only alters the extent and density of CENP-ACnp1 chromatin at centromeres but also influences the recruitment of other kinetochore proteins , resulting in defective chromosome segregation ., We have found that CENP-ACnp1 in fission yeast can associate with noncentromeric sequences provided that they are placed in an environment with the required contextual cues for CENP-ACnp1 chromatin assembly ., Our analyses show that excess CENP-ACnp1 ( or H4 ) allows the deposition of more CENP-ACnp1 at the expense of H3 while an excess of H3 allows the deposition of more H3 in place of CENP-ACnp1 and leads to defective chromosome segregation ., These analyses strongly support the concept that deposition of CENP-ACnp1 in the central kinetochore domain exhibits surprising plasticity in that it can be disturbed or enforced simply by changes in the ratios of H3 , H4 , and CENP-ACnp1 ., These observations indicate that maintenance of the unique chromatin composition of the central domain is vital in ensuring proper kinetochore assembly and function ., Thus , both the functional state of CENP-ACnp1 and its density are important for centromere function ., In metazoa , the assembly of CENP-A chromatin and kinetochores is plastic 1 , 2 , 13 ., In humans , CENP-A and kinetochores are normally associated with a subset of the centromeric alpha-satellite repeats ., However , CENP-A can assemble at locations on chromosome arms lacking alpha-satellite DNA , resulting in the formation of neocentromeres 14 ., In Drosophila , experiments with truncated minichromosomes derived from the X chromosome demonstrate that CENP-ACID and kinetochore proteins can spread into chromosomal regions where they do not normally associate and then act as a functional neocentromere when the X centromere is subsequently deleted 20 ., In addition , overexpression of CENP-ACID in Drosophila cells can attract other kinetochore proteins and even direct microtubule association at noncentromeric locations 22 ., The similarity in the organization of centromeric DNA at all fission yeast centromeres suggested that , like Saccharomyces cerevisiae , the assembly of CENP-ACnp1 chromatin might be more DNA sequence–dependent and thus less pliable than in metazoa 23 , 25 ., The identification of Ams2 , a GATA-like DNA-binding factor that affects CENP-ACnp1 deposition , supported this possibility 42 ., However , here we have shown that , as observed in Drosophila and human cells , fission yeast CENP-ACnp1 can spread into and associate with additional sequences inserted within the central domain ., Thus , provided the right contextual cues exist , fission yeast CENP-ACnp1 chromatin can potentially associate with any DNA sequence ., The establishment and continued loading of CENP-A in the central domain may be dependent on the central domain sequences themselves , but once it has been initiated , our analyses indicate that it can engulf inserted noncentromeric DNA ., In addition , the association of CENP-ACnp1 with exogenous sequences in the central domain is accompanied by the recruitment of other known kinetochore proteins ., It is also possible that the association of CENP-A with ura4 DNA in the central domain might reflect some higher order structure within the centromere ., In S . cerevisiae , additional CENP-ACse4 is degraded and only nondegradable mutant protein can be overexpressed 44 ., In fission yeast there appears to be no inherent difficulty in overexpressing CENP-ACnp1 since increased CENP-ACnp1 is detected by western analysis upon overexpression ( Figure S3 ) ; and this is the cause of increased CENP-ACnp1 incorporation into the central domain ., Conversely , elevated levels of H3 lead to more H3 and less CENP-ACnp1 in central domain chromatin ., In addition , mutations in CENP-ACnp1 are antagonized by H3 but rescued by H4 overexpression , and defects exhibited by H3 > H4 cells are suppressed by additional CENP-ACnp1 ., This indicates that perturbation in the relative levels of CENP-ACnp1 , H3 , and H4 can adversely affect the composition of CENP-ACnp1 chromatin assembled at centromeres , affecting the relative density of CENP-ACnp1/H4 to H3/H4 nucleosomes ., Disturbing the normal CENP-A:H3:H4 ratio clearly results in defective kinetochore function and chromosome segregation ., Consistent with this , decreased CENP-ACnp1 at centromeres in cells lacking Ams2 is suppressed or antagonized by overexpression of H4 and H3 , respectively 41 , 42 ., In fission yeast , it is likely that H3 and H4 mRNAs are coordinately expressed early in S phase from a common regulatory element residing between their divergent promoters , whereas CENP-ACnp1 mRNA is expressed from late M , peaking prior to S phase 11 , 41 , 45 , 46 ., This suggests that normally CENP-ACnp1 is available before maximal H3 and H4 expression ., As fission yeast centromeres are replicated early in S phase 48 , 49 , this difference in expression timing may enhance the effective concentration of CENP-ACnp1 relative to H3 at the start of S phase , and thus allow it to compete more effectively with the initially low levels of newly produced H3 for assembly with new H4 ( Figure 8 ) ., Our data suggest that altering the ratio of H3:H4 either by altering the normal gene number from equivalency ( 3:3 or 2:2 ) to 2:1 , or overexpressing H3 , provides more H3 to titrate out the pool of H4 available for assembly with CENP-ACnp1 into chromatin ., Conversely , excess H4 relative to H3 ( H3:H4 1:2 ) increases the pool of H4 available for assembly into chromatin with CENP-ACnp1 , leading to more CENP-ACnp1 at centromeres ., These data suggest that the relative amounts of H3 , H4 , and CENP-A are normally delicately balanced in the cell to allow normal CENP-ACnp1 chromatin and kinetochore assembly ., Changes in this balance , the timing , or amount of expression can lead to increased or decreased CENP-ACnp1 chromatin at centromeres ( Figure 8 ) ., Precise regulation of histone levels , e . g . , coregulation of H3 and H4 , is likely to be important in other organisms ( including human ) where CENP-A overexpression has been detected in the tissue from colorectal tumors 50 ., In S . cerevisiae , elevated levels of histone H3 and H4 cause defective chromosome segregation 51; it is possible that this is due to defects in deposition of CENP-ACse4 at centromeres ., The extent of sequence occupied by CENP-ACnp1 chromatin appears to be flexible and increases in response to higher levels of CENP-ACnp1 ., The tRNA genes flanking the central domain act as a barrier preventing heterochromatin seeping into the kinetochore domain 37 ., Here , we detect more CENP-ACnp1 on ura4+ between the tRNAala and tRNAglu when CENP-ACnp1 is overexpressed but not distal to tRNAala in the heterochromatin domain ( Figure 3C ) ; thus , CENP-ACnp1 chromatin is still mainly confined to the central domain ., The mechanism of CENP-A assembly into centromeric chromatin is not known ., In mammalian cells , CENP-A is synthesized in G2 and thereby separated from bulk histone synthesis 52 , 53 ., Such analyses suggested that CENP-A chromatin assembly is uncoupled from replication ., One model suggests that the CENP-A nucleosomes are randomly segregated at S phase and are subsequently recognized by the CENP-A chromatin assembly machinery resulting in the deposition of neighboring new CENP-A during interphase 2 , 52 , 53 ., In fission yeast , in addition to a replication-independent pathway , there appears to be a replication-coupled pathway that allows CENP-ACnp1 chromatin assembly 11 , 41 ., Deposition during S phase is coupled with the expression of new histones ( dependent on Ams2 ) while G2 assembly requires Mis6 11 , 41 , 42 ., Higher levels of H3 interfere with CENP-ACnp1 incorporation at centromeres , suggesting that excess H3 can overwhelm the normal assembly pathways to predominate in the central domain ., Surprisingly , there is no inherent impediment to deposition of H3 in the central domain ., If a mechanism exists to prevent the over-incorporation of H3 in the central domain , it is easily overwhelmed ., It is possible that H3 is incorporated during S phase but can be subsequently replaced by CENP-ACnp1 via a replication-independent mechanism operating throughout the cell cycle ., We envisage a scenario in which excess H3 results in overloading of the central domain with H3 nucleosomes , which in turn interferes with the recognition of central domain chromatin by activities that evict H3 and replace it with CENP-ACnp1 ., Thus , the balance between H3 , H4 , and CENP-ACnp1 levels is critical to the incorporation of new CENP-ACnp1 nucleosome and kinetochore function ., S . pombe strains are listed in Table S1 ., Standard procedures were used for growth and genetic manipulations 54 ., To construct FY4638 ( cnt1:bigura4+ ) , a 4 . 7-k

Introduction, Results, Discussion, Materials and Methods

The histone H3 variant CENP-A assembles into chromatin exclusively at centromeres ., The process of CENP-A chromatin assembly is epigenetically regulated ., Fission yeast centromeres are composed of a central kinetochore domain on which CENP-A chromatin is assembled , and this is flanked by heterochromatin ., Marker genes are silenced when placed within kinetochore or heterochromatin domains ., It is not known if fission yeast CENP-ACnp1 chromatin is confined to specific sequences or whether histone H3 is actively excluded ., Here , we show that fission yeast CENP-ACnp1 can assemble on noncentromeric DNA when it is inserted within the central kinetochore domain , suggesting that in fission yeast CENP-ACnp1 chromatin assembly is driven by the context of a sequence rather than the underlying DNA sequence itself ., Silencing in the central domain is correlated with the amount of CENP-ACnp1 associated with the marker gene and is also affected by the relative level of histone H3 ., Our analyses indicate that kinetochore integrity is dependent on maintaining the normal ratio of H3 and H4 ., Excess H3 competes with CENP-ACnp1 for assembly into central domain chromatin , resulting in less CENP-ACnp1 and other kinetochore proteins at centromeres causing defective kinetochore function , which is manifest as aberrant mitotic chromosome segregation ., Alterations in the levels of H3 relative to H4 and CENP-ACnp1 influence the extent of DNA at centromeres that is packaged in CENP-ACnp1 chromatin and the composition of this chromatin ., Thus , CENP-ACnp1 chromatin assembly in fission yeast exhibits plasticity with respect to the underlying sequences and is sensitive to the levels of CENP-ACnp1 and other core histones .

The DNA of all genomes is organized into chromosomes that are packaged in chromatin in which DNA is wrapped around nucleosomes composed of the histones H2A , H2B , H3 , and H4 ., Centromeres are the specialized regions on chromosomes that attach them to spindle microtubules , and this process is required to allow each daughter cell to receive one copy of each chromosome after they have duplicated ., Centromere regions are distinguished from other parts of the chromosome by the incorporation of the histone H3 variant CENP-A instead of histone H3 into specialized nucleosomes ., This CENP-A chromatin allows the machinery ( the kinetochore ) responsible for attaching chromosomes to microtubules to assemble ., Fission yeast centromeres contain a central domain where CENP-A is prevalent ., This study shows that CENP-A can associate with “foreign” DNA placed in the central domain ., Therefore , CENP-A appears to associate with any DNA placed in this environment , independent of its DNA sequence ., Increasing the relative level of H3 allows H3 to be assembled in place of CENP-A in this critical central domain and results in defective centromere/kinetochore function and chromosome segregation ., This study highlights the plasticity of centromeric chromatin .

schizosaccharomymes, cell biology, yeast and fungi, eukaryotes, molecular biology, genetics and genomics

journal.pcbi.1006207

Network supporting contextual fear learning after dorsal hippocampal damage has increased dependence on retrosplenial cortex

Brain lesions provide evidence primarily about the extent to which a brain function can persevere in the absence of the damaged region ., A preserved cognitive function or behavior after lesion is generally interpreted as alternative ‘brain routes’ being still able to meet the cognitive demands 1 , 2 ., In contextual fear conditioning ( CFC ) , hippocampal lesions reveal a complex relation to contextual memory , as they result in profound retrograde amnesia ( of pre-lesion events ) , but often result in no anterograde amnesia post-lesion events; 3 , 4 , 5 under defined circumstances 6 , suggesting that new learning can be supported by the reminiscent regions ., Evidence for hippocampal participation in CFC acquisition have been provided by manipulations ranging from pharmacological injections such as muscarinic 7 and NMDA receptors blockade 8 , to optogenetic approaches 9 ., Thus , although hippocampus participates in CFC learning if it is functional during acquisition , CFC learning can occur after hippocampal loss ., The CFC learning after hippocampal lesion inspired cognitively-oriented hypotheses about the content learned by non-hippocampal regions ., Although these accounts differ on whether the contextual representation without hippocampus is fragmented elemental; 10 , 11 or is—still—a configural representation reviewed in 12 , it is well accepted that learning under hippocampal loss is likely to be, 1 ) different in terms of content learned ,, 2 ) less efficient and, 3 ) more prone to generalization and decay over time 12–14 ., It is also accepted that the hippocampus has preference over the non-hippocampal regions ., This accounts for the impaired CFC observed after temporary manipulations , during which hippocampus inhibits the non-hippocampal regions while unable form long-term memory 12 ., However , little attention has been given to the remaining regions supporting CFC learning ., Although parahippocampal cortices were pointed out as putative candidates 11 , the regions supporting CFC learning after hippocampal lesion have not been empirically addressed ., Investigating how these regions learn and store CFC information can help to understand the dynamics of hippocampal function and its interactions within the memory systems ., There is evidence for a large number of regions to compose the neural circuits involved in CFC 15 , 16 and spatial/contextual memory 17 ., Understanding how a function/behavior can still be supported after lesion requires assessing complex interactions among the remaining regions and the changes in their engagement ., Network approaches assess complex brain interactions based on the representation of elements ( i . e . brain regions , neurons ) and connection concepts ( i . e . projections , functional connectivity ) , and offer quantitative tools for a data-driven assessment of network characteristics related to brain structure and function 18 ., Large-scale network studies based on structural and functional MRI data have been paving a solid ground in cognitive neuroscience 19 , 20 ., They have explored functional network topology in the brain 21 and its importance to learning 22 and emotion 23 ., Network studies have also been useful in identifying crucial brain regions ( hubs ) for network function 24 , and to identify functional network changes after traumatic brain injuries 25 and in psychiatric disorders 26 , 27 ., Some studies took advantage of rodent models and employed network analysis in the expression of the activity-dependent gene c-fos after remote CFC retrieval 28 and later empirically interrogated the network hubs given by the model 29 ., Here , we used a similar rationale to investigate how the brain support CFC learning after hippocampal lesion ., We used the phosphorylated cAMP response element binding ( pCREB , active form of CREB ) , which is critical to learning-induced synaptic plasticity 30 , as our marker of brain region engagement; and examined activation and co-activation of brain regions of hippocampectomized rats after a CFC session ., Using network analysis , we examined how the post-lesion network might support CFC learning and memory ., We hypothesized that different network attributes in the ‘damaged network’ could be underlying CFC learning after hippocampal lesion ., Further , we performed double lesions to empirically validate group differences found in the network analysis ., These double lesions tested whether the group differences revealed network changes supporting CFC learning after hippocampal loss ( Fig 1 ) ., Experiment 1 aimed to explore how CFC learning under dHPC damage changes other brain regions activity and interactivity compared to CFC learning in control rats ., We compared pCREB expression levels between the groups in 30 brain regions ( Fig 2 ) and modelled functional networks based on pCREB expression correlations ., Then we employed network tools to explore differences between damaged and control groups ., In Experiment 1 , the rats initially underwent bilateral electrolytic lesions in the dHPC or SHAM surgery ., After surgical recovery , the rats underwent a CFC training session ., Half the cohort was perfused 3 h after the training session , and their brains processed for pCREB immunolabelling ., CREB is a transcription factor involved in neuronal plasticity 30 ., CREB is phosphorylated into pCREB after neuronal activity , which is taken in this study as a proxy of brain region engagement ., The other half of the cohort was returned to the homecage and tested for contextual fear memory 48 h later ., A group of immediate shock controls ( Imm ) was added to the cohort that was tested for contextual fear memory ., The histological examination of dHPC lesions revealed that the cellular loss was overall confined to the dorsal part of the hippocampus , with occasional lesion to the overlaying cortex due to the electrode insertion ( Fig 3A ) ., The cohort tested for contextual memory had the freezing behavior measured as memory index , and was compared among the groups ., The sample size in the memory test cohort was 32 ( SHAM: N = 12; dHPC: N = 12; Imm: N = 8 ) ., A bootstrapped one-way ANOVA showed a significant group effect ( F2 , 29 = 8 . 822 , p = 0 . 0011 ) ., Multiple comparisons performed with p-corrected t-tests showed higher freezing time in both SHAM ( p = 0 . 0001 ) and dHPC ( p = 0 . 0178 ) groups compared to Imm group , but not statistically different from one another ( p = 0 . 4044; Fig 3B ) ., A KS test confirmed these results , showing no difference between SHAM and dHPC samples ( D24 = 0 . 3333 , p = 0 . 2212 ) and both different from Imm group sample ( SHAM: D20 = 0 . 917 , p = 0 . 0001; dHPC: D20 = 0 . 667 , p = 0 . 0070; Fig 3C ) ., A Cohen’s d showed a medium effect size between SHAM and dHPC means ( d = 0 . 630 ) and large effect sizes between these two groups and Imm ( SHAM: d = 2 . 226; dHPC: d = 1 . 275 ) ., These results show no effect of dHPC lesion in CFC learning and are in agreement with past studies 6 ., To ensure the consistency of this lack of effect , we performed two additional experiments in which we test dHPC lesions in two paradigms that are known to be impaired by dHPC lesions ( i . e . water maze and CFC with post-training dHPC lesions ) followed by CFC with pre-training dHPC lesion ( S1 Text ) ., These experiments , which are replications of past studies 6 , 31 , confirm the consistency of this ( lack of ) effect ., In the pCREB immunolabelling cohort , we tested whether the dHPC lesion altered pCREB expression after CFC learning in any of the studied regions by comparing the pCREB expression in each region between dHPC and SHAM groups ., The Fig 3D shows the pCREB expression in each region and each group ., The sample size in the pCREB expression cohort was 19 ( SHAM: N = 9; dHPC: N = 10 ) ., We analyzed the pCREB-positive nuclei density by comparing each region between the groups using t-tests with bootstrap resampling ., There was only one marginally significant difference showing a higher level in the SHAM group in the vSub ( t = 3 . 699 , fdr-corrected p = 0 . 053 ) ., All other regions did not present a significant difference ., This result indicates that dHPC damage diminish the pCREB expression in the vSUB , but otherwise does not alter the overall pCREB-positive nuclei density compared to the SHAM group ., We used the pCREB data to generate correlation-based networks for the SHAM and dHPC groups ., As the SHAM groups has three regions absent in the dHPC group ( dCA1 , dCA3 and dDG ) , a third network was generated as “SHAM with no dorsal hippocampus” , SHAM-nH , to allow for direct comparisons between the networks ( Fig 4 ) ., For each matrix , three networks were generated considering correlations with p-values under the thresholds of 0 . 05 , 0 . 025 or 0 . 01 , respectively ., The networks had a very similar connectivity density in all thresholds ( SHAM networks had 92 , 64 and 40 edges respectively , SHAM-nH had 74 , 53 and 35 edges , and dHPC had 77 , 53 and 32 edges ) ., All networks had initially one bigger component with occasional disconnected regions and fragmented across the thresholds ., As the threshold increased in rigor , the SHAM network remained with one component in the 0 . 05 and 0 . 025 thresholds , and fragmented into three components and 5 disconnected regions in the 0 . 01 threshold ., The SHAM-nH network presented the same behavior seen in the SHAM network ., The dHPC network had a bigger component in the 0 . 05 threshold , but fragmented into four components in the 0 . 025 and 0 . 01 thresholds ( S1 Fig ) ., Although in our study negative correlations were included as absolute values in the edge weights , no negative correlations survived the thresholds ., Overall , the networks presented some visual differences in their pattern of connectivity , which we formally tested in the analyses that follow ., We first tested whether the empirical networks ( SHAM , SHAM-nH and dHPC ) were small-world by comparing their global ( Geff ) and local ( Leff ) efficiencies to those of randomized null hypothesis networks ., We also tested if the dHPC lesion changed any of the efficiencies or the network small-worldness ., The Fig 5 depicts the distribution of the empirical/randomized ratios of Geff and mean Leff for all networks and thresholds ., In all cases , Geff ratios are roughly around 1 , with a slight decay on the 0 . 01 threshold ., Similarly , the mean Leff ratios are consistently above 1 , with the mean and upper range of ratios increasing as the threshold increased in rigor ., Equivalent integration ( Geff ) and robustly higher segregation ( Leff ) values in empirical networks compared to randomized networks is consistent with small-world networks accounts 32 , 33 ., These results suggest that the networks engaged in CFC learning are small-world , which is in agreement with a previous work showing small-world organization in CFC retrieval networks 28 ., Further , dHPC lesion did not seem to change the dHPC network small-worldness or its levels of Geff and mean Leff compared to the SHAM and SHAM-nH networks , suggesting that the overall characteristic interactivity in the dHPC network still benefit from small-world architecture ., Hubs are defined as nodes positioned to confer the largest contributions to global network function , and are usually identified using multiple centrality metrics 24 ., We considered as hub any region among the 25% most central regions in at least three of the four centrality metrics used ( weighted degree , Wdg; eigenvector , Evc; closeness , Clo; and betweenness , Bet ) ., Regions that were hubs across all thresholds were considered stable hubs ., The Fig 6A and 6B shows the ranked centralities for the dHPC network and the metric intersection in the threshold 0 . 05 ( S2–S4 Figs show all networks and thresholds ) ., In this threshold , the SHAM network showed the regions IL and BLV as hub , whereas in the SHAM-nH the BLV and Por were hubs , and the dHPC network the hubs were the Per_36 , Per_35 , RSC and LAVL ., The Fig 6C shows which regions were considered stable hubs across the thresholds , in each network ., In the SHAM network , the IL was the only region identified as a stable hub across all thresholds ., In the dHPC network , the RSC , and the Per_36 were stable hubs across all thresholds , and in the SHAM-nH network , no hub was stable across the three thresholds , but the IL was the most stable region ( hub in the 0 . 025 and 0 . 01 thresholds ) , similar to the SHAM network ., Employing connection-based and distance-based metrics to identify a hub makes more likely that the identified well-connected regions are also inter-region or inter-modular connectors ., Noticeably , the dCA1 was in the upper quartile of both connection-based metrics , but not the distance-based ones , across the all thresholds ( S2 Fig ) ., These results suggest that different hubs emerged in the dHPC network ., However , as the identification was descriptive , with no hypothesis test , it does not allow a priori interpretations regarding differences in the hub score between the networks ., However , they are a first indication that there might be differences in the connectivity patterns between the SHAM and dHPC networks , as different regions emerged as hubs in these networks ., We addressed the hub score differences more formally and quantitatively by directly comparing the centralities between the groups in each region and each threshold using permutation test ., The Fig 7 resumes the results of the permutation tests for each region , metric and threshold ., Most importantly , we observed that the identified stable hubs were overall associated with significantly higher centrality levels in some metrics , comparing the dHPC SHAM-nH networks ., In the dHPC network , the RSC showed significantly higher Wdg and Evc in all thresholds , and the Per_36 showed higher Evc levels in the 0 . 025 and 0 . 01 thresholds , compared to SHAM-nH network ., In the SHAM-nH network , the IL showed higher Evc levels in the 0 . 025 and 0 . 01 thresholds , compared to the dHPC network ( S5 Fig ) ., Besides the stable hubs , some of the single-threshold or two-threshold hubs were also associated to significantly different centrality levels between the networks ., In the dHPC network , the RSGd presented a higher Evc across all thresholds and a higher Wdg in the 0 . 025 and 0 . 01 thresholds ., The LAVL had a higher Evc in the 0 . 025 threshold ., In the SHAM-nH network , the BLV presented a higher Wdg across all thresholds , higher Bet in the 0 . 05 and the 0 . 01 thresholds , and higher Evc in the 0 . 05 threshold ., Further , the CeM and PrL showed higher Evc , and the RSGv showed higher Bet , all in the 0 . 01 threshold ., Some significant differences were present in non-hub regions such as BLP , vCA1 , DLE and Por ( higher metrics in dHPC network ) , and LAVM , BLA , and Por ( higher metrics in the SHAM-nH network; S5 Fig ) ., Lastly , some single-threshold hubs did not show significantly different centrality metrics in the thresholds they were considered hubs , such as LAVL , Per_35 , Por and Cg1 ( dHPC network ) and CeL , Por ( SHAM-nH network ) ., These results provide evidence that when comparing SHAM-nH and dHPC networks , stable hubs in one network were associated to higher centrality levels relative to the other , and vice-versa ., These results suggest that the CFC learning network under dHPC lesion has an increased dependence on its new hubs ., The analysis so far focused on the nodes ., We next examined edge ( correlation coefficients ) differences between the dHPC and SHAM-nH networks ., First , we compared the distribution of correlations of each matrix between groups using a two-sample KS test ., We observed significantly different correlation coefficient distributions between dHPC and SHAM-nH networks in all thresholds ( threshold 0 . 05: D151 = 0 . 2527 , p = 0 . 0125; 0 . 025: D106 = 0 . 3396 , p = 0 . 0042; 0 . 01: D67 = 4795 , p = 0 . 0005; Fig 8A and 8B and S6 Fig ) ., Next , we compared each correlation coefficient between the groups ., We computed the Z-score of the group difference for each correlation coefficient and considered a score of |2| to be significant within the distribution ., We observed 21 correlation differences with Z-scores above |2| ( Fig 8B ) ., In nearly 2/3 of the significant differences ( 15 out of 21 ) , the stronger correlation coefficients belonged to the SHAM-nH network , and 9 of them belonged to SHAM-NH hubs in that threshold; whereas only 6 differences the stronger correlation coefficient belonged to the dHPC network , one of which belonged to a hub ( Fig 8C ) ., These results were similar across thresholds ., In the 0 . 025 threshold , 19 out of 26 differences were higher in the SHAM-nH network ( 3 belonging to SHAM-nH hubs; S6 Fig ) , and in the 0 . 01 threshold , 20 out of 28 differences were higher in SHAM-nH network ( 9 belonging to SHAM-nH hubs ) ., Overall , these results show that the SHAM-nH network presented a higher number of significantly stronger correlations compared to the dHPC network , many of which belonged to SHAM-nH hubs for that threshold ., The different correlation distributions and the differences in correlation strengths between the networks add support to the hypothesis of different connectivity patterns in the dHPC network ., Further , it suggests that dHPC indirectly influences interactions between other regions , most of which were observed to be weakened ., The network analysis revealed some differences between the dHPC and the SHAM ( or SHAM-nH ) networks ., Particularly , the alternative hubs emerging in the dHPC network ( Per_36 and RSC ) and their statistically higher centralities compared to the SHAM-nH network suggest that these regions may increase in their importance to CFC learning in the absence of hippocampus ., We empirically tested this hypothesis in the next two experiments by damaging both the dHPC and one of these hubs pre-training to CFC ., Our hypothesis is whether further insult to the network would compromise the necessary structure of the network to promote CFC learning ., In Experiment 2 , because it was technically difficult to damage specifically the Per_36 and most animals had a significant part of the Per_35 damaged , we considered animals with lesions extending to both Per_36 and Per_35 , denominating it Per ., Henceforth , Per will be mentioned when Per_35 and Per_36 are considered together ., During histological analysis , we excluded four rats from the dHPC-Per , two from the Per and one from the dHPC groups due to either extensive bilateral lesions to the regions surrounding Per ( Temporal , Auditory , Parietal , Visual cortices , ventral CA1 or Lateral Amygdala ) , or no detectable dHPC and/or Per cellular loss in most slices examined ., The final sample in this experiment was 38 ( SHAM , dHPC and Per: N = 10/each; dHPC-Per: N = 8 ) ., In the remaining sample , cellular loss was mostly confined to the Per_36 , Per_35 and to dHPC ., In the dHPC and dHPC-Per groups , slight occasional damage was observed in the secondary Visual and Medial Parietal cortices overlying dHPC due to needle insertion ( Fig 9A ) ., In the behavioral analysis , the bootstrapped ANOVA showed no group difference ( F = 0 . 842 , p = 0 . 479; Fig 9B and 9C ) ., The KS test showed no significant differences among groups’ distributions and the Cohen’s d values did not show any considerable effect size ( Fig 9 bottom ) ., These results indicate that neither Per or dHPC-Per lesions affect CFC learning and memory ., Previous studies observed no pre-training Per lesion effect on CFC 34 , 35 , despite some contradictory evidence 36 ., Our results support the hypothesis that pre-training Per and dHPC-Per lesions do not affect CFC learning and memory ., During histological analysis , three rats from the RSC and one from the dHPC-RSC group were excluded from the analysis due to non-detectable cellular loss in most slices ., The final sample in this experiment was 39 ( SHAM: N = 10 , dHPC and RSC: N = 9/each , dHPC-RSC: N = 11 ) ., The lesions affected mainly the dHPC and RSC , with frequent lesions to RSGd and occasional minor unilateral lesions of RSGv and secondary visual cortex ., In the behavior analysis , the bootstrapped ANOVA revealed a main effect of group ( F3 , 35 = 3 . 691 , p = 0 . 01975 ) , which the p-corrected t tests showed to be due to a lower freezing in the dHPC-RSC compared to that of the SHAM group ( t20 = 3 . 315 , p = 0 . 0270; Fig 10B and 10C ) ., No other significant differences were observed ., This result was further confirmed by the KS test , which revealed significantly different distributions between the dHPC-RSC and the SHAM samples ( D = 0 . 609 , p = 0 . 0303 ) ., No other differences were observed ( SHAM vs dHPC: D = 0 . 378 , p = 0 . 330; SHAM vs RSC: D = 0 . 367 , p = 0 . 377; dHPC vs RSC: D = 0 . 333 , p = 0 . 316; dHPC vs dHPC-Per: D = 0 . 485 , p = 0 . 098; Per vs dHPC-Per: D = 0 . 374 , p = 0 . 289 ) ., The Cohen’s d values also confirmed the above results showing a large effect size between SHAM and dHPC-RSC means ( d = 1 . 469 ) ., Lesser effect size values were observed in the other comparisons ( SHAM vs dHPC: d = 0 . 463; SHAM vs RSC: d = 0 . 75; dHPC vs Per: d = 0 . 338; dHPC vs dHPC-Per: d = 1 . 056; Per vs dHPC-Per: d = 0 . 598; Fig 10 bottom ) , although the effect size between dHPC and dHPC-RSC was somewhat large ., These results show that both dHPC and RSC contribute to CFC learning , although single lesion of these regions was not sufficient to impair CFC ., Further , it supports the network analysis in Experiment 1 that RSC becomes a key region in the dHPC network engaged in CFC learning ., However , a careful analysis of the lesion extensions raises two possible alternative explanations for the observed results ., First , the RSC lesion extension seems to be larger in the dHPC_RSC group than in the RSC group , which raises the possibility of the unimpaired behavior in the RSC be due to a smaller lesion ., Second , as lesion frequently extended to RSGd in the dHPC_RSC group , there is the possibility that RSGd lesion had an effect in the impaired behavior observed in this group ., To test the first alternative possibility , we measured the percentage of lesion in the dHPC , RSC and RSGd regions in the RSC and dHPC_RSC groups , and compared them between these groups ., A t test with bootstrap resampling showed that the RSC and RSGd lesion extensions were , in fact , larger in the dHPC_RSC group than in RSC group ( RSC: t = 2 . 252 , p = 0 . 042; RSGd: t = 2 . 582 , p = 0 . 021; S7A Fig ) ., Next , we tested whether this difference in the lesion extension had any influence in the observed behavior ., We compared the freezing among the groups just as above using the percentage of damage as co-variables by an ANCOVA ., The ANCOVA still showed the group main effect ( F3 , 32 = 3 . 777 , p = 0 . 0199 ) , but none of the lesion extensions showed any co-varying effect ( RSC: F1 , 32 = 0 . 872 , p = 0 . 3575; RSGd: F1 , 32 = 0 . 001 , p = 0 . 9728; dHPC: F1 , 32 = 2 . 937 , p = 0 . 0962 ) ., These results suggest that the lesion extension of these regions had no effect on the observed behavior ., One thing to be considered however , is that both RSC and dHPC had a minimum of 50% of damage because of our exclusionary data ., We do not rule out the possibility that , in an unbiased sample , there could be a relation between lesion extension and behavior , but this relation seems not to be evident when lesion extension exceeds 50% of the target regions ., Therefore , it is unlikely that the lack of effect in the RSC group could have been due to the observed smaller RSC lesion observed in the group in comparison to the dHPC_RSC group ., To address the second possibility , we selected dHPC_RSC individuals that had their lesions more constrained to the RSC , with minor to undetectable damage to RSGd ( less than 30% , N = 4 ) , and compared them to the other groups ., If the dHPC and RSC double lesions have an effect that is independent from the RSGd lesions , the differences observed should still be observable ., Bootstrapped t tests showed a lower freezing time in the strict dHPC_RSC subgroup compared to the SHAM group ( t13 = 2 . 532 , p = 0 . 049 ) , whereas it did not differ from any of the other groups ( dHPC: t12 = 1 . 717 , p = 0 . 135; RSC: t12 = 0 . 992 , p = 0 . 349; dHPC_RSC: t14 = 0 . 024 , p = 0 . 971; S7B Fig ) ., These results replicate the results above where samples included significant RSGd damage , suggesting that double lesions of the dHPC and RSC impair CFC learning and memory independently of the RSGd damage ., It is important to note , however , that the RSGd presented almost as many higher centrality values as the RSC in the dHPC network compared to the SHAM-nH network ., Thus , we do not rule out the possibility of RSGd also possesses some increased influence in the dHPC network and in the behavior ., There are some points about the present study that need attention when interpreting the results ., First , we do not show both behavioral and brain activation data in the same subjects in Experiment 1 , therefore not directly showing that the altered brain networks are linked to preserved CFC learning ., The purpose of our study was to investigate how learning and memory formation about CFC occurs in the absence of the dHPC , however , a limitation of the imaging methods used in rodents to assess experience-driven proxies of brain activation ( i . e . IEGs such as c-fos and arc ) is that they are acquired post-mortem and , thus , require to be the last step in the experimental design ., Previous studies used post-retrieval IEG acquisition in order to acquire both brain and behavior data from the same subjects 28 ., However , such design would need us to assume that brain activation during CFC learning and retrieval are equivalent , which there is evidence showing the contrary 87 , 88 ., Further , IEGs and pCREB expressions are primarily related to neuronal plasticity processes 30 , and acquiring post-retrieval pCREB expression would not necessarily reflect a measure brain activation during retrieval , but possibly processes that could induce ‘post-retrieval plasticity’ i . e . new encoding , reconsolidation; 89 ., Given this limitation , we ensured that our brain network and behavior results are not product of variability by doing additional experiments challenging this hypothesis ., Our Experiments 2 and 3 test whether the network analysis could be a product of chance ( variability included ) , and we ran two additional experiments that replicate previous studies 5 , 6 showing paradigms impaired by dHPC lesion followed by unimpaired CFC in rats with pre-training dHPC lesions ( S1 Text ) ., Despite this caveat , the confirmation experiments show the consistency of our observations and the relation between the altered brain networks and preserved CFC learning in dHPC damaged rats ., Second , when comparing networks , one seeks differences attributable solely to the network structure , but network attributes such as number of nodes , edge density and mean degree might add confounding effects if they are not equivalent between the networks ., However , altering networks so that they match in these attributes ( ex . proportional thresholding , node removal ) might change the network topology and drive false positive and/or negative differences 90 ., As no formal solution exists for comparing networks of differing sizes , in our study , we removed nodes from the SHAM network ( producing SHAM-nH ) in order to compare it to dHPC network ., This node removal did not affect the overall network topology or its small-worldness ( Fig 5 ) , minimizing possible biases from this procedure ., Additionally , the thresholds applied in the present study intended to remove correlations that were no different from chance , which depended solely on the pCREB co-variation among the regions within each condition ., Therefore , there was no proportional thresholding to control edge density and mean degree across conditions ., The fact that we did observe equivalent edge densities across conditions was due to their similar co-variations ., This characteristic reflected in our analysis ( similar efficiencies ) and can be observed when plotting the networks based on Euclidean distances ( S1 Fig ) ., Importantly , the thresholding could have produced networks with differing edge densities or mean degree , which would require alternative approaches to the analysis ., Third , the lesion method used in Experiment 1 ( electrolytic lesion ) does not spare fibers of passage , which may have affected connections between other regions ., Whilst this could have altered the network more than intended , the behavior data suggests that the network is likely to contain the elements required in CFC learning and memory since no impairment was observed ., Furthermore , the networks studied here , which are based on pCREB expression , identified similar hubs to recent anatomical studies based on larger tract-tracing databases 91 , 92 , making a confounding effect of fiber lesion unlikely ., Fourth , the Experiment 1 differs from Experiments 2 and 3 in number of shocks during the training session ., Single shock CFC sessions is generally a weaker experience and tend to yield more variable levels of behavior ., We used the three shocks procedure to ensure a robust performance level in Experiments 2–3 such that impairments would be more detectable ., Additionally , the performances of SHAM controls and dHPC groups were very similar , ruling out the possibility of a ‘hidden’ memory impairment in the dHPC group in Experiment 1 ., There is growing interest in the use of network approaches to predict cognitive performance from brain imaging data 22 , 44 , 45 ., However , formally testing predictions in human experimentation is still a challenge called for attention 93 ., We applied network analysis in rodent models and were able to empirically test the validity of these models subsequently ., We found that CFC network under dHPC damage increase its dependence on new hubs , and further damaging these new hubs may compromise the formation of the functional network necessary for CFC learning and memory ., Future employment of finer techniques ( i . e . optogenetics , transgenic animals ) may provide sophisticated ways to test network predictions ., A hundred and thirty nine male Wistar rats weighting 300-370g were obtained from the university vivarium ( CEDEME , SP ) ., They were housed in groups of 4–5 and maintained on a 12h light/dark cycle , room temperature of 22 ± 2°C , with free access to food and water ., This research made use of one hundred and thirty nine male Wistar rats obtained from the university vivarium ( CEDEME , SP ) ., All experiments were approved by the University Committee of Ethics in Animal Research ( CEUA , approval numbers #0392/10 , #409649 and #7683270116 ) ., The guidelines used by CEUA are in accordance with National Institutes of Health Guide for the Care and Use of Laboratory Animals in the USA ., The present study involved stereotaxic surgeries ., In these procedures , the rats were anesthetized with Ketamine ( 90mg/kg , Ceva , Paulínia , Brazil ) and Xilazine ( 50mg/kg , Ceva , Paulínia , Brazil ) given in intraperitoneal injections ., In the end of the behavioral procedures , the rats were anesthesized with 10% chloral hydrate and perfused transcardially with 4% paraformaldehyde ., The rats were anesthetized with Ketamine ( 90mg/kg , Ceva , Paulínia , Brazil ) and Xilazine ( 50mg/kg , Ceva , Paulínia , Brazil ) , and mounted into a stereotaxic frame ( David Kopf Instruments , Tujunga , CA ) ., Each animal had their scalp incised , retracted and the bregma and lambda horizontally adjusted to the same plane ., Small holes were drilled in the skull in the appropriate sites ., The rats received bilateral electrolytic lesions in the dHPC by an anodic current ( 2 mA , 20 s ) passed through a stainless ste

Introduction, Results, Discussion, Materials and methods

Hippocampal damage results in profound retrograde , but no anterograde amnesia in contextual fear conditioning ( CFC ) ., Although the content learned in the latter have been discussed , alternative regions supporting CFC learning were seldom proposed and never empirically addressed ., Here , we employed network analysis of pCREB expression quantified from brain slices of rats with dorsal hippocampal lesion ( dHPC ) after undergoing CFC session ., Using inter-regional correlations of pCREB-positive nuclei between brain regions , we modelled functional networks using different thresholds ., The dHPC network showed small-world topology , equivalent to SHAM ( control ) network ., However , diverging hubs were identified in each network ., In a direct comparison , hubs in both networks showed consistently higher centrality values compared to the other network ., Further , the distribution of correlation coefficients was different between the groups , with most significantly stronger correlation coefficients belonging to the SHAM network ., These results suggest that dHPC network engaged in CFC learning is partially different , and engage alternative hubs ., We next tested if pre-training lesions of dHPC and one of the new dHPC network hubs ( perirhinal , Per; or disgranular retrosplenial , RSC , cortices ) would impair CFC ., Only dHPC-RSC , but not dHPC-Per , impaired CFC ., Interestingly , only RSC showed a consistently higher centrality in the dHPC network , suggesting that the increased centrality reflects an increased functional dependence on RSC ., Our results provide evidence that , without hippocampus , the RSC , an anatomically central region in the medial temporal lobe memory system might support CFC learning and memory .

When determined cognitive performances are not affected by brain lesions of regions generally involved in that performance , the interpretation is that the remaining regions can promote function despite of ( or compensate ) the damaged one ., In contextual fear conditioning , a memory model largely used in laboratory rodents , hippocampal lesions produce amnesia for events occurred before , but not after the lesion , although the hippocampus is known to be important for new learning ., Addressing how the brain can overcome lesion in animal models has always been challenging as it requires large-scale brain mapping ., Here , we quantified 30 brain regions and used mathematical tools to model how a brain network can support contextual fear learning after hippocampal loss ., We described that the damaged network preserved general interactivity characteristics , although different brain regions were identified as highly important for the network ( e . g . highly connected ) ., Further , we empirically validated our network model by performing double lesions of the hippocampus and the alternative hubs observed in the network models ., We verified that double lesion of the hippocampus and retrosplenial cortex , one of the hubs , impaired contextual fear learning ., We provide evidence that without hippocampus , the remaining network relies on alternative important regions from the memory system to coordinate contextual fear learning .

learning, cognitive neurology, medicine and health sciences, neural networks, brain damage, brain, social sciences, neuroscience, learning and memory, cognitive neuroscience, cognitive psychology, cognition, network analysis, memory, computer and information sciences, cognitive impairment, centrality, psychology, hippocampus, anatomy, neurology, biology and life sciences, cognitive science

journal.pgen.1000403

Death and Resurrection of the Human IRGM Gene

Immunity Related GTPases ( IRG ) , a family of genes induced by interferons , are one of the strongest resistance systems to intracellular pathogens 1–4 ., The IRGM gene has been shown to have a role in the autophagy-targeted destruction of Mycobacterium bovis BCG 5 ., Recently , whole genome association studies have shown that specific IRGM haplotypes associate with increased risk for Crohns disease 6 , 7 ., The IRG gene family exists as multiple copies ( 3–21 ) in most mammalian species but has been reduced to two copies , IRGC and a truncated gene IRGM , in humans 8 ., Analysis of mammalian genomes ( dog , rat and mouse ) has shown that all IRG genes except IRGC are organized in tandem gene clusters mapping to mouse chromosomes 11 and 18 ( both syntenic to human chromosome 5 ) 8 ., A comparison of the mouse and human genomes identified 21 genes in mouse but only a single syntenic truncated IRGM copy and IRGC in human 8 ., We investigated the copy number and sequence organization of the IRG gene family in multiple nonhuman primate species in order to reconstruct the evolutionary history of this locus ., Sequence analysis of two different prosimian species ( Microcebus murinus and Lemur catta ) confirmed the mammalian archetypical organization with three IRGM paralogs in each species ( Figure 1 ) ., FISH analysis showed that genes in these species are organized as part of a tandem gene family similar to the organization observed within the mouse genome ( Figure 2 ) ., In contrast , FISH and sequence analysis of various monkey and great ape species ( see Text S1 ) confirmed a single copy in each of these species ., Based on the estimated divergence of strepsirrhine and platyrrhine primate lineages , we conclude the IRGM gene cluster contracted to a single truncated copy 40–50 million years ago within the anthropoid lineage of evolution ., We next compared the structure of the IRGM gene in various primate species ., One of the three mouse lemur IRGM genes ( IRGM9 ) preserves a complete ORF based on the mouse model and shows the greatest homology to mouse Irgm1 ., The ORF encodes a putative 47 kD protein including a classical N-terminal region as well as classical motifs at the end of the carboxyl-terminus associated with most functional murine IRGM loci 8 , 9 ( see Text S1 ) ., The second mouse lemur gene , IRGM8 , is likely a pseudogene because of a mutation generating a stop codon within the G domain and a frameshift mutation at the C terminus ., The third mouse lemur gene , IRGM7 , is atypical because it has substitutions in the G domain that disrupt the G1 motif that interacts with the nucleotide phosphates and is highly conserved in P-loop GTPases 10 ( Figure S1 and Text S1 ) ., In contrast to mouse and prosimian species , all anthropoid primate lineages show the presence of an AluSc repeat immediately after the splicing acceptor that disrupts the ORF of the sole remaining IRGM gene ( Figure 1 and S2 ) ., Sequencing of the IRGM locus in four New World monkey species revealed the presence of the same two stop codons disrupting the ORF of IRGM in all species ., We similarly identified a common frameshift mutation resulting in premature stop codons within the IRGM locus in eleven diverse Old World monkey species suggesting that IRGM had become pseudogenized before the radiation of these species ., Sequencing of the gene in multiple individuals in the same species ( five unrelated Rhesus macaque and baboon ) suggested that the frameshift mutations were fixed ( Figure S3 and Text S1 ) ., In total , these data argue that the IRGM locus has been nonfunctional since the divergence of the New World and Old World monkey lineages ( 35–40 million years ago ) likely as a result of an Alu repeat integration event that disrupted the ORF of the gene in the anthropoid ancestor ( Figure 1 ) ., In contrast to New World and Old World monkeys , sequencing of the IRGM locus in humans and African great ape species reveals a restored , albeit truncated , ORF of ∼20 kD in length ., This is consistent with an antiserum raised against peptides from the human IRGM protein that detected a specific signal at ∼20 kD by Western blot 11 ., In contrast to humans and the African great apes , analysis of the orangutan genome assembly predicted a nonfunctional protein ( C to T transition at nucleotide position 150 with respect to the start codon resulting in a premature shared stop codon in the ORF ( Figure 1 and Text S1 ) ., This is the same substitution identified among all Old World monkey genomes suggesting that ancestral ape species carried a pseudogene ., We resequenced the IRGM gene in twelve different orangutans and five different gibbon species ., Six of the twelve individuals from orangutan and one of the five species from gibbon are heterozygous for the C to T substitution ., In addition , we noted that all ape IRGM copies also shared a new translation initiation codon with a preferred Kozak sequence immediately after the Alu integration ., These data indicate that the gene can exist as either a pseudogene or as a complete 20 kD ORF among these Asian ape lineages as a result of either balancing selection or recurrent mutational events ., It will be necessary to examine a larger number of individuals within each species to establish the evolutionary history of this locus among the Asian apes ., We noticed an important structural difference in the gene organization for species that regained putative IRGM function when compared to those primates with a pseudogenized version ., In the common ancestor of humans and great apes , an ERV9 retroviral element integrated within the 5′ end of the IRGM gene ( Figure 1 ) ., We reasoned that this structural difference may have conferred expression differences and analyzed the RT-PCR expression profile of IRGM in human , macaque and marmoset ., Full-length cDNA sequencing and 5′ RACE revealed that the human transcription start signal mapped specifically within the ERV9 repeat element ( Figure 1 and Figure S4 ) resulting in the addition of a novel 5′UTR exon and an alternative splice form ., Although there are five distinct , alternative splice forms of human IRGM , all human copies share this first intron ., In humans , we observe constitutive levels of expression of IRGM in all tissues examined , with the highest expression of IRGM in the testis ( Figure 3A ) 8 ., Although IRGM does not encode a functional protein in marmoset and macaque , we find evidence of low levels of expression , albeit in a more restricted manner ( Figure 3B ) ., Macaque and marmoset , for example , show no expression in the kidney with marmoset IRGM expression restricted to testis and lung ., Furthermore , we find no evidence in macaque of splicing of the first intron based on the human IRGM gene model ( Figure 3C ) but rather evidence that the first intron remains as a continuous unspliced transcript ., We also failed to confirm 3′ downstream splicing events of macaque IRGM suggesting that even if stop codons were reverted , a full-length cDNA ( comparable to human ) could no longer be produced ., These data strongly suggest that ERV9 integration significantly reshaped the expression and splicing pattern of IRGM in the common ancestor of humans and apes ( Figure 3 ) ., We note that structural changes of the human IRGM locus continues to occur within the human lineage with a 20 . 1 kb LTR-rich deletion polymorphism , recently identified and sequenced , located 2 . 82 kb upstream of the ERV9 promoter region 12 ., Our preliminary data suggest that this deletion polymorphism alters the relative proportion of alternative splicing of IRGM transcripts ( Figures S5 and S6 ) ., We tested for natural selection on IRGM coding sequence using maximum likelihood models to estimate evolutionary rates for individual branches in the phylogeny as well as specific codon changes 13 , 14 ., Based on the structural differences in IRGM organization , we first divided our species into three groups: Group 1 consists of species that carry a single copy of IRGM with the ERV9 element ( human ( Hs ) , chimpanzee ( Ptr ) , gorilla ( Ggo ) and orangutan ( Ppy ) ) ; Group 2 consists of species that carry a single copy of IRGM but lack the ERV9 element ( Macaque ( Rh ) , baboon ( Pha ) and marmoset ( Cja ) ) ; while Group 3 was formed by species ( dog and mouse lemur ) that had multiple copies in a tandem orientation ( Figure 4 ) ., Phylogenetic branch estimates of dN/dS revealed striking differences between Group 2 ( ω\u200a=\u200a0 . 9254 ) and Group 3 ( ω\u200a=\u200a0 . 3866 ) with an intermediate value for Group 1 ( ω\u200a=\u200a0 . 6073 ) ., Group 3 was found to be under constrained evolution ( ω\u200a=\u200a0 . 3866 ) and it was significantly different ( P\u200a=\u200a6 . 09E−12 ) from a model of neutral evolution ., In contrast , Group 1 and 2 gene evolutions were indistinguishable from a model of neutral evolution ( see Text S1 ) ., There are two possible interpretations of our results ., First , the IRGM gene is not functional in humans having lost its role in intracellular parasite resistance ∼40 million years ago when the gene family experienced a contraction from a set of three tandem genes to a sole , unique member whose ORF was disrupted by an AluSc repeat in the anthropoid primate ancestor ., In light of the detailed functional studies 11 and the recent associations of this gene with Crohns disease 6 , 7 , we feel that this interpretation is unlikely ., For example , McCarroll and colleagues recently demonstrated that a 20 . 1 kb deletion upstream of IRGM associates with Crohns disease as well as the most strongly associated SNP and that the deletion haplotype showed a distinct pattern of IRGM gene expression consistent with its putative role in autophagy and Crohns disease ., An alternate scenario is that the IRGM gene became nonfunctional ∼40 million years ago ( leading to pseudogene copies in Old World and New World monkeys ) but was resurrected ∼20 million years ago in the common ancestor humans and apes ( Figure 5 ) ., In addition to the genetic and functional data , several lines of evidence support this seemingly unusual scenario ., First , we find evidence of a restored ORF in humans and African great apes ., Second , this change coincided with the integration of the ERV9 element that serves as the functional promoter for the human IRGM gene ., Such retroposon-induced alterations of gene expression are not without precedent in mammalian species 15 , 16 ., Third , we find that ape/human codon evolution is consistent with a model of nucleotide constraint resulting in depressed dN/dS ratios in the hominid branch ( Figure 4 ) when compared to the Old World and New World species ., It is intriguing that the orangutan and gibbon populations possess both a functional and nonfunctional copy of IRGM , which would open the possibility to long-term balancing selection or recurrent mutations ( see Text S1 ) ., The inactivating stop codon is shared with all Old World monkey species suggesting an ancestral event ., Moreover , we and others 17 find that the structure of the locus is continuing to evolve in humans altering the expression profiles of IRGM transcripts in different tissues ., These structural changes are thought to underlie the strong association with Crohns disease , perhaps , by modulating the efficiency of the autophagic response 17 ., Our data suggest remarkable functional plasticity where alleles experience diverse evolutionary pressures over time ., Such dynamism in structure and evolution may be critical for a gene family locked in an arms race with an ever-changing repertoire of intracellular parasites ., We retrieved whole genome shotgun sequence of the IRGM locus for chimpanzee ( Pan troglodytes ) , gorilla ( Gorilla Gorilla ) , orangutan ( Pongo pygmaeus ) , rhesus macaque ( Macaca mulatta ) , marmoset ( Callithrix jacchus ) , baboon ( Papio hamadryas ) , and Gray Mouse Lemur ( Microcebus murinus ) from NCBI Trace Archive ( http://www . ncbi . nlm . nih . gov/Traces/trace . cgi ? ) and constructed local sequence assemblies using PHRAP ( http://www . phrap . org ) ., We sequenced and confirmed the IRGM genome organization based on DNA samples from four different New World monkey species and from eleven different Old World monkey species ., We also resequenced the IRGM gene in unrelated macaques ( n\u200a=\u200a5 ) , baboons ( n\u200a=\u200a5 ) , orangutans ( n\u200a=\u200a12 ) and gibbons ( n\u200a=\u200a7 ) ., For Microcebus murinus with multiple copies of IRGM , we first isolated large-insert BAC clones , subcloned and sequenced PCR amplicons corresponding to the different copies ., Metaphase spreads were obtained from lymphoblast or fibroblast cell lines from human ( Homo sapiens ) , rhesus macaque ( Macaca mulatta ) , marmoset ( Calithrix jacchus ) and lemur ( Lemur catta ) ., FISH was performed using either human IRGM probe WIBR2-3607H18 or lemur IRGM BAC DNA LB2-61D22 , LB2-77B23 and LB-61A22 , directly labeled by nick translation with Cy3-dUTP ( Perkin-Elmer ) ., Lemur BAC probes were obtained by library hybridization screening of a L . catta genomic library ( CHORI Resources: LBNL-2 Lemur BAC Library http://gsd . jgi-psf . org/cheng/LB2 ) ., Full-length human IRGM transcript was obtained by 5′RACE PCR followed by subcloning ( PGEM-T easy ) and sequencing ( EU742619 ) ., RT-PCR experiments were performed using cDNA synthesized ( Advantage RT-PCR , Clontech ) from mRNA extracted ( Oligotex isolation kit , Qiagen ) from total RNA ( RNA Easy , Qiagen ) ., Total RNA was obtained from tissues isolated from chimpanzee , rhesus macaque , marmoset and human ., IRGM splice variants were detected by a quantitative PCR assay using the LightCycler SYBR Green System ( Roche ) with primers IRGM, ( b ) -, ( c ) -, ( d ) and IRGM all primers ( Text S1 ) ., Transcript levels were normalized to the amount of the GAPDH and UBE1 transcript , which also served as positive controls for RT-PCR experiments ., We generated multiple sequence alignments using Clustal-W18 , 19 and constructed neighbor-joining phylogenetic trees ( MEGA 3 . 1 ) 20 ., Tests of selection ( ω\u200a=\u200adN/dS ) were performed by maximum likelihood using PAML 13 applying the Sites Model 14 to calculate the percentage of codons under positive , neutral evolution or purifying selection and the Branch model 21 to estimate evolutionary pressures at different times during evolution ., The Likelihood Ratio Test ( LRT ) was used to assess the significance of different values of ω for different groups .

Introduction, Results, Discussion, Methods

Immunity-related GTPases ( IRG ) play an important role in defense against intracellular pathogens ., One member of this gene family in humans , IRGM , has been recently implicated as a risk factor for Crohns disease ., We analyzed the detailed structure of this gene family among primates and showed that most of the IRG gene cluster was deleted early in primate evolution , after the divergence of the anthropoids from prosimians ( about 50 million years ago ) ., Comparative sequence analysis of New World and Old World monkey species shows that the single-copy IRGM gene became pseudogenized as a result of an Alu retrotransposition event in the anthropoid common ancestor that disrupted the open reading frame ( ORF ) ., We find that the ORF was reestablished as a part of a polymorphic stop codon in the common ancestor of humans and great apes ., Expression analysis suggests that this change occurred in conjunction with the insertion of an endogenous retrovirus , which altered the transcription initiation , splicing , and expression profile of IRGM ., These data argue that the gene became pseudogenized and was then resurrected through a series of complex structural events and suggest remarkable functional plasticity where alleles experience diverse evolutionary pressures over time ., Such dynamism in structure and evolution may be critical for a gene family locked in an arms race with an ever-changing repertoire of intracellular parasites .

The IRG gene family plays an important role in defense against intracellular bacteria , and genome-wide association studies have implicated structural variants of the single-copy human IRGM locus as a risk factor for Crohns disease ., We reconstruct the evolutionary history of this region among primates and show that the ancestral tandem gene family contracted to a single pseudogene within the ancestral lineage of apes and monkeys ., Phylogenetic analyses support a model where the gene has been “dead” for at least 25 million years of human primate evolution but whose ORF became restored in all human and great ape lineages ., We suggest that the rebirth or restoration of the gene coincided with the insertion of an endogenous retrovirus , which now serves as the functional promoter driving human gene expression ., We suggest that either the gene is not functional in humans or this represents one of the first documented examples of gene death and rebirth .

genetics and genomics/comparative genomics, evolutionary biology/evolutionary and comparative genetics, evolutionary biology/human evolution, genetics and genomics/functional genomics, immunology/innate immunity, molecular biology/bioinformatics, immunology/genetics of the immune system, evolutionary biology, genetics and genomics/bioinformatics, infectious diseases/gastrointestinal infections

journal.pgen.1000778

DNA Specificity Determinants Associate with Distinct Transcription Factor Functions

Transcriptional regulation of gene expression is programmed through DNA sequence elements , termed promoters and enhancers ., This genomic hard-wiring represents binding sites for transcription factors that have sequence specific DNA recognition and control development and homeostasis ., Although the fundamental properties of protein-DNA recognition are well established , the advent of powerful technologies that provide genome-wide occupancy data has only recently allowed observation of these interactions in vivo ., The emerging picture is that no single sequence motif fully explains all in vivo binding 1–3 ., Furthermore , the in vitro derived consensus motifs are often present in only a minority of bound regions ., These findings bring into question the purpose of binding site sequence variations ., Possibilities are illustrated by experimental analysis of subsets of sites gathered from genomic data ., For example , the PHA4/FOXO binding sites that program pharynx development in C . elegans differ in affinity , and thus carry developmental programming information dictating time of expression 4 ., In yeast , PHO4 responsiveness to phosphate levels is regulated by alterative sequence motifs that affect affinity and program different roles for binding sites 5 ., NF-κB and GR binding site variants can alter the repressing or activating transcriptional activity of the factor once it is bound 6 , 7 ., The challenge of genomic databases is how to take full advantage of the vast number of binding sites , yet parse out functional consequences of variation ., To realize their full potential , genomic approaches to transcriptional networks must go beyond a description of factor occupancy to include correlates of functionality ., We focus on the transcription factor ETS1 that provides a variety of contexts to address these central questions ., ETS1 is a member of the ETS family of transcription factors that display similar DNA binding properties , including the recognition of a core GGA ( A/T ) motif ., ETS family members are extensively co-expressed 8 , 9 ., For example mRNA of 17 ets genes , including ETS1 , is present in Jurkat T cells ., Despite overlapping expression and sequence preferences , experimental data indicate that individual ETS proteins have unique biological functions 10–13 ., For ETS1 , mouse deletion studies indicate a critical role in T cell activation 14 ., This specific genetic function implies that ETS1 has a unique mechanism that allows it , but not other ETS proteins , to bind the promoters or enhancers of genes important for T cell activation ., Finally , ETS1 functions , in part , by recruitment of the co-activator CBP to transcriptional control regions , presumably functioning to activate genes at which it binds 15–17 ., We utilize this in depth understanding of ETS1 at both a biochemical and biological level to inform our genomic approach and facilitate functional analysis ., Initial genomic occupancy studies with ETS1 have provided insight into the genomic dilemma and led to unexpected observations ., We previously identified two modes of ETS protein targeting to promoters in Jurkat T cells using chromatin immunoprecipitation and promoter microarrays ( ChIP-chip ) ., The surprising mode is redundant occupancy in which a sequence with the consensus CCGGAAGT is associated with occupancy of three different ETS transcription factors: ETS1 , GABPA ( GABPα ) , and ELF1 ., Because this sequence is consistent with the in vitro derived consensus sequences derived from multiple ETS family members , we concluded that it can alternately recruit various ETS transcription factors ., This redundant mode of binding generally occurs in the promoters of housekeeping genes and may represent shared function of the ETS family in the maintenance of constitutive expression ., The second ETS binding mode is specific occupancy ( e . g . ETS1 , but not GABPA or ELF1 ) , which requires a GGA core motif , but is not associated with a close match to the consensus ETS sequence ., We proposed that specific targets would mediate the specific biological functions of each ETS transcription factor ., However , the promoter-limited approach did not identify a significant correlation between the specific targets of ETS1 and genes important for the role of ETS1 in T cell activation ., Full investigation of this provocative dual role of the ETS family required an expansion to full genome analysis ., In this study we identified regions across the entire human genome occupied by ETS1 , a DNA binding partner RUNX , and co-activator CBP in Jurkat T cells to decipher sequence determinants and investigate the biological significance of sequence diversity ., We discovered a previously undescribed role for ETS1 at a large number of enhancers ., Enhancer occupancy of ETS1 was associated with a unique variant of the ETS binding site and in vitro DNA binding assays illustrated how this variant sequence functions as an ETS1 specificity determinant ., Enhancers co-occupied by ETS1 and RUNX contained a variant ETS sequence closely juxtaposed to a RUNX binding site – a composite sequence identical to that found in the T cell receptor enhancers ., These distinct enhancer sequences contrasted with prior observations of sequences at ETS1 bound promoters ., Importantly , ETS1 bound regions that contained the ETS/RUNX composite sequence were near genes important for T cell activation , thus establishing a tissue-specific , genomic dataset for a factor partnership ., Furthermore , ETS1 was closely associated with CBP occupancy at ETS1 specific enhancers , but not at redundantly occupied promoters ., By using genomic datasets for DNA binding factors , in addition to correlates of DNase I sensitive regions , histone marks , and co-factor binding , we decoded the functionality of in vivo binding sequences ., High-throughput sequencing coupled with chromatin immunoprecipitation ( ChIP-seq ) facilitates genome-wide searches for transcription factor binding sites 1 ., We detected 19 , 420 bound regions at an empirical false discovery rate of <0 . 01 for ETS1 in Jurkat T cells using this approach ., This included almost all ( 94% ) of the 1086 ETS1 bound promoters previously identified by ChIP-chip 18 plus an additional 6116 promoters , indicating a potential for higher sensitivity ., ETS1 bound promoter regions centered within 500 bp of a transcription start site ( TSS ) ( Figure S1 ) ; therefore , a 500-bp limit was used for promoter definition ., A large number of regions , 12 , 283 , were not in promoters ., We sought to establish the validity of these potential enhancer regions by comparison to other types of genome-wide datasets ., One powerful dataset from primary CD4+ T cells ( thus comparable to the CD4+ Jurkat cell line ) identifies DNase I accessible regions as mapped by high-throughput sequencing and ChIP-chip 19 ., Based on the long history of linkage of DNase I sensitivity to enhancers , we screened ETS1 bound regions for overlap ., 76% of ETS1 occupied regions overlapped with DNase I sensitivity ., ( Overlap was 98% for sites proximal to a TSS and 64% for distal sites . ), This represents a significant enrichment over the mean 4% overlap with datasets randomly derived from control sequences ( P<0 . 001 , Figure 1A ) ., Randomly selected DNase I sensitive , ETS1 bound regions were verified by quantitative PCR as ETS1 occupied ( 13 of 15 ETS1 bound , Figure S2 ) , whereas regions that were not DNase I sensitive included many apparent false positives ( 0 of 8 ETS1 bound , Figure S2 ) ., This strong correlation not only helped validate the ETS1 data , but also suggested that DNase I sensitivity is a strong correlate of robust ChIP signals ., We proposed that the 14 , 824 ETS1 bound regions that overlap DNase I sensitive regions represent functional regions , and only these were considered in further analysis ., Datasets for histone marks in primary CD4+ T cells 20 also provide a measure of the activity of promoters and enhancers and test for relevance of factor binding ., For example , H3K4 tri-methylation correlates with active promoters 21 ., 86% of ETS1 bound promoters had H3K4 tri-methylation compared to 28% of promoters without ETS1 ( P<0 . 0001 ) ( Table 1 ) ., Likewise , H3K4 mono-methylation is enriched in enhancers 21 , 22 ., For DNase I sensitive regions distal to a TSS , 52% with ETS1 , but only 18% without ETS1 , had an H3K4 mono-methyl mark ( P<0 . 0001 ) ( Table 1 ) ., Therefore , ETS1 occupancy is enriched at regions with histone marks that are indicative of enhancers and active promoters ., Contrary to the expectation that family members with unique genetic functions would have exclusive binding sites , we previously discovered the majority of proximal promoters bound by ETS1 , are also occupied redundantly by other ETS proteins ( e . g . GABPA and ELF1 ) ., The small number of ETS1 specific sites identified limited the robustness and fruitfulness of further analysis of this class of targets 18 ., Using a much larger ChIP-seq dataset that includes enhancer regions , we were poised to identify targets that mediate specific functions of ETS1 ., However , we first had to identify which of the 14 , 824 ETS1 bound regions were redundant sites , thus not strong candidates to mediate specific functions ., We analyzed genome-wide occupancy data reported for GABPA in Jurkat cells 23 with the same methodology as our ETS1 analysis ., There were 7724 GABPA bound regions , of which 6443 were redundantly bound by ETS1 ( Figure 1B ) , illustrated graphically on a genome section in Figure 1C ( left ) ., The remaining 8381 ETS1 specific regions were exemplified by the T cell receptor ( TCR ) α and β enhancer loci ( Figure 1C , right ) , which display ETS1 , not GABPA , occupancy ., 67% of GABPA and ETS1 co-occupied regions were proximal to a TSS consistent with previous findings that discovered this dominant class of promoter binding events 18 ., In contrast , 68% of ETS1 specific regions were distal to the nearest TSS indicating enhancer regions and are candidates to mediate the specific functions of ETS1 ., Genetic experiments have implicated ETS1 in T cell activation; therefore , we predicted that genomic sites that specifically bind ETS1 should be associated with genes necessary for T cell function ., The genes nearest to distal ETS1 bound regions were assumed to be a reasonable estimate of ETS1 regulated genes ., This gene set was compared to genes with 20-fold higher mRNA expression levels in CD4+ T cells than the median expression in multiple cell types ( 287 genes ) and , as a control , to 268 genes specific to pancreas ( Figure 2 ) ., Compared to all genes , or pancreas specific genes , genes that displayed T cell–specific expression were more likely to be near one or more distal ETS1 bound regions ., Furthermore , as the number of nearby distal ETS1 bound regions increased , the difference between the T cell–specific categories and the control categories became more apparent ., Thus , distal ETS1 binding was associated with a tissue-specific role of ETS1 in T cells , further validating the functionality of ETS1 specific regions ., At this point we had a dataset of ETS1 bound regions that tracked with distal enhancer marks and T cell function ., We sought to determine whether such bound regions displayed a unique sequence that would be responsible for ETS1 specific binding ., Unbiased searching for overrepresented sequences was performed with the MEME algorithm 24 ., ETS1 bound regions were grouped according to proximity to the nearest TSS ( proximal versus distal ) and specificity ( redundant: overlap with GABPA; specific: no overlap with GABPA ) to provide experimental and control datasets ., The most over-represented sequence motifs in redundant , proximal regions ( Motif, 1 ) and specific , proximal regions ( Motif, 2 ) were identical to the motifs previously identified in redundant and specific promoter proximal regions 18 ., In contrast , analysis of distal , specific ETS1 bound regions identified a third , distinct motif ( Motif, 3 ) as the most over-represented ( Figure 3 ) ., The major differences from the ETS family consensus ( CCGGAAGT ) present in Motif 1 were the almost exclusive presence of an A at the second position and the inclusion , in some instances , of a T at the sixth position ( CAGGA ( A/T ) GT ) ., Therefore , specific ETS1 binding to enhancers is associated with a sequence distinct from those found at ETS1 bound promoters ., A more directed bioinformatics approach assessed the importance of single nucleotide changes from the ETS consensus for enhancer and promoter occupancy of ETS1 ., ETS1 bound regions were partitioned into equally-sized sets of ETS1/GABPA redundantly occupied promoters and ETS1 specific enhancers ., The number of occurrences of 8-mer sequences was reported relative to the number of occurrences expected in a set of equally-sized random sequences ( Table 2 ) ., The ETS consensus sequence ( CCGGAAGT ) was enriched in redundant promoters , but not in ETS1 specific enhancers ., The enrichment of every possible single nucleotide change to the ETS consensus was then determined ., Only the change of the C at the second position to an A ( CaGGAAGT ) resulted in a significant enrichment ( P<0 . 0001 ) in ETS1 specific enhancers ., However , this sequence was also enriched in redundant promoters ., The A to T change at the sixth position ( CCGGAtGT ) reduced significantly the enrichment in redundant proximal regions ( P<0 . 0001 ) , but not distal regions ( P\u200a=\u200a0 . 3 ) ., Furthermore , the combination of both nucleotide changes ( CaGGAtGT ) resulted in enrichment at ETS1 specific enhancers ( P<0 . 0001 ) , but not redundant promoters ( P\u200a=\u200a0 . 3 ) ., Strikingly , a change of only two nucleotides inverted the enrichment pattern at redundant promoters and ETS1 specific enhancers ., We concluded that the sequence CAGGATGT is a specificity element for ETS1 ., In considering specificity within families of transcription factors the preference for a particular sequence may be due to intrinsic DNA binding properties of different family members ., To test the ability of the two nucleotide changes to act alone or in combination to select for ETS1 we measured the relative binding affinity of ETS1 and second ETS factor , ELF1 which is also reported to be active in T cells 25 ., Indeed , ELF1 is present at redundant , but not ETS1 specific promoters in Jurkat T cells in a similar manner to GABPA 18 ., Binding affinity for an ETS consensus sequence , each single nucleotide variant , and the two nucleotide variant was interrogated in vitro with purified proteins by quantitative gel shift ( Figure 4 ) ., The A to T change resulted in a loss of affinity for ELF1 ( 3 . 6-fold loss ) , but not for ETS1 ., The C to A change showed a modest effect on affinity and no discrimination between ELF1 and ETS1 ( 1 . 9-fold versus 1 . 5-fold loss ) ., In contrast , the change of both nucleotides caused an 18 . 3-fold loss of affinity for ELF1 , but only a 2 . 4-fold loss for ETS1 ., We concluded that the two nucleotide variant sequence CaGGAtGT serves as a specificity determinant for ETS1 versus ELF1 due to an intrinsic binding property of ETS1 ., Unique cooperative DNA binding between closely apposed binding proteins can also drive specific occupancy of transcription factors ., A well-characterized partnership for ETS1 is with the RUNX factors 26 , 27 ., Interestingly , the consensus derived from the most frequent nucleotides at each position of Motif 3 ( CAGGAAGTGG ) is similar to the sequences at the TCRβ ( CAGGATGTGG ) and TCRα ( GAGGATGTGG ) enhancers that support cooperative binding of ETS1 with RUNX1 through an ETS/RUNX composite sequence ( RUNX consensus YGYGGY ) ., To test whether ETS1 bound enhancers were co-occupied by RUNX factors , genome-wide occupancy of RUNX1/3 ( RUNX ) was determined ., Again , only regions that co-localized with DNase I sensitivity were considered bound ., Strikingly , 64% of the 1075 RUNX bound regions were co-occupied by ETS1 ., In contrast , only 14% of RUNX bound regions were co-occupied by GABPA ( Figure 5A ) ., 77% of the ETS1/RUNX co-occupied regions were ETS1 specific and distal to a TSS ( compared to 37% of ETS1 bound regions lacking RUNX ) , suggesting a role in T cell enhancer function ., An unbiased search with MEME for overrepresented sequence motifs in regions co-occupied by ETS1 and RUNX identified a motif ( Motif 4 ) similar to Motif 3 , but with the RUNX binding site more strongly represented ( Figure 3 ) ., Like many sequence identification algorithms , MEME is biased towards strongly preferred spacing distances between two binding sites ., To test whether other spacings of ETS and RUNX sites were also over-represented in ETS/RUNX bound regions , the distance from each ETS sequence to the nearest RUNX sequence was plotted ( Figure 5B ) ., This analysis indicated that only the spacing found by MEME was over-represented in these regions ., Therefore , ETS1 and RUNX co-occupy enhancer regions in T cells through a composite ETS/RUNX binding site similar to those found in the T cell receptor enhancers ., These findings indicate that pairing with a neighboring DNA binding motif , in conjunction with intrinsic DNA binding properties , can drive specificity ., ETS1 occupancy of enhancers is associated with T cell–specific genes ( Figure, 2 ) and ETS1 specific motifs ( Motifs 3 and 4 ) , whereas promoter occupancy is associated with housekeeping genes 18 and shows enrichment for sequences ( Motif, 1 ) that cannot distinguish family members ., The value of these motifs will be in their predictive accuracy ., All ETS1 bound regions were searched for Motif 1 and Motif 4 with PATSER 28 ., Regions containing Motif 1 were more likely to be found in promoters , and regions containing Motif 4 were more likely to be found in enhancers ( Table 3 , Table S1 ) ., Associated genes , as determined by nearest TSS , were searched for over-represented ontologies with the GoMiner program 29 ., Genes with Motif 4 were associated with T cell activation categories , whereas those with Motif 1 were associated with housekeeping ontologies ( Table 3 ) ., Therefore , each motif is predictive of the type of transcriptional control element and class of ETS1 target gene ., The emerging differences for ETS1 at promoters versus enhancers opened the possibility of distinct functions of ETS1 at these loci ., One mechanism of transcriptional activation by ETS1 is the recruitment of the co-activators CBP and p300 16 ., Identification of p300 occupancy within the 30 mb ENCODE region of the human genome revealed a greater proportion at distal sites than at promoters 21 , and p300 has been shown to mark tissue specific enhancers in mice 30 ., Thus , we proposed that ETS1 would recruit CBP/p300 to T cell–specific enhancers , but not promoters ., Genome-wide occupancy for CBP detected 14 , 374 CBP bound/DNase I sensitive regions in Jurkat T cells ., CBP bound regions showed a surprisingly high overlap with ETS1 bound regions at both redundant promoters ( 75% , P<0 . 0001 ) and ETS1 occupied enhancers ( 68% , P<0 . 0001 ) compared to regions not bound by ETS1 ( Table 1 ) ., The strong presence of CBP corroborated the general functionality of ETS1 binding sites ., Due to the unexpected equivalence of CBP overlap at both enhancers and promoters we investigated the connection between CBP binding and ETS1 function by a more detailed mapping method that presented the two types of sites as a class average ( Figure 6A ) ., At redundantly occupied promoters , the location of ETS1 , CBP , GABPA , H3K4 tri-methylation , and Motif 1 were plotted relative to the TSS ., At ETS1 occupied enhancers the location of CBP , RUNX , H3K4 mono-methylation , and Motif 4 were plotted relative to the center of the ETS1 bound region ., At promoters ETS1 and GABPA binding were coincident with the consensus ETS binding site at a position 25–30 bp upstream of the transcription start site ., This extremely TSS proximal location and the location of histone H3K4 tri-methylation on either side of the ETS1 bound region indicated that redundant ETS binding occurs in the nucleosome-free region 31 ., CBP occupancy was co-incident with the downstream H3K4 tri-methyl , but not ETS1 and GABPA binding , suggesting that CBP is not directly bound by ETS factors at promoters ., In contrast , at enhancers , CBP , ETS1 , and RUNX binding overlapped , suggesting that ETS1 and/or RUNX may directly bind CBP ., Again , ETS1 occupied a region between histone marks , in this case H3K4 mono-methyl , indicating that ETS1 binds between nucleosomes ., To test directly whether ETS1 was necessary for occupancy of RUNX and CBP , ETS1 protein levels were knocked-down by two independent shRNAs and occupancy was monitored by ChIP ( Figure 6B and 6C ) ., Decreased ETS1 protein levels correlated with a loss of ETS1 , CBP , and RUNX ChIP enrichment at the TCRβ enhancer ( containing Motif 4 ) ., We concluded that ETS1 is critical for recruitment or stable binding of CBP in enhancers important for T cell activation ., In contrast , reduction of ETS1 occupancy in a redundantly occupied promoter ( containing Motif, 1 ) did not affect CBP enrichment ., Thus , distinct sequence motifs at ETS1 binding sites correlate with not only different types of regulatory elements , but also distinct histone marks and co-activator binding ., We conclude that these sequences mediate unique functions of ETS1 ., Selecting regions at which ChIP-seq enrichment coincided with DNase I sensitivity improved the accuracy of a dataset of transcription factor bound regions ., The fraction of regions removed ( 24% of ETS1 bound regions , 39% of CBP bound regions and 70% of RUNX regions ) may reflect the quality of the antibodies used for ChIP-seq ., Removed regions had ETS1 binding properties ( presence of binding motifs , correlations with other factors and histone marks ) , but at markedly lower levels than retained regions ., Thus , we propose that the use of DNase I sensitivity screening improves the quality of a ChIP-seq dataset and may be particularly useful for the interpretation of data generated with suboptimal antibody reagents ., The genome-wide set of ETS1 binding sites showed sequence variants that distinguish enhancer versus promoter binding events ., Specific ETS1 occupancy of enhancers was associated with a sequence that varies by two nucleotides from the ETS consensus sequence used for redundant binding at promoters ., Our bioinformatics analysis indicated that these two nucleotide changes are not equivalent ( Table 2 ) ., The C to A change at the second position appeared to be required for specific ETS1 binding to enhancers , but also occurs at redundant promoters ., In contrast , the A to T change at the sixth position appeared to restrict occupancy of redundant promoters , but not specific enhancers ., The A to T change has previously been shown to provide specificity for ETS1 versus the ETS protein ELF1 in vitro 32 ., Our in vitro comparison confirmed the role of this single nucleotide change and identified a dramatic specificity difference between ETS1 and ELF1 when both nucleotides were changed ( Figure 4 ) ., However , the in vitro data did not explain why the C to A change alone appeared necessary for genomic enrichment in ETS1 specific enhancers ( Table 2 ) ., Therefore , the nucleotide preferences at these ETS binding sequences are likely due to a combination of the intrinsic differences in DNA binding attributes of ETS proteins and other in vivo factors ., A striking difference between the ETS consensus sequence , CCGGAAGT , and the C to A variant , CaGGAAGT , is the susceptibility to DNA methylation ., Indeed , methylation of this CpG dinucleotide within the consensus has been shown to block the binding of ETS proteins 33 , 34 ., We have previously observed a very strong correlation between redundant ETS occupancy of promoters , the sequence CCGGAAGT and CpG islands 18 ., The CpG islands at housekeeping promoters are generally hypomethylated , whereas CpG dinucleotides distributed in lower density throughout the genome are likely to be methylated 35 , 36 ., Thus , ETS sites may be shielded from methylation at CpG island-containing promoters , but not at enhancers ., Therefore , the C to A change in the ETS binding sequences at enhancers may have evolved to protect these sites from the repressive effects of DNA methylation ., Because other transcription factors whose binding sites bear a CpG dinucleotide ( NRF-1 , BoxA , SP1 , CRE , and E-box ) are also enriched in housekeeping promoters 37 , the use of an alternate sequence in tissue-specific enhancers may also extend to these transcription factor families ., Another factor that might influence the ETS1 binding sequence observed in vivo is the presence of closely juxtaposed binding sites for other transcription factors ., A subset of the ETS1 specific enhancers were co-occupied by RUNX and had a composite ETS/RUNX binding sequence ( Motif 4 ) ., In the context of this sequence , the A to T change at the sixth position of the ETS sequence allows the RUNX sequence to be a closer match to the RUNX consensus ( YGYGGT , sixth position underlined ) ., This factor could contribute to the enrichment of the A to T change in specific enhancers ., However , a T at the sixth position was no more likely in regions co-occupied by ETS1 and RUNX ( Motif 4 ) than at ETS1 specific enhancers in general ( Motif 3 ) ., This indicates that either an A or a T at this position can support ETS1 and RUNX co-occupancy ., Furthermore , only 55% of the ETS1 bound regions that had the sequence CAGGATGT , had the full ETS/RUNX composite CAGGATGTGG ., We propose that the remaining 45% of regions recruit ETS1 either only through the ETS binding site , or in cooperation with other unidentified transcription factors ., In conclusion , the sequences associated with ETS1 specific occupancy of enhancers reflect intrinsic differences in DNA binding or interactions with other factors and may be influenced by susceptibility to DNA methylation ., Mice with an ETS1 gene disruption have reduced numbers of NK and NKT cells and show defects in T cell activation 14 , 38 , 39 ., RUNX genes are essential for NK , NKT , and T cell differentiation 40–42 ., However , the role of ETS1 and RUNX in these immune functions has not been fully understood on the level of individual target genes ., Our data suggest that ETS1 and RUNX regulate genes important for T cell activation pathways by direct occupancy of nearby enhancers via a particular ETS/RUNX binding site ., The gene categories presented in Table 3 suggest that the primary role of these transcription factors is not the direct activation of genes downstream of T cell receptor signaling , but rather the control of expression of the signaling machinery ., This conclusion is consistent with the finding that ETS1 null T cells are defective in activation upon receptor stimulation , but respond normally to pharmacological stimulation , which bypasses membrane proximal signaling events 39 ., The strongest determinant of ETS1 specificity , Motif 4 , fixed the sequence of both ETS1 and RUNX binding sites as well as spacing and relative orientation of the two sites ( Figure 3 and Figure 5 ) ., This strict conservation was somewhat surprising because ETS1 and RUNX1 bind to DNA cooperatively in vitro at a variety of other spacings and orientations 26 , 43 ., Alternate spacing can also function in transcription activation in vivo ., For example , the MMLV enhancer is activated by ETS1 and RUNX1 at a sequence in which the ETS and RUNX sites are four nucleotides further apart than in Motif 4 44 ., Furthermore , Motif 2 can also support ETS1 and RUNX1 cooperativity in vitro 18 ., Motif 2 utilizes much more divergent ETS and RUNX sequences set two nucleotides further apart than in Motif 4 ., ( Our analysis in Figure 5 does not identify this spacing because these sequences are too divergent from the Transfac ETS1 and RUNX motifs . ), However , only Motif 4 , not Motif 2 or the MMLV enhancer motif , was associated with ontologies aligned with T cell–specific functions ( Table 3 , and data not shown ) ., We speculate that this spacing could have a function in addition to the simple recruitment of ETS1 and RUNX ., This may reflect a requirement for a specific conformation of ETS1 and RUNX for the transcriptional activation function of these enhancers ., For example , because both ETS1 and RUNX can bind the co-activator CBP 15 , 16 , 45 , and CBP occupies the same position as ETS1 and RUNX at enhancers ( Figure 6A ) , cooperative CBP recruitment may require this particular configuration of ETS1 and RUNX ., This first picture of genome-wide occupancy of a transcription factor in combination with the co-factor CBP presented surprising diversity ., In spite of the general picture that CBP occupancy was strongly correlated with ETS1 binding at both tissue specific enhancers and at active promoters in Jurkat T cells ( Table 1 ) , fine mapping uncovered a more complex picture ( Figure 6A ) ., The coincident binding observed at enhancers and the sensitivity of CBP occupancy at the TCRβ enhancer to an ETS1 knockdown ( Figure 6C ) is consistent with direct recruitment of CBP by ETS1 and/or RUNX ., These data are consistent with reports that CBP/p300 has a strong correlation with enhancers 21 , 30 , 46 ., This not only supports the functionality of ETS1 bound distal enhancers , but also strongly demonstrates the role of DNA factors in CBP recruitment ., In contrast , the lack of concordance of ETS1 and CBP binding events at promoters suggests that CBP is associated with other factors at these sites ., Potential CBP recruitment mechanisms include interaction with general transcription factors 47 and enhancer-promoter looping 48 , 49 ., Either of these mechanisms could contribute to the location of CBP at ETS1 bound promoters ., We note , however , that the CpG island-containing promoters of housekeeping genes , at which we observe redundant ETS occupancy , are thought to lack enhancers ., Thus , we suggest that CBP is brought to these promoters by enhancer-independent interactions with the transcriptional machinery ., One possibility is that ETS1 participates in recruitment , but maintenance at these constitutively active sites relies on cooperation with basal machinery or modified histones ., Like many cellular proteins , transcription factors can have multiple roles that vary based on cell type and condition ., Transcription factor function can also vary based on the context of other proteins present at each genomic locus ., Here we show that the type of genes that are near ETS1 binding events , and the location of the co-activator CBP differ based on the sequence that recruits ETS1 to DNA ., Thus , two different roles of ETS1 in T cells – a role at housekeeping promoters , and one at tissue specific enhancers – can be defined by distinct sequence motifs ., The sequence variation for different functions of a transcription factor provides an explanation for the lack of a single binding sequence in many genome-wide occupancy studies ., Our investigation provides a route to sort genome-wide binding data by the presence of such sequence motifs and other correlative data to define the distinct functions of a transcription factor ., ChIP from Jurkat T cells was performed as described previously 18 ., In brief , 5×107 cells were crosslinked with 1% formaldehyde and sheared chromatin extract was prepared ., Dynabeads ( Invitrogen ) coupled to the appropriate secondary antibody were used to immunoprecipitate extracts treated with one of the following antibodies; polyclonal ETS1 , sc-355; polyclonal CBP , A-22; ( Santa Cruz Biotechnology ) , or monoclonal RUNX , α3 . 2 . 3 . 1 ., Crosslinks were reversed by heating and DNA was purified ., Input controls were prepared in parallel , but with no immunoprecipitation step ., qPCR anal

Introduction, Results, Discussion, Materials and Methods

To elucidate how genomic sequences build transcriptional control networks , we need to understand the connection between DNA sequence and transcription factor binding and function ., Binding predictions based solely on consensus predictions are limited , because a single factor can use degenerate sequence motifs and because related transcription factors often prefer identical sequences ., The ETS family transcription factor , ETS1 , exemplifies these challenges ., Unexpected , redundant occupancy of ETS1 and other ETS proteins is observed at promoters of housekeeping genes in T cells due to common sequence preferences and the presence of strong consensus motifs ., However , ETS1 exhibits a specific function in T cell activation; thus , unique transcriptional targets are predicted ., To uncover the sequence motifs that mediate specific functions of ETS1 , a genome-wide approach , chromatin immunoprecipitation coupled with high-throughput sequencing ( ChIP-seq ) , identified both promoter and enhancer binding events in Jurkat T cells ., A comparison with DNase I sensitivity both validated the dataset and also improved accuracy ., Redundant occupancy of ETS1 with the ETS protein GABPA occurred primarily in promoters of housekeeping genes , whereas ETS1 specific occupancy occurred in the enhancers of T cell–specific genes ., Two routes to ETS1 specificity were identified: an intrinsic preference of ETS1 for a variant of the ETS family consensus sequence and the presence of a composite sequence that can support cooperative binding with a RUNX transcription factor ., Genome-wide occupancy of RUNX factors corroborated the importance of this partnership ., Furthermore , genome-wide occupancy of co-activator CBP indicated tight co-localization with ETS1 at specific enhancers , but not redundant promoters ., The distinct sequences associated with redundant versus specific ETS1 occupancy were predictive of promoter or enhancer location and the ontology of nearby genes ., These findings demonstrate that diversity of DNA binding motifs may enable variable transcription factor function at different genomic sites .

Genomes contain sequences that encode both gene products and the instructions for where and when each gene is expressed ., This gene expression code is critical for normal development and goes awry in disease processes such as cancer ., The gene expression code is interpreted by proteins called transcription factors that bind to particular DNA sequences and carry instructions for gene activation or repression ., This recognition code is challenged by the presence of highly-similar transcription factors that prefer almost identical DNA sequences ., In addition , studies in living cells indicate that individual transcription factors have significant flexibility in sequence recognition ., Here , we identify thousands of positions in the genome of human T cells that are bound by the transcription factor ETS1 ., These data , along with comparisons to other genomic datasets , allow us to identify DNA sequences that specify ETS1 binding while excluding binding of other related transcription factors ., Furthermore , we discover that ETS1 binds more than one sequence and that these sequence variants can predict distinct biological functions of ETS1 ., Thus , this work contributes to our understanding of the gene expression code by addressing both how a transcription factor can bind unique genomic locations and why a transcription factor binds multiple DNA sequences .

molecular biology/chromatin structure, biochemistry, cell biology/leukocyte signaling and gene expression, molecular biology/histone modification, molecular biology/transcription initiation and activation, genetics and genomics/gene expression, genetics and genomics/genetics of the immune system, evolutionary biology/genomics, biochemistry/bioinformatics, biochemistry/transcription and translation, genetics and genomics/bioinformatics

journal.pcbi.1000686

Spatial Simulations of Myxobacterial Development

Bacteria are able to sense their surroundings in order to adapt to environmental change ., Most bacteria live in dense populations , therefore other cells constitute a major part of their physical and chemical environment allowing regulatory interactions between cells to be established ., The benefits of coordinated behaviour include: more efficient proliferation resulting from a cellular division of labour , access to resources that cannot be utilised by isolated cells , defence against antagonists and population survival by differentiation into distinct cell types 1 ., Myxobacteria are Gram-negative , ubiquitous , soil dwelling bacteria that are semi-flexible , and rod-shaped ., Cells glide across a surface using the adventurous ( A ) and the social ( S ) motility systems 2 ., S-motility is coordinated at the leading pole; cells extend type IV pili which can adhere to the surface of other bacteria or polysaccharides , and upon retraction the cell is pulled forward ., A-motility is coordinated at the lagging pole; cells are thought to extrude a slime which expands and generates a propulsive force to push cells forward 3 , 4 ., Myxobacteria display distinct social phenotypes and multicellular behaviours ., Myxococcus xanthus is the most commonly studied species of myxobacteria ., In response to starvation , cells undergo multiple phases of behaviour culminating in the formation of fruiting bodies and myxospores ., The developmental process involves a series of macroscopic changes in colony morphology ., A key regulator of development is C-signal ling which occurs when C-signal , a cell surface-associated signal encoded by csgA , is exchanged between cells in close contact with one another ., C-signal stimulates the expression of csgA leading to a rise in C-signal ling throughout development from positive feedback ., Different colony morphologies are a consequence of different C-signal ling levels 5 ., C-signal ling is thought to affect the reversal frequency of individual cells in a contact-dependent fashion allowing the synchronisation of behaviour 5–7 ., During vegetative growth cells move in the direction of their long axis , reversing typically once every ten minutes 8 ., Under starvation conditions , C-signal accumulates within a cell 5 reducing its reversal frequency 9 ., The reduction in the reversal frequency and the effects of A and S motility causes cells to form streams and increases the likelihood of aggregation; cells which cannot reverse tend to remain stuck in one location since their ability to move around obstacles is limited by only being able to move forward 10 ., M . xanthus cells begin to form fruiting bodies after a prolonged starvation period of approximately 24 h ., Starved cells form into large , intricate multicellular aggregates containing between 50 , 000 and 100 , 000 cells 11 ., The fruiting body is the precursor to sporulation where cells undergo morphogenesis and physically change shape from rods to nearly spherical cells 12 ., Inside the nascent fruiting body , a percentage of the cells differentiate into dormant myxospores ., This process requires both temporal and spatial coordination in three dimensions , making it one of the most complex and least understood phases of the life-cycle ., Relatively little is known about the spatial dynamics of fruiting body construction with research primarily devoted to understanding the signalling mechanisms required to coordinate development rather than the actual physics 13 ., There is some disagreement over how fruiting actually begins ., OConnor and Zusman 11 , 14 observed that cells appear to orbit around a largely stationary aggregation centre ., This led to the traffic jam model , which proposes that streams of cells collide together causing the formation of a kernel of stationary cells ., Cells move around and over the static centre leading to a mound formation 15 ., Work on Stigmatella fruiting body formation showed that cells form circular orbits around a base and then move upwards in a spiral fashion around the base , building the stalk on top of it 14 , 16 ., It was presumed other myxobacteria , including M . xanthus , form fruits in a similar way; however , Kuner and Kaiser 17 did not observe the spiralling patterns suggesting that this behaviour is possibly non-essential and may not be intrinsically important to fruiting development ., Recent work by Curtis et al . 13 suggests that fruiting bodies are formed using a stepped layer building approach; large streams of cells forming sheets collide causing a rapid build up in density at the meeting point ., Cells in one of the opposing streams are forced upwards and over the other , similar to tectonic plate movements ., The displaced cells are supported on top of the dense layer of cells and extra-cellular polysaccharide ( EPS ) underneath and begin to spread out forming a new layer ., As the new layer becomes more dense itself , cells at the centre start to get pushed upwards to form a new layer and the process repeats causing the formation of an expanding mound of cells ., Previous computational models of fruiting body development 18–20 are based on the orbiting traffic jam model and rely upon the artificial induction of an aggregation centre to start fruiting body development , typically by making a subset of cells stationary ., In this paper , we take a different approach and use an off-lattice Monte Carlo simulation to show how cells can spontaneously aggregate to form layers and fruiting bodies based on the observations of Curtis et al . 13 ., The motivation of this work is to gain an increased understanding of fruiting , by examining the physical properties driving cells to engage in fruiting , using mathematical and computational modelling ., Periodic boundary conditions were disabled in the -plane since it does not make sense for cells to be able to push through the floor nor move through the ceiling for which there is no physical interpretation ., Boundaries are maintained with a boundary energy term which severely penalises a cell for attempting to cross a particular domain boundary ., The energy penalty is several orders of magnitude larger than the value any of the other energy terms might produce so it is nearly impossible for a configuration with these domain crossings to be favourable ., Fruiting body formation requires a highly dense region of cells to seed aggregation ., To achieve such a density at the start of simulation would require cells to be placed so that they fill all available space on the floor of the simulation volume ., Even under these conditions , the cell density is usually not sufficient to seed fruiting , and the lack of space for movement would inhibit cell motility ., Biologically , fruiting bodies are thought to form from the confluence of streams of cells resulting in the cell density increasing over time 4 , 22 ., To capture this behaviour in the simulations , entry zones were placed around the edges of the simulation volume ( see Figure S3 ) ., Entry zones allow new cells to be introduced into the simulation over time to model cell influx ., A maximum influx rate ( ) can be specified to govern how quickly new cells enter the simulation volume ., The actual influx rate is stochastic and less than the maximum influx rate , , and is determined by the amount of free space within the entry zones where new cells can be placed ., New cells are placed at random locations by periodically sampling the entry zones to see if there is free space to place a cell and then placing a cell if the maximum influx rate ( ) has not been exceeded ., Fruiting requires a high cell density and simulating a finite number of cells makes it problematic to assemble enough cells in an area to form a fruit; the cell density is never high enough ., A finite number of cells may clump and partially aggregate but they are unlikely to form a fruiting body ., With the entry zone model , a constant cell density can be maintained to sustain fruiting body growth ., Cell reversals are thought be controlled by C-signal stimulating the complex Frz pathway , however the exact function of each component has yet to be determined 23 ., We therefore model the macroscopic behaviour of the pathway , where an internal phase switch is used as an abstract representation of C-signal ., The switch increments until it reaches at which point it resets and the cell reverses ., The switch can be perturbed by signalling between neighbouring cells to make reversals happen more quickly , by a factor proportional to the number of collisions a cell experiences with its neighbouring cells ., The function is therefore: ( 1 ) ( 2 ) where is the new cumulative value , is the current value , is a basal increase factor , is the signal strength , and is the level of C-signal ling a cell experiences at time , defined by the collisions a cell experiences with each of its neighbours and a collision factor ., In this work we keep the model of C-signal ling quite simple , as our goal is to explore other factors which can facilitate the formation of aggregates and fruiting bodies ., Experiments indicate that even 15 hours into starvation , levels of C-signalling are sufficient to reduce the rate of reversals to once every 22 minutes 24 ., Moreover , these experiments show that the slowdown in reversal induces a 15-fold increase in travel distance , in what could be considered a ‘unidirectional behaviour’ ., We approximate this low frequency of reversals by considering cells which have come near to an aggregation as non-reversing , reflecting the approach taken in other simulations 18 , 19 ., Nevertheless , in simulations of fruiting C-signal ling levels and collisions are monitored , enabling the imposition of a threshold C-signal ling level governing the induction of sporulation ., Figure 1 describes the program used for simulation ., The Metropolis algorithm 25 is used to determine the acceptance probability of making any particular change ., Simulations were carried out using a volume equivalent to m ., The model captures the physical dynamics of the cells using the method proposed by Glazier and Graner 26 ., A Cellular Potts Model is a probabilistic Cellular Automata with Monte-Carlo updating , where the next state of the lattice is chosen by evaluating a Hamiltonian equation used to calculate the probability of accepting lattice updates ., The original Potts model 27 was developed to capture behaviour at the level of statistical mechanics but has been successfully generalized for a variety of domains ., The tuning of a Cellular Potts Model is based on finding an appropriate Hamiltonian function and appropriate parameters for this function ., The heart of our model is the development of a set of terms that correctly describes important physical characteristics of the M . xanthus cell ( see Figure 2 ) ., The level of detail used needs to be balanced with the computational cost of these calculations ., The following Hamiltonian function , inspired by the approach of Izaguirre et al . 28 , describes the energy components of M . xanthus we use: ( 3 ) A separate collision resolution algorithm such as used by Wu et al . 29 was not required since collision avoidance is a feature of the Hamiltonian ., In the following presentation of each of the components of the Hamiltonian , we use boldface fonts to indicate vectors , and the cap operation to denote an average or mean vector ., M . xanthus cells are modelled as having a finite volume and stable shape; cells can be squashed to an extent but they maintain a rod shaped structure except during sporulation ., Cell length governs a cells length and is analogous to the spring constant in Hookes Law ., ( 4 ) where is a dimensionless stretching coefficient , is the number of segments in cell , is the optimal distance between segments , is the vector position of segment in cell and a dimensionless stretching coefficient ., Stretching energy is defined as a squared sum which compares the distance between the centres of neighbouring segments and to and penalises a cell for allowing segments to get either too close or too far apart ., In close proximity , cells tend to align with each other reflecting the effect of the S-motility engine ., Cells extend Type IV pili from their leading pole which grab onto neighbouring cells ., Upon retraction this pulls a cell closer to the neighbour it latched onto 4 ., The natural consequence of this movement is the alignment of cells 30 ., ( 5 ) ( 6 ) ( 7 ) ( 8 ) ( 9 ) where is a dimensionless alignment coefficient , is the normalised average direction of the cell , is the average direction of all the cells in a local neighbourhood surrounding cell ., reflects that cells tend to turn through the acute angle to align with other cells in either direction ., Each cell in the model has a semi-flexible body which must maintain a certain stiffness , otherwise the cell would fold up upon itself ., Incorporating bending energy in the Hamiltonian ensures that the radius of curvature of a cell does not exceed a threshold causing the cell to flail uncontrollably and unnaturally ., ( 10 ) ( 11 ) ( 12 ) ( 13 ) ( 14 ) ( 15 ) ( 16 ) where is a dimensionless bending coefficient , returns the angle between and , is the average direction of segment of cell and is the vector between the segment ahead of ( ) and the segment behind ( ) ., The A-motility system provides myxobacteria cells with propulsion ., Cells extrude a polysaccharide slime from nozzles at their lagging pole , which is thought to expand when hydrolysed and push a cell forward 4 ., This effect is modelled using a propulsion term which causes cells to move preferentially in the average direction of the cell simulating the slime pushing a cell along ., ( 17 ) ( 18 ) ( 19 ) where is a dimensionless propulsion coefficient , is the normalised average direction of the cell and is the update direction of segment of cell ., Each segment moves towards where its head segment was previously , unless this causes segments to become unaligned ., As well as extruding slime to move , cells can also detect slime trails left by other cells and preferentially follow them ., This allows cells to follow other adventurous cells and leads to the formation of streams that can break away from the main colony ., Slime following is complementary to A-motility ., As each cell moves , it deposits a slime trail ., Early evidence of the presence and effect of slime trail following is provided by the videos created by Reichenbach 31 ., This effect of slime following is represented in our model as a set of normalised vectors representing the average direction of a cell ., The slime ages over time and is eventually removed ., Cells can sense slime trails within a limited neighbourhood around them ., Using a weighted sum of the all slime trail directions based upon their age , the average slime direction is calculated and cells preferentially follow that ., We use a weighted sum to account for the fact that a cell is more likely to follow a large slime trail than a small one ., ( 20 ) ( 21 ) ( 22 ) ( 23 ) ( 24 ) where is a dimensionless slime following coefficient , is average direction of the slime trails in a neighbourhood and is the normalised direction vector of the slime trail at location ., Curtis et al . 13 observe that cells appear to move in sheets towards each other and , upon impact during a collision , cells from one sheet can move up and on top of the other ., This is consistent with OConnor and Zusman 11 who suggest that cells appear to behave as independent sheets ., This effect has been modelled so that it is somewhat analogous to a snow plow , which is pushed forwards into the snow pushing the snow up and away ., In a similar way it is proposed that the oncoming force of a sheet of cells is sufficient to push oncoming cells up and direct them over and on top ., Each cell monitors the number of head-on collisions it has , and the more the collisions the greater the chance of it being pushed up ., Cells are not forced to always be pushed upwards , as this would be imposing an artificial constraint on the system , instead cells prefer regions of lower cell density where they are freer to move ., Some cells will be pushed outwards away from the stream , but the majority will be pushed upwards since this is the only region of free space available ., Curtis et al . 13 propose that when two sheets of oncoming cells encounter each other , individual cells have a proclivity to move out of the potential “traffic jam” that can ensue and typically this is upwards so one sheet of cells effectively moves over the other ., A simulation of climbing cells which form layers can be seen in Figure 3 ., The energy term we use , described below , encourages cells to move upwards , proportionally to the number of oncoming cells they interact with ., ( 25 ) ( 26 ) ( 27 ) ( 28 ) ( 29 ) where is a dimensionless climbing coefficient , determines the number of oncoming cells , determines if two cells are moving in opposing directions by examining the dot-product between the normalised average direction ( ) of each pair of interacting cells , and compares the direction cell tries to move in ( ) with a normal vector ( typically a normal to the -plane ) ., In a three-dimensional model , cell movement in the -axis needs to be controlled so that cells do not randomly climb into empty space and defy gravity ., The other energy terms do not prevent cells from climbing so gravity is therefore introduced as an energy penalty for trying to climb; the steeper the climb the greater the penalty ., An object acting under gravity requires the greatest amount of energy to directly oppose the force and move in the opposite direction ( upwards ) ., It should be noted that the use of the dot-product means there is no net effect of this term for a cell moving horizontally in the -plane; since gravity is a constant , there is no change in energy from moving between two positions with a direction vector perpendicular to the direction of the gravitational force ., ( 30 ) ( 31 ) where is a sensitivity parameter , is the normalised update direction of the head segment , a normalised direction vector pointing towards the ground , is a location below the centre of segment of cell and is a local neighbourhood surrounding ., In order to capture natural elasticity and bending , each cell is modelled as a number of segments each with a finite volume ., Segments must exert a repulsive force between themselves to prevent cells colliding ., This force contributes to the Hamiltonian as follows ., ( 32 ) ( 33 ) where is the position of segment of cell and is the minimum distance allowed between segments of difference cells ., The collision energy compares the distance between a segment and the neighbouring segments around it and severely penalises a cell for getting too close to another ., Although the centres of segments cannot occupy the same space , a small overlap is allowed to model deformation effects of cells in close proximity ., This is required because of the rigid segment shape which would otherwise not allow for this type of effect ., Extracellular polysaccharide ( EPS ) secreted by the cells during aggregation formation appears to play an important role in the formation of the physical structure of the fruiting body 11 , 17 ., The exact role of the slime has yet to be elucidated due to the methods used to collect data and the very high cell densities within the fruit , making it difficult to resolve individual cells ., Electron microscopy can resolve cells at higher resolutions 11 , 16 but this can only take a snapshot of a dynamic process and is unsuitable for tracking cells over a relatively long time period ., In a dense region , cells generate a lot of EPS with a fruit being a large amalgamation of cells within an EPS matrix ., The EPS is likely to exert a surface tension effect causing cells to stick together rather than drifting apart ., This is separate from the slime trail following effect as it is non-directional , acting over the whole cell area ., If two cells are close to each other and encased in slime , breaking them apart requires extra energy to counter the adhesive effects of the slime ., In contrast to the climbing effect , here cells experience an energy penalty for breaking apart ., It is a form of non-specific attraction and operates over short ranges since two cells several cell lengths apart will not affect each other; only cells in close proximity experience adhesion ., The high density of cells in a swarm and fruiting body means there is a large amount of polysaccharide slime produced which encases all of the cells in a slime matrix 11 , 16 , 32 ., The slime casing prevents cells coming apart , for example even with a rotary shaker ., This matrix effects an adhesive force on the cells making it harder for cells to move apart from each other ., Cells typically aggregate at a colony edge due to surface tension effects making it difficult to escape the colony 20 ., This effect is different from the effects of A-motility and is a global property of a large mass of cells ., ( 34 ) where is a dimensionless adhesion coefficient , determines the number of oncoming cells , determines if two cells are moving in opposing directions by examining the dot product between the normalised average direction ( ) of each pair of interacting cells , and compares the direction cell wants to move in ( ) with a normal vector ( typically a normal to the -plane ) ., As a fruiting body matures , 65–90% of cells lyse , with the remaining cells going on to form myxospores 33 , 34 ., Spores appear to migrate to the centre of the fruiting body with motile cells remaining on the outside and the periphery 11 ., The fruiting model was extended to incorporate sporulation and its effects on fruit formation ., Each cell is given a type: motile or spore ., Motile cells accumulate C-signal from collisions with other motile cells ., Once C-signal exceeds a threshold ( ) , cells convert to non-motile spores ., Spores can be moved by motile cells pushing them ., Each cell type has its own Hamiltonian governing its behaviour ., Normal cells continue to use the Hamiltonian defined in Equation 3: ( 35 ) However spores are non-motile cells with a fixed size and shape , and the Hamiltonian controlling them loses terms associated with autonomous cell motion and is therefore simpler: ( 36 ) Although spore cells are immobile , other motile cells can move them during collisions when they collide and through adhesive effects between cells ., ( 37 ) ( 38 ) ( 39 ) where is a dimensionless coefficient , is the normalised average direction of the cell , is the average direction of all the cells in a local neighbourhood surrounding cell ., The term reflects the tendency of cells to turn through an acute angle to align with other cells in either direction ., The parameters used in the simulation are listed in Table 1; the same parameters were used in our previous model of rippling behaviour 35 ., Some of these parameters reflect the level of abstraction that approximate the level of behaviour observed through video microscopy ., The 7∶1 length to width ratio reflects evidence from 36 ., The volume of each cell was set so that thousands could be fit into a volume large enough to hold a fruiting body without requiring unreasonable amounts of computational memory ., Representing cells via eight segments seemed to provide a reasonable approximation of the degree of flexibility observed in various phases of the lifecycle ., Motility parameters were based initially on experimental evidence 37 , 38 to get an idea of the speed of cells , and then tuned so that cells moved at the correct speed given their size and volume in the simulation environment ., Likewise , parameters governing the flexibility of cells were based initially on 39; other parameters were tuned relative to these to emulate the cell motion patterns observed in nature ., Simulations were carried out in a three-dimensional environment using a model based on our previous stochastic model of myxobacteria motility 35 ., M . xanthus is approximately 5–7 long and 0 . 5 in diameter so the model cells were given a length to width ratio of 7∶1 ., Each cell was composed of eight three-dimensional segments ( see Figure 2 ) with each segment being composed of 27 segment nodes arranged in a cube formation ., Segments were allowed to overlap so that cells maintained a continuous volume and the correct aspect ratio despite being made of multiple separate segments ( see Figure 2, b ) ., The physical behaviour of cells was described using a Hamiltonian function whilst the internal state was described using ordinary differential equations ( ODEs ) ., The EPS surrounding cells is rarely considered in models; however , in our model we found that slime can have an essential role in fruit formation ., The Hamiltonian includes an adhesion term , which generates energy proportional to the inverse square of the distance between any two cells in a neighbourhood ., It is more energetically favourable for cells to remain close to other cells otherwise there is a severe penalty for moving apart that increases exponentially with distance ., An inverse relationship was chosen so that long range interactions are weak; cells towards the perimeter of the local neighbourhood should not exert the same influence as cells in close proximity ., Adhesion acts to control the viscosity of the slime determining how easy it is for cells to move through it ., The amount of slime and thickness varies depending on the stage of fruiting and the cell density 16 ., We note that Curtis et al . reported that a pilA mutant produces less EPS and this inhibits fruiting body formation 13 ., It is not known biologically whether the effect of the pilA mutation is a consequence of reduced EPS production , or due to altered motility properties ., Therefore , a direct comparison with the pilA mutant described by Curtis et al . is not possible ., Figure S1 shows the effect of varying the strength of adhesion on a stream of cells ., When there is no adhesion , cells at the front of the stream are able to move adventurously , causing the stream to break down into a number of smaller streams which diverge ., As the adhesion strength ( ) is increased , cells remain much closer ., When cells tend to stay as one or possibly two large coherent streams ., When , the slime is so viscous that cells are no longer able to move ., Fruiting begins with streaming and the confluence of streams to form aggregation centres ., The fruiting model presented here allows cells to spontaneously form streams and aggregation centres ( see Figure 4 ) ., A simulation consisting of 300 cells was run three times to determine the efficacy of streaming and aggregation ( model parameters are given in Table 1 ) ., Cells were initially randomly distributed ., After approximately 100 min of simulated real time , cells formed into streams regardless of their initial configuration ., Cells aligned and formed small streams which joined other streams when they came into contact ., After 300 min cells typically formed an aggregate , which expanded as the the majority of cells joined it ., The effects of motility along with cell adhesion causes model cells to form streams ., As the streams approach an aggregation centre , cells will attempt to avoid collision and alter course ., They begin to move around the aggregate causing the stream to change direction and form the characteristic spiral patterns observed by OConnor and Zusman 11 ., In a model with a finite number of cells , it is difficult to achieve a high enough cell density to maintain aggregates ., There is an upper bound on the density of cells in a mono-layer above which cells will not have enough space to move and be able to engage in activity ., With a high cell density which still allows cells to move , it is possible to get aggregation , but once the fruit starts to form , the number of cells moving into the aggregate will not be sustainable and the fruit will simply dissociate ., This type of model is also unrealistic because in reality , an aggregate would be surrounded by other cells and not sit in isolation as more cells join it ., Although it would be ideal to model a vast mono-layer of cells to ensure there were a sufficient number of cells to a form a fruit , computational limitations ( typically memory ) restrict the size of a simulation ., Figure S4 shows the output of a fruiting body simulation using a finite number of cells ., 1600 cells were arranged into two opposing sets of streams with one set perpendicular to the other ., The streams move into each other and collide ., In the aggregation centre , some cells push upwards and move over others forming new layers and the base of a stalk ., The effect of using a finite number of cells becomes apparent after 300 time steps when the stalk begins to disassociate ., The cells organise themselves into a stack four layers thick , but since there are no more cells to expand the base layer , the upper cells begin to climb down and move away from the fruit ., Once a few cells move away , a mass exodus is triggered causing all of the cells to move away ., The formation and subsequent rapid dispersion of fruits will occur at any point where an aggregate forms ., This effect will be more apparent on subsequent aggregation formations since the number of cells within the mass is unlikely to be as high as in the initial formation so the deterioration will be more pronounced ., Curtis et al . 13 observed that during the initial stages of fruiting , small fruiting bodies would sometimes repeatedly start to form and then dissipate before a stable fruiting body finally formed ( see Figure 1 in 13 ) ., The formation of transitory aggregates can be explained by adjusting the cell influx rate ., The simulations maintained the same initial conditions as the previous fruiting simulation , except the rate of influx was altered ., Figure S2 shows a snapshot of a simulation where the influx rate was reduced by 90% ., Although a fruiting body begins to form it rapidly dissociates over time ., Cells accumulate and the stack expands outwards from the centre for approximately 200 min after which the fruit collapses and the cells begin to disperse ., The cell density remains too low for cells to attempt a new fruit formation suggesting that influx could be a primary driving factor behind development ., The base influx rate was selected to ensure a constantly high cell density to enable fruit formation ., Lower influx rates promote transitory fruiting body formation and dissociation ., Figure 5 shows three-dimensional snapshots of fruiting development when the influx rate was reduced to 25% of the base value ., After 500 min , three small mounds have formed; however , they dissociate and new mounds form ., This agrees with experimental evidence showing transitory aggregates 13 ., If the simulation volume were enlarged by several orders of magnitude ( which has not been computationally feasible ) , we predict that as fruiting bodies disperse , a cohesive layer of cells would form and drift off ., This would meet other disparate layers from other dispersed fruits and further fruiting development would be initiated where they collide ., The process would repeat leading to multiple transitory fruit formations 13 ., The prerequisite for this to occur is a sufficiently high cell influx that allows a fruit to form but at a sub-optimal rate such that development cannot be sustained ., The fruiting body must be sufficiently large so that , when cells leave it , they form a layer of equal density to the initial layers so that fruiting can occur spontaneously at other locations ., The influx rate appears to be the rate limiting step in controlling fruiting growth; there is a point where the number of cells forming new layers will begin to exceed the number of cells flowing into the system so the development o

Introduction, Model, Results, Discussion

Many bacteria exhibit multicellular behaviour , with individuals within a colony coordinating their actions for communal benefit ., One example of complex multicellular phenotypes is myxobacterial fruiting body formation , where thousands of cells aggregate into large three-dimensional structures , within which sporulation occurs ., Here we describe a novel theoretical model , which uses Monte Carlo dynamics to simulate and explain multicellular development ., The model captures multiple behaviours observed during fruiting , including the spontaneous formation of aggregation centres and the formation and dissolution of fruiting bodies ., We show that a small number of physical properties in the model is sufficient to explain the most frequently documented population-level behaviours observed during development in Myxococcus xanthus .

Understanding how relatively simple , single cell bacteria can communicate and coordinate their actions is important for explaining how complex multicellular behaviour can emerge without a central controller ., Myxobacteria are particularly interesting in this respect because cells undergo multiple phases of coordinated behaviour during their life-cycle ., One of the most fascinating and complex phases is the formation of fruiting bodies—large multicellular aggregates of cells formed in response to starvation ., In this article we use evidence from the latest experimental data to construct a computational model explaining how cells can form fruiting bodies ., Both in our model and in nature , cells move together in dense swarms , which collide to form aggregation centres ., In particular , we show that it is possible for aggregates to form spontaneously where previous models require artificially induced aggregates to start the fruiting process .

computational biology/synthetic biology, computer science/applications, cell biology/cell signaling, biophysics/theory and simulation, cell biology/microbial growth and development, microbiology/microbial growth and development, computational biology/systems biology

journal.pcbi.1002339

A Viral Dynamic Model for Treatment Regimens with Direct-acting Antivirals for Chronic Hepatitis C Infection

Chronic hepatitis C ( CHC ) affects approximately 180 million people worldwide and is a frequent cause of increased risk for hepatic fibrosis , cirrhosis , hepatic failure , and hepatocellular carcinoma 1 , 2 , 3 , 4 ., The treatment objective for CHC is SVR , or viral eradication , which is considered to be a virologic cure of the infection ., The previous treatment for patients with genotype 1 CHC , 48 weeks of therapy with PR ( PR48 ) ; results in SVR for 42%–50% of treatment-naïve patients 5 , 6 ., In clinical trials , a combination therapy of telaprevir and PR treatment ( TPR ) achieved potent antiviral activity and higher SVR rates compared to treatment with PR alone 7 , 8 , 9 , 10 , 11 , 12 , 13 ., As a consequence of its high replication rate and its error-prone polymerase , the HCV population in a patient exists as quasispecies ., At the start of treatment with direct-acting antivirals such as telaprevir , the HCV population must be considered as a mixed population , consisting predominantly of wild-type HCV ( WT ) and a small population of HCV variants with varying levels of resistance to telaprevir ., The resistant variants generally exist at a lower frequency than WT prior to the start of treatment 14 because they are less fit ( have lower replicative capacity ) 15 , 16 , 17 , 18 , 19 ., Variants with lower-level resistance ( 3 to 25-fold increase in telaprevir IC50 in vitro: V36A , V36M , R155K , R155T , T54A , A156S ) have higher fitness than variants with higher-level resistance ( 25-fold increase in telaprevir IC50 in vitro: A156T , A156V , V36M/R155K ) 18 ., These variants retain sensitivity to PR treatment in vitro 20 and in patients 16 , 21 , 22 ., WT virus was eliminated more rapidly in the presence of telaprevir than with interferon-based regimens alone in clinical trials 23 , 24 ., The treatment duration required to achieve SVR is based on the time to eradicate all HCV , including WT and all variants ., For PR treatment , models of viral dynamics have successfully predicted SVR rates by calculating the percentage of patients whose on-treatment HCV RNA levels reach the viral eradication limit 25 , 26 , 27 ., For TPR treatment , because of the presence of multiple variants in the quasispecies , the time when the level of each variant within a patient reaches the viral eradication limit may vary depending on the variants fitness and resistance , and individual patient responses to treatment ., The importance of these different eradication times to treatment strategies has not been elucidated ., Here , we describe a viral dynamic model of response to TPR treatment ., The model incorporates the presence of viral variants of differing degrees of resistance and fitness , and the diversity in patient responses to treatment ., The viral dynamic model was improved from the previously published model 18 , with 2 novel features:, 1 ) the model integrated TPR pharmacokinetics into viral dynamics , and, 2 ) viral dynamic parameters were estimated using a population-approach method ., The model was developed using in vitro and clinical data in early studies obtained from 28 patients treated with 2 weeks of telaprevir monotherapy and 478 treatment-naïve patients treated with PR and TPR regimens ., Model predictions were evaluated from the outcome data of 2380 patients ., Eradication of each viral variant was simulated as a discrete event occurring at variable times during treatment: when eradicated , variants were assumed to stop replicating ., If eradication of all viral variants within a simulated patient was achieved , the patient was deemed to have reached SVR ., Model parameters were estimated from HCV RNA and drug concentration data from 478 patients who participated in early phase clinical studies ( study regimens are described in Supplementary Table S1 ) ., The goodness-of-fit plot was provided in Supplementary Figure S1 and examples of fits in representative patients were provided in Supplementary Figure S2 ., The estimated parameters were provided in Supplementary Table S2 ., The estimated replicative fitness among all the variants ( Figure 1 ) showed that the R155K variant has the highest fitness ( with estimated fitness of about 50% of WT fitness ) , and the A156T variant has the lowest fitness ( with estimated fitness of about 10% of WT fitness ) ., Some lower-level telaprevir resistant variants ( R155K , V36M , and V36A ) had higher fitness than the higher-level telaprevir resistant variants ( V36M/R155K , A156T ) ., The other lower-level telaprevir resistant variants ( A156S , R155T , and T54A ) had lower fitness than the higher-level telaprevir resistant variants ., The individual contributions of telaprevir and PR to antiviral blockage and infected-cell clearance rates estimated from treatment-naïve population are provided in Table 1 ., Telaprevir contribution to production blockage ranged from −2 . 51−log10 to −2 . 27−log10 for WT and lower-level telaprevir resistant variants and −0 . 01−log10 to 0 . 00−log10 for higher-level telaprevir resistant variants , while PR treatment contributed −1 . 09−log10 for all variants ., Compared to WT , lower-level telaprevir resistant variants have similar median blockages but reduced blockage in the extreme ( 95th percentile ) , which occurred in patients with lower telaprevir concentrations ., Infected-cell elimination rates were higher for WT and lower-level telaprevir resistant variants ( 0 . 62 d−1 ) than for higher-level telaprevir resistant variants ( 0 . 29 d−1 ) ., The higher elimination rates were mainly driven by higher antiviral blockage against WT and lower-level telaprevir resistant variants by telaprevir than by PR ., These results suggest that the primary role of telaprevir is to block viral replication of WT and lower-level telaprevir resistant variants , and the primary role of PR is to block viral replication of higher-level telaprevir resistant variants ., The model prediction capability was validated by comparing predicted and observed SVR rates in study regimens in which on-treatment data were used to estimate the model parameters ( 478 patients ) and in which the model was used strictly in prediction mode ( 2380 patients , Supplementary Table S1 ) ., Predicted SVR rates were generated based on these inputs:, ( a ) simulated drug concentrations and HCV RNA dynamics using parameter values re-sampled from the estimates;, ( b ) the actual number of patients treated;, ( c ) the number of patients who prematurely discontinued treatment;, ( d ) the number of patients who failed to reach SVR because of other reasons ( lost to follow-up , noncompliance , and withdrawal of consent ) ;, ( e ) the timing of treatment discontinuations; and, ( f ) the distribution of HCV genotype ( 1a and 1b ) for each regimen/patient population ., Figure 2 shows the correspondence between observed and predicted SVR rates ., In the early studies in which the on-treatment data were used to develop the model , all observed SVR rates were within the 90% confidence intervals ( CIs ) of the predicted rates ., In subsequent studies , observed SVR rates were also consistent with predicted rates ., In the subsequent Phase 2 studies , the majority of the observed SVR rates ( 13/14 treatment groups ) were within the 90% CI bounds of the predicted rates; the other group had a rate within 3% of the nearest 90% CI bounds ., In the Phase 3 treatment-naïve Studies ADVANCE and ILLUMINATE , the observed rates were within the 90% CI bounds in 4/5 groups; the other group had an observed rate that was 1% of the nearest CI bounds ., In the Phase 3 treatment-experienced Study REALIZE , the observed SVR rates were all lower ( by up to 7% ) than the 90% CI lower bounds of the predicted rates ., The discrepancy was greatest in the prior PR48-nonresponder population ., In all regimens in this study , observed SVR rates were lower than predicted rates; therefore , the comparison of rates among regimens within the study is comparable between observed and predicted rates ., Despite being trained only for the treatment-naïve population , the model produced consistently predictive results even for different patient populations such as prior PR48-nonresponders and prior PR48-relapsers ., The predicted SVR rates by prior PR48 responses were calculated from a subset of simulated treatment-naïve patients by classifying these patients based on their simulated HCV RNA dynamics in response to PR48 treatment , using the standard definition of PR responsiveness: prior PR48-SVR , if patients would reach SVR with PR48 treatment; prior PR48-relapser , if patients have undetectable HCV RNA at the end of PR48 treatment but not reached SVR; prior PR48-partial responder , if patients have more than 2-log10 HCV RNA decline at week 12 but detectable HCV RNA throughout PR48 treatment , prior PR48-null responder , if patients have less than 2-log10 HCV RNA decline at week 12 during PR treatment ., Using the assumption that each subgroup of prior PR responses was a narrower subset of the diverse PR responsiveness of treatment-naïve population , the model was able to predict the observed higher SVR rates in prior PR48-relapser and lower SVR rates in prior PR48-nonresponders compared to rates in treatment-naïve patients ., To examine how viral eradication is affected by variant fitness , resistance , antiviral inhibition of each drug in the combination regimen , and patients diversity in responses to treatment , simulations were performed for patients with 3 levels of PR responsiveness treated with 12 weeks of telaprevir in combination with 48 weeks of PR ( T12PR48 , Figure 3 ) : 1 ) typical patient who would achieve SVR if treated with PR48 ( left panel ) , 2 ) typical prior PR48-relapser ( middle panel ) , and 3 ) typical prior PR48-null-responder ( right panel ) ., Simulated patients were assumed to have subtype 1a or 1b infection to provide a representative illustration ., These simulations illustrate only representative examples with median responses , as there is variable PR responsiveness even within each respective group of prior PR response ( the predicted SVR rates by groups of prior PR responses are provided elsewhere 28 ) ., Patients in each HCV subtype were assumed to have the same set of major variants: for subtype 1a: WT , R155K , V36M/R155K , and A156T; for subtype 1b: WT , V36A , A156T; variants with intermediate fitness or resistance were not included ( see methods ) ., The PR responsiveness of the first 2 simulated patients with subtype 1a succeeded in eliminating all variants , but that of the last patient failed to eliminate the higher-level telaprevir resistant variant V36M/R155K ., Both WT and the lower-level variant R155K were eliminated by about 6 weeks of telaprevir treatment in these 3 patients; however , the higher-level telaprevir resistant variant V36M/R155K was eliminated only in patients with better PR responsiveness ( the first 2 simulated patients ) ., In contrast , the 3 simulated patients with subtype 1b were able to reach eradication because the PR responsiveness of these patients overcame the relatively poor fitness of A156 variants ( V36M/R155K variants were not present at baseline in the subtype 1b patients ) ., The simulation above illustrates that the variability in PR responsiveness affects the time needed to eradicate higher-level telaprevir resistant variants ., For these 3 simulated patients , the time to eradicate was similar for WT and lower-level telaprevir resistant variant R155K ., However , the time to eradicate higher-level telaprevir resistant variants differed by PR responsiveness: for variant A156 , eradication times were 8 , 11 , and 13 weeks for the 3 patients; for variant V36M/R155K , the eradication time was 5 and 8 weeks for the first 2 patients , and was undefined in the last patient ( because this variant was never eradicated ) ., For the simulated null-responder patient ( which as noted above , represents a median response for the null responder population ) , the increase in the level of V36M/R155K resulted in re-generation of R155K variant after completion of 12-week telaprevir treatment , resulting in a viral population with R155K-dominant quasispecies at week 48 ( because of the higher fitness of R155K compared to V36M/R155K ) ., However , a telaprevir duration longer than 12 weeks would also result in virologic failure but with different predominant variant in the quasispecies when failure is detected ( V36M/R155K variant predominant instead of R155K variant predominant ) ., To examine the contribution of the eradication assumption—that a variant stops replicating when its level is below the eradication limit—a simulation was performed with and without the eradication assumption ., In the simulation without eradication , all variants were allowed to continue replicating even when their levels were below the eradication limit ., The simulations were performed for simulated patients with 2 levels of PR responsiveness treated with T12PR: 1 ) typical treatment-naïve patient ( Figure 4 left panels ) , and 2 ) typical patient who would not reach SVR with PR48 treatment ( Figure 4 right panels ) ., In the typical treatment-naïve patient , the predicted outcomes were the same with and without the eradication assumption: Week 48 HCV RNA levels were below the eradication limit ., However , for the patient who failed to reach SVR with PR48 treatment , the outcomes differed ., The dynamics in the first 12 weeks were the same: WT and lower-level telaprevir resistant variant levels reached the eradication limit by week 6 ., With the eradication assumption , the quasispecies left were residual higher-level telaprevir resistant variants with reduced fitness that continued to be eliminated by PR treatment , resulting in a Week 48 HCV RNA level below the eradiation limit ., Without the eradication assumption , the WT HCV RNA level returned back to the baseline value around week 24 after the level reached the eradiation limit during the first 12 weeks of TPR treatment ( telaprevir was only administered in the first 12 weeks ) ., The return of HCV RNA levels after the completion of 12 weeks of telaprevir treatment with quasispecies predominately WT is rarely observed in clinical trials 8 , 10 , 29 , supporting the eradication assumption ., The predicted treatment outcomes with and without the eradication assumption for a population of simulated treatment-naïve patients completing a T12PR24 regimen are shown in Figure 5 ., Virologic outcomes were categorized as virologic failure at weeks 1–12 when TPR treatment was administered; virologic failure at Weeks 13–24 when PR treatment was administered; virologic failure after Week 24 when no treatment was administered ( relapse ) ; and SVR ., Comparing simulations with and without the eradication assumption , the largest difference was observed for virologic failure between Weeks 13–24: 4 . 4% with eradication and 16 . 5% without eradication ., The virologic failure rate with the eradication assumption is more consistent with rates observed in clinical trials ( see discussion ) , supporting the eradication assumption ., An integrated model of viral dynamic responses to treatment with telaprevir and PR has been developed and validated by comparing predictions against observed outcomes in late-phase clinical trials ., It provides a framework to integrate multi-faceted information related to this novel CHC regimen , including in vitro resistance and fitness , pharmacokinetics , viral sequencing , and viral dynamics ., The framework supported decisions pertaining to treatment strategies and optimizing regimens during clinical development ., The model that was based on data from early-phase trials was predictive of observed SVR rates in subsequent studies that were not used in model building ., The model also aided understanding of a novel CHC treatment regimen consisting of telaprevir and PR ., It provided a consolidated picture of the interplay between the fitness and resistance of variant populations , antiviral inhibition by telaprevir and by PR treatment , and patient diversity in PR responsiveness , and connected these factors to the ultimate treatment outcome of SVR ., The model suggested that the primary role of telaprevir in a TPR regimen is to eradicate WT and lower-level telaprevir resistant variants , and the complementary role of PR is to eradicate higher-level telaprevir resistant variants ., Accordingly , virologic failure during the telaprevir-treatment phase has been associated predominately with higher-level telaprevir resistant variants , indicating a failure of PR to inhibit higher-level telaprevir resistant variants in some patients 9 , 29 ., Modeling results and analysis of viral populations derived from patients who stopped treatment prior to viral eradication 28 have led to the working hypothesis that a successful regimen should have ( 1 ) a telaprevir treatment duration sufficient to eradicate WT and most lower-level telaprevir resistant variants , and ( 2 ) a PR treatment duration sufficient to eradicate any remaining variants , particularly higher-level telaprevir resistant variants ., Once WT and lower-level telaprevir resistant variants have been eradicated and higher-level telaprevir resistant variants are the dominant residual viral population , telaprevir adds no additional antiviral effect ., The PR duration required to eradicate higher-level telaprevir resistant variants depends greatly on the PR responsiveness of a given patient and likely the number of residual higher-level telaprevir resistant variants ., Because higher-level telaprevir resistant variants pre-exist at lower frequency than WT and have reduced fitness , a greater percentage of patients can be treated with a shorter duration of PR treatment in the TPR regimen than in the PR regimen ., The personalization of PR durations for patients treated with T12PR treatment has been demonstrated in those who achieved early virologic response in clinical trials 11 , 12 ., Data and modeling analyses suggest different eradication times for variants with varying fitness and resistance , leading to different optimal treatment durations of telaprevir and PR treatment ., Modeling analysis showed that a higher percentage of patients would be expected to have virologic failure during PR treatment after the completion of 12 weeks of telaprevir treatment if simulated without viral eradication , a phenomenon that has rarely been observed in clinical trials: the virologic failure rates after 12-week of telaprevir treatment in treatment-naïve patients were 1% for the T12PR24 arm of Study PROVE2 8 and 4 . 4% in the T12PR24-48 arms of ADVANCE 28 , 29 ., Moreover , the shorter eradication times of sensitive variants as compared to resistant variants are also consistent with the observed more rapid elimination of WT HCV in patients dosed with telaprevir as compared to those typically observed in PR treatment 23 , 24 ., The model produced consistently predictive results for different prior PR48-treatment-failure populations despite being trained only for the treatment-naïve population ., This finding supports the hypothesis that a treatment-naive population contains several types of patients with differing PR responsiveness , and suggests that a model estimated from the treatment-naive population can be used to predict results for populations with different PR responsiveness ., In the 2 studies in the treatment-experienced population ( Studies PROVE3 and REALIZE ) , the predicted and observed SVR rates were generally consistent: comparable SVR rates in PROVE3 and slightly higher predicted SVR rates compared to those rates observed in REALIZE ., The discrepancy is greatest in the prior nonresponder population ., The discrepancy in the REALIZE study may arise from a limitation of the model: that the underlying parameters constituting PR responsiveness were assumed to be continuously distributed in treatment-naïve population , while the actual parameters may be more discrete and based on other factors such as the IL28B genotypes 30 , which has been reported to produce different viral dynamics in response to PR treatment 31 , 32 ., Alternatively , the discrepancy may be attributed to a higher proportion of patients with adverse prognostic factors for achieving SVR ( e . g . , advanced liver disease ) enrolled in REALIZE , whereas the predictions were generated from the dataset that contained treatment-naïve patients with fewer of these adverse factors ., In the modeling described here , adverse factors were not formally examined as covariates because of the limited data available from the early studies ., In summary , the proposed model served as a framework in integrating information from multiple sources and was useful in supporting decision-making for the optimization of treatment strategies during clinical development ., The model provided insights to help design novel treatment regimens of combination therapy with telaprevir , peginterferon alfa-2a and ribavirin for CHC treatment , and may be useful for evaluating future CHC treatment regimens that include direct-acting antiviral agents ., The study protocols and informed consent forms were reviewed and approved by ethics committees or institutional review boards for each clinical research site before initiation of studies at that site ., Written informed consent was obtained in accordance with the Declaration of Helsinski ., The model was developed from HCV RNA and drug concentration from a total of 478 patients treated with PR and TPR regimens in early studies of telaprevir ., The model was validated using outcomes from 2380 patients in later studies ., The list of studies is provided in Supplementary Table S1 ., The study design , enrollment criteria , and primary results have been published elsewhere 7 , 8 , 9 , 10 , 11 , 12 , 13 , 33 ., Only quantifiable HCV RNA data were used in the estimation ., Additional limitations were implemented:, 1 ) for PR regimens , only HCV RNA data up to time when the first dose modifications of either peginterferon or ribavirin were used ( or end of the treatment ) to evaluate the PR responses with one dose level; and, 2 ) for TPR regimens , only patients with WT-dominant quasispecies ( 98% of patients ) were included because few patients ( 2% ) had resistant-variant dominant quasispecies ., While the model can be applied to the patients with resistant-variant dominant quasispecies , the small number of patients in this dataset prevented us from making accurate conclusion regarding the comparability of the fitness of resistant variants in these patients to those in patients with WT-dominant quasispecies ., The model structure is given in Equations 1–8 , and the descriptions of symbols are given in Table 2 ., Drug pharmacokinetics were estimated from time-concentration data in early studies ., Telaprevir and peginterferon alfa-2a pharmacokinetics were described by one-compartmental models and provided in Equation 8 ., Ribavirin pharmacokinetics were described by a 3-compartmental model , with parameters estimated using empirical Bayesian feedback from published distributions of parameter estimates 34 ., Model-predicted drug concentrations were simulated based on the dosing records and pharmacokinetic model parameters and were entered into the viral dynamic model ., ( 1 ) ( 2 ) ( 3 ) ( 4 ) ( 5 ) ( 6 ) ( 7 ) ( 8 ) A schematic of the viral dynamic model is provided in Figure 6 ., Viral populations were represented as a mixture of quasispecies with varying fitness and sensitivities to telaprevir ., Variant V represents a virion with characterized amino-acid substitution ( s ) in the NS3/4A protease ., Variant Vi infects target cells ( T ) to form V-infected cells ( I ) at rate βTV ., Each variant competes for the same target cells T . Target cells T also represent limited “replication space” shared by all variants; target cell T has a synthesis rate s and a first-order elimination d ., In 18 , a model with different representation of T ( which maintain T+I ) resulted in comparable estimates ., The maximum target cells were assumed to be 1011 35 ., Each infected cell ( I ) produces a population of variants at production rate pf , with a m-fraction of this production mutating to produce variant j ( V ) ., The mutation rate was assumed to be 1 . 2 10−4 base−1 cycle−1 36 ., The production rate ratio ( f ) quantifies variant replicative fitness in the absence of any drug ., Different production rates ( pf ) , but the same infection ( β ) and clearance ( c ) rates , are assumed for different variants ., This assumption is consistent with the function of the NS3/4A protease to cleave a precursor polyprotein as a crucial step in the HCV replication cycle 37 ., Each drug ( telaprevir , peginterferon alfa-2a , ribavirin ) assumes a dual role in clearing HCV ., First , each drug blocks viral production by a factor ( 1- ε ) ., Telaprevir antiviral blockage εi , T is constrained to be consistent with in vitro sensitivity assay of variant i to telaprevir 38 , 39 , 40 ., Blockage by peginterferon alfa-2a and ribavirin are assumed to be equal among variants , consistent with in vitro sensitivity assay ., While the antiviral mechanism of ribavirin ( of whether ribavirin blocked viral production or changed infectious into noninfectious viral strains ) remained controversial , our data were unable to distinguish a model with a simple production blockage from a model with infectious and noninfectious viral strains 25 , and therefore , a simpler model with production blockage was chosen instead of the alternative model because the alternative model would need twice as many number of variants ., The blockage factors were calculated as a function of plasma concentrations of each drug ( multiplied by a factor κ to convert plasma to effective concentrations ) , and the sensitivities of each variant as measured in HCV replicon cells ( represented by parameters IC50 , and hill-power values h ) ., Overall blockage in the combination regimen assumed additive ( in logarithmic scale ) blockages of each drug ., The second role of each drug is to enhance the infected-cell clearance δ ., WT δWT values were up to 10-times higher in patients dosed with telaprevir than in patients treated with interferon-based regimen alone 41 , 42 ., These observations were represented into the model by assuming that δ increased proportionally with log10 ( 1-ε ) 18 ., The enhanced δ may be attributed to increases in infected-cell clearance or uncovering of intracellular viral RNA 27 ., Consistently , as these mechanisms may not be specific to direct-acting antivirals , the enhancement may also be observed for interferon if its effectiveness is high enough ., The HCV variants used in this model was based on the major variants detected in clinical studies: one major variant with the highest fitness for each of the resistant groups ( lower-level and higher-level resistance ) and the nucleotide changes from WT ., Subtypes 1a and 1b were modeled separately because when telaprevir was administered in monotherapy , different sets of resistant variants emerged 15 , 16 ., All patients with the same subtype were assumed to have the same set of major variants: for subtype 1a: WT , R155K , V36M/R155K , and A156T/V; for subtype 1b: WT , V36A , A156T/V ., The frequency of these variants prior to treatment was calculated by assuming a steady-state condition ., The intermediate-resistant variants R155T/I and other minority variants observed in a few patients were not included in the model used to generate predictions because of lack of data to estimate their fitness ., Including these variants in the model was expected to result in only small changes in the SVR rates , because these variants appeared to be less fit than the variants used in the model 18 ., The parameters related to the antiviral activity of peginterferon alfa and ribavirin were correlated because the current training dataset contained data from regimens where peginterferon and ribavirin were administered simultaneously ., Because of the data limitation , the proportionality constant related to the enhanced infected-cell clearance for ribavirin is assumed to be equal to the constant for peginterferon ., SVR rates were predicted by evaluating simulated HCV RNA dynamics and entering the observed patient disposition into the model ., The predicted HCV RNA dynamics for treatment-naïve patients were generated by simulations , with parameters re-sampled from the distributions of estimates in Supplementary Table S2 , truncated by lower and upper bounds ( bounds were obtained from the extreme values of the observed individual estimates ) ., Dosing compliance was assumed to be 100% ., Ribavirin dose modification followed the observed modification in the training dataset ., A simulated patient was considered to achieve eradication ( or SVR ) if the overall HCV RNA level by the end of treatment was below 1 copy in the body 25 ( or reached a 12-log decline from baseline in HCV RNA , assuming a baseline value of 107 IU/mL ) ., Predicted SVR rates for different categories of PR responsiveness ( SVR with PR48 , prior PR48-non-SVR , prior PR48-relapser , prior PR48-nonresponder , prior PR48-null responder ) were generated by simulating HCV RNA dynamics to PR48 treatment , and by filtering the responses with the respective PR responsiveness criteria ., The categories of PR responsiveness followed these criteria: SVR with PR48 , if patients viral load reached eradication by the end treatment; prior PR48-non-SVR , if patients viral load did not reach eradication by the end of treatment; prior PR48-relapser , if patients viral load was undetectable by the end of treatment but did not reach eradication; prior PR48-nonresponder , if patients viral load was always detectable during treatment; prior PR48-null-responder , if patients viral load at week 12 declined <2−log10 ., Drug concentrations were estimated or simulated using a Bayesian approach implemented in NONMEM version 6 ., Viral dynamic model was implemented in Jacobian® software version 4 . 0 ( RES group , Inc . , Cambridge , MA ) .

Introduction, Results, Discussion, Methods

We propose an integrative , mechanistic model that integrates in vitro virology data , pharmacokinetics , and viral response to a combination regimen of a direct-acting antiviral ( telaprevir , an HCV NS3-4A protease inhibitor ) and peginterferon alfa-2a/ribavirin ( PR ) in patients with genotype 1 chronic hepatitis C ( CHC ) ., This model , which was parameterized with on-treatment data from early phase clinical studies in treatment-naïve patients , prospectively predicted sustained virologic response ( SVR ) rates that were comparable to observed rates in subsequent clinical trials of regimens with different treatment durations in treatment-naïve and treatment-experienced populations ., The model explains the clinically-observed responses , taking into account the IC50 , fitness , and prevalence prior to treatment of viral resistant variants and patient diversity in treatment responses , which result in different eradication times of each variant ., The proposed model provides a framework to optimize treatment strategies and to integrate multifaceted mechanistic information and give insight into novel CHC treatments that include direct-acting antiviral agents .

Hepatitis C virus chronically infects approximately 180 million people worldwide ., The treatment aim for patients chronically infected with hepatitis C is viral eradication or sustained viral response ( SVR ) ., Historical standard of care for HCV treatment was peginterferon-alfa and ribavirin ., Recently , approved HCV protease inhibitors , in combination with peginterferon-alfa and ribavirin , have demonstrated higher SVR rates compared to peginterferon-alfa and ribavirin alone ., As members of a novel class of compounds directly targeting hepatitis C virus , HCV protease inhibitors have different mechanisms of actions and are affected by resistance and fitness of HCV variants ., The significance of these different mechanisms of action , and the interplays between resistance and viral fitness to the treatment outcome has not been elucidated ., Here , we developed and validated an integrative , mechanistic model of viral dynamics in response to a combination regimen including telaprevir , peginterferon-alfa , and ribavirin ., The model was developed from early studies in 478 treatment-naïve patients and its SVR rate predictions were verified in 2380 patients in subsequent studies ., These results provide an example of the use of mechanistic information to the development of viral dynamic model that has been useful in the design of optimal treatment regimens .

bioengineering, medicine, hepatitis c, biological systems engineering, infectious diseases, biotechnology, hepatitis, viral diseases, engineering

journal.pcbi.0030187

Heat Shock Response in CHO Mammalian Cells Is Controlled by a Nonlinear Stochastic Process

Complex biological systems are built out of a huge number of components ., These components are diverse: DNA sequence elements , mRNA , transcription factors , etc ., The concentration of each component changes over time ., One way to understand the functions of a complex biological system is to construct a quantitative model of the interactions present in the system ., These interactions are usually nonlinear in terms of the concentrations of the components that participate in the interaction process ., For example , the concentration of a dimer is proportional to the product of the concentrations of the molecules that dimerise ., Besides being nonlinear , the interactions are also stochastic ., The production process of a molecule is not deterministic , and it is governed by a probability rate of production ., In what follows , a nonlinear stochastic model for the response to heat shocks in CHO mammalian cells will be developed ., Heat stress is just one example of the many ways a molecular system can be perturbed ., From a general perspective , the structure of a molecular system is uncovered by imposing different perturbations ( input signals ) on the system under study , and then the responses of the system ( output signals ) are measured ., From the experimental collection of pairs of input–output signals , laws that describe the system can be uncovered ., This is the fundamental idea in Systems and Synthetic Biology 1–5 and has long proved to be successful in the field of electronics ., The input signals are applied through the use of signal generators 6–8 ., An input signal that is easy to manipulate is a heat pulse , the parameters to change being the pulse temperature and its duration ., Members of the stress protein family such as the heat shock protein 70 ( HSP70 ) are highly responsive to temperature variations ., This protein is a molecular chaperone and is a critical component of a complex genetic network that enables the organism to respond to deleterious effects of stress 9–11 ., Since Hsp70 is thus an important regulator in a complex system , our goal was to find if it is possible to develop a mathematical model of the regulation of its expression in mammalian cells exposed to heat shock ., Our specific objectives were, 1 ) determine an equation representing the average expression of Hsp70 over time in a cell population after an initial heat shock ,, 2 ) determine how the physical parameters of heat shock ( temperature and duration ) influence the parameters of this equation , and, 3 ) determine the mathematical model that describes the expression of Hsp70 at the single-cell level ., We first describe the process of inferring the mathematical model from the experimental data ., Then a mathematical study of the model will follow ., To acquire the experimental data , we elected to use a system using a reporter gene where the expression of the green fluorescent protein ( GFP ) is under the control of the promoter region of the mouse Hsp70 gene ., The GFP reporter proved useful for quantitative analysis 12 and was used before in connection with Hsp70 in different biological systems 13–17 ., The Hsp70-GFP fusion gene was integrated into a plasmid and transfected in Chinese hamster ovary ( CHO ) cells ., Stable transfectants were selected for their low level of basal expression of GFP and their capacity to upregulate GFP effectively and homogenously after exposure to heat shock ., Flow cytometry was used to make precise quantitative measurements of the fluorescence of a large cell population ., Since the quality of the experimental data was critical to the feasibility of the mathematical analysis , steps were taken to minimize sample-to-sample and experiment-to-experiment variability and to maintain the experimental noise to a minimum ., To that effect , temperature and time were tightly controlled for heat shocks , the cells were treated as a batch in a single tube for each condition ( combination of temperature and time ) , and aliquots were taken at each time point ., All samples were fixed for at least 24 h before analysis by flow cytometry so that changes of fluorescence due to fixation would not be a factor , and all the samples from the same experiment were analyzed at the same time ., Flow cytometry was chosen for analysis because it allows a very accurate quantitative measurement of the fluorescence of a large number of events , independently of the actual size of the sample ., Within the same experiment and between experiments , the same instruments settings were used for the flow cytometer , and at least 1 × 104 cells were analyzed per sample ., Detailed protocols and experimental conditions are available in the Materials and Methods section ., First , we will follow a description of the time course of the mean response to a heat shock ., At elevated temperatures ( 39 °C to 47 °C ) , the heat shock promoter HSP70 is active and GFP starts to be synthesized ., The input signals were chosen in the form of a pulse at a temperature ( T ) and duration in time ( D ) ( Figure 1A ) ., In the first experiment , the dynamic response of GFP after a heat pulse at 42 °C for 30 min was monitored by taking samples each 30 min for 18 h ., Before and immediately after the heat shock , the GFP intensity remains at approximately the same level; this phenomenon was observed in all subsequent experiments ., The fold induction of GFP with respect to a reference ( GFP0 ) was then determined:, The reference is the first measured sample away from the end of the heat shock ( 30 min after the shock in Figure 1A ) ., Our finding is that the logarithm of the fold induction of GFP follows an exponential saturation trajectory ( Figure 1B ) , with tight confidence bounds for the estimated parameters and tight prediction bounds for nonsimultaneous observations ., The tight prediction bounds appear even when almost half of the data is not used during fitting ( Figure 1B ) ., The time t is measured relative to the reference time t0 . The initial fold induction at t = 0 ( or equivalent t0 after the end of heat shock ) is 1 ., This value of 1 for the initial fold induction is consistent with the entire time evolution if a fit with the expression, will give a value for parameter, very close to the value for parameter a ., Theoretically ,, must be equal with a to have a fold induction of 1 at t = 0 ., The result of the fit ( Figure 1B ) shows this consistency ., From now on we will take, ., The empirical law for the response of the cells to the heat pulse can be thus cast into the form:, The same law appeared in repeated measurements of pulses at 42 °C for 30 min duration ( unpublished data ) ., Parameter b describes the quickness of the response ., As b increases , the saturation value of the response is reached in less time ., Parameter a specifies the saturation value of the response ., The plateau reached by the fold induction is ea and thus grows exponentially with parameter a ., These findings suggest that the same law is valid for other heat shock pulses , parameters a and b being dependent on the heat pulse height T and its duration D ( Figure 1C ) ., To find the range of validity for the empirical law , measurements were taken for the responses to heat shocks at various heat pulse parameters T and D in a series of three experiments that partially overlapped Figure 2 ., The law was again present in all responses for temperatures between 41 . 5 °C and 42 . 5 °C , ( examples selected in Figure 3A , fit 3 , 4 ) ., For lower temperature ( 39 . 5 °C to 40 . 5 °C ) , the law was valid , but with poor 95% confidence intervals for estimated parameters a and b , as in Figure 3A , fit 1 , 2 ( the activity of the Hsp70 promoter was low ) ., At high temperatures or long durations ( Figure 3B ) , the double exponential law still explains the main characteristic of the stress response and is valid after a few hours from the end of the heat shock ., In the following , a theoretical model will be developed to explain the experimentally discovered law ., The exponential accumulation of the GFP shows that the derivative with respect to time of the mean GFP is proportional with itself:, There must be thus a molecular process , described by the exponential term abe−bt , which controls the heat shock response ., This theoretical suggestion is confirmed by previous studies of the heat shock system which revealed that the accumulation and subsequent degradation of the heat shock transcription factor 1 ( HSF1 ) regulates Hsp70 18–22 ., Experimental results 18 show that HSF1 activation is characterized by a rapid and transient increase in hsp70 transcription which parallels the kinetics of HSF1–DNA binding and inducible phosphorylation ., This rapid increase in HSF1–DNA binding activity reaches a maximal level and thereafter attenuates to a low level ., This rapid increase in activity followed by attenuation will form the starting point for our theoretical model ., An activation–accumulation two-component model will be developed as a minimal theoretical description of the empirical law ., The “activation” variable ( X1 ) represents the first phase of the heat shock response and includes components like HSF1–DNA binding activity ., X1 will increase during the duration of the heat shock and then , after the shock , will decrease with a lifetime proportional to parameter b ( Figure 4 ) ., The “accumulation” variable ( X2 ) includes the products of transcription and translation ., This second variable , at low levels before the shock , will gain momentum after the shock ., To connect the model with the experimental data , the GFP will be considered to be proportional with X2 ., The speed of accumulation of X2 , that is , dX2/dt , will be proportional to the product X1X2 ., Immediately after the shock , X1 has a big value ( the activation is high ) , and thus the speed of X2 is high ( the accumulation is in full thrust ) ., This will trigger an initial fast accumulation of GFP , which is proportional with X2 ., Later on , the activity X1 disappears , nullifying the product X1X2 and thus the speed of X2 ., The process is then terminated ( the accumulation stops ) ( Figure 4 ) ., The empirical law follows directly as a solution of the activation–accumulation system of equations:, with b and c as some constants ., Indeed , given the initial conditions X1 ( 0 ) and X2 ( 0 ) at a zero time reference t0 = 0 , the solution to this system of differential equations is, With the notation, the empirical law follows from X2 ( t ) :, The theoretical model contains two parameters: b and c ., Parameter b is directly accessible to experimental measurements , whereas parameter c is not; however , the product cX1 ( 0 ) which equals the product of a and b can be measured ., It is interesting to notice that the above time evolution can be re-expressed as a conservation law which is independent of any reference time ., For any two time points t1 and t2 , the following holds, At this point , there is no more information in the activation–accumulation description above than is in the empirical law ., However , one can search for more information hidden in the above two-component description by turning attention to the full data available , not only to the mean value of GFP ., For each sampled time , the full data available consists of measured GFP levels for at least 10 , 000 single cells ., These 10 , 000 single-cell measurements are typically distributed as in Figure 5 . There is a long tail at high values of GFP ., This biological variation in response to the stress is explained by turning the deterministic two-component system into a stochastic two-component system 6 , 7 ., The stochastic description must be completely enforced by ideas behind the deterministic two-component system ., The stochastic model is simple ., X1 is the mean value of a stochastic activation variable which will be denoted by q1 , X1 = 〈q1〉 ., After the heat shock , q1 will decrease with a probabilistic transition rate bq1 ., The activation–accumulation stochastic model is based on the same relation as before ( compare bq1 with bX1 ) , but now it describes the probabilistic transition rate and not a deterministic speed of attenuation ., By the same token , X2 is the mean value of q2 and its probabilistic accumulation rate is cq1q2 ., One notices that the transition probability rate cq1q2 is nonlinear in the variables q1 and q2 ., The stochastic two-component description is thus a mirror image of the deterministic two-component system ., However , the probabilistic system is more powerful as it predicts that the histograms of GFP ( proportional with q2 ) obtained from the flow cytometry measurements follow a Gamma distribution, with GFP ≡ x ., This prediction is confirmed experimentally ( Figure 5 ) ., The fact that the levels of proteins in gene networks tend to follow a Gamma distribution , which is a continuum version of a discrete negative-binomial distribution , was presented in 23 , 24 ., The papers 23 , 24 develop theoretical models describing the steady-state distribution of protein concentration in live cells ., Our interest lies in the non–steady-state behavior of these distributions ., Namely , the aim is to find the time evolution of the parameters that characterize these distributions ., The entire time evolution of the distributions is presented in Figure 6 . The distributions become wider as time passes ., The experimental data reveal that parameter ρ remains constant in time and only θ changes ., These experimental findings are theoretically explained in detail in the section Analysis of the Theoretical Model ., What follows summarizes the theoretical conclusions that are useful in understanding the experimental results of Figures 5 and 6 . The probability distribution for the discrete molecule number q2 , predicted by the stochastic activation–accumulation model , is the negative-binomial distribution ., This distribution appeared in earlier theoretical studies of genetic networks 23 , 24 and in physics 25 , 26 ., The GFP intensity is proportional with q2 and appears in measurements as a decimal number and not as a pure integer ., Thus , to describe the probability distribution of the GFP intensity , a continuous version of the discrete negative-binomial distribution is necessary ., This continuous version is the Gamma distribution observed experimentally in Figures 5 and 6 . The physical interpretation of parameter ρ will now be discussed ., At initial time t0 , immediately after the heat shock , there will be at least one cell from the entire cell population which contains the minimum number of molecules q2 ., Denote this number by N0 ., As the time passes , the molecule number q2 will grow , following the described stochastic process ., However , there is a nonzero probability , though extremely small , that the process of accumulation in one cell does not start even after 24 h ., This can happen in one of those cells that contain the minimum number of molecules q2 at the initial time t0 ., Thus , at any later time t > t0 , the lowest possible number of molecules q2 in a cell is N0 as it was at the initial time t0 ., It can be shown ( see the section Analysis of the Theoretical Model ) that ρ = N0 ., This explains the time independence of the experimental values of ρ; it also gives a physical meaning to ρ as being proportional to the minimum number of GFP molecules in a cell ., Parameter θ contains the time evolution of the stochastic accumulation of the GFP molecules ., This evolution can be again expressed as a time conservation property, valid between any two time points t1 and t2 ., The above relation Equation 10 contains parameters a and b and can be used to check the consistency of the model ., Using the data from Figure 6 , it follows that a = 3 . 159 with a 95% confidence interval ( 3 . 074 , 3 . 244 ) and b = 0 . 2572 with a 95% confidence interval ( 0 . 2358 , 0 . 2785 ) ., From the mean value for GFP , it results that a = 2 . 423 with a 95% confidence interval ( 2 . 351 , 2 . 496 ) and b = 0 . 2579 with a 95% confidence interval ( 0 . 2344 , 0 . 2814 ) ., Parameter a is sensitive to the estimation procedure , a phenomenon connected with the fact that parameter ρ is not perfectly constant but decreases a bit with time ., The mean value of the Gamma distribution is ρθ ., For a perfectly constant ρ , the estimated value for a would be the same using either the θ values or the ρθ data ., Contrary to parameter a , parameter b is independent of the way it is estimated , and the estimation is highly reliable ., To further check the reality of the Gamma distribution for heat shock response , a comparison of the Gamma fit with the lognormal fit is presented in Figure 7 . The lognormal was chosen because it can be viewed as a result of many random multiplicative biological processes ., A loglikelihood ratio less than 1 favors the Gamma distribution against the lognormal ., Moreover , at 37 °C the Gamma distribution is not a good fit ( loglikelihood ratio is bigger than 1 ) as it should be because the promoter is not active ., The law, is useful in making predictions for the fold induction to many other heat shock pulses ., For a heat pulse of a given temperature and duration , parameters a and b can be read out from Figure 8 . The constant level contours were inferred from the experimental data ., The level patterns differ; parameter a increases monotonically with the temperature and duration of the heat pulse ( Figure 8A ) , while the levels of parameter b form an unstable saddle shape pattern ( Figure 8B ) ., The conclusion of this section will be rephrased using a control theory perspective ., The end result of this paper is an input–output relation for the response of the CHO cells to heat shocks , together with a theoretical model that explains it ., The input signals are pulses of a precise time duration D and temperature height T . The output measured signals are the GFP intensity ., The input–output relation is given by the time-dependent probability density for GFP intensity, with, Parameters a and b are functions of the input signal , that is a = a ( T , D ) and b = b ( T , D ) ., The dependence of parameters a and b on temperature T and duration D is given by the contour plots of Figure 8 . The functional forms of a = a ( T , D ) and b = b ( T , D ) is a consequence of biological phenomena that take place during the heat shock ., We do not have a theoretical model for the phenomena that take place during the heat shock ., To explain the time evolution of the output variable ( GFP intensity ) , we developed a coarse-grained model for the heat shock response ., This coarse-grained model is valid for the biological phenomena that takes place after the end of the heat shock ., The model predicts the existence of a molecular factor that controls the GFP accumulation ( variable q1 ) ., We associated this theoretical factor with the heat shock factor HSF1-DNA binding activity ., The theoretical model is based on an activation variable q1 and an accumulation variable q2 ., The state of this two-component model is thus ( q1 , q2 ) , and any pair of positive integer numbers can be a possible state ., The main goal is to find the mean value and standard deviation for the activation and accumulation variable , respectively ., These quantities will be obtained from the equation for the probability P ( q1 , q2 , t ) that the system is in the state ( q1 , q2 ) at the time t ., The equation for P ( q1 , q2 , t ) depends on the multitude of transitions which can change a state ( q1 , q2 ) ., The experimental results suggest that two possible transitions change the state ( q1 , q2 ) ., One transition represents the decreasing of the activation variable from q1 to q1 − 1 ., On the state ( q1 , q2 ) , this attenuation appears as ( q1 , q2 ) → ( q1 − 1 , q2 ) , with an unaffected accumulation variable q2 ., The second transition will describe the accumulation of the accumulation variable from q2 to q2 + 1 ., On the state ( q1 , q2 ) , this accumulation appears as ( q1 , q2 ) → ( q1 , q2 + 1 ) , with the activation variable q1 now being unaffected ., A notation for the transition direction can be introduced: ɛ−1 = ( −1 , 0 ) ., The degradation transition can thus be written as ( q1 , q2 ) → ( q1 , q2 ) + ɛ−1 ., The negative sign in the index −1 is just a reminder of the fact that the transition reduces the number of molecules; the 1 in the subscript tells us that the transition is on the first variable ., Likewise , the accumulation transition can be expressed as ( q1 , q2 ) → ( q1 , q2 ) + ɛ2 and ɛ2 = ( 0 , 1 ) ., The index 2 is positive ( accumulation ) and is associated with the second component ., To find the probability P ( q1 , q2 , t ) , the transition probabilities per unit time are needed ., The experiment suggests we use, as the transition probability rate for the attenuation of the activation component , and, as the transition probability rate for the increasing of the accumulation component ., The stochastic model can be represented with the help of a molecular diagram 7 ( Figure 9 ) ., The components q1 and q2 are represented by ovals and the transitions by squares ., The lines that start from the center of a transition square represent the sign of that transition and point to the component on which the transition acts ., The transition ɛ−1 is negative , so the line ends in a bar and acts on q1 ., The transition ɛ2 is positive and so the line ends with an arrow; it acts on q2 ., The lines that stop on the edges of the transition squares represent the transition probability rates ., The line that starts from q1 and ends on ɛ−1 represents the transition probability rate bq1 ., In other words , the transition ɛ−1 is controlled by q1 ., The lines that start on q1 and q2 and merge together to end on ɛ2 represent the product cq1q2 , ( the merging point represents the mathematical operation of taking the product ) ., At this point , the theoretical model is fixed and what comes next is a sequence of computations to extract information out of it ., This information will be compared with the experimental results ., Given the transition rates , the equation for the probability P ( q1 , q2 , t ) is given by the following equation 7 , 25 ., The above equation for P ( q1 , q2 , t ) is not easy to solve ., We will use the method outlined in 6 , 7 and work with the function X ( z1 , z2 , t ) defined by, The equation for the function X ( z1 , z2 , t ) is a consequence of the equation for P ( q1 , q2 , t ) :, The goal is to find the time variation of the mean value and standard deviation for the activation and accumulation variable: 〈q1〉 , 〈q2〉 ,, ,, , 〈q1 , q2〉 , etc ., Here 〈〉 is a notation for the mean value with respect to the probability distribution P ( q1 , q2 , t ) ., From X ( z1 , z2 , t ) , the above mean values can be obtained by taking partial derivatives of X ( z1 , z2 , t ) at z1 =1 , z2 = 1 ., These partial derivatives are actually the factorial cumulants of the probability distribution P ( q1 , q2 , t ) ., In what follows , the sign =: means that the right side is introduced as a notation ., The equations for X1 ( t ) , X2 ( t ) , X11 ( t ) , and higher factorial cumulants result from the equation for X ( z1 , z2 , t ) :, The activation–accumulation model being nonlinear , the equations for the factorial cumulants cannot be reduced to a finite system of equations , unless some approximation technique is employed ., All third-order cumulants were discarded to obtain the above system of equations ., In 7 it was shown , using simulations , that the effect of discarding higher-order factorial cumulants is negligible ., The finite system thus obtained contains X1 , X2 , X12 , X11 , and X22 as variables ., Although it can be solved for X1 and X2 , we found that the influence of the correlation term X12 is small and cannot be experimentally detected in the GFP response ., Taken thus , X12 = 0 , and the system of equations is reduced to:, The solution to X22 from the four-equation system is, with k a constant determined from the initial value X2 ( t0 ) at some time t0 after the heat shock ., The solution can be restated in terms of the variance , Var , of the variable q2 ., The transformation from the factorial cumulants to Var is, And , thus , remembering that the mean value of q2 is X22 , it follows that, Such a relation between Var and Mean is satisfied by the negative-binomial distribution , a point to which we will return later ., Employing the general procedure , we continue to solve the system of equations for X1 , X2 , X11 , and X22 ., However , for the case of negligible X12 , the stochastic process is decoupled in two stochastic processes , each of which is exactly solvable ., It is thus useful to solve directly for the probability distribution of q2 at this point ., The transition probability rate for the first stochastic process ( for the activation component q1 ) is the same as before:, ., For the second one , it changes from, to, ( the coupling between q1 and q2 is through the mean value of q1 now ) ., This simplifies the problem of finding the distribution of q2 ., Denote the mean value of cq1 with g ( t ) , which acts actually like a signal generator on q2 6 , 7 ., The time variation of g ( t ) from the first equation in Equation 20 is, so the stochastic process for q2 now has an accumulation transition rate, The origin of time , t = 0 , is taken at the end of the heat shock , so X1 ( 0 ) represents the mean value of the activation variable at the end of the heat shock ., The probability P ( q2 , t ) to have q2 number of molecules at time t can be found from the master equation for this process, To find the solution , an initial condition P ( q2 , t0 ) must be specified ., The time t = t0 is some time taken after the heat shock pulse ( t0 > 0 ) , when the effects of the shock start to be detectable; it can be , for example , 30 min or 2 h after the pulse ., The probability distribution P ( q2 , t0 ) can be obtained , in principle , from the experimental values of GFP since GFP = fq2 ., There is an obstacle though: the proportionality factor f is unknown ., The factor f converts the number of molecules q2 into the laser intensity which is the output of the flow cytometry machine ., The conversion from the molecule numbers to the laser intensity can be more complicated than the proportionality relation GFP = fq2 ., For example , a background B can change the relation into: GFP = fq2 + B . We measured the GFP in regular CHO cells ( no HFP70-GFP construct ) and found that the background B is about 50 times less than the minimum intensity of GFP in the transfected CHO cells ., The settings of the flow cytometry instrument were set in a linear response range , and thus we will use the scaling relation GFP = fq2 to connect the flow cytometry readings with the number of molecules ., To conclude this initial condition discussion , in a perfect setting we would know the scaling factor f and then get P ( q2 , t0 ) from the measured data ., Because the scaling factor f is unknown , the problem will be solved in two steps ., The first step in choosing P ( q2 , t0 ) is based on a simple assumption: all cells have the same number of molecules q2 = N at the time t = t0 ., That is P ( q2 , t0 ) = δ ( q2 , N ) where δ is the Kronecker delta function ., The solution to Equation 26 with this initial condition is, Here q2 can take only values greater than N , q2 = N , N + 1 , ··· ., This distribution appeared in the study of cosmic rays 26 , and in the context of protein production was presented in 23 ., In terms of the variable x = q2 − N , it is known as the negative-binomial distribution , with interpretations that are not connected with the present problem ., The number N also represents the minimum possible number of molecules q2 in any cell ., This physical interpretation of N will be helpful in what follows ., The variable p ( t ) in the distribution is time-dependent , since the signal generator g ( t ) acts on q2:, The mean and variance for q2 are, from which follows Equation 23, Although the assumption that all the cells contain the same number of molecules at t = t0 is unreal , it produces a valuable outcome ., The negative-binomial distribution implies a Gamma distribution for the GFP intensity ( through the scaling relation GFP = fq2 ) , a fact to be discussed shortly ., Because the Gamma distribution is a good fit for the experimental data , we conclude that the negative-binomial is the correct solution for the distribution of the accumulation variable q2 ., The second step in choosing the probability distribution P ( q2 , t0 ) will be guided by the experimental results ., The experimental results show that the biological system passes through a chain of events from an unknown distribution of GFP before the heat shock , to a Gamma distribution at some time t0 after the heat shock ( 2 h , for example ) ., Also , the experiment shows that the distribution of GFP is Gamma at later times t > t0 ., In other words , the distribution of q2 becomes a negative-binomial at some time t0 after the heat shock and then afterward remains negative-binomial ., These experimental observations are mathematically explained by showing that a solution to Equation 26 with a negative-binomial distribution at t0 remains negative-binomial for all later times t > t0 ., Indeed , the solution to Equation 26 with a negative-binomial initial condition, is, which is a negative-binomial at all times t > t0 ., The number N0 is the minimum number of molecules q2 to be found in a cell at t0 and also at all later times t > t0 ( because q2 cannot decrease ) ., The time evolution of the mean 〈q2〉 is, and represents , using Equation 24 , the same empirical law ( Equation 8 ) as before ., To conclude , the dynamical system is such that once the cells enter into a negative-binomial distribution at some time after the heat shock , the distribution remains negative-binomial at later times ., As the time passes , all the distributions will have the same parameter N0 but different parameters p ( t ) ., To connect the theory with the experimental results , the probability distribution for the GFP intensity is needed ., This distribution is the continuum limit of the distribution for q2 ., It is a well-known fact that the continuum limit of a negative-binomial distribution is the Gamma distribution ., This continuum limit is presented here in order to find parameters ρ and θ , which can be experimentally measured ., The change from the integer variable q2 to the real variable fq2 is simple if advantage is taken of the fact that the common parameter N0 is a small number ., Parameter N0 is less than any possible molecule number q2 present in the system after the time t0 , q2 ≫ N0 ., Then , writing for simplicity p ( t ) as p ,, In the last step , we used the approximation 1 − y ≅ e−y for small values of y ., To go from the discrete variable q2 to the continuous variable GFP , we write the above relation as an equation for the probability density, with Δq2 = 1; then scale to GFP , ( GFP = f q2 ) ., The probability density, P℘ for GFP is then, This is a Gamma distribution for GFP ≡ x, with, From Equation 24 and 28 , we get, The mean value of the Gamma distribution is ρθ from which the empirical law Equation 8 follows ., The way the material is organized and presented in this paper is an outcome of a series of guiding principles imposed upon the project ., These guiding principles were formulated to keep in balance the experimental data with both the mathematical and biological models ., The guiding principles are:, 1 ) start from experimental measurements and discover an empirical law from data using signal generators as input into the system;, 2 ) build a simple mathematical model with as few parameters as possible to explain the empirical law;, 3 ) check the mathematical model using additional experimental information;, 4 ) use a general math

Introduction, Results, Discussion, Materials and Methods

In many biological systems , the interactions that describe the coupling between different units in a genetic network are nonlinear and stochastic ., We study the interplay between stochasticity and nonlinearity using the responses of Chinese hamster ovary ( CHO ) mammalian cells to different temperature shocks ., The experimental data show that the mean value response of a cell population can be described by a mathematical expression ( empirical law ) which is valid for a large range of heat shock conditions ., A nonlinear stochastic theoretical model was developed that explains the empirical law for the mean response ., Moreover , the theoretical model predicts a specific biological probability distribution of responses for a cell population ., The prediction was experimentally confirmed by measurements at the single-cell level ., The computational approach can be used to study other nonlinear stochastic biological phenomena .

The structure of an unknown biological system is uncovered by experimentally perturbing the system with a series of input signals ., The response to these perturbations is measured as output signals ., Then , the mathematical relation between the input and the output signals constitutes a model for the system ., As a result , a classification of biological molecular networks can be devised using their input–output functional relation ., This article studies the input–output functional form for the response to heat shocks in mammalian cells ., The Chinese hamster ovary ( CHO ) mammalian cells were perturbed with a series of heat pulses of precise duration and temperature ., The experimental data , taken at the single-cell level , revealed a simple and precise mathematical law for the time evolution of the heat shock response ., Parameters of the mathematical law can be experimentally measured and can be used by heat shock biologists to classify the heat shock response in different experimental conditions ., Since the response to heat shock is the outcome of a transcriptional factor control , it is highly probable that the empirical law is valid for other biological systems ., The mathematical model explains not only the mean value of the response but also the time evolution of its probability distribution in a cell population .

mus (mouse), computational biology

journal.pbio.1002210

The Discovery, Distribution, and Evolution of Viruses Associated with Drosophila melanogaster

Viral infections are universal , and virus-mediated selection may play a unique role in evolution 1 ., Viruses are also responsible for highly pathogenic diseases , and the detection , treatment , and prevention of viral disease are important research goals ., The model fly , Drosophila melanogaster , provides a valuable tool to understand the biology of viral infection 2 , 3 and antiviral immune responses in invertebrates 4 , 5 , as well as the interaction between viruses and their vectors 6 ., Drosophila has also helped elucidate the role of RNA interference ( RNAi ) as an antiviral defence 7 , 8 and has shown that endosymbiotic Wolbachia can protect against viruses 9 , 10 ., Recently , the Drosophilidae have been used to address important questions in virus evolution , including determinants of host-range and disease emergence 11–13 ., However , although Drosophila virus research has a long history , few D . melanogaster viruses are known in the wild 4 , 14 , and experiments using non-natural Drosophila pathogens may bias our understanding of immune function and its evolution 12 ., Following the discovery of Sigma Virus in D . melanogaster ( DMelSV , Rhabdoviridae; reviewed in 15 ) , classical virology surveys in the 1960s and 1970s uncovered Drosophila C Virus ( DCV , Dicistroviridae ) , Drosophila A Virus ( DAV , related to Permutotetraviridae ) , Drosophila X Virus ( DXV , Birnaviridae ) , DFV ( Reoviridae ) , DPV , and DGV ( unclassified ) in D . melanogaster 4 , 14 ., Subsequent transcriptomic studies of D . melanogaster identified D . melanogaster Nora Virus ( unclassified Picornavirales; 16 ) , and analyses of small RNAs from D . melanogaster cell culture 17 identified Drosophila Totivirus , American Nodavirus ( closely related to Flockhouse Virus ) , and Drosophila Birnavirus ( closely related to DXV ) ., However , only four of these viruses have been isolated from wild flies , have genome sequences available , and are available for experimental study ., These include DCV 18 , DMelSV 19 , DAV 20 , and Nora Virus 16 , while DXV is reported to be a cell culture contaminant 14 , 21 ., Of these four , only DMelSV has been widely studied in the field 15 , 22 , 23 ., Our limited knowledge of D . melanogaster’s natural viruses reflects a historically tight research focus on viruses with direct medical and economic impact ., While high-throughput “metagenomic” sequencing has broadened our knowledge of viral diversity in general 24 , 25 , most studies focus on vertebrate faeces 26 , 27 , potential reservoirs of human and livestock disease 28–30 , or crop plants 31 ., Relatively few studies have performed metagenomic virus discovery in invertebrates ( for a review , see 32 ) ., Recently , a large survey identified an exceptional and unsuspected diversity of negative sense RNA viruses associated with arthropods , suggesting that we may only have been scratching the surface of viral diversity 33 ., However , aside from some lepidopteran pests 34 and hymenopteran pollinators 35 , we still know little about the biology of most invertebrate viruses ., Although we have much to learn about Drosophila viruses , experiments using both natural and non-natural Drosophila pathogens have given us a better understanding of viral infection and immunity in Drosophila than in any other invertebrate 2 , 5 ., DCV , DXV , and Nora Virus have all been used to study the molecular biology of host–virus interaction 7 , 36–39 , and classical genetic approaches have elucidated the basis of host resistance to DCV and DMelSV 40–45 ., Many insect viruses—notably Cricket Paralysis Virus , Flock House Virus ( from beetles ) , Sindbis Virus ( mosquito-vectored ) , Vesicular Stomatitis Virus ( mosquito-vectored ) , and Invertebrate Iridovirus 6 ( from mosquitoes ) —have helped characterise the roles of the RNAi , IMD , Toll , autophagy , and Jak-Stat pathways in antiviral immunity ( see 5 for a review ) ., These studies show that Drosophila has a sophisticated and effective antiviral immune response , and both molecular 12 and population genetic 46–48 studies suggest that this immune system may be locked into an evolutionary arms race with viruses ., However , until we understand the diversity , distribution , or prevalence of viral infection in D . melanogaster , it is hard to put these results into their evolutionary or ecological context ., Here we use a metagenomic approach to identify more than 20 novel viruses associated with D . melanogaster , including the first DNA virus to be identified in D . melanogaster ., Based on the presence of virus-derived 21 nucleotide ( nt ) small RNAs ( which are characteristic of an antiviral RNAi response in Drosophila 49 ) , we argue that these sequences represent active viral infections ., Using a survey of individual wild-collected flies , we give the first quantitative estimates of prevalence for 15 different viruses in D . melanogaster and its close relative D . simulans , and rates of co-occurrence with the Wolbachia bacterial endosymbiont ., In addition , by examining publicly available RNA datasets , we catalogue the presence of these viruses in D . melanogaster stock lines and cell culture ., Our results provide an unprecedented insight into the virus community of Drosophila , and thereby provide the evolutionary and ecological context needed to develop Drosophila as a model for virus research ., We used metagenomic sequencing of ribosome-depleted total RNA to identify virus-like sequences in five large collections of wild-caught adult Drosophilidae ., Three collections ( denoted E , K , and I ) were sampled from fruit baits in Kilifi ( Kenya ) and Ithaca ( New York , United States ) ., The aliquots pooled for total RNA sequencing represented around 2 , 000 individuals of D . melanogaster , D . ananassae , D . malerkotliana , and Scaptodrosophila latifasciaeformis ., Two collections ( S and T ) were sampled from fruit baits in southern England , and aliquots pooled for total RNA sequencing represented around 3 , 000 D . melanogaster ( <1% other Drosophila ) ., In total 0 . 5% of all reads mapped to DAV , DCV , DMelSV , and Nora Virus ( S1 Fig and S1 Table ) ., Viral read numbers varied dramatically between samples , with DCV absent from pool EIK and DAV absent from pool ST ( verified by qPCR; S2 Fig ) ., Only three reads mapped to DXV , and no reads mapped to Drosophila Totivirus , Drosophila Birnavirus , or American Nodavirus ( S1 Fig and S1 Table ) ., As DXV is reported to be a cell culture contaminant 14 , and the other unmapped viruses were described from cell culture 17 , this suggests that these viruses may be rare in wild flies ., To discover novel viruses , we assembled all RNA-seq reads de novo using Trinity 50 and Oases 51 , and used the Basic Local Alignment Search Tool ( BLAST ) 52 against the Genbank protein reference sequences 53 to identify virus-like sequences ., Raw de novo metagenomic contigs are provided in S2 Data ., Virus-like contigs were supplemented with PCR and targeted Sanger sequencing to improve completeness , and in total we identified more than 20 partial viral genomes ., Those sequences that could be unambiguously associated with D . melanogaster rather than other Drosophila species were provisionally named according to collection locations ., Based on sequence similarity , these “BLAST-candidate” viruses included two Reoviruses ( “Bloomfield Virus” and “Torrey Pines Virus” ) , three Flaviviruses ( “Charvil Virus” , and two others ) , a Permutotetravirus ( “Newfield Virus” ) , a Nodavirus ( “Craigie’s Hill Virus” ) , a Negevirus , a Bunyavirus , two Iflaviruses ( “La Jolla Virus” and “Twyford Virus” ) , two Picorna-like viruses ( “Thika Virus” and “Kilifi Virus” ) , a virus related to Chronic Bee Paralysis Virus ( “Dansoman Virus” ) , a virus related to Sobemoviruses and Poleroviruses ( “Motts Mill Virus” ) , six Partitiviruses , and a Nudivirus ( “Kallithea Virus” ) ., These novel viruses constituted a further 3 . 3% of RNA-seq reads , taking the viral total to 3 . 8% of all reads ( S1 Fig ) ., Further details of the new viruses are given in Table 1 and S2 Table , and virus sequences have been submitted to Genbank as KP714070-KP714108 and KP757922- KP757936 ., We note that fragments of Bloomfield Virus and Thika Virus were simultaneously identified in 54 , there denoted DRV and DUV , respectively ., To place the virus sequences in a phylogenetic context , we subjected conserved regions to phylogenetic analysis along with known viruses and uncurated viral sequences from the NCBI Transcriptome Shotgun Assemblies 55 ., Kallithea Virus , a DNA virus , was closely related to Nudiviruses from Drosophila innubila 56 and the beetle Oryctes rhinoceros 57 ., Most of the new RNA viruses were relatives of known or suspected insect pathogens ( Table 1 and S3 Fig ) ., For example , based on polymerase sequences , Torrey Pines Virus is distantly related to Aedes pseudoscutellaris reovirus and several lepidopteran Cypoviruses , while Dansoman Virus is related to Chronic Bee Paralysis Virus ., Others were distantly related to arthropod viruses , but were close to uncurated transcriptome sequences ( Fig 1 ) ., For example , Bloomfield Virus is closely related to transcriptome sequences from the flies Delia antiqua and Teleopsis dalmanni ( 63% AA identity in the replicase ) , and distantly related to Nilaparvata lugens reovirus ., Similarly , Motts Mill Virus is related to the recently described Ixodes scapularis ( tick ) associated viruses 1 and 2 58 , but is closer to uncurated transcriptomic sequences from three bees and the bobtail squid Euprymna scolopes ., Together these sequences appear to represent a novel clade related to plant Sobemoviruses and Poleroviruses ( Fig 1 ) ., Strikingly , the six unnamed Partitivirus-like polymerases were not related to any known insect pathogens ( known Partitiviruses are pathogens of plants and fungi ) , but were related to uncurated sequences from arthropod transcriptomes—again suggesting a novel lineage of insect viruses ( Fig 1 ) ., The close relationship between the newly identified virus-like sequences and known arthropod viruses suggests that they are likely to be Drosophila pathogens , and some may be previously described viruses that lack sequence data ., For example , Kilifi Virus or Thika Virus ( related to the Picornavirales; S3 Fig ) may correspond to DPV—which was reported to have a 25–30 nm particle and infect gut tissues 59 ., Similarly , either Torrey Pines Virus or Bloomfield Virus could correspond to Drosophila F Virus 14 , Drosophila K Virus or other reoviruses 21 ., However , as those viruses were described only from capsid morphology , density , and serology , it would be challenging to conclusively link them with the novel sequences presented here ., Although phylogenetic analyses show that these sequences are viral in origin , they may represent Endogenous Viral Elements integrated into the Drosophila genome ( “fossil” viruses , or EVEs 60 ) , or they may derive from gut contents or surface contamination rather than active infections ., To exclude the possibility that these sequences represent EVEs segregating in D . melanogaster , we mapped the raw genomic reads from 527 distinct D . melanogaster genomes 61 to our set of BLAST-candidate viruses and confirmed that no genome mapped at a rate high enough to be consistent with a genomic copy of any virus in that individual ., As this test for EVE status was not possible for other Drosophila species , only sequences associated with D . melanogaster were named as viruses , and others ( which could potentially represent EVEs in other taxa ) were denoted “virus-like . ”, To test if these sequences derive from active viral infections , we additionally sequenced all 17–29 nt small-RNAs from the EIK and ST pools , reasoning that virus-like sequences will only be processed into 21 nt siRNAs ( viRNAs , derived from replicative intermediates by Dicer-2 ) if they represent active infections within host cells 49 , 62 ., In total ca ., 7% of all 17–29 nt small-RNAs derived from DAV , DCV , DMelSV , and Nora Virus , and ca ., 9% derived from the new “BLAST-candidate” viruses ., As expected , for most viruses the viRNA size distribution was tightly centred on 21 nt reads and included reads from both the genomic and complementary strands , consistent with active viral infections processed by antiviral RNAi in Drosophila ( Figs 2 , S4 and S5 ) ., Even if viral sequences do not represent EVEs or inactive contaminants , they could instead be active infections of Drosophila-associated microbiota rather than Drosophila ., However , the 21 nt viRNAs observed for the majority of these viruses are inconsistent with viral infection of likely parasites or parasitoids such as hymenoptera , chelicerata , or nematodes , which have predominantly 22 nt viRNAs 63 , 64 ., And , while 21 nt viRNAs could derive from viral infections of some fungi 65 or from eukaryotes with uncharacterised antiviral RNAi , in most cases the phylogenetic position of these viruses , high read numbers , and/or their appearance in laboratory fly stocks and cell culture ( below ) argue in favour of Drosophila as the host ., In addition to the viral candidates identified by BLAST , we reasoned that contigs which lack BLAST similarity to reference sequences , but which display a signature of Dcr-2 processing ( high levels of 21–23 nt siRNAs and low levels of 25–29 nt piRNAs ) , may also be viral in origin ( Fig 3; This approach was very recently advocated in 54 , but see also 17 , 65 ) ., Using these small RNA criteria we identified a list of “siRNA-candidate” contigs that are potentially viral in origin , but could not be placed within a phylogeny of known viruses ., Several siRNA-candidate viruses identified in this way were subsequently attributable as fragments of the BLAST-candidate viruses by other means ( below ) ., Of those that could not be attributed to viral genomes , two were provisionally named ( Chaq Virus and Galbut Virus ) and the remaining 57 contigs were submitted to Genbank as uncultured environmental virus sequences ( KP757937–KP757993 ) ., These unnamed siRNA-candidate viruses contribute 0 . 2% of all RNAseq reads , and 1% of 17–29 nt small RNAs in the EIK and ST pools ( S1 Fig and S1 Table ) ., As with the BLAST-candidate viruses , these siRNA-candidate viruses were absent from D . melanogaster genomic reads and their siRNAs were predominantly 21 nt in length and derived from both strands ( S6 Fig ) , again suggesting that they represent active viral infections of Drosophila ., Although the siRNA-candidate viruses displayed no strong BLAST similarity to known viruses , Galbut Virus and siRNA-candidate 24 display weak similarity to Nilaparvata lugens Commensal X Virus ( a satellite virus with unknown helper 66 ) , and Galbut and Chaq viruses appear to be related to uncurated sequences present in a diverse set of arthropod transcriptome shotgun datasets ( Fig 1 and Table 1; S2 Table and S3 Fig ) ., To test whether viruses vary in their small RNA profile , we analysed small RNAs from the EIK and ST pools , and from two libraries for each of the five collections ( E , I , K , S , and T ) , with and without “High Definition” ligation adaptors designed to reduce ligation bias 67 ., Overall , the number of viRNA reads per RNAseq read varied substantially among viruses ( “viRNA ratio;” Figs 3 and S7 ) , with Twyford Virus and Motts Mill Virus giving rise to more than a 1 , 000-fold additional 21–23 nt viRNAs than DCV , and DMelSV giving rise to nearly 7 , 000-fold more ( S7 Fig ) ., For viruses present in both EIK and ST , the viRNA ratios were highly correlated between pools ( rank correlation ρ > 0 . 99; S7 Fig ) , suggesting that they are repeatable ., This may reflect differences in the proportion of non-replicating viral genomes ( e . g . , encapsidated viruses in the gut lumen ) which can contribute to RNAseq but are not actively processed by Dcr-2 ., Alternatively , differences could result from the action of viral suppressors of RNAi ( VSRs ) , such as those encoded by DCV and Nora Virus 7 , 38 ., The sizes and strand bias of small RNAs also varied substantially among viruses , although small RNA reads from the majority of RNA viruses were biased toward the positive strand ( 50%–70% ) and to 21 nt in length ( Fig 2 ) , as expected for virus-derived small RNAs in Drosophila 49 ., Exceptions included Nora Virus , DCV , Kilifi Virus , and Thika Virus , which showed a stronger positive-strand bias ( 85%–95% of reads ) and a broad size range of positive-sense viRNAs peaking at 21 nt , with a wide “shoulder” from 23 to 27 nt ( Fig 2 ) ., This size distribution is not seen in most DCV infections of cell culture 49 , and although 26–30 nt viral piRNAs have been reported in OSS cell culture 17 , the 25–28 nt reads identified here did not display the 5′ U bias expected of piRNAs ( Fig 2 ) 68 ., As these four picorna-like viruses also displayed low numbers of viRNAs ( Fig 3; S1 Table and S4 Fig ) , this is consistent with a difference in the way that the viral genome and/or viRNAs are processed—perhaps reflecting a higher fraction of non-specific degradation products for these viruses ., However , because we found that viRNA properties were reproducible across sequencing libraries ( S4 Fig ) , and because viRNAs were sequenced from large pooled samples composed of mixed infections , these profiles cannot result from idiosyncrasies of RNA extraction or library preparation ., In contrast to the other RNA viruses , Twyford Virus displayed unusual viRNAs ., Although Twyford Virus is an Iflavirus ( S3 Fig ) and thus has a positive sense ssRNA genome , the strand bias was strongly negative and the viRNAs peaked sharply at 22–23 nt ( cf . 21 nt for other RNA viruses ) ., In addition , although other viruses displayed no strong 5′ base-composition bias except for a weak bias against 5′ G , most Twyford Virus viRNAs were 5′ U , as is seen for piRNAs ( but lacking the A at position 10 expected of Drosophila piRNAs; S5 Fig ) ., This bias is not due to small sample size or low sequence diversity , as we saw >9 , 000 unique sequences , and 3 , 500 of those were seen more than once ., Comparison with the virus genome showed the 5′ U bias was driven by differential production or retention of viRNAs , rather than subsequent editing or addition ., Interestingly , the 3′ position was also slightly enriched for U ( S7 Fig ) , and a substantial fraction of 3′ U were non-templated , indicative of 3′ uridylation as seen in D . melanogaster and other species in the absence of the Hen-1 methytransferase 69 ., These observations suggest that Twyford Virus may not be processed by the Drosophila Ago2-Dcr2 pathway , and could instead represent viral infection of an unknown eukaryotic commensal ., Nevertheless , potential arthropod parasites such as chelicerata have not been reported to display this pattern of viRNAs , and although the 5′ U is reminiscent of the 21U piRNA of C . elegans and related nematodes 70 , neither Rhabditid nor Tylenchid nematodes could be detected by PCR in individual wild-caught D . melanogaster carrying Twyford Virus ., Similarly , while 22 nt 5′-U small RNAs are known from the filamentous fungus Neurospora 71 , those are derived from the host genome and are not associated with viral infection ., Thus , although speculative , if these 22–23U viRNAs cannot be explained by a non drosophilid host , then it is possible that they reflect a previously unrecognised tissue-specific phenomenon in Drosophila , or one associated with the action of a novel suppressor of RNAi ( e . g . , suppression of Hen-1 ) ., The siRNA-Candidate 14 ( KP757950 ) shows a similar pattern ( 22–23 nt peak in viRNAs , 5′ U ) , suggesting it forms part of the same virus and/or is processed in the same way ( S6 Fig ) ., We also identified many 21 nt viRNAs widely dispersed around the Kallithea Virus genome ., This is consistent with an antiviral RNAi response against this DNA Nudivirus , as has been previously reported for Invertebrate Iridovirus 6 artificially infecting D . melanogaster 72 ., As DNA viruses often encode miRNAs 73 , 74 , and miRNAs have been implicated in the establishment of latency in Heliothis zea Nudivirus 75 , we screened small RNAs from Kallithea Virus for potential virus-encoded miRNAs ., One 22 nt RNA sequence was highly abundant ( S5 Fig ) , and is predicted to represent the 5′ miRNA from a pre-miRNA-like hairpin ( miRDeep2 76; S1 Text ) ., The predicted mature 5′ miRNA ( AUAGUUGUAGUGGCAUUAAUUG ) represented >35% of all small RNAs derived from Kallithea Virus , while the 3′ RNA “star” sequence represented 0 . 3% of reads ., This sequence was less highly represented , relative to 21nt viRNAs from Kallithea Virus , in an oxidised library , consistent with the absence of 2′O-methylation at the 3′ end , as expected for miRNAs in Drosophila ( S4 Fig ) 77 ., The seed region displays no obvious similarity to known miRNAs ( although positions 5–17 are similar to dme-miR-33-5p ) , but a survey of potential binding sites in D . melanogaster using miRanda 3 . 3a 78 identified 522 genes with at least one potential binding site in the 3′-UTR ( miRanda score ≥150 ) ., These were highly enriched for Gene Ontology terms that might be associated with viral function ( including , amongst others , Regulation of Gene Expression , Cell development , mRNA binding , and Plasma Membrane; S3 Table ) ., This miRNA could alternatively regulate virus gene expression 75 , and 21 potential binding sites were identified in the Kallithea Virus genome ., While predicted miRNA target sites include many false positives 79 , and experimental work would be required to confirm a biological role , it is interesting to note that Kallithea Virus’ closest relative ( Oryctes rhinoceros Nudivirus ) does not encode a detectable homolog of the miRNA and contains only four predicted binding sites ( miRanda score ≥150 ) ., To detect the BLAST-candidate and siRNA-candidate viruses in previously published Drosophila studies , we mapped up to 2 million reads from each of 9 , 656 publicly available fly and cell culture RNAseq and small-RNA datasets to the new and previously described Drosophila viruses ., Around 33% of “run” datasets contained viral reads above a threshold of 100 reads per million ( rpm ) , representing 39% of “samples” and 58% of submitted “projects” ( Figs 4 and S8 and S5 Data ) ., The proportion of positive samples varied with log-threshold , so that 53% had at least one virus at ≥10 rpm , but 17% of runs had at least one virus at ≥1 , 000 rpm ., These rates are slightly lower than , but not dissimilar to , previous estimates from serial passage of fly stocks , which found around 40% of fly stocks were infected 14 ., BLAST-candidate and siRNA candidate viruses were both found , and by noting their co-occurrence , we were able to identify several siRNA-candidates as component parts of other virus genomes ( e . g . , segments of Bloomfield Virus and the second segment of Craigie’s Hill Virus were initially identified as siRNA Candidate Viruses ) ., The presence of BLAST-candidate and siRNA-candidate viruses in laboratory cultures supports these as bona fide infections of Drosophila and demonstrates the utility of the viRNA signature as a marker for viruses ., Based on the 100 rpm threshold , DAV ( 1 , 025 of 9 , 656 datasets ) , DCV ( 979 datasets ) , Nora Virus ( 629 datasets ) , Newfield Virus ( 483 datasets ) , FHV , Drosophila Totivirus , American Nodavirus , Drosophila Birnavirus , DMelSV , Thika Virus , Kilifi Virus , La Jolla Virus , Craigie’s Hill Virus , Bloomfield Virus , Chaq Virus and Galbut Virus were all present in public datasets ( Figs 4 and S8 and S5 Data ) ., However , some viruses ( Kilifi Virus , Craigie’s Hill Virus , Chaq Virus , Galbut Virus , DMelSV ) were extremely rare , appearing in 12 or fewer datasets ., It is widely known that Drosophila cell culture harbours many viruses 21 , and we identified multiple viruses and occasionally high numbers of viral reads in cell culture datasets ., For example , reads mapping to nine different viruses were found in datasets SRR770283 and SRR770284 ( Piwi CLIP-Seq in OSS cells ) 80 and ≥70% of reads from SRR609669-SRR609671 were viral in origin ( total RNA from piRNA-pathway knock-downs ) 81 ., Virus presence/absence for widely used Drosophila Cell cultures 82 is presented in S9 Fig . RNAseq datasets were also available for species other than D . melanogaster , and these , too , included virus-like sequences ., We detected DAV in D . pseudoobscura , D . virilis , D . bipectinata , D . ercepeae and D . willistoni , DCV in D . simulans , D . ananassae and D . mojavensis , Nora Virus in D . simulans , D . ananassae , and D . mojavensis , Thika Virus in D . virilis and D . ficusphila , Kilifi Virus in D . bipectinata , La Jolla Virus in D . simulans , and Bloomfield Virus in D . virilis ., Given the presence of viruses in public RNAseq and siRNA datasets , we selected a subset of 2 , 188 datasets to perform viral discovery by de novo assembly ., In adult D . melanogaster this identified a novel Picorna-like virus ( present in three datasets among 9 , 656 ) and a novel Negevirus ( present in 10 datasets among 9 , 656 ) ., We have provisionally named these Berkeley Virus and Brandeis Virus , respectively ( Table 1 , S2 Table and S6 Data ) ., The survey also identified a novel Totivirus and several ( possibly fragmentary or non-coding ) Reovirus segments in cell culture , at least one of which is widespread ( 214 datasets; Figs 4 and S8 and S5 Data ) ., The small number of novel viruses we identified in fly stocks over and above those described previously , may suggest that few further Drosophila viruses remain to be found regularly infecting laboratory stocks ., To infer viral prevalence and distribution in wild flies we used reverse transcription PCR ( RT-PCR ) to assay for the presence of 16 different viruses in a total of 1 , 635 D . melanogaster and 658 D . simulans adults sampled from 17 locations across the world ( Fig 5 and S4 Table ) ., Excluding siRNA-candidate viruses , the most prevalent virus in large samples of D . melanogaster was La Jolla Virus ( 12/16 locations , 8 . 6% of flies averaged across locations ) and the rarest was Twyford Virus ( 1/16 locations , average 0 . 3% of flies ) ., Of those detected in large samples of D . simulans , the most prevalent was Thika Virus ( 4/7 locations , average 4 . 5% of flies ) while the rarest was Motts Mill Virus ( 1/7 locations , average 0 . 2% of flies ) ., Despite their high prevalence in the lab and presence in the metagenomic pools , DCV and Newfield Virus were not detected at any of the 17 locations , and Twyford Virus , DMelSV , Dansoman Virus , and Craigie’s Hill Virus were not detected in D . simulans ( Fig 5 and S4 Table ) ., Although sampling locations varied substantially in overall viral prevalence ( in Athens GA >80% of D . melanogaster carried a virus; in Marrakesh less than 10% ) there was no clear geographic structure in viral prevalence ( Figs 5 and S10 ) , and for most viruses prevalence was not correlated between D . melanogaster and D . simulans from the same location ., Excluding siRNA-candidate viruses , around 30% of D . melanogaster individuals and 13% of D . simulans individuals carried at least one virus , and over 6% of D . melanogaster individuals carried more than one virus ( S11 Fig ) ., We are unable to explain the unusually high viral prevalence in D . melanogaster sampled from Athens ( Georgia , US ) —flies were separated to individual vials within a few hours , and D . simulans and D . melanogaster were netted together , but D . simulans did not show unusually high virus prevalence ( S10 Fig ) ., In contrast to the other viruses , the novel siRNA-candidate viruses Galbut Virus and Chaq Virus often displayed extremely high prevalence ., In D . melanogaster , Galbut Virus ranged from 13% in Plettenberg to 100% in Accra , and Chaq Virus from <5% in Edinburgh to 35% in Porto ., In D . simulans , Galbut Virus ranged from 57% ( Athens ) to 76% ( Torquay , Australia ) , although Chaq Virus was not highly prevalent ( only 2/7 locations , at low prevalence ) ., These rates are much higher than for the other viruses , and the inclusion of siRNA candidate viruses in overall infection rates brings many populations to ≥70% of flies carrying at least one virus ( Figs 5 and S11 ) ., The high prevalence of these “siRNA-candidate” viruses is surprising and could perhaps imply an alternative ( non-virus ) origin for the sequences ., However , we believe a viral origin is supported by a combination of the viral-like siRNA signature , their absence from D . melanogaster genomic reads , their close relationship to unclassified insect transcriptome sequences ( Figs 1 and S3 ) , and their occasionally low prevalence ( S10 Fig ) ., For DMelSV , which is the only virus previously surveyed on a large scale 15 , our data agree closely with earlier estimates based on other assays: here 4 . 6% of D . melanogaster infected versus 2 . 8% in 22 and 5 . 0% in 23 ( DMelSV is absent from D . simulans ) ., Nevertheless , our estimates more generally should be treated with some caution as RT-PCR assays are unlikely to be reliable for all virus genotypes , leading to PCR failure for divergent haplotypes and thus potentially underestimation of prevalence ., Virus prevalence was strikingly different between publicly available RNA datasets and our field survey ( Fig 4 ) ., For example , we identified Newfield Virus and DCV in many fly stocks and cell cultures but very rarely in the field ( Fig 4; also compare Figs 5 and S8 ) , whereas Kallithea Virus and Motts Mill Virus were common in the wild ( 4 . 6% and 6 . 7% global average prevalence in D . melanogaster; Fig, 5 ) but absent from DNA and RNA public datasets ., The case of DCV is particularly striking as early surveys assayed by serial passage in laboratory cultures suggested DCV may be common in the field 14 , as did PCR surveys of recently established laboratory stocks 83 ., In the case of Newfield Virus and DCV , a downward collection bias ( for example , if high titre flies do not get collected ) may explain the result ., However , for viruses that have higher prevalence in the field than the lab such as Kallithea Virus and Motts Mill Virus , and also the siRNA candidates Galbut Virus and Chaq Virus , it seems likely that differences in ( e . g . ) transmission ecology between the lab and field may explain the disparity ., Infection with the bacterial endosymbiont Wolbachia pipientis has previously been shown to confer protection against secondary infection by some RNA 9 , 10 but not DNA 84 viruses in insects ., We therefore surveyed Wolbachia prevalence by PCR in the wild flies , and tested whether Wolbachia was correlated with virus presence ., We found Wolbachia at detectable levels in all populations , ranging from 1 . 6% of individuals ( Accra ) to 98% ( Edinburgh ) in D . melanogaster and from 82% ( Plettenberg ) to 100% ( Marseille ) in D . simulans ( Figs 5 and S10 ) ., As expected 85 , overall Wolbachia prevalence was higher in D . simulans ( approximately 90% ) than in D . melanogaster ( approximately 50% ) ., We could not detect any consistent correlation across populations between the prevalence of Wolbachia and that of any virus ( S12 Fig ) , nor could we detect any association between Wolbachia and viral infection status within populations ( contingency table tests with p-values combined across populations using Fisher’s method; all p > 0 . 05 ) ., While this may indicate that Wolbachia-mediated antiviral protection is not an important determinant of viral infection in the field , our relatively small sample sizes ( median n = 63 flies per location ) mean that our power to detect an interaction is low ., In addition , it is unclear whether a positive or negative association should be expected , as some Wolbachia-host combinations appear to protect via increased tolerance rather than increased resistance 86 , not all strains are highly pro

Introduction, Results and Discussion, Materials and Methods

Drosophila melanogaster is a valuable invertebrate model for viral infection and antiviral immunity , and is a focus for studies of insect-virus coevolution ., Here we use a metagenomic approach to identify more than 20 previously undetected RNA viruses and a DNA virus associated with wild D . melanogaster ., These viruses not only include distant relatives of known insect pathogens but also novel groups of insect-infecting viruses ., By sequencing virus-derived small RNAs , we show that the viruses represent active infections of Drosophila ., We find that the RNA viruses differ in the number and properties of their small RNAs , and we detect both siRNAs and a novel miRNA from the DNA virus ., Analysis of small RNAs also allows us to identify putative viral sequences that lack detectable sequence similarity to known viruses ., By surveying >2 , 000 individually collected wild adult Drosophila we show that more than 30% of D . melanogaster carry a detectable virus , and more than 6% carry multiple viruses ., However , despite a high prevalence of the Wolbachia endosymbiont—which is known to be protective against virus infections in Drosophila—we were unable to detect any relationship between the presence of Wolbachia and the presence of any virus ., Using publicly available RNA-seq datasets , we show that the community of viruses in Drosophila laboratories is very different from that seen in the wild , but that some of the newly discovered viruses are nevertheless widespread in laboratory lines and are ubiquitous in cell culture ., By sequencing viruses from individual wild-collected flies we show that some viruses are shared between D . melanogaster and D . simulans ., Our results provide an essential evolutionary and ecological context for host–virus interaction in Drosophila , and the newly reported viral sequences will help develop D . melanogaster further as a model for molecular and evolutionary virus research .

The fruit fly Drosophila melanogaster is extensively used as a model species for molecular biology and genetics ., It is also widely studied for its evolutionary history , helping us understand how natural selection has shaped the genome ., Drosophila research has been particularly valuable in determining how the insect immune system interacts with viruses and how co-evolution between hosts and viruses can shape the host immune system ., Understanding insect–virus coevolution is important because some viruses—such as those which cause dengue and yellow fever in humans—also infect their insect vectors , and because the viruses of bees and other pollinators are implicated in pollinator decline ., Although we have an increasingly good idea of how flies recognise and combat viral pathogens , we still have much to learn about the viruses they encounter and interact with in the wild ., In this paper , we sequence all of the genetic material from a large collection of wild fruit flies and use it to identify more than 20 new viruses ., We then survey individual wild flies and laboratory stocks to find out which viruses are common , which are rare , and which species of fruit fly they infect ., Our results provide valuable tools and an evolutionary and ecological perspective that will help to improve Drosophila as a model for host–virus biology in the future .

Sequencing of metagenomic RNA and small RNA identifies more than 20 new viruses associated with the fruit fly Drosophila melanogaster, and large-scale surveys show that many are common in the lab and in the field.

journal.pgen.1002091

Genome-Wide Scan Identifies TNIP1, PSORS1C1, and RHOB as Novel Risk Loci for Systemic Sclerosis

Systemic sclerosis ( MIM181750 ) is a connective tissue disease characterized by generalized microangiopathy , severe immunologic alterations and massive deposits of matrix components in the connective tissue ., Being an orphan disease , SSc presents a major medical challenge and is recognized as the most severe connective tissue disorder with high risk of premature deaths 1 ., Epidemiological data on SSc vary in different parts of the world and depend on selection criteria for the study population ., Inasmuch , the prevalence of the disease fluctuates across global regions and population-based studies result in higher prevalence than do hospital records-based studies ., In North America , the prevalence of SSc has been reported as 0 . 7–2 . 8 per 10 , 000 in a Canadian study , whereas in the U . S . figures of 2 . 6 per 10 , 000 versus 7 . 5 per 10 , 000 were reported by medical records - versus population-based studies , respectively ., In Europe , a prevalence of 1 . 6 per 10 , 000 was reported in Denmark , 3 . 5 per 10 , 000 in Estonia , 1 . 58 per 10 , 000 adults ( 95% confidence interval , 129–187 ) in Seine-Saint-Denis in France 2–4 ., The risk of SSc is increased among first-degree relatives of patients , compared to the general population ., In a study of 703 families in the US , including 11 multiplex SSc families , the familial relative risk in first-degree relatives was about 13 , with a 1 . 6% recurrence rate , compared to 0 . 026% in the general population 5 ., The sibling risk ratio was about 15 ( ranging from 10 to 27 across cohorts ) ., The only twin study reported to date included 42 twin pairs 6 ., The data showed a similar concordance rate in monozygotic twins ( 4 . 2% , n\u200a=\u200a24 ) and dizygotic twins ( 5 . 6% ) ( NS ) and an overall cross-sectional concordance rate of 4 . 7% ., However , concordance for the presence of antinuclear antibodies was significantly higher in the monozygotic twins ( 90% ) than in the dizygotic twins ( 40% ) suggesting that genetics may be important for the auto-immune part of the disease ., The aetiology of SSc is still unclear but some key steps have been described , in particular early endothelial damage and dysregulation of the immune system with abnormal autoantibody production 7 ., At the cellular level , early events include endothelial injury and perivascular inflammation with the release of a large array of inflammatory mediators 8 , 9 ., In the advanced stage , a progressive activation of fibroblasts in the skin and in internal organs leads to hyperproduction of collagen and irreversible tissue fibrosis 9 ., Epidemiological investigations indicate that SSc follows a pattern of multifactorial inheritance 10 ., Previous candidate-gene association studies have only identified a handful of SSc risk loci , most contributing to the genetic susceptibility of other autoimmune diseases 9–16 ., So far , two genome-wide association studies of SSc have been conducted 17 , 18 ., The studies differ according to the ancestry of the studied population ( Korean vs US/European ) and the genome-wide association data: map density ( ∼440 K vs 280 K SNPs ) and sample size ( ∼700 vs ∼7300 subjects ) ., They provided evidence of association with known MHC loci , but only one ‘new’ locus was identified at CD247 in the US/European dataset , variants at CD247 being known to contribute to the susceptibility of systemic lupus erythematosus 18 ., The diagnosis of SSc is based on recognized clinical criteria established decades ago however , these do not include specific autoantibodies or recent tools for assessment of the disease 19 , 20 ., Therefore , phenotypic heterogeneity is a concern for SSc and genetic heterogeneity is also highly probable with regards to data obtained in other connective tissue disorders ., Given these considerations , and previous findings in other autoimmune diseases , it is apparent that additional risk variants for SSc remain to be discovered ., Therefore , to identify further common variants that contribute to SSc risk in the European population , we conducted a two-stage GWAS , in two case-control samples ( total >8 , 800 subjects ) ., We established a collaborative consortium including groups from 4 European countries ( France , Italy , Germany and Eastern-Europe ) from which we were able to draw upon a combined sample of over 8 , 800 subjects ( before quality control ) and conducted a two-stage genome-wide association study ., In stage 1 , we genotyped 1 , 185 samples on Illumina Human610-Quad BeadChip and genotypes obtained using the same chip from 2 , 003 control subjects were made available to us from the 3C study 21 , 22 ., After stringent quality control , we finally tested for association in stage-1 , 489 , 814 autosomal SNPs in 2 , 340 subjects ( 564 cases and 1 , 776 controls ) ( Table 1 ) ., We tested for association between each SNP and SSc using the logistic regression association test , assuming additive genetic effects ., The quantile-quantile plot and estimation of the genomic inflation factor ( λ\u200a=\u200a1 . 035 ) indicated minimal overall inflation ( Figure 1A ) ., The genome-wide logistic association results are presented in Figure 1B ., Table S1 provides details for all SNPs with P<10−4 , including one SNP exceeding P<10−7 , the Bonferroni threshold for genome-wide significance ., The three top SNPs were located on 6p21 , in the HLA-DQB1 gene: rs9275224 , P\u200a=\u200a9 . 18×10−8 , OR\u200a=\u200a0 . 69 , 95%CI0 . 60–0 . 79; rs6457617 , P\u200a=\u200a1 . 14×10−7 and rs9275245 , P\u200a=\u200a1 . 39×10−7 ( Figure 1B and Table 2 ) ., Several associated SNPs in HLA-DQB1 have already been reported but rs6457617 was also identified as the most associated SNP in the previous US/European GWAS study 18 ., Of note , the three SNPs in HLA-DQB1 are in strong LD ( r2>0 . 97 ) ., Within the MHC region , the next most associated SNP ( rs3130573 , P\u200a=\u200a1 . 86×10−5 , OR\u200a=\u200a1 . 361 . 18–1 . 56 ) is located in the psoriasis susceptibility 1 candidate 1 ( PSORS1C1 ) gene ( Table 2 ) , a candidate gene for psoriasis 23 ., Conditional analyses of susceptibility variants within MHC showed that there were two independent association signals at rs6457617 ( HLA-DQB1 ) and at rs3130573 ( PSORS1C1 ) ., Indeed , the association at PSORS1C1 remained significant ( P<2 . 1×10−5 ) after controlling for the association at HLA-DQB1 and the association at HLA-DQB1 remained also significant ( P<1 . 5×10−7 ) after controlling for the association at PSORS1C1 ( Table S2 ) ., Outside the MHC region , our GWAS analysis revealed 7 top SNPs ( P<10−5 ) that spanned 6 independent genomic regions ( Figure 1B and Table S1 ) ., Conditional analyses of each of them on HLA-DQB1 showed no significant drop in the association signals ( Table S3 ) ., The 6 loci having at least one SNP with a P<10−5 were selected for follow-up analysis ., Within each locus we selected the SNPs with the strongest ( P<10−4 ) association signals to be genotyped in a post-QC replication sample of 1 , 682 SSc cases and 3 , 926 controls ( Table 1 ) ., To this list we added two top SNPs in HLA-DQB1 and the SNP in PSORS1C1 ., Finally , we further included 4 SNPs at the two known loci ( STAT4 and TNPO3-IRF5 ) and at the newly identified locus ( CD247 ) by Radstake et al 18 ., Out of a total set of 21 SNPs submitted for replication , 20 passed the quality-control analyses ., Stratified association analyses in stage 2 data ( Table 2 ) , confirmed the strong association for HLA-DQB1 ( rs6457617 , P\u200a=\u200a1 . 35×10−28 ) at 6p21 . 3 and also with the PSORS1C1 variant ( rs3130573 , P\u200a=\u200a4 . 98×10−3 ) at 6p21 . 1 ., Of the 6 remaining loci selected in stage 1 , only 2 were replicated with nominal P<5% and with same direction of effect ., They mapped at 2p24 ( rs342070 , P\u200a=\u200a0 . 026; rs13021401 , P\u200a=\u200a0 . 024 ) and 5q33 ( rs3792783 , P\u200a=\u200a4 . 14×10−5; rs2233287 , P\u200a=\u200a4 . 38×10−3; rs4958881 , P\u200a=\u200a2 . 09×10−3 ) ., None of the replicated SNPs showed evidence for heterogeneity of effects among the 4 geographical origins ( Breslow-day P>0 . 10 ) ., As expected , ORs estimated in the discovery tended to be higher than those obtained in the replication stage data ., Afterwards , association signals from joint analyses of the 2 datasets ( Table 2 ) consistently showed highly significant association for HLA-DQB1 ( P\u200a=\u200a2 . 33×10−37 ) , PSORS1C1 ( P\u200a=\u200a5 . 70×10−10 ) and TNIP1 ( P\u200a=\u200a4 . 68×10−9 ) , and also showed some evidence of association for RHOB ( P\u200a=\u200a3 . 17×10−6 ) ., All populations showed same direction of effects ( Figure 2 ) ., Finally , we also replicated association signals at IRF5 ( P\u200a=\u200a3 . 49×10−5; combined-P\u200a=\u200a4 . 13×10−7 ) , at STAT4 ( P\u200a=\u200a1 . 9×10−10; combined-P\u200a=\u200a2 . 26×10−13 ) and at the recently identified new SSc risk locus , CD247 ( P\u200a=\u200a2 . 90×10−5; combined-P\u200a=\u200a1 . 30×10−6 ) ( Table 2 ) ., In our combined data , the locus-specific PAR estimates were 24% for HLA-DQB1 , 4% for TNIP1 , 8% for PSORS1C1 , 7% for CD247 , 8% for STAT4 and 3% with IFR5/TNPO3 ., The combined PAR estimate was 47 . 4% ., As secondary analyses , we assessed homogeneity of SNPs effect between sub-categories of SSc ( cutaneous sub-types and auto-antibodies ) ., Case-only analyses revealed no significant evidence for heterogeneous ORs between cutaneous sub-types of SSc patients for any of the 5 replicated SNPs at 2p24 or 5q33 loci ( Table S4A ) ., Indeed , similar association signals were obtained from case-category association analyses ( Table S4B ) ., Altogether , the results did not suggest that the association signals in the newly identified 5q33 locus were driven by a specific sub-type of SSc ., Conversely , for HLA-DQB1 and PSORS1C1 we found evidence of heterogeneity in OR estimates in positive vs negative ACA or TOPO auto-antibody SSc patients ( Table S4A ) ., Yet , the association signals in each of these sub-types of patients remained strong ( Table S4B ) ., These results support the previously reported hypothesis that the magnitude of the HLA-DQB1 effect on SSc susceptibility may depend on auto-antibody status 11 ., The GWAS stage had 78% power to detect loci of the effect sizes observed in the discovery sample for TNIP1 variants ( OR\u200a=\u200a1 . 50 ) at a significance of P<10−5 ., However , it is widely acknowledged that effect sizes of significant GWAS loci are overestimates of true effects and other genes of lower effect sizes are unlikely to reach stringent significant thresholds ., Our GWAS analysis revealed strong association with PSORS1C1 , which is ∼1 Mb of HLA-DQB1 ., Notably , PSORS1C1 is known to be involved in autoimmune response 23 ., In the combined data , association with PSORS1C1 was highly significant ( P\u200a=\u200a5 . 70×10−10 ) and remained significant after controlling for the association at HLA-DQB1 ., Altogether , our results suggest that this region is likely to contain more than one gene playing a role in the pathogenesis of autoimmune disorders 23 , 24 ., Fine mapping at this locus is warranted to identify causal variants ., The three strongly associated SNPs at the 5q33 locus are located within the TNFAIP3 interacting protein 1 ( TNIP1 ) gene ., TNIP1 is a very interesting new candidate gene for SSc ., The protein encoded by this gene exerts a negative regulation of NF-kappaB via two sequential activities: deubiquitination of Lys63-based chains and synthesis of Lys48-based chains on the TNF receptor-interacting protein and also inhibition of NF-KappaB processing 25 ., TNIP1 interacts with A20 ( TNFAIP3 ) to negatively regulate NF-kappaB ., Several recent studies have suggested that the activation of some inflammatory factors may upregulate fibrotic mediators through Toll-like receptors ( TLRs ) , thereby contributing to SSc pathogenesis 8 ., It has been shown that TLR engagement leads to A20 induction in macrophages and that TNIP1/A20 is essential for the termination of TLR-induced NF-kappaB activity and proinflammatory cytokine production 26 ., Although interactions between TNIP1 and A20 are not well known , A20 also acts as a deubiquitinating enzyme , suggesting a molecular link between deubiquitinating activity and the regulation of TLR signals 26 ., Therefore , TNIP1 and A20 may play a critical role in the regulation of downstream TLR signals , and this issue will have to be addressed in SSc ., Interestingly , variants at TNIP1 have been shown to be implicated in systemic lupus erythematosus susceptibility 27 , 28 and in psoriasis 29 ., Furthermore , we have recently reported an association of one TNFAIP3 variant with SSc 30 ., In our stage-1 data , evidence of association at TNFAIP3 was nominal ( lowest P\u200a=\u200a0 . 047 ) and no pairwise interaction was found ( P>0 . 06 ) between TNFAIP3 and TNIP1 variants ., Analysis of the LD structure across the TNIP1 gene revealed that the 3 strongly associated SNPs belong to the same LD-block ( Figure 3 ) ., No residual association signals were observed when rs3792783 and each of the other 2 SNPs were paired in conditional analyses ., Therefore , any of them , or other variants yet to be identified , could be the causal variant ( s ) ., Interestingly , rs3792783 is located upstream from the transcription start site in exon 2 ( Figure 3 ) ., It is noteworthy that previously reported lupus TNIP1 variants were located in the same LD-block 27 , 28 ., Because of the compelling evidence of the potential role of NF-kappaB in autoimmune diseases and our raised new signal association for SSc at TNIP1 ( a negative regulator of this pathway ) we performed ex vivo and in vitro investigations to assess TNIP1 expression in SSc patients and healthy controls ., For SSc patients , the results showed a strikingly reduced expression of TNIP1 in skin tissue ( Figure 4A ) , and of both mRNA ( Figure 4B ) and protein ( Figure 4C ) synthesis by cultured dermal fibroblasts ., Addressing the question of the potential link between the NF-kappaB pathway and the fibrotic propensity that characterizes SSc , we next assessed the influence of pro-inflammatory cytokines and TNIP1 on the synthesis of extra-cellular matrix by dermal fibroblasts in culture ., Using cells from the skin of healthy controls ( Figure, 5 ) and SSc patients ( Figure 6 ) , we showed that recombinant TNIP1 abrogated collagen synthesis induced by inflammatory cytokines both at the mRNA and protein levels ., It must be acknowledged that TNIP1 is described as an intra-cellular protein whereas we used recombinant protein added to cell supernatant in these experiments ., The observed effects may be related to different hypotheses ., TNIP1 has been described as a nuclear shuttling protein and it could have a chaperon-like activity , highly interacting with other protein that could result in engulfment of TNIP1 through interaction with a cell surface protein ., Such intra-cellular effects of extra-cellular proteins has been shown also for the S100 family of proteins that have no leader sequence and for clusterin for which it is postulated that the protein could be taken up by interacting with either a yet unidentified receptor or by a mechanism related to their chaperon-like activity 31 ., More work is needed to determine which of these hypotheses has to be retained and to investigate more in depth soluble TNIP1 In this first attempt to explore TNIP1 functional disturbances , we could not investigate a relationship between specific TNIP variants and in vitro or in vivo changes; this will need to be addressed ideally after the identification of the causal variant and using a much larger sample size ., Nevertheless , our results raise a potential relationship between inflammation and fibrosis and open a new and highly relevant field of investigation in SSc pathogenesis and in fibrotic disorders ., Our next most associated SNPs at 2p24 are in strong LD ( r2\u200a=\u200a0 . 98 ) and map ∼30 kb from RHOB ., RHOB is the Ras homolog gene family member B that regulates protein signalling and intracellular protein trafficking ., RhoB is essential for activity of farnesyltransferase inhibitors and also statins that are two strong potential future drugs in SSc 9 , 32 ., To our knowledge , association to RHOB has never been reported so far ., The signal for association was weaker at this locus and therefore will need to be confirmed in other samples and more investigations are warranted to assess RHOB implication in this disease ., In conclusion , we have conducted a large genome-wide association study of SSc and identified two new SSc-risk loci , PSORS1C1 and TNIP1 ., We also confirmed the association of SSc with variants at STAT4 , IRF5 and CD247 , in the European population ., We also found compelling evidence of association to a putative new SSc risk locus on 2p24 , close to the RHOB gene ., None of the newly identified 3 loci have been previously reported associated to SSc ., The TNIP1 variants identified do not have precise functional implications; however , their localization within a regulatory region strongly suggests an impact on transcription of the gene ., This is supported by our ex vivo and in vitro investigations ., Altogether , our results are consistent with a reduced inhibition of NF-kappaB , therefore favoring inflammatory/immune responses and potentially contributing to the overproduction of extra-cellular matrix ., This raises a new clue for a link between inflammation and SSc that could also be of importance in other fibrotic disorders ., Stage-1 included 654 SSc patients and 531 controls recruited through the French GENESYS project 11 , 13 , 30 and 2 , 003 controls from the French Three-City ( 3C ) cohort 21 , 22 ., The stage-2 data included an independent collection of 4 , 492 samples ( pre quality controls ) from several University Hospitals in France , Italy , Germany and Eastern Europe ., It also included 721 Italian controls recruited through nationwide efforts by HYPERGENE consortium and 481 Illumina HumanHap550 for the KORA S4 study 33 , recruited in the city of Augsburg , Southern Germany ., In both stage 1 and stage 2 samples , SSc patients fulfilled ACR criteria 34 and were classified in cutaneous subsets according to LeRoys criteria 19 ., Table 1 shows the main characteristics of the post-QC SSc patients and controls ., All participants gave written informed consent , and approval was obtained from the relevant local ethical committees ., Association analysis of the genotype data was conducted with PLINK ( v1 . 07 ) software 35 ., All reported P values are two sided ., In stage 1 , we applied logistic regression assuming an additive genetic model ., The quantile-quantile plot was used to evaluate overall significance of the genome-wide association results and the potential impact of residual population substructure ., A conservative genome-wide significance threshold of 0 . 05/489 , 918\u200a=\u200a1 . 02×10−7 was used ., Stage 2 association and combined analyses were carried out with the Mantel-Haenszel test to control for differences between geographical groups ., A Breslow-Day test was performed to assess the heterogeneity of effects in different populations ., In the replication analysis , P values<0 . 05 and direction of effect as observed in the stage-1 data , were considered to indicate statistical significance ., Secondary statistical analyses were conducted to assess independency of multiple association signals within and between loci and homogeneity of effects between subgroups of SSc patients ., Case-only association analyses were conducted using the three main clinical variables ( Table 1 ) ., The LD structure of the identified loci was analyzed using Haploview 4 . 1 37 and LD blocks delimited using the D′-based confidence interval method 38 ., The locus-specific Population attributable risk ( PAR ) was calculated for each of the 6 replicated loci ( HLA-DQB1 , TNIP1 , PSORS1C1 , STAT4 , IFR5/TNPO3 and CD247 ) according to the following formula: PAR\u200a=\u200aRAF× ( OR-1 ) / ( RAF× ( OR-1 ) +1 ) , where RAF is the frequency of the associated allele in the controls , and OR is the odds ratio associated with the risk allele ., The combined PAR was computed as 1−Pj ( 1−PARj ) ., Fibroblast cultures were prepared by outgrowth cultures from lesional skin biopsy specimens of eleven SSc patients and from twelve healthy controls matched for age and sex ., The median age of SSc patients was 49 years old ( range: 22–67 years ) and their median disease duration was 7 years ( range: 1–17 years ) ; seven had the limited cutaneous subset and four the diffuse ., Immunohistochemistry was performed on paraffin-embedded skin sections from 5 SSc patients and 5 controls using mouse anti-human TNIP1 antibodies ( eBioscience , Frankfurt , Germany ) ., Total RNA , issued from cultured dermal fibroblasts , isolation and reverse transcription into complementary DNA were performed as previously described 39 ., Gene expression was quantified by SYBR Green real-time PCR , with a specific primer pair available upon request ., Protein assessment was performed on western blots , as previously described 40 using mouse anti-human TNIP1 antibodies ( eBioscience , CA , USA ) ., In selected experiments , dermal fibroblasts from healthy control subjects and patients with SSc were treated for 24 hours with recombinant TNIP1 ( 2 µg/ml , Abnova , Tapei City , Taiwan ) in the presence or not of the following proinflammatory cytokines: TNFα ( 20 ng/ml , R&D systems , Abingdon , UK ) , IL1β ( 1 µg/ml , Immunotools , Friesoythe , Germany ) or IL6 ( 1 µg/ml , Immunotools ) ., mRNA levels of human α1 ( I ) and α2 ( I ) procollagen were quantified by quantitative real-time PCR , specific primers are available upon request ., The collagen content in cell culture supernatants was analyzed with the SirCol collagen assay ( Biocolor , Belfast , UK ) 41 ., Comparisons were performed using Students T test .

Introduction, Results/Discussion, Material and Methods

Systemic sclerosis ( SSc ) is an orphan , complex , inflammatory disease affecting the immune system and connective tissue ., SSc stands out as a severely incapacitating and life-threatening inflammatory rheumatic disease , with a largely unknown pathogenesis ., We have designed a two-stage genome-wide association study of SSc using case-control samples from France , Italy , Germany , and Northern Europe ., The initial genome-wide scan was conducted in a French post quality-control sample of 564 cases and 1 , 776 controls , using almost 500 K SNPs ., Two SNPs from the MHC region , together with the 6 loci outside MHC having at least one SNP with a P<10−5 were selected for follow-up analysis ., These markers were genotyped in a post-QC replication sample of 1 , 682 SSc cases and 3 , 926 controls ., The three top SNPs are in strong linkage disequilibrium and located on 6p21 , in the HLA-DQB1 gene: rs9275224 , P\u200a=\u200a9 . 18×10−8 , OR\u200a=\u200a0 . 69 , 95% CI 0 . 60–0 . 79; rs6457617 , P\u200a=\u200a1 . 14×10−7 and rs9275245 , P\u200a=\u200a1 . 39×10−7 ., Within the MHC region , the next most associated SNP ( rs3130573 , P\u200a=\u200a1 . 86×10−5 , OR\u200a=\u200a1 . 36 1 . 18–1 . 56 ) is located in the PSORS1C1 gene ., Outside the MHC region , our GWAS analysis revealed 7 top SNPs ( P<10−5 ) that spanned 6 independent genomic regions ., Follow-up of the 17 top SNPs in an independent sample of 1 , 682 SSc and 3 , 926 controls showed associations at PSORS1C1 ( overall P\u200a=\u200a5 . 70×10−10 , OR:1 . 25 ) , TNIP1 ( P\u200a=\u200a4 . 68×10−9 , OR:1 . 31 ) , and RHOB loci ( P\u200a=\u200a3 . 17×10−6 , OR:1 . 21 ) ., Because of its biological relevance , and previous reports of genetic association at this locus with connective tissue disorders , we investigated TNIP1 expression ., A markedly reduced expression of the TNIP1 gene and also its protein product were observed both in lesional skin tissue and in cultured dermal fibroblasts from SSc patients ., Furthermore , TNIP1 showed in vitro inhibitory effects on inflammatory cytokine-induced collagen production ., The genetic signal of association with TNIP1 variants , together with tissular and cellular investigations , suggests that this pathway has a critical role in regulating autoimmunity and SSc pathogenesis .

Systemic sclerosis ( SSc ) is a connective tissue disease characterized by generalized microangiopathy , severe immunologic alterations , and massive deposits of matrix components in the connective tissue ., Epidemiological investigations indicate that SSc follows a pattern of multifactorial inheritance; however , only a few loci have been replicated in multiple studies ., We undertook a two-stage genome-wide association study of SSc involving over 8 , 800 individuals of European ancestry ., Combined analyses showed independent association at the known HLA-DQB1 region and revealed associations at PSORS1C1 , TNIP1 , and RHOB loci , in agreement with a strong immune genetic component ., Because of its biological relevance , and previous reports of genetic association at this locus with other connective tissue disorders , we investigated TNIP1 expression ., We observed a markedly reduced expression of the gene and its protein product in SSc , as well as its potential implication in control of extra-cellular matrix synthesis , providing a new clue for a link between inflammation/immunity and fibrosis .

medicine, rheumatology, public health and epidemiology, cytokines, immunology, rheumatoid arthritis, genetic epidemiology, systemic lupus erythematosus, inflammatory and psoriatic arthritis, disease mapping, scleroderma, connective tissue diseases, major histocompatibility complex, epidemiology, immune system, psoriasis, genetics of the immune system, clinical immunology, autoimmune diseases

journal.pgen.1006352

Smad2 and Smad3 Regulate Chondrocyte Proliferation and Differentiation in the Growth Plate

The cartilage growth plate is the primary driver of endochondral bone growth ., In the growth plate , resting , columnar , prehypertrophic and hypertrophic chondrocytes are arrayed in discrete zones ., Resting chondrocytes , located at the top of the growth plate , are small and relatively quiescent ., Upon stimulation by extracellular signals , cells near the bottom of the resting zone transition to columnar chondrocytes , which exhibit a higher rate of proliferation and a flatter morphology ., These cells form stacks along the long axis of the developing skeletal element ., Columnar cells at the bottom of this zone eventually exit the mitotic phase and become prehypertrophic chondrocytes ., Prehypertrophic cells further differentiate into enlarged hypertrophic cells , comprising a zone adjacent to the site of replacement of cartilage by bone ., Chondrocyte proliferation and differentiation in the growth plate is tightly regulated by Indian hedgehog ( Ihh ) and parathyroid hormone-related peptide ( Pthrp ) ., Ihh , a secreted protein expressed in prehypertrophic chondrocytes , stimulates cell proliferation and differentiation ., Its role in proliferation is mediated in part by inducing Pthrp expression in epiphyseal resting chondrocytes ., Secreted Pthrp maintains columnar cells in a mitotic state , preventing their transition to the pre-hypertrophic phase , and hence negatively regulating Ihh expression ., Once cells escape the zone of influence of Pthrp , they exit the cell cycle , become prehypertrophic , and upregulate Ihh expression , which promotes hypertrophy and matrix mineralization ., This feedback loop thus controls the transition of chondrocytes through each zone of the growth plate ., Transforming growth factor βs ( TGFβs ) and activins are secreted proteins that are members of the TGFβ superfamily of growth factors ., TGFβs and activins bind to distinct receptor complexes , but activate similar signal transduction pathways ., Binding of TGFβs or activins to their receptors leads to activation of the kinase activity of the receptor ., The activated receptor complexes then transduce signals through multple pathways ., These pathways can be broadly divided into Smad-dependent and Smad-independent pathways 1–3 ., In the canonical Smad-dependent pathway , activated receptor complexes phosphorylate the receptor-activated Smads ( R-Smads ) , Smad2 and Smad3 ., Smads 2 and 3 are transcription factors; once phosphorylated , they form hetero-oligomeric complexes with the transcription factor Smad4 ., These complexes enter the nucleus , bind promoters , and regulate target gene expression ., In addition , there exist numerous non-canonical Smad-independent pathways for transduction of TGFβ signals , such as MAP kinases , RhoA and mTOR 4–8 ., TGFβs play critical roles in growth plate chondrocyte proliferation and differentiation , and in the maintenance of articular chondrocytes 9–17 ., There are three TGFβ isoforms in mammals: TGFβ1 , 2 , 3 ., Only Tgfb2 mutants exhibit obvious skeletal dysplasia ., Tgfb2-/- mice exhibit embryonic lethality , accompanied by skeletal defects that include shortened limbs , axial , and craniofacial defects 18 ., The role of the TGFβ signaling pathway in cartilage has been studied most extensively using mice lacking the type II TGFβ receptor TGFβRII ., This receptor is required for transduction of the TGFβ pathway by TGFβs 1–3 , but is not used by activin ligands ., Tgfbr2;Prrx1Cre mice , in which TGFβRII is ablated in limb bud mesenchyme , exhibit growth plates with decreased proliferation and accelerated onset of hypertrophy , but delayed terminal maturation 19 , 20 ., In contrast , conditional deletion of the TGFβRII in committed Col2a1-expressing chondrocytes did not lead to obvious defects in appendicular elements 21 ., These findings suggest that TGFβRII transduces TGFβ signals at prechondrogenic stages and/or in perichondrium , but may not have a substantial role in cartilage once a growth plate forms ., The extent to which Smad2/3-dependent signaling mediated by TGFβ and activins is required in developing cartilage is unknown ., Smad2 , Smad3 and Smad4 are co-expressed throughout the growth plate 12 , 22–24 ., Smad2 , 3 and 4 are all present in articular cartilage 12 , 25 ., Smad3-/- mice are born with a normal skeletal phenotype , but develop postnatal dwarfism and osteoarthritis-like pathologies in adulthood 12 , 26–28 ., The function of Smad2 in cartilage during embryogenesis has not been characterized ., Smad2 and Smad3 have some distinct roles in mediating TGFβ/activin signaling ., Smad3 can bind DNA directly , whereas Smad2 regulates gene expression by interacting with Smad3 or other transcription factors 29 ., Mice lacking Smad2 die before 8 . 5 days of development ( E8 . 5 ) 30 , precluding a genetic analysis of its function in chondrogenesis ., It is not known whether Smad2 partially compensates for the loss of Smad3 in the growth plates of Smad3-/- mice ., Studies in Smad2-/- vs . Smad3-/- epiblast , epithelial cells and fibroblasts show that Smads 2 and 3 regulate some common and some distinct target genes 31 , 32 ., Overexpression of Smad2 or Smad3 can block the spontaneous maturation observed in Smad3-deficient chondrocytes 33 , providing support for the hypothesis that Smad2 and Smad3 may have some overlapping functions in the growth plate ., To define the role of Smad2/3-mediated signaling in cartilage , we generated mice with conditional deletion of Smad2 using Col2a1-Cre ( Smad2fx/fx;Col2Cre , hereafter referred to as Smad2CKO ) , mice with Smad3 globally deleted ( Smad3-/- ) , and the corresponding double mutants ( Smad2/3 ) ., Loss of Smad2 leads to a growth plate phenotype in neonates that is more severe than that seen in Smad3-/- mice ., At the neonatal stage , Smad2/3 ( Smad2CKO;Smad3-/- ) double mutants exhibit more severe defects in the hypertrophic zone than do Smad2CKO or Smad3-/- mice ., Defects include elevated levels of proliferation in resting zone chondrocytes at neonatal stages , leading to enlarged columnar and hypertrophic zones ., This may be a consequence of depletion of the resting zone due to the accelerated entry of resting zone chondrocytes into the columnar zone ., Overall , our results suggest that Smad2 inhibits proliferation of resting zone chondrocytes during embryogenesis , and acts as a negative regulator of Ihh expression , and that Smad2 and Smad3 have some overlapping functions in cartilage ., Our results show that both of these Smads are required at neonatal stages for normal chondrocyte proliferation and differentiation in the growth plates ., We generated mice lacking Smad2 and Smad3 in cartilage in order to study the role of canonical Smad2/3-mediated signaling in chondrogenesis ., Smad2fx/fx mice were intercrossed with Col2a1Cre mice to generate Smad2fx/fx;Col2a1Cre ( Smad2CKO ) mice ., Smad2CKO mice are viable and fertile ., The Smad3-/- allele we used has a LacZ cassette and internal ribosomal entry site ( IRES ) inserted in the second exon , leading to a loss-of-function allele 27 ., We crossed Smad3+/- mice with Smad2CKO mice to generate Smad2fx/fx;Col2a1Cre;Smad3-/- ( Smad2/3 double mutant ) mice ., To confirm efficient Cre-mediated excision , levels of activated C-terminal phosphorylated Smad2 and Smad3 were examined by IHC using an antibody that recognizes both pSmad2 and pSmad3 ( S1 Fig ) ., No pSmad2/3 was detected in E16 . 5 or P0 double mutant growth plates , verifying efficient loss of Smads 2 and 3 ( S1D and S1H Fig ) ., Consistent with previous studies 34 , at E16 . 5 and P0 , pSmad2/3 is present throughout the resting , columnar , and prehypertrophic zones of the E16 . 5 Smad2fx/fx control growth plate , and this pattern of expression persists at least until birth ( S1A and S1E Fig ) ., No obvious differences were noted in the distribution of pSmad3 in Smad2CKO or pSmad2 in Smad3-/- mice compared with control Smad2fx/fx littermates ( S1B , S1C , S1F and S1G Fig ) ., These findings confirm previous reports that pSmads 2 and 3 have extensively overlapping distributions in the growth plate 12 , 22–24 ., Consistent with previous studies 12 , 26 , 27 , skeletal preparations revealed no apparent skeletal defects in Smad3-/- mice at P0 ., Analysis of P0 Smad2CKO mice also revealed no obvious defects ( Fig 1A–1C ) ., However , Smad2CKO;Smad3-/- ( Smad2/3 ) double mutant mice exhibited subtle defects in axial and craniofacial elements ( Fig 1A and 1D ) ., Lateral views of P0 skulls revealed reduced ossification of the bones of the inner ear in double mutants ( Fig 1E–1H ) ., The length of the body ( from top of the skull to the proximal end of the tail ) of double mutants is about 9 . 7% shorter than that of control Smad2fx/fx littermates ( n = 4 , P = 0 . 04 ) ( Figs 1A , 1D and S2A ) ., Ventral views revealed that double mutants have slightly smaller skulls , with a shorter nasal to occipital length ( 94% of control ( Smad2fx/fx ) , n = 3 , p = 0 . 05 ) , narrower cranial base ( 93% of control , n = 3 , p = 0 . 04 ) , and reduced ossification of the occipital condyle ( arrow , S2B Fig ) ., Dorsal views of the rib cage revealed no obvious differences between Smad2CKO or Smad3-/- mice compared with control Smad2fx/fx littermates ( S2C Fig ) ., However , double mutants have a shorter sternum and a bifurcated xiphoid process ( S2C Fig ) ., Moreover , localized defects were seen in the vertebral columns of Smad2/3 mutants ., Ventral views of the lumbar spine revealed shorter vertebrae and smaller ossified vertebral bodies in double mutants; reduced ossification was not observed in Smad2CKO or Smad3-/- mice ( S2D Fig ) ., No differences were evident in the cervical vertebrae in Smad3-/- mice compared to control Smad2fx/fx mice , but the ossification center of the axis ( C2 ) was reduced in Smad2CKO mice and in double mutants ( S2E Fig ) ., Cleared skeletal preparations revealed no clear differences in appendicular elements in Smad2CKO , Smad3-/- mice , or Smad2/3 double mutants ( Figs 1I–1L and S2F ) ., As discussed above , Smad2/3 double mutants had only subtle skeletal alterations at birth ., However , these mice developed progressive postnatal dwarfism , which was not seen in Smad2CKO and Smad3-/- mice ( S3 Fig ) ., At 1 month , double mutants were 12 . 5% ( n = 3 , p = 0 . 05 ) shorter than control Smad2fx/fx littermates ( S3C Fig ) ., These data indicate that Smad2 and Smad3 have compensatory roles in the regulation of axial skeletal growth after birth ., The above analysis revealed no clear impact of loss of Smad2 or Smad3 on appendicular length at P0 ., A previous study demonstrated that loss of Smad3 led to appendicular defects beginning in the early postnatal period 12 ., However , we used a different Smad3 allele ., We therefore performed a histological analysis of P0 appendicular cartilage to investigate whether the Smad3 mutant allele we used 27 exhibits a similar phenotype ., We also examined whether loss of Smad2 exerts any effect on growth plate architecture ., This analysis revealed that the lengths of the resting zones in Smad2CKO and Smad2/3 double mutants were shorter than in control Smad2fx/fx mice ., In contrast , the columnar and hypertrophic zones of Smad2CKO , Smad3-/- , and Smad2/3 double mutant mice were longer than those of control Smad2fx/fx littermates ( Figs 2A and S4 ) ., Although both Smad2CKO and Smad3-/- mice exhibited elongated columnar zones , the effect was greater in Smad2CKO mice than in Smad3-/- mice , and the degree of elongation did not differ between Smad2/3 double mutants and Smad2CKO mice , suggesting that Smad2 has a more prominent role than Smad3 in elongation of the columnar zone ( S4 Fig ) ., Both Smad2CKO and Smad3-/- mice exhibited elongated hypertrophic zones compared to Smad2fx/fx controls , and there was a significant elongation of the hypertrophic zone in Smad2/3 double mutants compared to either Smad2CKO or Smad3-/- mice ., Immunohistochemistry for PCNA was performed to test whether the increases in lengths of the colunmar zones were correlated with increased proliferation ., A 2 to 3-fold increase in PCNA-positive cells was seen in the resting zones of Smad2CKO , Smad3-/- and Smad2/3 double mutant mice compared with control Smad2fx/fx littermates; the degree of PCNA staining in the resting zone was similar in Smad2/3 double mutants and Smad2CKO mice ( Fig 2C ) ., No differences were detected in the columnar zone ( Fig 2B and 2C ) ., TUNNEL assays were performed to evaluate whether differences in cell survival contribute to the longer hypertrophic zones in mutants ., No differences were detected ( Fig 2D ) ., These results indicate that loss of Smad2 and/or Smad3 promotes the entry of resting chondrocytes into the highly proliferative columnar phase , and that Smad2 appears to have a more prominent role than does Smad3 ., The increased lengths of the columnar and hypertrophic zones in mutants is consistent with an increased pool of chondrocytes transiting out of the resting zone and eventually undergoing hypertrophy ., These findings suggest that Smad2 and Smad3 function to maintain the pool of resting chondrocytes in a quiescent state in neonatal growth plates ., Ihh is expressed in prehypertrophic chondrocytes and is a critical regulator of chondrocyte proliferation and differentiation ., Since pSmads 2 and 3 are present in prehypertrophic chondrocytes ( S1 Fig ) , RNA in situ hybridization was performed to assess Ihh RNA levels ., The zone of Ihh expression is increased ( Fig 3A ) , and qPCR quantification of RNA isolated from P0 growth plate cartilage showed that the level of Ihh RNA is increased in all three mutant strains compared with control Smad2fx/fx littermates ( S5 Fig ) , suggesting that Smad2 and Smad3 normally inhibit Ihh expression in the growth plate ., Smad2/3 double mutants had elevated levels of Ihh RNA and protein compared to both Smad2CKO and Smad3-/- mice , suggesting that both Smad2 and Smad3 contribute to elevated Ihh expression ., As shown previously 34 , Ihh protein was found in the control Smad2fx/fx growth plate , with highest levels in the prehypertrophic and hypertrophic zones ( Fig 3B ) ., Ihh protein was detected in these regions in Smad2 and Smad3 single and double mutants , but unlike control Smad2fx/fx littermates , was also detected in the resting zones ( Figs 3B , 4C and S5 ) ., This suggests that the elevated Ihh RNA level in the prehypertrophic zone leads to increased diffusion of Ihh protein to the resting zone in mutants ., Patched1 ( Ptch1 ) is a direct transcriptional target of Ihh signaling ., Consistent with elevated levels of Ihh RNA and protein , immunostaining for Ptch1 demonstrated increased levels throughout the growth plates in mutants; this was most evident in the resting zones of Smad2CKO and Smad2/3 double mutant mice ( Figs 3D and S5 ) ., Ihh is a downstream target of BMP signaling 35 , 36 , raising the possibility that the increased Ihh RNA expression is a consequence of increased BMP signaling in the growth plates of Smad2CKO , Smad3-/- and Smad2/3 double mutant mice ., However , immunohistochemical examination in P0 growth plates revealed no obvious change of pSmad1/5/8 in mutant mice compared with control Smad2fx/fx mice ( Figs 3E and S5 ) ., In summary , the elevated level of Ihh RNA in the prehypertrophic zone , and increased domain of Ihh protein localization to the resting zone is correlated with an increase in the level and distribution of Ptch1 ., As Ihh promotes chondrocyte proliferation in the growth plate 37 , the increased Ihh level and activity seen in Smad2CKO , Smad3-/- , and double mutants may contribute to the increased rate of proliferation in the resting zone in mutants ., The above findings indicate that Smads 2 and 3 act to decrease Ihh RNA levels in the neonatal growth plate ., To investigate whether this effect is direct , primary rib chondrocytes were isolated , matured to the prehypertrophic phase by maintenance in chondrogenic differentiation medium , and then treated with TGFβ1 or TGFβ2 ., TGFβ1 and TGFβ2 decreased Ihh RNA levels in control Smad2fx/fx chondrocytes ( Fig 4A ) ., However , the ability of TGFβ to inhibit Ihh expression was impaired in Smad3-/- mutant chondrocytes , and abolished in Smad2CKO and Smad2/3 double mutant chondrocytes ., However , a caveat of these findings is that although levels of Ihh RNA in primary chondrocytes from Smad2CKO and Smad3-/- mice were not reduced compared to control Smad2fx/fx chondrocytes under basal conditions ( no serum and no growth factor addition ) , levels of Ihh RNA were reduced in prehypertrophic chondrocytes from Smad2/3 double mutants under basal conditions ( Fig 4A ) ., It is unclear whether this reflects a defect in the ability of Smad2/3 double mutant primary chondrocytes to undergo timely differentiation in vitro ., Alternatively , Smads2/3 may play a role in maintaining basal levels of Ihh RNA ., Overall however , the in vivo ( Fig 3A ) and in vitro ( Fig 4A ) results indicate that Smad2 and Smad3 are required to inhibit Ihh expression in prehypertrophic chondrocytes in the neonatal growth plate ., To test whether Smads 2/3 play a direct role in regulating Ihh promoter activity , a luciferase reporter containing the proximal 742 bp of the mouse Ihh promoter 38 was transfected into ATDC5 chondrocytic cells ., After culture in differentiation medium to induce prehypertrophy , the cells were treated with TGFβ1 for 24 hours ., To test whether Smad2 and/or Smad3 mediate Ihh inhibition , Smad2 and/or Smad3 levels were knocked down by transfection of verified siRNAs ., Reporter assays showed that TGFβ1 inhibits Ihh promoter activity in control chondrocytes by > 50% ( Fig 4B ) ., P3TP-Luc activity was used as a positive control for TGFβ activity , and showed robust activation under the same conditions ( Fig 4C ) ., Consistent with the analysis in vivo ( Figs 3A and S5 ) , knockdown of Smad2 and/or Smad3 blocked the inhibitory effect of TGFβ on Ihh promoter activity ., Activated Smads 2 and 3 bind to Smad binding elements ( SBEs ) that contain ( C ) AGAC motifs 39–42 ., ChIP-chip/ChIP-seq studies have confirmed that the SBE is enriched in Smad2/3 binding regions ., 43–47 ., In silico examination of the 742bp proximal Ihh promoter identified 5 putative SBEs , designated S1 to S5 ( Fig 5A ) ., ChIP analysis performed in ATDC5 cells for Smad2 and Smad3 binding showed differential occupation of S1-S3 in the presence of TGFβ; neither Smad2 nor Smad3 associated with S4 or S5 ( Fig 5B ) ., In addition , comparative Genomic Analysis using the UCSC Genome Browser ( http://genome . ucsc . edu ) 48 showed that S1 , S2 and S3 are 100% conserved in the mouse , rat , human and dog genomes ( S6 Fig ) ., At S1 , TGFβ increased binding of Smad2 but not Smad3 ( Fig 5B ) ., In contrast , at S2 , TGFβ increased the association of Smad3 , but had no effect on Smad2 binding ., Association of both Smad2 and Smad3 is increased by TGFβ at S3 ., These results reveal 3 binding elements for pSmad2/3 within the 742 bp Ihh promoter region , and demonstrate that Smad2 and Smad3 have both common and distinct binding patterns ., To examine whether S1-S3 mediate the inhibitory effect of TGFβ on Ihh , six nucleotides covering the conserved SBE regions in S1 , S2 and S3 were replaced with PsiI recognition sites in pIhh742-Luc to generate mutated constructs M1 , M2 and M3 , respectively ., Reporter assays revealed no significant inhibitory effect of TGFβ on activity of the M1 and M3 constructs , and the inhibitory effect of TGFβ on activity of the M2 construct was decreased compared with of the control construct ( Fig 5C ) ., These results indicated that S1 and S3 exert more inhibitory function than S2 ., Similar to the results in Fig 4B showing lower basal activity of the Ihh promoter in Smad2/3 double mutant primary chondrocytes , M1 and M3 exhibited decreased basal activity compared to the control Ihh promoter ( Fig 5C ) , indicating that S1 and S3 also play a role in mediating basal activity of the Ihh promoter ., The basis for the lower basal activity of the Ihh promoter is unclear ., However , BMP signaling enhances Ihh promoter activity and association of Smad4 with SBEs in mouse teratocarcinoma P19 cells 35 ., This raised the possibility that S1-S3 might recruit Smad4 to the Ihh promoter in response to BMPs , and that this recruitment is required for basal Ihh promoter activity ., We therefore compared the activities of M1 , M2 and M3 in response to BMP treatment in ATDC5 cells ., We observed slightly lower basal levels of activity as in Fig 5C , but found no significant differences in BMP-mediated induction between the control and mutant promoters ( S7 Fig ) ., Together , our data indicate that S1 and S3 are important for Smad2 and Smad3-mediated inhibition of Ihh expression ., We also find that S1 and S3 are important for maintaining basal levels of Ihh expression , but that this activity is not due to a role for these SBEs in mediating BMP responsiveness in chondrocytes ., Smad2 and Smad3 interact with a variety of DNA-binding proteins in different contexts ., Hdac4 is expressed in prehypertrophic and hypertrophic zones of the growth plate and represses hypertrophy by binding to and blocking Runx2 activity 49 ., Although Runx2 is a potent inducer of Ihh expression 50 , whether or not Hdac4 directly regulates Ihh expression is unknown ., Immunostaining confirmed that Hdac4 is present in the lower proliferative , prehypertrophic and hypertrophic zones of control Smad2fx/fx P0 growth plates , and is localized in the nucleus , as reported previously 49 ( Fig 6A ) ., The level of Hdac4 protein was greatly diminished in the growth plates of Smad2CKO and double mutant mice ( Figs 6A and S8A ) ., The percentage of cells expressing nuclear Hdac4 was significantly decreased at the border of the prehypertrophic and lower columnar zones of Smad2/3 double mutants compared to either Smad2CKO or Smad3-/- mice ( S8A Fig ) , indicating that both Smad2 and Smad3 contribute to decreased Hdac4 nuclear localization ., To test whether the effect on Hdac4 expression is transcriptional , RNA levels were examined in growth plate cartilage from control Smad2fx/fx , Smad2CKO , Smad3-/- , and double mutant neonatal mice ., No significant differences were observed ( S8B Fig ) ., These results suggest that both Smad2 and Smad3 regulate Hdac4 localization and/or stability at the border of the lower columnar and prehypertrophic zones , but Smad2 has a greater impact than Smad3 ., Next , we tested whether Smad2 and/or Smad3 associate directly with Hdac4 in chondrocytes ., Immunoprecipitation assays revealed weak association between Hdac4 and Smad2 without TGFβ , and this association was increased by TGFβ; no association of Hdac4 with Smad3 was detected ( S8C Fig ) ., Because this analysis revealed that Hdac4 associates with Smad2 , we used ChIP to test whether Smad2 and Hdac4 interact on the Ihh promoter ., TGFβ increased the level of Hdac4 associated with SBE1 , the site exhibiting the highest level of pSmad2 binding; we did not detect Hdac4 binding to SBE3 , in spite of the fact that Smad2 binds to this site ( Fig 6B ) ., These findings suggest that Smad2 regulates Hdac4 protein expression or stabilization , and also recruits Hdac4 to a repressive complex at SBE1 in the Ihh promoter ., SnoN is a repressor that can be induced by TGFβ 51 , 52 ., It is expressed in prehypertrophic chondrocytes , and inhibits chondrocyte hypertrophy 53 ., Immunostaining of P0 growth plates showed that SnoN was expressed and localized within the nucleus in prehypertrophic and hypertrophic zones in control Smad2fx/fx mice ( Fig 6C ) ., There was no obvious change in SnoN protein levels in prehypertrophic and hypertrophic chondrocytes in Smad2CKO , Smad3-/- and double mutant mice ( Fig 6C ) ., However , there was an increase in SnoN protein levels in the proliferative zones in all mutant strains ( Fig 6C ) ., This increase in protein levels was not due to increased levels of SnoN RNA; in fact , SnoN RNA levels were decreased in Smad3-/- and Smad2/3 mutant growth plates ( S8D Fig ) ., TGFβ signaling leads to rapid degradation of SnoN in a Smad2/Smad3-dependent manner 54 ., Hence the elevated protein levels in mutants in spite of reduced mRNA levels may reflect increased SnoN stablity in the absence of Smads 2 and 3 ., Ski proteins can suppress TGFβ and BMP signaling by binding to R-Smads and Smad4 55 , 56 ., Ski was localized to the prehypertrophic and hypertrophic zones in Smad2fx/fx mice ( Fig 6D ) ., There was no apparent change in protein levels in prehypertrophic and hypertrophic zones in any mutant strain , but there were increased levels of immunostaining in the proliferative zones in Smad3-/- and Smad2/3 double mutant strains ., As is the case for SnoN , this effect appears to be posttranscriptional because loss of Smad3 led to reduced Ski RNA levels in spite of the elevated protein levels ( S8E Fig ) ., ChIP was performed to test whether SnoN and Ski might mediate the repressive effects of Smads 2 and 3 on Ihh expression in chondrocytes ., This analysis showed that TGFβ increased the association of SnoN and Ski with SBE1 , SBE2 , and SBE3 in the Ihh promoter in ATDC5 cells ( Fig 6E ) ., SBE1 exhibited greater association with SnoN than with Ski , whereas more Ski than SnoN bound to SBE2; SnoN and Ski associated with SBE3 at similar levels ( Fig 6E ) ., These results parallel the differential binding of Smads 2 and 3 to these sites ( Fig 5B ) , suggesting that SnoN and Ski may interact preferentially with Smad2 and Smad3 , respectively ., In addition , siRNA knockdown of SnoN abolished the ability of TGFβ to repress Ihh742-luc reporter activity; siRNA against Ski and Hdac4 significantly reduced the ability of TGFβ to repress Ihh742-luc reporter activity ( S9 Fig ) ., Overall , the data suggest that both Smad2 and Smad3 mediate suppression of Ihh transcription by binding to distinct SBEs and associating with different repressors , but that Smad2 has a greater impact on Ihh RNA levels than Smad3 ., Canonical TGFβ and activins transduce signals through Smad2 and Smad3 57 , 58 ., Smad2 and Smad3 interact with different transcriptional regulators on DNA and can bind to distinct sites 32 , 58 ., Although Smad2 and Smad3 have similar functions in a number of contexts 32 , 59 , 60 , they exert distinct , and even opposing , effects in others 32 , 61 , 62 ., Whether this is the case in cartilage was unknown ., Analysis of Smad3-/- mice demonstrated previously that Smad3 suppresses chondrocyte hypertrophy 12 ., The finding that overexpression of Smad2 can block the accelerated chondrocyte maturation seen in Smad3-/- chondrocytes suggested that Smad2 and Smad3 exert at least some similar functions in vitro 33 ., A limitation of the previous study as well as ours is the use of global Smad3 mutants , raising the possibility that some aspects of the Smad3 mutant growth plate phenotype are due to effects on other cell types , such as the perichondrium ., However , direct effects on chondrocytes seem plausible based on in vitro studies 63 , 64 and our results , which document direct effects in Smad3-deficient neonatal chondrocytes ., Nonetheless , loss of Smad2 appears to have a greater impact in the non-hypertrophic zone chondrocytes of the growth plate during embryogenesis than does loss of Smad3 ., We found that both Smad2 and Smad3 are essential for chondrogenesis in vivo to inhibit proliferation and hypertrophy ., Some of the alterations in proliferation , maturation , and gene expression were more pronounced in Smad2CKO;Smad3-/- double mutants than in single mutants , indicating that Smads 2 and 3 exert similar functions in the growth plate ., This may be mediated in part by the ability of Smads 2 and 3 to repress Ihh expression ., Elevated Ihh RNA levels in mutants were correlated with increased levels of Ihh protein and its direct target Ptch1 in the resting zone ., Although altered matrix properties in Smad2/3 mutants may contribute to increased Ihh protein diffusion , it is likely that the elevated Ihh mRNA levels seen in these mutants are also responsible for the elevated Ihh protein levels in the resting zone and elevated Ptch1 levels throughout the growth plate ., In accordance , the growth plates of mutants are lengthened at midgestation stages and P0 ., This is consistent with previous studies of Ihh function in the growth plate , where depletion of Ihh in prenatal cartilage caused a loss of columnar structure and dwarfism 65 ., In spite of the longer growth plates at these stages , double mutants exhibit postnatal dwarfism ., This can be explained if the accelerated rate of entry of resting chondrocytes into the proliferative columnar phase leads to premature depletion of the pool of resting chondrocytes and an inability to sustain growth plate elongation at postnatal stages ., The precise role of Ihh in the postnatal dwarfism in Smad2/3 double mutants is unclear; postnatal loss of Ihh in cartilage leads to dwarfism 66 ., On the other hand , Ihh can promote terminal hypertrophic maturation , which could lead to dwarfism , in cells outside the range of PTHrP 67 , 68 ., Additional studies at postnatal stages would be needed to identify the mechanism underlying the postnatal dwarfism phenotype in Smad2/3 mutants ., We speculate that direct effects of Smad2 and Smad3 on genes regulating cell cycle progression may contribute ., Interestingly , the growth plate phenotype at the neonatal stage in Smad2/3 double mutants is distinct from that seen in mice lacking the type II TGFβ receptor TβRII ( Tgfbr2 ) in cartilage 21 ., TβRII is required for responsiveness to all TGFβs ., Tgfbr2CKO mice , which were generated using the same Col2a1-Cre allele used here , exhibit defects in formation of intervertebral discs ( IVD ) , but no apparent alterations in chondrocyte differentiation in axial or appendicular elements 21 ., Obvious defects in IVD formation were not observed at birth in Smad2/3 double mutants , but there were clear defects in the tibial growth plates ., There are several possible explanations for these differences ., Loss of Tgfbr2 impacts both canonical Smad2/3 and non-canonical pathways ., Hence , the axial defects seen in neonatal Tgfbr2CKO mice but not in neonatal Smad2/3 double mutants could reflect the actions of non-canonical TGFβ pathways ., Furthermore , the growth plate defects at neonatal stages seen in Smad2/3 double mutants but not in Tgfbr2CKO mice could reflect a role for Smads 2 and 3 in signaling mediated by ligands other than TGFβs , such as activins 69 ., Our analysis suggests that Smads 2 and 3 may regulate Ihh RNA levels by binding to distinct elements in the Ihh promoter ., Candidate SBEs can be predicted in the promoter regions of many genes , but these motifs are common , and the majority of them are not occupied by R-Smads when examined using ChIP-chip/ChIP-seq 70 ., We found three SBEs within the proximal Ihh promoter that bind Smads 2 and 3 ., These sites exhibit differential recruitment of Smads 2 and 3 , and differential association with distinct co-repressors ., S1 and S3 in the Ihh promoter mediate more of the repressive activity of TGFβ on Ihh expression than does S2 ., A caveat of this study is that the Ihh promoter sequences that regulate Ihh expression in the growth plate have not yet been identified; in vivo mutagenesis studies will be required to identify these ., However , the proximal promoter and enhancer region we investigated is the most highly conserved region among 60 mammalian species ( S6 Fig ) , and this region was shown previously to mediate Smad4 effects on Ihh expression 35 as well as the impact of multiple transcription factors on Ihh expression in chondrocytes 38 , 39 , 50 ., Our studies revealed that Smad2 and Smad3 associate with transcriptional inhibitors Hdac4 , SnoN and Ski after TGFβ stimulation ., Hdac4 inhibits Runx2 activity and Ihh expression 49; SnoN and Ski repress the transcriptional activation activities of Smad2 and Smad3 51 , 52; Sn

Introduction, Results, Discussion, Materials and Methods

TGFβs act through canonical and non-canonical pathways , and canonical signals are transduced via Smad2 and Smad3 ., However , the contribution of canonical vs . non-canonical pathways in cartilage is unknown because the role of Smad2 in chondrogenesis has not been investigated in vivo ., Therefore , we analyzed mice in which Smad2 is deleted in cartilage ( Smad2CKO ) , global Smad3-/- mutants , and crosses of these strains ., Growth plates at birth from all mutant strains exhibited expanded columnar and hypertrophic zones , linked to increased proliferation in resting chondrocytes ., Defects were more severe in Smad2CKO and Smad2CKO;Smad3-/- ( Smad2/3 ) mutant mice than in Smad3-/- mice , demonstrating that Smad2 plays a role in chondrogenesis ., Increased levels of Ihh RNA , a key regulator of chondrocyte proliferation and differentiation , were seen in prehypertrophic chondrocytes in the three mutant strains at birth ., In accordance , TGFβ treatment decreased Ihh RNA levels in primary chondrocytes from control ( Smad2fx/fx ) mice , but inhibition was impaired in cells from mutants ., Consistent with the skeletal phenotype , the impact on TGFβ-mediated inhibition of Ihh RNA expression was more severe in Smad2CKO than in Smad3-/- cells ., Putative Smad2/3 binding elements ( SBEs ) were identified in the proximal Ihh promoter ., Mutagenesis demonstrated a role for three of them ., ChIP analysis suggested that Smad2 and Smad3 have different affinities for these SBEs , and that the repressors SnoN and Ski were differentially recruited by Smad2 and Smad3 , respectively ., Furthermore , nuclear localization of the repressor Hdac4 was decreased in growth plates of Smad2CKO and double mutant mice ., TGFβ induced association of Hdac4 with Smad2 , but not with Smad3 , on the Ihh promoter ., Overall , these studies revealed that Smad2 plays an essential role in the development of the growth plate , that both Smads 2 and 3 inhibit Ihh expression in the neonatal growth plate , and suggested they accomplish this by binding to distinct SBEs , mediating assembly of distinct repressive complexes .

The cartilage growth plate regulates the size and shape of nearly every skeletal element in the body ., TGFβs are potent inducers of cartilage formation , but the mechanisms by which they transduce their signals in cartilage during development are poorly understood ., Similarly , there is strong evidence that dysregulation of the TGFβ pathway increases the risk for osteoarthritis ( OA ) in humans , but the underlying mechanisms are unknown ., TGFβs transduce their signals through a canonical pathway involving Smad2 and Smad3 , and through several non-canonical pathways ., However , the roles of canonical vs . noncanonical signaling are unknown in cartilage because the combined roles of Smad2 and Smad3 have not been determined ., We generated mice lacking both Smad2 and Smad3 in cartilage in order to determine the role of canonical TGFβ signaling during embryonic development ., We determined that Smad2 has a more prominent role than Smad3 in non-hypertrophic chondrocytes in the growth plate , and identified elevated levels of Ihh RNA in neonatal cartilage in Smad2 and Smad3 mutants ., These findings may be important because Ihh is a vital regulator of cartilage proliferation and differentiation during cartilage development ., More generally , the studies identify how Smad2 and Smad3 can regulate a common target gene through distinct mechanisms .

growth plate, medicine and health sciences, chondrocytes, gene regulation, dna-binding proteins, bone, connective tissue cells, cartilage, small interfering rnas, animal cells, proteins, gene expression, connective tissue, biological tissue, smad signaling, biochemistry, rna, signal transduction, anatomy, cell biology, nucleic acids, genetics, tgf-beta signaling cascade, biology and life sciences, cellular types, non-coding rna, cell signaling, signaling cascades

journal.pcbi.1003585

Computational Prediction of Alanine Scanning and Ligand Binding Energetics in G-Protein Coupled Receptors

G-protein coupled receptors ( GPCRs ) are an important group of membrane proteins that mediate physiological signals from the outside to the inside of cells ., They are targets for approximately 30% of all prescribed drugs and of major interest to the pharmaceutical industry 1 ., The understanding of GPCR structure , function and ligand binding has traditionally advanced through a combination of biochemical experiments and computationally generated 3D structure models 2 ., Common experimental approaches include site-directed mutagenesis , generation of chimeric receptors and the substituted-cysteine accessibility method , while 3D models are used for design and interpretation of such experiments ., In recent years , the field has benefitted enormously from breakthroughs in membrane protein crystallography , with a steadily increasing number of GPCR crystal structures determined since 2007 3 ., These structures not only enable structure-based drug design for crystallized targets but also make modelling of homologous GPCRs for the same purpose feasible 4 ., Computational modelling is of optimal use in combination with site-directed mutagenesis data and structure-activity relationships for series of ligands 5 , but requires careful validation ., Reliable free energy calculations based on molecular dynamics ( MD ) simulations can provide the missing links between experimental binding affinities and 3D structures of protein-ligand complexes 6 ., In particular , approaches based on the free energy perturbation ( FEP ) method enable the evaluation of relative binding free energies between different ligands binding to a given receptor as well as to mutant versions of it 7 , 8 ., These techniques can yield accurate and convergent results provided that the complexes compared are not too dissimilar 9 , 10 ., However , when ligands differ by larger substituents , or receptors differ by more drastic mutations ( e . g . , tryptophan to alanine ) , the methodology becomes considerably less reliable due to convergence and sampling problems associated with the simulations ., Hence , reliable FEP schemes for the systematic prediction of ligand binding and mutagenesis effects are rather scarce , and particularly so in the field of GPCRs where they would have a large impact 11 ., The basic problem with applying free energy calculations to complexes that differ substantially in chemical structure is both that numerical instabilities can arise and that conformational sampling becomes more critical , when large groups of atoms vanish or appear during the computational “alchemical” transformations used 8 ., To overcome this limitation , we present here a new FEP scheme for accurate calculation of the energetics of alanine scanning , which is applied to characterize the binding of antagonists to the human neuropeptide Y ( NPY ) receptor type 1 GPCR ., The NPY system is comprised in mammals by three neuronal and endocrine peptides ( NPY , peptide YY and pancreatic polypeptide ) which activate receptors belonging to the rhodopsin-like ( class A ) GPCRs ., Four functional receptors named Y1 , Y2 , Y4 and Y5 exist in humans and are all expressed in the peripheral and central nervous system ., The NPY system has broad biological functions , including involvement in control of feeding behavior , cortical neural activity and emotional regulation ., As a consequence , this system has been implicated in several human diseases such as obesity , alcoholism and depression 12 ., However , until now no effective drugs have been developed for the NPY system , an area that would definitely benefit from structural insights into receptor-ligand interactions ., With no crystal structures yet determined for any of the Y receptors , homology modelling in combination with site-directed mutagenesis has proven extremely useful for characterization of receptor-ligand interactions 13 ., BIBP3226 is a competitive and Y1-selective antagonist which is widely used as a pharmacological tool for studying the physiological role of the Y1 receptor ., For therapeutic application , however , the compound has drawbacks with regard to toxicity as well as low oral availability and brain penetration 14 ., There is extensive experimental data available in the literature for this particular receptor-ligand pair , with binding studies for BIBP3226 to both wild-type ( wt ) and alanine mutants of Y1 15 , 16 , as well as Y1 wt binding data for numerous BIBP3226 analogs 17 , 18 ., We apply our new free energy perturbation scheme to a combined data set of alanine scanning for thirteen amino acids in the binding site region of Y1 and the binding of seven analogs of BIBP3226 , and show how this methodology can be efficiently used to validate structural models of the hY1-BIBP3226 complex ., The structural insights obtained further demonstrate the applicability of the approach in ligand design projects aimed at structure-based development of new GPCR ligands ., In this work thirteen amino acids in the binding site region of Y1 are mutated to alanine using the free energy perturbation technique , namely Y2 . 64 , N3 . 28 , S4 . 57 , F4 . 60 , Y5 . 38 , T5 . 39 , Q5 . 46 , W6 . 48 , T6 . 52 , N6 . 55 , T6 . 56 , F6 . 58 and D6 . 59 ( Figure 1 and Table S1 , Supporting Information ) ., Experimental relative binding free energies for the hY1 mutants compared to the wt receptor were derived from BIBP3226 Ki values 15 , 16 , whereas relative binding free energies between the reference compound BIBP3226 and the seven analogs ( Figure 1 , Table S2 ) were estimated from experimental IC50 values 17 , 18 for wt hY1 ( Methods ) ., The hY1-BIBP3226 complex that was used as starting structure for all FEP calculations is shown in Figure 1A ., The system was generated by homology modelling of hY1 with the program Modeller 19 , followed by insertion of the model in a lipid bilayer and refinement by MD equilibration using GROMACS4 . 0 . 5 20 , as implemented in the GPCR-ModSim web server 21 ., Then both automated docking with Glide 22 and mutagenesis-guided docking of BIBP3226 into the hY1 model were carried out , and the resulting complexes were subject to a final round of MD equilibration using a spherical simulation system using the program Q 23 , which also allows for very efficient FEP calculations 6 ., Based both on structural stabilities of the wt hY1− BIBP3226 complexes and subsequent free energy calculations , the mutagenesis-guided docking approach was found to provide the best starting model ( see below ) ., In this complex BIBP3226 is positioned at the bottom of the hY1 orthosteric binding cavity ., The deep pocket between F4 . 60 and W6 . 48 is occupied by the phenol moiety of BIBP3226 , which places the hydroxyl group at hydrogen bond distance to both Q5 . 46 and N6 . 55 ., The guanidinium group of the ligand forms a salt bridge with the key NPY receptor residue D6 . 59 15 , 16 , 24 and hydrogen bonds to N6 . 55 ., The pocket between transmembrane ( TM ) helices TM2 , TM3 and TM7 and extracellular loop 2 accommodates the biphenyl moiety of BIBP3226 ., The position of the ligands and their interactions with the receptors were generally very stable throughout the MD simulations ., As an example , the BIBP3226 heavy atom RMSD was only 0 . 3 Å between the initial structure and the average wt structure from a total of ( 13+7 ) ×6\u200a=\u200a120 independent equilibration runs ( 60 ns ) for this complex ., Analogously , the RMSD of the side chain heavy atoms belonging to the binding site ( defined as all residues within 5 Å of the ligand ) was also very low ( RMSD\u200a=\u200a0 . 5 Å ) ., The only exceptions to this stability were two types of mutations ., The first includes the N6 . 55A and D6 . 59 receptor mutations which both involve the deletion of a key polar interaction with the D-arginine moiety of BIBP3226 , thereby rendering the ligand more flexible and shifting its position somewhat in the binding pocket ., The second type is ligand modifications that remove the hydroxyl group from BIBP3226 , which provides the hydrogen bonds responsible for attachment to both N6 . 55 and Q5 . 46 ., Free energy simulations of single point mutations where larger residues are mutated to alanine ( alanine scanning ) involve the annihilation of a substantial number of atoms ., The conformational states of the native ( wt ) protein and a given alanine mutant are then often too dissimilar for standard FEP protocols to yield accurate and convergent results ., The most common ways to computationally transform the protein from wt to mutant is either to simultaneously change both electrostatic and van der Waals interaction potentials or to do it separately in two stages ., It has been established that in the annihilation of repulsive atomic centers , an intermediate stage with so-called soft-core potentials ( that avoid singularities ) is beneficial for convergence 25 ., However , the main problem with these approaches is still that the transformation between each stage is carried out via linear combinations of the end state potentials for all atoms involved ., To overcome this problem , we instead constructed a smooth scheme based on successive fragment annihilation , which is illustrated for the case of a Tyr→Ala mutation in Figure, 2 . The basic idea is to divide the whole transformation into a series of smaller “subperturbations” between a number of additional intermediate states , which are designed to be similar enough to ensure convergent free energy differences ., Each subperturbation is as usual divided into a series of even finer grained FEP windows , yielding a total number of perturbation steps of several hundred ( Figure 3 ) ., This strategy is not to be confused with the nowadays outdated “slow growth” method 26 in which only the two end states are used together with a transformation potential that changes in every MD step ., In our scheme we defined groups of atoms in the wt residue ( Figure 2 shows the Tyr example ) , based on their distance to the Cβ atom ., Each group will undergo three consecutive types of transformations during its annihilation: charge annihilation , regular van der Waals ( Lennard-Jones ) potential transformation to soft-core and , finally , annihilation of the soft-core potential ., In the Tyr→Ala case five atom groups are defined and eight independent subperturbations are used ( Figure 2 ) ., For cases where new atoms are instead created , as in the BIBP3226 ligand perturbations discussed below , the scheme is simply reversed and annihilation and creation of groups can also , of course , be treated simultaneously ., We assessed the precision of our method for every protein and ligand mutation from six independent MD/FEP simulations , each corresponding to a total length of 4–6 ns including all subperturbations ., Besides the precision , a critical convergence measure is the hysteresis resulting from applying the FEP formula ( see Methods section ) in the forward and reverse summation direction for each individual simulation ., The average hysteresis obtained in this way from the six replicate trajectories for each alanine scan FEP calculation was in the range 0 . 0–0 . 5 kcal/mol , with an overall average for all mutations of 0 . 25 kcal/mol ., The corresponding hysteresis range for the BIBP3226 ligand mutations was 0 . 0–0 . 1 kcal/mol , with an average over all ligands of 0 . 06 kcal/mol ., These hysteresis errors are , in fact , remarkably small and clearly demonstrate the efficiency of our FEP scheme ., As an illustration , Figure 3A shows the forward and reverse progression of the free energy change for a Tyr→Ala mutation in the hY1 apo structure corresponding to the upper row of the thermodynamic cycle in Figure, 2 . Furthermore , the precision of the different free energy calculations , in terms of standard errors of the mean ( s . e . m . ) based on the six independent trajectories , is very satisfactory and typically about 0 . 5 kcal/mol for the different protein simulations and ≤0 . 2 kcal/mol for the BIBP3226 mutations in water ( Table 1 and Table S3 ) ., The above results can be compared to those of less intricate reference protocols as shown in Figure, 3 . The first of these ( Figure 3B ) transforms electrostatic and van der Waals parameters simultaneously with no extra intermediate states ., The second reference scheme utilizes intermediate soft-core 25 van der Waals interactions and separate transformations of electrostatic and van der Waals potentials , but performs the operations on the entire sidechain simultaneously ( Figure 3C ) ., Intermediate states with soft-core potentials clearly reduce the hysteresis error to some extent ( Figure 3C ) , but it is evident that the stepwise elimination of atoms , with many extra intermediate states , is key to the superior performance of our method ( Figure 3A ) ., As an additional control , Figure 4 shows analogous FEP curves for our scheme and the second reference protocol , extracted from a transformation where one phenyl group is created and one simultaneously annihilated in water ., This is a useful benchmark since the correct free energy change is exactly zero and both hysteresis errors and accuracy ( in this case based on ten independent simulations ) can be assessed ., The result of the FEP calculations utilizing our new method is ΔG\u200a=\u200a−0 . 06±0 . 07 kcal/mol with an average hysteresis error of 0 . 13 kcal/mol ( Figure 4A ) ., Hence , convergence ( hysteresis ) , precision and accuracy are all excellent ., In contrast , the performance of the reference protocol is considerably worse with ΔG\u200a=\u200a3 . 8±0 . 2 kcal/mol with a hysteresis of 0 . 4 kcal/mol ( Figure 4B ) ., The relative binding free energies calculated from the MD/FEP simulations are generally in good agreement with experimental values , thus supporting the validity of the underlying structural model ., For the alanine mutations the mean unsigned error with respect to experimental BIBP3226 binding free energies is 0 . 9 kcal/mol and the method is generally successful in discriminating mutations that have large effects on ligand binding from those that have only minor effects ( Figure 1C ) ., If only the data from Sjödin et al . is considered , which has smaller relative experimental errors 16 , the performance of the FEP calculations improves ( <|error|>\u200a=\u200a0 . 6 kcal/mol ) and better agreement is observed in this case for the two independently measured mutations 15 , 16 F4 . 60A and T5 . 39A ( Figure 1C ) ., Moreover , for the six mutations for which has been determined with an uncertainty of less than 0 . 2 kcal/mol , the mean unsigned error of the calculations is only 0 . 5 kcal/mol ( Table 1 ) ., Comparison of binding free energy differences between calculations and experiment can thus be used to validate the structural model ., Here , the agreement is very good in most instances indicating that this GPCR-antagonist model has a close resemblance to the correct structure ., The binding pocket between TM3 , TM4 , TM5 and TM6 and its interactions with the 4-hydroxybenzylamine and D-arginine groups of BIBP3226 are the part of the structure that is most thoroughly validated ., In our structure , six of the thirteen mutated amino acids - F4 . 60 , T5 . 39 , Q5 . 46 , W6 . 48 , N6 . 55 and D6 . 59 - line the wall of this subpocket and the ligands differ only in this region ( Figure 1A ) ., The FEP calculations reproduce the large positive ΔΔGbind associated with mutating D6 . 59 , N6 . 55 and Q5 . 46 to alanine ( Figure 1C ) ., In the hY1 structure these three residues have ionic and polar interactions with the guanidinium and hydroxyl groups of the ligand ( Figure 1A ) ., It can be clearly seen from the FEP calculations that the large ΔΔGbind is primarily due to considerably more favourable electrostatics for the D6 . 59 , N6 . 55 and Q5 . 46 sidechains in the holo structure compared to the apo structure ( ΔΔGFEP1 in Table 1 ) ., Further , the large effect of the W6 . 48A mutation is also well reproduced by the simulations ., When this tryptophan residue is mutated to alanine a cavity is created deep in the binding site and gradually filled with water , with the total change in binding free energy accumulating gradually over the series of smaller perturbations ( Table 1 ) ., As mentioned , the experimental data for the two mutants F4 . 60A and T5 . 39A is ambiguous ., One report indicates that F4 . 60 has a significant effect on BIBP3226 binding but that T5 . 39A has a negligible effect 15 ., In contrast , the higher precision data say the opposite 16 which is also supported by the present FEP calculations ( Figure 1C ) ., In the structural model of the hY1 complex both of these residues are in contact with the ligand ., Residues Y2 . 64 and N3 . 28 face another part of the binding cavity , namely the pocket between TM2 , TM3 and TM7 ( Figure 1A ) ., Y2 . 64 contacts one of the phenyl groups of the ligand and the FEP calculations yield a lower binding affinity for Y2 . 64A to BIBP3226 in accordance with experimental measurements ., N3 . 28 , on the other hand , is not in direct contact with the ligand and the calculations in this case predict no change in affinity of N3 . 28A for the antagonist , again in agreement with experiment ., The five remaining mutated residues are situated in interfaces between TM helices ., Among these , S4 . 57A , T6 . 52A and T6 . 56A were shown in the experimental assays to bind BIBP3226 with essentially wt affinity 15 ., The FEP calculations reproduce this pattern for S5 . 47A and T6 . 56A , while the binding free energy difference for T6 . 52A is overpredicted by 2 . 7 kcal/mol ( Figure 1C ) ., This is the only real outlier among the 13 alanine mutations examined , which might indicate that the conformation of this sidechain and/or its interaction network is not properly modeled ., Finally , the calculations also reproduce the detrimental effect on BIBP3226 binding affinity for alanine mutations of the two aromatic residues F6 . 58 and Y5 . 38 ., The overall results of the simulations for the relative binding free energies of the BIBP3226 ligand series are remarkably good , with a mean unsigned error of 1 . 2 kcal/mol ., Moreover , the method is clearly successful in discriminating the best binders from the low affinity ligands ( Figure 1D ) ., The calculations closely reproduce the weaker affinity of the dehydroxylated analog ( 2 ) as well as the larger effect of the combined dehydroxylated and ( S ) -methylated compound ( 9 ) ., Although ΔΔGbind for the ( R ) -enantiomer of the latter compound ( 8 ) is somewhat underestimated by the FEP simulations , it is noteworthy that the structural model still correctly discriminates between the two enantiomers ( 8 vs . 9 ) ., Furthermore , the enantiomeric compounds 11 and 12 , which differ in the stereochemistry of their hydroxymethyl substituent at the same chiral center , are both correctly ranked and predicted to be low affinity ligands , in agreement with the experimental binding data ., From the FEP calculations it is also clear that the low affinity of the hydroxymethyl compounds 11 and 12 is due to unfavorable desolvation in the hY1 binding pocket ( see corresponding ΔΔGFEP4 values in Table S3 ) ., The calculations further yield diminished affinities for both the pyridine analog ( 18 ) and the tertiary amide compound ( 25 ) ., As a useful control of the ability of the free energy calculations to discriminate against suboptimal structural models , all of the above FEP simulations were also carried out for the top-ranked solution resulting from the automated docking of BIBP3226 to the hY1 model ( Figure S1 ) ., This docking solution essentially has the ligand rotated 180° around its arginine sidechain thereby interchanging the binding cavities for the phenol and biphenyl groups ., The conformation is intuitively unrealistic since it places the biphenyl moiety in the vicinity of a number of polar groups ., With this ligand orientation the correlation with the experimental binding data for the series of analogs is completely lost , indicating that the substituted phenol moiety must be in the wrong place ., Also the alanine scanning results deteriorate although the effect is not as pronounced , probably due to the fact that the ligand is still occupying the same cavities even though it is flipped ., It is , however , noteworthy that both the N6 . 55A and Q5 . 46A mutations now become outliers , most likely because the hydrogen bonding interactions with the phenol have been lost ., Although the prediction for T6 . 52A mutation is actually better for this model this probably just reflects our suspicion that this receptor sidechain is in the wrong conformation , as discussed above ., Thermodynamic cycle free energy perturbation methods , or alchemical free energy calculations as they are sometimes called , have been around for quite some time 27 and were early applied to biochemical problems such as ligand binding 28 , 29 , protein stability 30 and enzyme catalysis 31 ., These applications were clearly of more exploratory character and it is only recently that more systematic use of the FEP technique has been made , particularly in studies of aqueous solvation 32 , 33 , but also for ligand design purposes 34 and other key biochemical problems dealing with molecular recognition 6 ., However , reliable computational schemes for systematically quantifying the effects of protein mutations on ligand binding have largely been lacking ., In particular , the feasibility of carrying out larger scale computational alanine scanning simulations would be of great importance in connection with such mutagenesis experiments , as these are one of the major experimental routes for probing protein-ligand interactions in the absence of 3D structures ., This is especially true for membrane protein interactions with ligands , such as ion channel blocking and ligand binding to GPCRs , given the limited availability of structural information for these systems ., The free energy calculation scheme developed here turns out to be very efficient for systematically modelling the effect of single-point alanine mutations on protein-ligand binding , even for the complex case of a membrane receptor ., The smooth stepwise transformation procedure overcomes the long-standing convergence problem with FEP simulations that involve the creation or annihilation of many atoms 9 , 10 ., When applied to the hY1-BIBP3226 system , the agreement between calculated and experimental binding free energies is remarkably good for the thirteen alanine mutations and the series of eight receptor antagonists ., These results thus serve to validate the 3D model of the complex and , conversely , a severely erroneous model could immediately be identified as such based on the loss of correlation between calculations and experiment ., It is also noteworthy that even for the most complex Trp→Ala mutation , which involves the annihilation of a complete indole ring , a precision within 1 kcal/mol can be attained with only about 35 ns simulation time for each of the holo and apo states ., A key aspect with regard to efficiency when dealing with many mutants and/or ligand molecules is also the size of the simulation system ., Hence , while the common practice in MD studies of membrane proteins is to set up large simulation systems encompassing lipid bilayer patches with lateral dimensions of a hundred Å or more and a large number of solvent molecules 11 , 35 it is not clear that this strategy is optimal for doing many independent free energy calculations ., After all , the goal in this case is not to simulate conformational changes distal to the binding site but to obtain as reliable free energy estimates as possible at a computational cost that allows many mutants or ligands to be evaluated ., In this respect , reduced models that still yield correct local structural fluctuations of the binding site 36 may be significantly more efficient than larger scale models , precisely because they do not sample large scale conformational motions that require much longer timescales for convergence ., A case in point here is large ribosome complexes where reduced models allow for extensive free energy calculations 6 at a low computational cost ., As far as GPCRs are concerned there has been considerable recent progress with virtual screening strategies using homology models , as exemplified by the D3 dopamine 37 and A2A adenosine 38 receptors ., These cases seem particularly favorable in terms of availability of experimental data ., The D3 receptor both has structural templates with high homology and the existence of well-defined dopamine anchoring points , which is true for aminergic receptors in general ., The A2A homology model , on the other hand , was validated using a unique proprietary technology to generate and characterize hundreds of mutants in vitro 5 , together with large amounts of available binding data ., For systems that are structurally less well characterized it is questionable to what extent virtual screening based on docking to homology models is really meaningful ., In this respect , the combination of experimental and computational alanine scanning , as well as free energy calculations of structure-activity relationships for a series of ligands , can provide the necessary validation needed for model refinement and subsequent virtual screening efforts ., We have shown here that a computationally derived model of the Y1-antagonist complex , obtained from homology modeling and docking simulations , rationalizes the existing mutagenesis and binding data while a suboptimal model of the same complex clearly fails to do so ., Experimental relative binding free energies for the hY1 mutants compared to wt hY1 were derived from BIBP3226 Ki values as ., For the F4 . 60A and T5 . 39A mutants there are two sets of experimental values available from independent reports 15 , 16 , resulting in values that differ by at least 1 . 4 kcal/mol between the two sources ., In these cases we used the average of the two measurements to assess the errors between the calculations and experiment ., Further , mutations that have a BIBP3226 Ki value outside the concentration interval screened in the binding assay were not considered when calculating mean unsigned errors ., Relative wt hY1 binding free energies between the reference compound BIBP3226 and the seven analogs were estimated from experimental IC50 values 17 , 18 as ., The sequence of the hY1 receptor ( Swiss-Prot accession number: P25929 ) was aligned with a multiple sequence alignment of all the inactive-like GPCRs of known structure using the GPCR-ModSim ( http://gpcr . usc . es ) web-server 39 ., The human C-X-C chemokine receptor type 4 ( hCXCR4 ) was considered the best template for modeling of the hY1 receptor because it is a peptide binding GPCR with high homology to hY1 in the C-terminal part of extracellular loop 2 ., This loop segment ( Cys5 . 25-Ser5 . 31 in hY1 ) constitutes part of the orthosteric binding cavity wall and is often involved in ligand binding ., Further , the hCXCR4 structures are determined in the inactive state in complex with antagonists 40 ., This is important since BIBP3226 binds inactive state hY1 ., The sequence identity between hCXCR4 and hY1 in the transmembrane region is 29% ., A chimeric template receptor was assembled making use of the structural alignment of the X-ray structures available from GPCR-ModSim ., The chimeric template consisted mainly of hCXCR4 in complex with a cyclic peptide antagonist 40 ( PDB entry 3OE0 ) , but with some poorly defined intracellular parts extracted from two alternative templates: the intracellular loop 1 and the N-terminal end of TM6 from the hCXCR4 structure in complex with a small antagonist 40 ( PDB entry 3ODU ) while TM8 and the C-terminal end of TM7 were adopted from the hA2AR in complex with ZM241385 41 ( PDB entry 3EML ) ., This chimeric structure was used as template for homology modeling of the hY1 receptor using the program Modeller 9 . 9 19 ., The hY1-hCXCR4 sequence alignment was manually refined in the longer loop regions and the N-terminus was discarded from hY1 modeling due to lack of sequence similarity ., Five hundred homology models of the hY1 receptor were generated and the best candidate model was selected on the basis of low DOPE-HR assessment score 42 and orientation of Asp6 . 59 towards the binding crevice , a residue shown by mutagenesis to be important for both agonist and antagonist binding 15 , 16 , 24 ., The hY1 model was treated with the membrane insertion and equilibration protocol implemented in the GPCR-ModSim web-server 21 ( Figure S2A ) ., Briefly , the system is embedded in a pre-equilibrated POPC ( 1- palmitoyl-2-oleoyl phosphatidylcholine ) membrane model so that the TM bundle is parallel to the vertical axis of the membrane ., The system is then soaked with bulk water and inserted into a hexagonal prism-shaped box of dimensions 118×121×100 Å , consisting of slightly more than 60 . 000 atoms ., The system is energy minimized and equilibrated for 5 ns in a MD simulation with periodic boundary conditions ( PBC ) using GROMACS4 . 0 . 5 20 ., In the equilibration , a first phase of 2 . 5 ns where positional restraints for the protein atoms are gradually released is followed by 2 . 5 ns where positional restraints are only applied to the α-carbons 43 ., The OPLS all-atom ( OPLS-AA ) force-field 44 was used with Berger united-atom parameters for the POPC lipids 45 ., The binding mode of the antagonist BIBP3226 in the equilibrated homology model of hY1 was explored with two alternative docking strategies ., First , automated docking with Glide SP ( Glide , version 5 . 7 , Schrödinger , LLC , New York , NY , 2011 ) was carried out , using default settings and a grid dimension of 30 Å×30 Å×30 Å centered on a point in the binding cavity halfway between T2 . 61 and S5 . 39 , where the top ranked binding mode by GlideScore 22 was selected ., Second , mutagenesis-guided docking was performed with PyMOL ( Version 1 . 4 . 1 , Schrödinger LCC , New York ) , using the extensive mutagenesis and structure-activity relationship data available 15–18 to guide placement of the ligand in the binding site ., Here , we particularly required a salt bridge between D6 . 59 and the D-arginine moiety of BIBP3226 as well as hydrogen bonds between the ligand and the two residues Q5 . 46 and N6 . 55 , as the experimental data indicate these interactions to be important ., Briefly , the manual docking started from a lower ranked docking solution from Glide which had these polar contacts with the receptor ., Manual adjustments of torsion angles and translation displacement of the ligand were performed in PyMOL to enhance the hydrogen bonds ., The structural stability of the obtained ligand-receptor complex was evaluated using the MD equilibration protocol described below ., The final mutagenesis-guided docking pose was generated after two iterative rounds of MD and manual adjustments ., BIBP3226 binding modes from both strategies were further evaluated using MD and FEP calculations ., The hY1-BIBP3226 system was further equilibrated using the MD software Q 23 ., A 40 Å radius spherical system was used , containing the predicted receptor-ligand complex with surrounding lipids and water molecules extracted from the equilibrated PBC simulation system described above ( Figure S2B ) ., Water molecules with oxygen atoms within 2 . 6 Å of any ligand heavy atom were removed ., This spherical GPCR system was equilibrated for 2 . 1 ns using the MD settings described in detail below ., From the final structure of this equilibration a 24 Å radius spherical simulation system was extracted and used as starting structure for all free energy calculations ., MD simulations were carried out using Q with the OPLS-AA force-field 44 ., Simulations of the holo and apo states of the hY1 receptor as well as free BIBP3226 in water were conducted with spherical systems with a radius of 24 Å ( Figure S2C ) ., The GPCR simulation systems were cen

Introduction, Results, Discussion, Methods

Site-directed mutagenesis combined with binding affinity measurements is widely used to probe the nature of ligand interactions with GPCRs ., Such experiments , as well as structure-activity relationships for series of ligands , are usually interpreted with computationally derived models of ligand binding modes ., However , systematic approaches for accurate calculations of the corresponding binding free energies are still lacking ., Here , we report a computational strategy to quantitatively predict the effects of alanine scanning and ligand modifications based on molecular dynamics free energy simulations ., A smooth stepwise scheme for free energy perturbation calculations is derived and applied to a series of thirteen alanine mutations of the human neuropeptide Y1 receptor and series of eight analogous antagonists ., The robustness and accuracy of the method enables univocal interpretation of existing mutagenesis and binding data ., We show how these calculations can be used to validate structural models and demonstrate their ability to discriminate against suboptimal ones .

G-protein coupled receptors constitute a family of drug targets of outstanding interest , with more than 30% of the marketed drugs targeting a GPCR ., The combination of site-directed mutagenesis , biochemical experiments and computationally generated 3D structural models has traditionally been used to investigate these receptors ., The increasing number of GPCR crystal structures now paves the way for detailed characterization of receptor-ligand interactions and energetics using advanced computer simulations ., Here , we present an accurate computational scheme to predict and interpret the effects of alanine scanning experiments , based on molecular dynamics free energy simulations ., We apply the technique to antagonist binding to the neuropeptide Y receptor Y1 , the structure of which is still unknown ., A structural model of a Y1-antagonist complex was derived and used as starting point for computational characterization of the effects on binding of alanine substitutions at thirteen different receptor positions ., Further , we used the model and computational scheme to predict the binding of a series of seven antagonist analogs ., The results are in excellent agreement with available experimental data and provide validation of both the methodology and structural models of the complexes .

membrane proteins, biomacromolecule-ligand interactions, biochemistry, biochemical simulations, transmembrane proteins, cell biology, proteins, biology and life sciences, computational biology, cellular structures and organelles, cell membranes

journal.pgen.1005069

Multi-locus Analysis of Genomic Time Series Data from Experimental Evolution

There is a small but growing literature on the analysis of evolve-and-resequence data ., Feder et al . 30 present a statistical test for detecting selection at a single biallelic locus in time series data ., ( Although it is not a major focus , their method can also be used to estimate the selection parameter . ), Similar to our method , they model the sample paths of the Wright-Fisher process as Gaussian perturbations around a deterministic trajectory in order to obtain a computable test statistic ., However , their aim is slightly different from ours in that they analyze yeast and bacteria data sets where the population size is both large and must be estimated from data ., Here we focus on population sizes which are smaller and more typical of experiments performed on higher organisms , for example mice or Drosophila ., We generally assume that the effective population size is known but also test our ability to estimate it from data ., Also , because of the increased amount of drift present in the small population regime , we necessarily restrict our attention to selection coefficients which are somewhat larger than those considered by Feder et al . Finally , although Feder et al . do study the performance of their method when time series data are corrupted by noise due to finite sampling ( as in e . g . a next-generation sequencing experiment ) , they do not model this effect ., Here we properly account for the effect of sampling by integrating over the latent space of population-level frequencies when computing the likelihood ., Another related work is Baldwin-Brown et al . 31 , which presents a thorough study of the effects of sequencing effort , replicate count , strength of selection , and other parameters on the power to detect and localize a single selected locus segregating in a 1 Mb region ., Results are obtained by simulating data under different experimental conditions and comparing the resulting distributions of allele trajectories under selection and neutrality using a modified form of t-test ., Because it is not model-based , this method is incapable of performing parameter estimation ., As a result of their study , Baldwin-Brown et al . present a number of design recommendations to experimenters seeking to attain a given level of power to detect selection ., In a related work , Kofler and Schlötterer 32 carried out forward simulations of whole genomes to provide guidelines for designing E&R experiments to maximize the power to detect selected variants ., Illingworth et al . 33 derive a probabilistic model for time series data generated from large , asexually reproducing populations ., The population size is sufficiently large ( on the order of ∼ 108 ) that population allele frequencies evolve quasi-deterministically ., The deterministic trajectories are governed by a system of differential equations describing the effect of a selected ( “driver” ) mutation on nearby linked neutral ( “passenger” ) mutations ., Randomness arises due to the finite sampling of alleles by sequencing ., The main difference between the setting of Illingworth et al . ’s and our own concerns genetic drift ., While drift may be ignored when studying a large population of microorganisms , we show that it confounds our ability to detect and estimate selection in populations of order ∼ 103 ., Thus , for E&R studies on ( smaller ) populations of macroscopic organisms , methods which assume that allele frequencies evolve deterministically may not perform as well as those which explicitly take drift into account ., Topa et al . 34 present a Bayesian model for single-locus time series data obtained by next-generation sequencing ., In each time period , the allele count is modeled as a draw from a binomial distribution with number of trials equal to the depth of sequencer coverage , and success probability equaling the population-level allele frequency ., The posterior allele frequency distribution is used to test for selection by comparing a neutral model to one in which unobserved allele frequencies to depend on time ., In the non-neutral case , a Gaussian process is used to allow for directional selection acting on the posterior allele frequency distributions ., Finally , Lynch et al . 35 derive a likelihood-based method for estimating population allele frequency at a single locus in pooled sequencing data ., The method allows for the possibility of sequencing errors as well as subsampling the population prior to sequencing ., Using theoretical results as well as simulations , the authors give guidelines on the ( subsampled ) population size and coverage depth needed to reliably detect a difference in allele frequency between two populations ., Unlike the other methods surveyed here , the approach of Lynch et al . is not designed to analyze time series data ., Hence the data requirements needed to reliably detect allele frequency changes using their method—for example , sequencing coverage depth of at least 100 reads—are potentially greater than for methods are informed by a population-genetic model of genome evolution over time ., Our method differs from the above-mentioned approaches in several regards ., To the best of our knowledge , ours is the first method capable of analyzing time series data from multiple linked sites jointly ., We find that this is advantageous when studying selection in E&R data ., Furthermore , it enables us to analyze features of these data which cannot be studied using single-locus models , such as local levels of linkage disequilibrium and the effect of a recombination hotspot ., Additionally , because our model is based on a principled approximation to the Wright-Fisher process , it can numerically estimate the selection coefficient , dominance parameter , recombination rates , and other population genetic quantities of interest ., In this way it is distinct from the aforementioned simulation-based methods 31 , 32 , methods which only focus on testing for selection 30 , 31 , 34 , or methods based on general statistical procedures which are not specific to population genetics 34 , 35 ., Source code implementing the method described in this paper is included in S1 Code ., The experimental data analyzed in Analysis of a real E&R experiment data are from Franssen et al . 36 and are available on the Dryad digital repository http://dx . doi . org/10 . 5061/dryad . 403b2 ., We consider the following model of an E&R experiment ., A sexually reproducing population of N diploid individuals is evolved in discrete , non-overlapping generations ., Pooled DNA sequencing 37 , 38 is performed T times at generations t1 < t2 < ⋯ < tT ., At each segregating site in the resulting data set , we assume that there are two alleles , denoted A0 and A1 ., ( As will be seen below , up to a change in the sign of the selection coefficient associated with each site , the model is agnostic to which allele is called A0 or A1 . ), Let L and R denote the number of loci and the number of experimental replicates , respectively ., The array D ∊ 0 , 1T×L×R counts relative frequency with which the A1 allele was observed for each combination of generation , locus and replicate ., Given D and a vector of underlying population-genetic parameters θ , let ℙ ( D∣θ ) denote the model likelihood ., In an idealized E&R experiment , generations are discrete and non-overlapping , mating is random , and the population size is fixed , so the likelihood is well approximated by the classical Wright-Fisher model of genome evolution 39:, ℙ ( D ∣ θ , G 0 ) = ∑ G 1 ∊ 𝒢 ⋯ ∑ G T ∊ 𝒢 ℙ ( D ∣ G 0 , ⋯ , G T ) ℙ θ ( G T ∣ G T - 1 ) ⋯ ℙ θ ( G 1 ∣ G 0 ) , ( 1 ), where ℙθ ( Gi∣Gi−1 ) is the transition function of the discrete , many-locus Wright-Fisher Markov chain from genomic configuration Gi−1 to Gi given parameters θ , 𝒢 is the set of all possible genotypic configurations in a diploid population of size N , and ℙ ( D∣G0 , … , GT ) is the probability of the sequencer emitting D conditional on G0 , … , GT ., ( Here , G0 represents the haplotypic configuration of the founding experimental population . In order to take advantage of linkage information we assume that this is known , although as described in Methods this is not necessary in order to use a single-locus version of our model . ), For typical problems , evaluating ( 1 ) is intractable since ∣𝒢∣ is very large and the transition density ℙθ ( Gi∣Gi−1 ) is difficult to compute and store ., Asymptotic ( i . e . , diffusion ) approximations to the transition density may be inaccurate if the population size N and/or scaled generation time 2Nt are small , as is common in an E&R experiment ., Hence , alternative approximations to ℙ ( D∣θ ) are needed to perform inference ., The approximation we make is as follows ., Let X ≡ ( Xijk ) ∊ 0 , 1T×L×R denote the ( unobserved ) population frequency of the A1 allele at each data point ., Conditional on knowing X , and assuming that the DNA sequencer samples each site independently and binomially from the population , we have Dijk ∼ Binomial ( cijk , Xijk ) where cijk is the depth of sequencing coverage observed at this site ., ( Although sequencer coverage is random , we assume that it is independent of all other variables in the experiment and treat it as constant conditional on the observed data . ), Marginalizing over X , we obtain, 𝓛 ( D ∣ θ ) = ∫ 0 , 1 T × L × R ∏ i , j , k 𝓑 ( D i j k ; c i j k , x i j k ) 𝑝 X ( x ∣ θ ) d x , ( 2 ), where 𝓑 ( d;c , x ) = ( cd ) xd ( 1−x ) c−d is the probability mass function of the binomial distribution and 𝑝 X ( x ) is the density of X . Note that if each cijk is large , as when the samples have been deeply sequenced , then the likelihood is ( approximately ) proportional to the density of X , i . e . , 𝓛 ( D∣θ ) ∝ 𝑝 X ( x ) , and the integral in ( 2 ) does not need to be evaluated ., This computational savings can be useful when performing simulations ., To perform inference we must approximate the density 𝑝 X , which represents the joint distribution of all allele frequencies observed in the experiment ., In Methods , we provide the details of the approximation we use ., Briefly , it is as follows: we assume that , conditional on the initial genome configuration G0 , the underlying allele frequencies Xijk are distributed according to a Gaussian process:, X ∣ G 0 , θ ∼ 𝒩 ( μ ( G 0 , θ ) , Σ ( G 0 , θ ) ) ( 3 ), where the first- and second-order moment functions μ ( ⋅ ) and Σ ( ⋅ ) are obtained by considering Wright-Fisher models on a small number of loci ., For example , the terms of Σ ( ⋅ ) correspond to the covariance between a pair of linked sites ( potentially at different time points in the experiment ) under the Wright-Fisher model ., To compute this we can “marginalize out” the remaining loci in the model and study simpler Wright-Fisher model on only two loci ., ( A slightly more elaborate approximation is needed in the case when there is a nearby selected locus , as detailed in Methods . ), Thus , we are essentially approximating the complex joint distribution of allele frequencies using a sequence of simpler one- and two-locus distributions ., This approximation enables us to capture the correct mean and covariance structure in the random variable X while omitting higher order correlations ., Using this approximation we can perform tractable , likelihood-based inference while capturing salient aspects of the linkage-induced correlation present in the data ., Indeed , by ( 2 ) , ( 3 ) and the preceding discussion we have, 𝓛 ( D ∣ θ ) ≈ ∫ 0 , 1 T × L × R ∏ i , j , k 𝓑 ( D i j k ; c i j k , x i j k ) ϕ θ ( x ) d x = : 𝓛 ˜ ( D ∣ θ ) , ( 4 ), where ϕθ denotes the density function of the Gaussian distribution in equation ( 3 ) ., This expression may then be maximized over θ to perform inference ., Alternatively , by placing a prior on θ a Bayesian approach may be adopted , but we do not explore that in this work ., We tested our method on simulated data designed to capture the essential features of an E&R experiment ., See Methods for the details on simulation ., Briefly , it consisted of cloning a set of F homozygous founder lines ( whose haplotypes are assumed to be known ) to form an experimental population of N diploid organisms , which were then simulated forwards in time for T generations according to the Wright-Fisher random mating model ., The experiment was repeated using the same starting conditions to form R experimental replicates ., After the simulation terminated , the frequency of allele A1 was recorded for each combination of segregating site , time period and replicate , possibly with introduced sampling error; this setup mimics pooled sequencing ., The input to the model consisted of these time series allele frequency data along with the haplotypes of the founder lines ., Certain aspects of the simulation were varied to test different aspects of the model; these changes are described more fully in their respective sections below ., Unless otherwise noted , the simulations were performed using F = 200 founder lines , census population size N = 1000 , sampling at generations ti ∊ {10 , 20 , 30 , 40 , 50} , R = 3 experimental replicates and a region of size L = 105 sites ., These values were chosen to reflect a typical E&R experiment and we refer to them in the sequel as the “default” parameter values ., Expected sequencing coverage depth is denoted by C , with C = ∞ corresponding to having perfect knowledge of the population allele frequencies ., We use C = ∞ in the default parameter setting to upper bound the performance of our method , but also consider C ∊ {10 , 30} to investigate the effect of uncertainty in allele frequency estimation ., In these scenarios , coverage at each site was Poisson distributed with mean C . Lastly , scenarios with coverage “Ĉ” denote simulations in which each segregating site had a random level of coverage drawn from the empirical coverage depth distribution observed in actual E&R sequencing data ( see Analysis of a real E&R experiment data for further details . ) The average coverage depth observed in this experiment was 84 short-reads , but the distribution has a heavy left tail which leads to a small percentage of sites having little to no coverage ( S1 Fig ) ., A common objective in E&R experiments is to detect genetic adaptation ., For example , a population may be partitioned , with one subgroup placed in a new environment ., Upon running an E&R experiment , one wishes to 1 ) determine whether a fitness difference exists between the control and subject groups; 2 ) find the alleles responsible for the adaptation; and 3 ) estimate the strength of selection acting on these alleles ., To test our model’s ability to perform each of these tasks , we simulated E&R experiments in which a segregating site in the founding population was chosen uniformly at random and placed under selection ., The relative fitnesses of A0/A0 and A1/A1 homozygote genotypes are respectively given by 1 and 1+s , while the relative fitness of the heterozygote A0/A1 is 1+hs ., In what follows , we assume h = 1/2 unless stated otherwise ., Let si denote the coefficient of selection at segregating site i = 1 , … , K , where K is the total number of segregating sites in the region being considered ., We wish to test the following null and alternative hypotheses:, H 0 : s 1 = ⋯ = s K = 0 , versus H A : s j ≠ 0 for some j , ( 5 ), which can be implemented using a standard likelihood-ratio ( LR ) test ., As the number R of experimental replicates grows large , the distribution of the test statistic under the null hypothesis tends to a χ2 distribution ., However , since R was set to a realistic ( i . e . , small ) value in our experiments , we found that the test performed better if the null distribution was determined empirically ., The null distribution was calculated by performing additional simulations under neutrality ( s = 0 ) , computing the maximum likelihood estimate ŝ for each simulation , and then using these estimates to compute the empirical null distribution of the LR test statistic, - 2 log 𝓛 ˜ ( D ∣ s = 0 ) - sup u log 𝓛 ˜ ( D ∣ s = u ) , ( 6 ), where 𝓛 ˜ ( D | s = u ) is defined in ( 4 ) ., Using the default parameter settings mentioned earlier , Fig . 1 displays the test’s estimated receiver operating characteristic ( ROC ) curve for various strengths s of selection and various numbers of founding haplotypes ( F ) ., Larger values of F correspond to increased haplotypic diversity in the start of the E&R experiment ., Each curve was estimated from 200 simulations ., Some overall trends are apparent: stronger selection is easier to detect than weaker selection , and increased haplotypic diversity makes it more difficult to confidently reject the null hypothesis of neutrality ., With a small number of initial haplotypes ( F = 20 ) , strong selection ( s = 0 . 1 ) is easily distinguished from neutrality ., Moderate selection ( s = 0 . 05 ) is more challenging to detect , but the test still has 75% power with a false positive rate of ∼ 6% ., Weaker selection ( s = 0 . 02 ) poses more of a challenge; in this case achieving 50% power would entail a false positive rate of approximately 30% ., As the number of founding lineages increases , it becomes harder to test for selection ., This occurs because many sites are segregating at low initial frequencies , increasing the chance that some are lost due to drift ., Detecting weakly selected variants may be confounded by genetic drift , which can cause low-frequency alleles to be lost even if they are under positive selection ., One option for improving sensitivity to weaker selection is to reduce the effect of drift by increasing the effective population size used in the experiment ., To study how this influences our ability to detect weaker selection , we ran additional simulations with larger population sizes N ∊ {2000 , 5000} while holding the remaining experimental parameters fixed ., Results from these experiments are shown in Fig . 2 ., The top panel ( N = 1000 ) is reproduced from the middle panel of the preceding figure for ease of comparison ., We see that reducing the amount of genetic drift in the data improves the performance of the test , particularly when it comes to distinguishing weak selection ( s = 0 . 02 ) ., Once selection has been detected in a region , it is desirable to map the selected site as accurately as possible ., An obvious estimator in this case is to declare the site with the highest likelihood-ratio ( versus a neutral model ) from the preceding test to be the selected site ., Table 1 shows how this estimation procedure performed for different strengths of selection ., We also studied how varying the number of founding lines affected the ability to precisely locate the selected site by allowing F to take on the values F ∊ {20 , 200 , 2000} ., Since the minimum minor allele frequency ( MAF ) in an E&R experiment is 1/F , a low number of founding lines ensures that sites are segregating at intermediate frequencies , while a large value of F decreases LD and improves the ability to map the selected site accurately ., Note that under our default parameter regime , setting F = 2000 amounts to sampling each founder from a panmictic population of size , so that the patterns of diversity reflect what would be seen in a ( neutrally evolving ) region in nature ., Two measures of the accuracy are displayed in Table 1 ., The first set of columns examines the distribution of the distance ( in base pairs ) between the estimated and true selected site ., The second set of columns examines the distribution of the rank of the true selected site when all segregating sites in the region are sorted according to their likelihood ratio ., As the table shows , selection becomes easier to localize as it becomes stronger and as the number of founder haplotypes grows ., With strong selection ( s = 0 . 1 ) and 20 founding haplotypes , the method correctly pinpointed the exact location of the selected site in over 50% of the simulations ., Additionally , the correct selected site was among the top four in 75% of the simulations ., With F = 200 founder lines , the true selected site ranked among the top two overall in over half the simulations ., The top rows of Table 1 indicate that weak selection ( s = 0 . 02 ) is difficult to localize precisely using this method; the median estimated distance from the true selected site was 27–29 kb in these cases ., Since increasing the number F of founder lines diminishes linkage disequilibrium , it may seem counterintuitive that our results suggest that localizing selection actually becomes more difficult as F increases ., In S1 Table , we have displayed the same statistics as Table 1 for the restricted subset of simulations where the selected site was segregating at an initial frequency of at least 0 . 1 ., Compared to the unrestricted data set , these sites are more likely to rise in frequency by the action of positive selection , and less likely to be lost due to drift ., Here we see that increasing F does improve the ability to map the selected site for s ∊ {0 . 02 , 0 . 05}; for strong selection ( s = 0 . 1 ) , essentially all cases of F performed equally well ., Interestingly , an intermediate number of founding lineages ( F = 200 ) seems to outperform both other regimes , suggesting that there is a trade-off between improving localizability by increasing F and limiting the number of segregating sites which must be considered by decreasing the number of founding lineages ., We also studied how coverage depth affects the ability to map the selected site ., For F = 200 , Table 2 repeats the analysis of Table 1 when the data are sampled at simulated coverage depths of 10 and 30 short-reads , as well as from the empirical coverage distribution discussed above ., Comparing the two tables , we see that the additional noise introduced by sequencing makes the problem of localizing the selected site more difficult; the modal estimate is often separated from the true site by tens of kilobases ., Nevertheless , in more than half the trials performed we observed that a strongly selected site would be among the top five segregating sites ( in terms of likelihood ratio; see Table 2 , last two rows ) ., For medium selection , increasing coverage depth from 10 to 30 short-reads improved our ability to map the selected site by several kilobases , and more than halved the number of segregating sites we would need to examine before encountering the selected site ., Weaker selection , already difficult to detect without sampling , is even more so when noise is introduced ., Once a selected site has been located , it is desirable to numerically quantify the fitness of the A1 allele ., Table 3 describes the distribution of these estimates for various combinations of selective strength , coverage depth , and model complexity ( i . e . , the number of loci in the Gaussian process approximation ) ., For each of the simulations above we estimated s by maximum likelihood ., To separate the ability of our model to estimate selection from its ability to locate the selected site , we assumed that the selected site was already known when performing these estimates ., Aside from varying selection strength , we also examined how coverage depth and the number of loci used for estimation affected the quality of the estimates ., For each parameter combination , the table displays the mean , median and inter-quartile range ( IQR ) of the distribution of the maximum likelihood estimate ŝ of s ., Several interesting features emerge from the table ., Inter-quartile range is of roughly the same order across scenarios , so that estimation error shrinks relatively as selection become stronger ., For one-locus models , IQR shrinks as coverage depth increases ., For multi-locus models the effect of increasing the number of sites used to perform estimation is interesting ., When the data are observed without noise , we saw little improvement in the accuracy of ŝ when using a single-locus model fit only to data from the selected site versus a multi-locus model which also took the trajectories of linked sites into account ., In fact , in several cases this cause the estimates to become more dispersed as the trajectory of the selected allele had relatively less weight in the likelihood calculation ., On the other hand , when allele frequencies are sampled with noise we see that estimates ŝ obtained from a five-locus model generally have smaller IQR , particularly in the low-coverage-depth case C = 10 ., These findings are confirmed in Fig . 3 , which displays density estimates for the residual s−ŝ for each of these cases presented in the table ., Compared with the one-locus model , the five-locus model which takes additional data from linked sites into account produces estimates which are more concentrated around the true parameter value ., Thus , when the data are noisy ( i . e . , when C is small ) , the trajectories of nearby linked sites provide useful information concerning the ( unobserved ) population frequency of the selected allele as it evolves over time ., We observed a slight negative bias for weaker selection and a slight positive bias for medium and strong selection , which can be attributed to loss or fixation of the selected allele ., Indeed , estimated selection may be negative when a weakly selected allele segregating at low frequency is lost due to drift; similarly , there is a tendency to overestimate the strength of selection acting on a high-frequency allele which fixes quickly ., It is also interesting to consider the effect of study design on estimation accuracy ., In Table 4 we examine how parameter estimates are affected by sequencing effort and experimental duration ., We focus on the limited-coverage case ( C = 10 ) since it is most sensitive to adding or removing sequence data from additional generations ., For ease of comparison , the first set of rows reproduces data from Table 4 , where generations {10 , 20 , 30 , 40 , 50} were sequenced ., The next subsection examines the case when sequencing effort is reduced to two time periods {25 , 50} ., The final subsection studies estimation quality when the experimental duration is halved , and only one round of sequencing is performed at generation 25 ., In all cases we see that the estimators are approximately unbiased , 𝔼 ( ŝ ) ≈s , but that their dispersion about the true parameter value is greatly affected by data availability ., Sampling genomic data at just a single time period t = 25 roughly doubles the IQR of the estimator in each case ., Interestingly , with two time periods ( t ∊ {25 , 50} ) performance is improved , and the estimator is only somewhat less precise than when sampling at every tenth generation ., Finally , as in the previous table we see again that , at least for data sampled at low coverage , estimation performance is unilaterally improved by fitting a multi-locus model versus a single-locus model ., In the preceding discussion , the dominance parameter was fixed at h = 1/2 , so that selection acted additively ., Our method is capable of handling general diploid selection ., In our experiment , we tested our method’s ability to estimate the effect of overdominance , in which case heterozygotes are fitter than either homozygote ., We simulated populations under the conditions h > 1 and s ≪ 1 such that heterozygotes had a relative fitness of 1+hs where hs ∊ {0 . 02 , 0 . 05 , 0 . 10} ., Thus , heterozygotes have a fitness advantage of the same order as that which we were able to detect in the additive case ., Results for jointly estimating h and s are shown in Table, 5 . A fixed value of s = 0 . 01 was used for fitness in all cases , while h was varied ., We found that estimating overdominance is difficult when both alleles are initially segregating near their limiting frequency of ½ , since the resulting allele trajectories appear very similar to those generated by a neutral model with drift ., The results in the table are therefore conditioned on the initial allele frequency residing outside of the interval 0 . 4 , 0 . 6 ., When considered individually , the estimators ĥ and ŝ are highly variable ( see Table 5 , columns 3–6 ) ., This behavior is expected since , as witnessed in the previous subsections , small values in s ( specifically , s = 0 . 01 ) are difficult to detect in experimental data ., Encouragingly , a different picture emerges when we consider the product estimator ĥ⋅ŝ ( see Table 5 , columns 7–8 ) ., The estimator is close in expectation to the true value hs ( column 2 ) and also more tightly concentrated around that value ., Density estimates of the product estimator ĥŝ are shown in Fig . 4 and confirm this finding ., Each density estimate has a mode at the true parameter value hs and is reasonably concentrated around that value ., Our multi-locus model can also be used to study phenomena which alter covariance between linked alleles ., For example , in a region containing a recombination hotspot , covariance decreases markedly as increased recombination breaks down linkage disequilibrium ., Using the same likelihood-based approach as above , the recombination rate within the hotspot can be estimated from E&R data ., To test this , we simulated a region of length L = 100 kb in which the middle 2 kb region had an elevated recombination rate rH = α ⋅ r , where r = 10−8 is the background recombination rate and α ∊ {10 , 102 , 103} ., For simplicity , we focused on the case of C = ∞ and assumed that the hotspot boundaries are known ., For each simulation , a 30-locus model was fit using 10 randomly-selected loci from within the hotspot and 20 outside of it ., Density estimates for the residual log10 ( r̂H ) −log10 ( rH ) are shown in Fig ., 5 . In all cases , the mode of the density occurs close to zero ., A 3-order increase in the recombination rate is easily detected in experimental data , and a 2-order increase can also be estimated to well within an order of magnitude of accuracy ., Increasing the recombination rate by only a factor of 10 leads to a fairly dispersed estimator , and it would be difficult to detect using the default experimental parameters ., As a final application of our method , we consider estimating the effective population size Ne from experimental data ., Up to now we have assumed that the ( census ) size N of the experimental population is fixed at a known value ., In practice , the effective and census population sizes may differ due to various factors , including nonrandom mating and population structure ., It could be interesting to quantify this effect by estimating Ne in experimental data using the same likelihood-based procedures described above ., Since our model approximates the Wright-Fisher process , in which Ne = N , and simulations were carried out also assuming the Wright-Fisher model , we expect our estimate N̂e to be close to N . Fig . 6 shows a scatter plot of N̂e versus N for 1 , 000 simulated E&R experiments ., In each experiment , the population size N was chosen uniformly at random from the interval 10 , 104 ., We see that the estimator is quite accurate for small population sizes and becomes more variable as N grows ., This is expected since N̂e is essentially measuring genetic drift , which is of order O ( 1/N ) as N grows ., Thus , the inverse map taking drift to population size is well-conditioned for small N and becomes ill-conditioned as N grows ., Finally , we tested our method on data from an actual E&R experiment of D . melanogaster adapting to a new laboratory environment involving an alternating cycle with 12-hrs of cold ( 18∘C ) and 12-hrs of hot ( 28∘C ) temperature conditions ., The experiment has been described previously 25 , 36 , so we give only a brief summary here ., The experiment consists of three D . melanogaster populations each of N ≈ 1000 individuals ., The populations were founded by gravid females from isofemale lines , and then evolved forward in discrete generations ., Pooled sequencing was performed at generations 15 , 37 , and 59 on three experimental replicates ., The observed coverage distribution for a sele

Introduction, Results, Discussion, Methods

Genomic time series data generated by evolve-and-resequence ( E&R ) experiments offer a powerful window into the mechanisms that drive evolution ., However , standard population genetic inference procedures do not account for sampling serially over time , and new methods are needed to make full use of modern experimental evolution data ., To address this problem , we develop a Gaussian process approximation to the multi-locus Wright-Fisher process with selection over a time course of tens of generations ., The mean and covariance structure of the Gaussian process are obtained by computing the corresponding moments in discrete-time Wright-Fisher models conditioned on the presence of a linked selected site ., This enables our method to account for the effects of linkage and selection , both along the genome and across sampled time points , in an approximate but principled manner ., We first use simulated data to demonstrate the power of our method to correctly detect , locate and estimate the fitness of a selected allele from among several linked sites ., We study how this power changes for different values of selection strength , initial haplotypic diversity , population size , sampling frequency , experimental duration , number of replicates , and sequencing coverage depth ., In addition to providing quantitative estimates of selection parameters from experimental evolution data , our model can be used by practitioners to design E&R experiments with requisite power ., We also explore how our likelihood-based approach can be used to infer other model parameters , including effective population size and recombination rate ., Then , we apply our method to analyze genome-wide data from a real E&R experiment designed to study the adaptation of D . melanogaster to a new laboratory environment with alternating cold and hot temperatures .

A growing number of experimental biologists are generating “evolve-and-resequence” ( E&R ) data in which the genomes of an experimental population are repeatedly sequenced over time ., The resulting time series data provide important new insights into the dynamics of evolution ., This type of analysis has only recently been made possible by next-generation sequencing , and new statistical procedures are required to analyze this novel data source ., We present such a procedure here , and apply it to both simulated and real E&R data .

journal.pcbi.1002390

Effects of Electrical and Structural Remodeling on Atrial Fibrillation Maintenance: A Simulation Study

Atrial fibrillation ( AF ) is a cardiac arrhythmia characterized by rapid and irregular atrial activation ., Such desynchronized activation may occur when multiple waves circulate the atria ., Unlike ventricular fibrillation , where unsynchronized activation of the ventricles ( the main pumping chambers of the heart ) causes an immediate and typically fatal loss of blood pressure , atrial fibrillation may be a repetitive , even chronic , disease ., In fact , AF is the most common sustained cardiac arrhythmia in the United States and the rest of the developed world 1 , 2 , with more than 2 . 3 million sufferers in the U . S . 3 ., AF becomes increasingly common with age 2 and is associated with significant mortality and morbidity , such as heart failure and stroke 1 ., AF is more prominent in the context of alterations in atrial tissue properties – due to disease , arrhythmias , or age – known as remodeling ., In fact , AF itself leads to remodeling , causing electrophysiological ( “electrical” ) , contractile , and structural changes 4 ., Although AF can typically be reversed in its early stages , it becomes more difficult to eliminate over time due to such remodeling – hence the expression “AF begets AF” 5 ., A central hypothesis for why AF begets AF is that electrical and structural remodeling due to chronic or persistent AF shorten the action potential wavelength , which measures the spatial extent of the action potential ., Such wavelength shortening allows more waves to fit in the atria and maintain the arrhythmia 6 ., Electrical remodeling primarily shortens the refractory period and the action potential duration ( APD ) of the atrial action potential , while structural remodeling impedes propagation and hence decreases conduction velocity ( CV ) ., Since the wavelength is given as the product of APD and CV ( or , alternatively , the product of the effective refractory period and CV ) , electrical remodeling and structural remodeling both decrease the wavelength , thus potentially perpetuating AF ., Additionally , the stability of reentrant waves may be affected by remodeling ., Prior modeling work has shown that flattening APD restitution ( the dependence of APD on the previous resting interval or Diastolic Interval , DI ) , which typically occurs as a consequence of electrical remodelling 7 , may stabilize reentry 8 ., Likewise , diffuse fibrosis , which may occur during structural remodelling 9 may stabilize reentrant waves 10 ., Clinically , because electrical and structural remodeling typically present jointly in patients with chronic AF , their effects are difficult to separate ., Animal models of primarily electrical remodeling ( due to rapid atrial pacing ) and predominantly structural remodeling ( induced heart failure or mitral regurgitation ) exist , however the rapid atrial pacing models also typically develop some degree of structural remodeling while the heart failure animals undergo some concomitant electrophysiological changes 9 , 11 ., We therefore decided to use computer modeling as a means to investigate the mechanisms of how APD shortening due to electrical remodelling , and CV slowing due to structural remodelling , influence the duration and spatiotemporal dynamics of simulated AF in a computational multiscale model of human electrophysiological dynamics and substrates ., Because structural aspects of the complex atrial anatomy are important for , e . g . , anchoring waves to anatomical obstacles and thus influencing the duration of reentrant activity , we use an anatomically detailed structural model of the human atria ., We model the atria using the Courtemanche et al . cellular model of human atrial cell electrophysiology 12 , with computational cells diffusively coupled to their nearest neighbors in an anatomically derived , three-dimensional structural model of the human atria 13 ., As in previous work from our group 14 , we increase the conductance of the inward rectifier current , IK1 , here by 75% , in order to get the baseline action potential duration and resting membrane potential closer to experimentally observed values ., Further , as in our previous work 14 , we fix the intracellular concentrations of K+ and Na+ ( at 139 . 0 mM and 11 . 2 mM , respectively ) to avoid long-term drift ., The anatomical model incorporates heterogeneous coupling , resulting in different conduction velocities in different anatomical regions , in agreement with human data 13 , 15 ., Specifically , the model exhibits fast conduction in Bachmanns bundle , the pectinate muscle network , the crista terminalis , and the limbus of the fossa ovalis ( 120 cm/s during sinus pacing ) ; slow conduction in the isthmus and the fossa ovalis ( 36 cm/s ) , while the remaining ( bulk ) atrial tissue has intermediate conduction velocity ( 65 cm/s ) ., The different regions are shown in Fig . S1 in Text S1 ( online Supporting Information ) ., The model does not include anisotropy ., In human atria , the bulk atrial muscle has a more random fiber orientation than the fast conduction pathways ( and also more random than the ventricular myocardium ) , which have well-organized orientations along the bundles 13 ., However , due to the strip-like anatomy of the fast tissues ( Fig . S1 in Text S1 ) , anisotropy is predicted to play a minor role there ., This mathematical representation of the atria reproduces basic features such as depolarization time and spatial profile during normal ( sinus ) pacing 13 ., The three-dimensional anatomical model is discretized in a 300×285×210 grid of spatial nodes , with a spacing of Δx\u200a=\u200a0 . 025 cm , and no-flux boundary conditions ., The equations were solved using an operator-splitting method 16 with forward Euler integration of both operators ., We used a fixed time step of Δt\u200a=\u200a0 . 01 ms for the partial differential equation describing the diffusion of voltage ., The cellular model was integrated using an adaptive time step 16 ., The code was parallelized using OpenMP , and run on multi-core machines ., Simulating 60 s of reentrant activity took 5 days on a 24-core machine ( 2 . 66 GHz Intel® Xeon® X7460 processors , 128 GB memory ) ., Because the simulations are this computationally costly , simulations were stopped when reentry terminated or at 60 s ( in which case , the arrhythmia was classified as sustained ) , whichever came first ., Electrical remodeling due to chronic AF was simulated as in previous work , incorporating a 70% decrease in the conductance of the L-type calcium current ( ICaL ) , a 50% decrease in the conductance of the transient outward current ( Ito ) , and a 50% decrease in the conductance of the atrial-specific , ultra-rapid potassium current ( IKur ) 7 ., These values are based on current recordings in cells isolated from human atrial appendages ., We refer to this set of values as 100% , or full , electrical remodeling ., In order to simulate different degrees of electrical remodeling ( 10–90% ) , in some simulations the percentage changes in the three affected conductances were downscaled by the same factor ., We simulate structural remodeling by decreasing the diffusion coefficients ( i . e . , the coupling strengths between computational cells ) , which reduces conduction velocity ., The three different nominal diffusion constant values ( assigned to fast , bulk , and slow conducting tissue ) were scaled by the same factor ., Maximal structural remodeling was set to a 50% decrease in diffusion , causing the time for full activation of the atria with sinus pacing to increase from 108 ms to 149 ms . However , as experimental and clinical data show a large range of conduction impairment with structural remodelling 9 , we investigate two more levels of structural remodeling , using downscaling in diffusion of 70% and 83% ( increasing activation time to 119 ms and 130 ms , respectively ) ., Because the duration of reentrant episodes may depend on where the reentry is initiated , we simulated reentry initiated at three different locations: the left atrial free wall , the left atrium near the left pulmonary veins , and the right atrial free wall ., At each of these locations , reentry was initiated using a cross-gradient protocol , using a stimulus current of 80 nA/µF for 1 ms . Because the vulnerable window for reentry initiation is very small in the non-remodeled virtual tissue , we applied a brief hyperpolarizing clamp ( −80 mV for 1 ms ) to the region of the second wave excitation , 30 ms after its initiation ., This allows for earlier reentry into this region and increases the vulnerable window ., The coupling interval ( i . e . , the time between the first and the second excitation ) was varied systematically between simulations in steps of 10 ms within the vulnerable window ., The size of the vulnerable window varies with variation in electrical and structural remodeling parameters , but was in the range of 30–60 ms , such that 4–7 reentry simulations were initiated at each location ., Applied to three different locations , this means that for a given set of parameters describing the degree of electrical and/or structural remodeling , 12–21 simulations were run with different initiations ., DI and APD were recorded from 16 different locations , spread evenly throughout the atria ( see Fig . S2 in Text S1 ) ., The APD was measured as the time from the crossing of −70 mV on the upstroke to the crossing of −70 mV during repolarization ., Inversely , DI was measured as the time between the crossing of −70 mV during repolarization to the crossing of −70 mV on the next upstroke ., The wavelength ( WL ) is difficult to measure accurately during reentrant activity , even in computational studies 17 , due to wave collisions and irregular wave propagation ., We use a method similar to that employed by Graux et al . ( Ref . 18 ) and determine CV during periodic pacing in the left atrial free wall ., However , rather than obtaining CV for very few pacing rates as necessitated in the clinic , we systematically varied the pacing rate to establish the dependence of the CV on diastolic interval ., Such restitution curves were obtained for all the combinations of the different levels of electrical and structural remodeling simulated ., We later used these restitution curves to estimate local CV for DI values measured during reentry and , finally , to compute the local wavelength as WL\u200a=\u200aAPD×CV ., Note that this definition of the wavelength is more practical than WL\u200a=\u200aERP×CV , where ERP is the effective refractory period , since ERP measurement requires a series of stimuli and cannot be measured directly during simulated AF ., In paced tissue simulations , we found that APD underestimates ERP by 9–16 ms ( 4–10% ) depending on pacing rate and level of remodeling ., Hence , our calculations of the wavelength using APD are presumably 4–10% larger than estimates based on ERP ., Atrial fibrillation and flutter can be characterized by the number of wavelets present in the tissue ., As in our groups previous work 14 , we compute the location of wave tips ( filaments ) from the crossing of two isopotential curves ( the crossing of −30 mV on the upstroke ) , separated in time by 2 ms . The number of separate filaments is determined by applying a k-means clustering analysis to the filament location data ., To characterize individual simulations we use the maximal number of filaments present in that run ., Dominant waves were defined as waves existing for at least five rotations ., Their locations were determined directly from the filament location data or ( in the case of anchored waves without filaments ) based on periodicities in the transmembrane potential from the 16 recording sites , as well as visual inspection of isopotential surface maps ., As shown previously , electrical remodeling leads to shortening of the APD 7 ., In particular , in our simulations of tissue strands , full electrical remodeling reduces the APD at 1 Hz pacing from 228 ms to 135 ms , while at 5 Hz the APD is shortened from 134 ms to 103 ms . This reduction in the amount of APD shortening with faster pacing demonstrates the flatter APD restitution occurring with electrical remodeling ( see Fig . S3A in Text S1 ) ., The CV is unchanged with electrical remodeling ( Fig . S3C in Text S1 ) ., In contrast , structural remodeling decreases CV ( CV restitution slope remains largely unchanged; Fig . S3D in Text S1 ) , while the APD is unchanged ( Fig . S3B in Text S1 ) ., As a measure of the effects of electrical and structural remodeling , we focus primarily on the duration of reentrant activity ., In our three-dimensional model , in the absence of electrical and structural remodeling , reentrant activity is not sustained: in all simulations of normal tissue , with varying initiation time and location ( see Methods ) , reentry ended 1–3 seconds after initiation ., This is consistent with clinical findings in the normal human atria , where AF episodes typically self-terminate soon after initiation ., Fig . 1A shows an example of non-sustained reentrant activity in normal tissue ., The wavelength in this case is sufficiently long that the reentrant wave eventually runs into refractory tissue and dies out ., In contrast , when simulating full electrical plus structural remodeling , reentry was sustained for 60 s in 18 of 21 simulations ., With such electrical plus structural remodeling the wavelength is much shorter than in normal tissue ( Fig . 1B ) , and the reentrant wave in the left atrial free wall does not self-terminate ., Videos showing these dynamics with and without remodeling are available as Supporting Information ( Videos S1 and S2 ) ., To investigate whether this maintained reentrant activity results from more waves being present versus those present being more stable , we measured the maximal number of filaments in a simulation , as well as the mean duration of reentrant activity , while varying the level of electrical and structural remodeling ., Electrical and structural remodeling both lead to increases in the duration of reentrant activity , and the combination of electrical plus structural remodeling gives even longer sustained reentry ( Fig . 2A ) ., Structural remodeling also causes an increased number of filaments , while the level of electrical remodeling does not exhibit a clear correlation with the maximal number of filaments ( Fig . 2B ) ., Note that the number of filaments present at baseline is consistent with previous experimental and modeling studies 14 , 19 , 20 ., To test the hypothesis that a smaller wavelength perpetuates AF , we determined the dependence of the duration of reentrant activity on the estimated wavelength ., For both electrical and structural remodeling , a decrease in WL is associated with longer reentry duration when WL is below a threshold value of around 7 cm ( Fig . 3A ) ., Interestingly , the dependence of reentry duration on WL is similar for both electrical and structural remodeling , suggesting that WL is a more important determinant of duration than other factors , such as APD restitution slope , that vary between electrical and structural remodeling ., In contrast , the dependence of the maximal number of filaments on WL is very different for electrical vs . structural remodeling , with more filaments present during structural than electrical remodeling for the same WL ( Fig . 3B ) ., This indicates that more conduction block and wavebreaks occur during structural remodeling , but might also be the result of fewer waves being anchored to anatomical obstacles , since an anchored wave does not necessarily have a filament ., From analyzing the dynamics of the number of filaments , we found that with structural remodeling , frequent occurrences of conduction block and wavebreaks cause the larger maximal number of filaments relative to electrical remodeling ., These wavebreaks often heal , such that the increase in filaments is transient ( Figs . S4 and S5 in Text S1 ) ., Taken together , these results demonstrate that the wavelength determines the duration of simulated AF , despite differences in dynamics such as APD restitution , conduction block , and number of filaments ., As mentioned above , in our simulations of periodically paced tissue strands , electrical remodeling shortens APD without changing CV , and structural remodeling decreases CV without affecting APD ., If these dependencies hold during reentry in the three-dimensional atrial anatomy , then APD itself should be a marker for reentry duration during electrical remodeling , while CV should correlate with reentry duration during structural remodeling ., Such markers might be valuable given the methodological difficulties in determining WL ( see Methods ) ., However , during reentry in the anatomical model with simulated electrical remodeling , there is both a decrease in APD and a concomitant fall in CV ( Fig . 4A , C ) ., For structural remodeling , there is a primary decrease in CV ( Fig . 4D ) and a secondary increase in APD ( Fig . 4B ) ., These results show first of all that APD alone ( for electrical remodeling ) and CV alone ( for structural remodeling ) are not accurate surrogates for WL during reentry ., A similar finding was reported for AF/AFL inductance in a canine model 20 ., Importantly , the secondary changes also suggest that the dynamically induced differences in the wave characteristics ( APD and CV ) may be due to differences in preferred pathways during reentry in the anatomical model , since differences in pathway lengths would affect the size of the excitable gap , and hence cause dynamical changes in DI , APD , and CV ., Our different reentry initiation protocols allow a range of different spatiotemporal dynamics to occur ., In all simulations that ran the full 60 s , the reentrant activity settled into a relatively periodic rhythm , with dominant waves remaining in a particular location ( often circulating an anatomical obstacle ) ., However , as shown in Table 1 , the location of the dominant wave ( s ) varied significantly ., In general , with only electrical remodeling dominant waves tended to be located in the left atrium , while dominant waves were found in the right atrium when we simulated structural remodeling only ., With combined electrical and structural remodeling , dominant waves were in either or both atria ., More specifically , for electrical remodeling , 4 of 6 simulations resulted in reentry around the pulmonary veins ( Table 1 ) ., Fig . 5A shows an example of such dynamics ( see also Video S3 in the Supporting Information ) ., A wave is anchored to the left pulmonary veins during the entire rotation ( Fig . 5A , left ) ., Another wave front is circulating the right pulmonary veins , but does not remain completely anchored for the entire rotation ( Fig . 5A , right ) ., Excitation spreads from the left atrium to the right ., During structural remodeling , all simulations of sustained reentry result in dominant waves rotating around the tricuspid annulus and the inferior vena cava ( Table 1 ) ., In most cases ( 5 of 7 ) , the rotation is counter-clockwise around the tricuspid annulus ., An example is shown in Fig . 5B ( and Video S4 ) , with the wave anchored to the tricuspid annulus on the left , and the wave circulating the inferior vena cava on the right ., With electrical plus structural remodeling , the outcome is more varied ., In some cases waves are anchored ( to the pulmonary veins , the superior vena cava , or the tricuspid annulus ) ., However , in some simulations , the dominant waves are un-anchored but spiral around the left or the right atrial free walls ( Table 1 ) ., An example of such a scroll wave is shown in Fig . 5C , left ( and Video S5 ) ., In this example , there is also a wave anchored around the superior vena cava in the right atrium ( Fig . 5C , right , and Video S5 ) , while in other simulations the excitation of the right atrium by the scroll wave in the left atrium is more irregular ., Note that the dominant periods vary with the different interventions ., In the baseline ( no remodeling ) model , the mean period of ( non-sustained ) reentry is 176 ms . With full electrical remodeling , the shortening of the APD allows a faster mean rhythm of 137 ms . For full structural remodeling , the conduction slowing increases the main period to 205 ms . Interestingly , for full electrical plus structural remodeling , the period is the same as for only electrical remodeling ( 138 ms ) , suggesting that the preferred pathways are shorter on average for electrical plus structural remodeling ., These values fall within the range seen in patients with paroxysmal and persistent AF 21 , 22 ., The different rhythms and their occurrence patterns ( Table 1 ) correspond to clinical and experimental observations ., The pulmonary veins frequently act as triggers of AF 23 , and reentrant waves have been mapped in the pulmonary vein region 24 , 25 ., In canine models of structural remodeling , and in typical atrial flutter , excitation often occurs around the tricuspid annulus 26 , sometimes in concert with reentry around the inferior vena cava 27 ., Recent years have seen a large increase in modeling atrial-specific aspects of arrhythmogenesis ., In particular , there has been increased development of anatomical models , powered by increasing computational speed and data handling ( see , e . g . , 28 for a review on modeling and 29 for technical details on computational and visualization aspects ) ., At the cellular level , multiple mathematical models describe the same ionic currents in different representations of human atrial myocytes ., The Courtemanche et al . model 12 and the Nygren et al . model 30 are both well-established and have been compared in great detail 31 , 32 ., Although their simulated behavior can be quite different , we believe that neither is empirically better ., Rather they may represent intrinsic variability ., Given that neither is obviously better , we have opted to use the Courtemanche et al . model , largely because it is more widely used ., The Nygren et al . model was recently updated in terms of some of its potassium currents 33 and its intracellular calcium handling system 34 ., The effects of electrical remodeling on action potential morphology , in particular the role of the individual currents involved in remodeling 7 , 35 , have been studied at the cellular level ., In two-dimensional tissue , electrical remodeling accelerates spiral waves generated with both the Courtemanche et al . model 32 , 36 and the Nygren et al . model 32 ., With electrical remodeling there is also a decrease in spiral wave meandering in the Courtemanche et al . model 32 , 36 , but not with the Nygren et al . model 32 ., Such a decrease in spiral meandering with the Courtemanche et al . model is enhanced with increased IK1 37 and can indeed occur in simulated atrial tissue with increased IK1 in the absence of electrical remodeling 38 ., Increased IK1 also accelerates spiral waves 37 , as does another inwardly rectifying current IK , ACh , which is triggered in cholinergic AF 39 , 40 ., In simulated tissue ( Ref ., 36 , as well as in our simulations ( Fig . 4 ) ) , electrical remodeling decreases the wavelength in tissue strands and causes arrhythmias in anatomical models to be of longer duration ( Fig . 2; 32 , 36 ) ., Decreasing the calcium current conductance , which is a main component of electrical remodeling , has similar effects in a human anatomical structure of virtual guinea pig ventricular cells 17 , 19 ., Combining electrical remodeling and left atrial dilation leads to increased vulnerability to reentry in an anatomical model 41 , while combining electrical remodeling and decreased intercellular coupling causes shortened wavelength and sustained spiral wave activity in two-dimensional tissue simulations 42 , consistent with our results ., Other lines of study , pursued with complex atrial models , include the effects of myocardial stretch on conduction 43 , 44 , incorporation of intrinsic APD heterogeneity 45 , 46 , 47 , and simulated ablation 48 , 49 ., The mechanism underlying AF maintenance is not entirely clear ., There are two predominant theories:, ( i ) the multiple wavelet hypothesis and, ( ii ) the “mother rotor” hypothesis ., The multiple wavelet theory 50 hypothesizes that AF is composed of multiple interacting electrical wavelets and is maintained by the processes of wavebreak and reentry ., The mother rotor theory 51 hypothesizes that , rather than the multiple , equally important wandering wavelets of the multiple wavelet hypothesis , there is one dominant “mother” reentrant wave that sheds and initiates daughter waves as conduction block occurs at multiple sites away from its core ., The dynamics in our simulations are characterized by one or two reentrant waves , rotating fairly periodically around an anatomical obstacle or un-anchored in the ( left or right ) atrial free wall ., Hence , our simulations are more consistent with the mother rotor theory ., Further , in the presence of simulated structural remodeling , waves emanating from these dominant waves often exhibit local conduction block and transient fragmentation ., Recent high-density mapping studies of patients with persistent AF and structural heart disease also found increased occurrence of conduction block in the right atrium compared to patients without persistent AF and structural heart disease 52 ., However , the AF dynamics in the former patient group is characterized by more epicardial breakthroughs and more waves than our simulations , possibly due to increased decoupling of neighboring myocyte bundles 52 , 53 ., Using non-invasive imaging techniques , Cuculich et al . determined the number of wavelets in AF patients as 1–5 , with more wavelets in patients with long-term persistent AF ( average 2 . 6 ) than in patients with paroxysmal AF ( average 1 . 1 ) 25 , consistent with our findings ., However , these patients also had a considerable number of focal sites ( possibly due to spontaneous triggering , microreentry , or epicardial breakthroughs ) 25 , not seen in our simulations ., Atrial flutter ( AFl ) is usually associated with a single macroreentrant circuit and may be very regular ., However , AF can also exhibit a large degree of both temporal periodicity and spatial organization 54 , 55 ., Hence , our simulated arrhythmias demonstrate characteristics of both AF and AFl ., Clinically , AF and AFl may occur in the same patient , and AFl often converts to AF ., By initiating reentry at different locations and at different times , we were able to explore a large range of spatiotemporal dynamics in our simulations ., A large variety of ensuing dynamics is also observed clinically and experimentally ., While much of this variability may stem from differences among experimental AF models and patient-to-patient variation in underlying heart disease , age , and stage of remodeling , there is considerable variability in activation patterns even among patients with the same underlying condition 22 ., Importantly , many features observed in our simulations correlate directly with experimental and clinical characteristics ., One example is reentry around the pulmonary veins , seen in our model with electrical remodeling ., Another example is reentry around the tricuspid annulus , with 5 of 7 simulations of sustained activity with structural remodeling resulting in the direction of rotation being counter-clockwise ., Such right-atrial reentry , involving slow conduction through the isthmus , is typical in patients with AFl 26 , 56 often with the majority ( 19/26 ) of cases having counter-clockwise rotation around the tricuspid annulus 57 ., Further , with structural remodeling in our simulations , waves were also anchored around the inferior vena cava ., Such dual-loop reentry involving the tricuspid annulus and the inferior vena cava is often observed in typical atrial flutter 27 ., In addition to anatomical reentry , we also observed several examples of meandering scroll waves , which have been mapped experimentally and simulated computationally 58 ., Finally we observed incidences of double-wave reentry , also observed clinically 56 and experimentally 59 ., Rate gradients , with higher activation frequencies in the left atrium , exist in some animal models of AF 11 , 54 , 55 and have been observed in patients with both paroxysmal and persistent AF 22 , 60 , 61 , 62; however , other chronic AF patients do not exhibit such inter-atrial variability 21 ., In our simulations of remodeled tissue , there are no significant left-to-right gradients in activation frequency ., However , our simulated AF with electrical remodeling alone was primarily a left atrial phenomenon in the sense that most dominant waves were found in the left atrium ., This suggests that under some circumstances , the left atrium may be the driver of AF even in the absence of electrophysiological left-to-right heterogeneity ( see below ) ., The extent to which perpetuation of AF depends on the wavelength varies considerably among different studies using different AF models and different methods for wavelength estimation ., While several studies , both computational and clinical , have demonstrated a facilitation of AF maintenance with a decrease in wavelength 17 , 18 , others have not found such a dependence 63 ., We found a clear functional dependence of reentry duration on wavelength in our simulations , and found that the wavelength must be below a value of around 7 cm for reentry perpetuation ., However , the wavelength has to be considerably shorter for the majority of simulations to be sustained; this critical value for sustenance is about 5 cm ., This value is consistent with a previous report of 5 cm using a direct measurement of the wavelength during simulated AF 17 ., In contrast , clinical estimates for this threshold tend to be larger , with reported values of 12–13 cm 20 , 64 ., Importantly , as detailed in Ref ., 17 , these values are obtained through indirect methods , due to the inherent problem of insufficient spatial mapping , and tend to overestimate the wavelength ., AF and AFl become increasingly common with age , and are also particularly frequent in patients with existing structural heart disease , valvular heart disease , coronary artery disease , ischemic heart disease , hypertension , or a history of heart attacks ., Indeed , the majority of AF patients have one or more cardiovascular diseases in addition to AF ., Electrical and structural remodeling due to chronic AF occur in concert with substrate changes due to any existing condition ( s ) ., Changes in ionic currents due to chronic AF have been observed consistently in experimental models and in patients 9 ., Recent studies have shown differences in several of the outward potassium currents between the left and the right atrial appendages , with some current differences being present in sinus rhythm or paroxysmal AF and others in chronic AF 65 , 66 ., As it is presently unclear what type of inter-atrial gradients these appendage differences represent , we did not incorporate any left vs . right electrical heterogeneity in our model at this point ., Structural remodeling in chronic AF shows more variability among patients and animal models than does electrical remodeling ., Structural remodeling processes also occur on a much slower time scale than electrical remodeling , which may contribute to the substrate variability ., The processes include myocyte hypertrophy , fibrosis , and changes in the expression levels of connexin , the protein comprising the gap junctions that couple cells ., Fibrosis can manifest in both patchy patterns and diffuse morphologies ., Fibrosis and decreases in connexin levels both impede propagation , while atrial dilation and cell hypertrophy increase atrial activation times ., Our simulations using decreased coupling between virtual cells , which

Introduction, Methods, Results, Discussion

Atrial fibrillation , a common cardiac arrhythmia , often progresses unfavourably: in patients with long-term atrial fibrillation , fibrillatory episodes are typically of increased duration and frequency of occurrence relative to healthy controls ., This is due to electrical , structural , and contractile remodeling processes ., We investigated mechanisms of how electrical and structural remodeling contribute to perpetuation of simulated atrial fibrillation , using a mathematical model of the human atrial action potential incorporated into an anatomically realistic three-dimensional structural model of the human atria ., Electrical and structural remodeling both shortened the atrial wavelength - electrical remodeling primarily through a decrease in action potential duration , while structural remodeling primarily slowed conduction ., The decrease in wavelength correlates with an increase in the average duration of atrial fibrillation/flutter episodes ., The dependence of reentry duration on wavelength was the same for electrical vs . structural remodeling ., However , the dynamics during atrial reentry varied between electrical , structural , and combined electrical and structural remodeling in several ways , including:, ( i ) with structural remodeling there were more occurrences of fragmented wavefronts and hence more filaments than during electrical remodeling;, ( ii ) dominant waves anchored around different anatomical obstacles in electrical vs . structural remodeling;, ( iii ) dominant waves were often not anchored in combined electrical and structural remodeling ., We conclude that , in simulated atrial fibrillation , the wavelength dependence of reentry duration is similar for electrical and structural remodeling , despite major differences in overall dynamics , including maximal number of filaments , wave fragmentation , restitution properties , and whether dominant waves are anchored to anatomical obstacles or spiralling freely .

Atrial fibrillation is an abnormal heart rhythm characterized by rapid and irregular activation of the upper chambers of the heart ., Atrial fibrillation often shows a natural progression towards longer and more frequently occurring episodes and often occurs in patients with existing heart disease ( s ) ., Because atrial fibrillation has several variants , is complex in nature , and evolves over time , it is very difficult and expensive to study comprehensively in large-animal models , in part due to the inherent technical difficulties of imaging whole-atria electrophysiology in vivo ., Predictive multiscale computational modeling has the potential to fill this research void ., We have incorporated aspects of chronic atrial fibrillation to model some of its various disease states ., As such , this study represents the first comprehensive computational study of chronic atrial fibrillation maintenance in a biophysically detailed cell model in a realistic three-dimensional anatomy ., Our simulations show that disease-like modifications to cellular processes , as well as to the coupling between cells , perpetuate simulated atrial fibrillation by accelerating the rhythm and/or increasing the number of circulating activation waves ., Given the models ability to reproduce a number of clinically and experimentally important features , we believe that it presents a useful framework for future studies of atrial electrodynamics in response to , e . g . , ion channel mutations and various drugs .

medicine, physiology, biology, anatomy and physiology, biophysics, computational biology, cardiovascular

journal.ppat.1000741

Nucleoporin 153 Arrests the Nuclear Import of Hepatitis B Virus Capsids in the Nuclear Basket

Most DNA viruses depend on nuclear host factors for their replication ., Viruses infecting non-dividing cells have to pass the nuclear envelope through the nuclear pore complexes ( NPCs ) ., The NPC is large proteinaceous structure of ∼30 different proteins called nucleoporins ( Nups ) ., Due to the eight fold rotational symmetry of the NPC each Nup is present in 8–48 copies , forming a complex of ∼125 MDa ., On the cytoplasmic face of the NPC eight fibers extrude from a central ring-like framework , which is embedded in the nuclear envelope ., This ring forms openings in the nuclear envelope allowing translocation of cargos with a diameter up to 39 nm 1 ., On the karyoplasmic face of the NPC 8 fibers form the cage-like structure of the nuclear basket ( reviewed by 2 ) ., NPCs regulate the traffic of proteins and nucleic acids into and out of the nucleus ( reviewed by 3 ) ., While small molecules may diffuse between cytoplasm and nucleus karyophilic macromolecules are transported in a complex with soluble nuclear import receptors ., It is estimated that ∼1000 transport complexes pass each NPC per second 4 ., The best characterized transport receptors belong to the importin ( karyopherin ) β superfamily , comprising importin β , transportin 1 , 2 , transportin SR and exportins ., Nuclear import is initiated by binding of the receptors to a signal on the surface of the karyophilic cargo ., There is a variety of signals as e . g . M9 domains , interacting with transportins , importin β binding domains and “classical” nuclear localization signals ( NLS ) , which bind to importin β via the adapter molecule importin α ., The driving force of nuclear import and export is a gradient of the small GTPase Ran in its GTP-bound form across the nuclear envelope ., RanGTP is enriched in the karyoplasm , where it interacts with the transport receptors of the import complex , leading to its dissociation ., While the cargo diffuses deeper into the karyoplasm , the RanGTP-receptor complex is exported into the cytoplasm where it dissociates ., A key component of nuclear import is Nup153 to which the import complex of cargo and importin ( s ) binds and subsequently dissociates ., Nup153 is a 1445 amino acid ( aa ) protein of the nuclear basket 5 , which comprises three domains ( reviewed by 6 ) ., The N terminus ( aa 1–670 ) at the nuclear ring of the NPC contains an NPC targeting domain and an RNA binding domain ., The N terminus interacts with other proteins of the NPC as Nup107 and Tpr and is required for proper NPC architecture 5 , 7 , 8 ., The zinc finger region ( aa 650–880 ) at the distal ring of the NPC facilitates interactions with Ran ., The ∼30 FXFG repeat-containing C terminus is part of the hydrophobic meshwork that forms the “sieve” through which karyophilic cargos have to pass 9 ., The participation of Nup153 in vital cellular processes makes it difficult to analyze its functions without interfering with cell viability 10 ., Viruses with a nuclear phase in their life cycle make use of this import machinery ., Best characterized are adeno- , herpes , influenza and the human immune deficiency virus ., The latter two viruses disassemble in the cytoplasm and release the genome in complex with karyophilic viral proteins ., These complexes fall below the limit of the NPC diameter and pass the pore like cellular macromolecules ( reviewed by 11 ) or by transiently interacting with Nups 12 In contrast the genomes of adenoviruses and herpes viruses remain encapsidated within their capsids ., Their diameters of 90 and 120 nm exceed the diameter of the nuclear pore ., Adenoviruses bind to the cytoplasmic face of the NPC where they disassemble ., Genome translocation through the pore involving viral and cellular karyophilic proteins is not well understood 13 , 14 , 15 , 16 ., Herpes virus capsids also bind to the exterior of the NPC using importin β 17 but they open at the penton facing the NPC and inject the DNA through the pore ., For both viruses the trigger of genome release is unknown ., Similar investigations have been performed in digitonin-permeabilized cells on capsids of the medically important hepatitis B virus ( HBV ) 18 , 19 ., Hepatitis B is endemic in large parts of the world ., Approximately 350 million people are chronically infected , accounting for 1 million deaths per year ., HBV is an enveloped virus comprising an icosaedral capsid , which contains the partially double stranded DNA genome ( relaxed circular , rcDNA ) ., The capsid exists in T\u200a=\u200a4 ( major form ) and T\u200a=\u200a3 symmetry 20 , composed of 240 or 180 copies of the core protein , respectively ., The two forms differ in diameter ( 36 and 32 nm ) but functional differences are not known ., Entry of the virus into the hepatocyte is not well understood due to low efficiency of the available cell culture systems for HBV infection 21 ., Circumstantial evidence however suggests that the capsid enters the cytosol after fusion of the viral envelope with a cellular membrane 22 ., In fact lipofection of capsids , which by-passes the rate-limiting natural entry causes productive HBV “infection” of hepatoma cells with in vivo-like efficiency 23 ., Like other DNA viruses ( with the exception of baculoviruses 24 ) HBV capsids are transported towards the NPCs using the microtubule transport system 23 ., HBV is a pararetrovirus replicating via an RNA pregenome that is transcribed from the nuclear covalently-closed circular form of the viral genome ., Consequently the viral rcDNA has to enter the nucleoplasm upon infection where the rcDNA is converted to the covalently-closed circular form ., To allow access of the unknown cellular DNA repair factors the viral genome has to be liberated from the capsid either prior to , during or after transport through the NPC ., Transport and genome release are obviously highly efficient and well-coordinated , since ∼80% of virions are infection-competent in vivo 25 ., After export to the cytoplasm , the RNA pregenome is translated to core protein and the viral polymerase which binds in situ to an encapsidation signal on the pregenome ., This complex is specifically encapsidated within the assembling core protein ( reviewed by 26 ) , which forms an immature capsid ( Immat-C ) ., The polymerase converts the RNA into rcDNA , which is found in mature capsids ( Mat-C ) ., It is worth noting that genome maturation occurs exclusively inside the capsid and is a prerequisite for envelopment of the capsids by the surface proteins for virion formation 27 ., Core proteins assemble spontaneously to HBV capsids , e ., g ., upon expression in E . coli ., The first 140 aa of the core protein are essential for assembly and exhibit an ordered structure ., Within a resolution limit of 16 Å E . coli-expressed capsids show the same morphology as virion-derived capsids or capsids from infected cells 28 ., The arginine-rich C-terminus ( aa 141–185 ) is flexible 29 and exhibits phosphorylation sites for a cellular protein kinase , which has not yet been unequivocally identified ., In E . coli-expressed capsids , which contain unspecific E . coli RNA and which are not phosphorylated , the C terminus is localized within the lumen of the capsid 30 ., Some steps of genome maturation depend upon core protein phosphorylation , notably pregenome encapsidation and plus strand DNA synthesis 31 , 32 ., Both phosphorylation and genome maturation lead to translocation of the core protein C termini , harboring a NLS , to the exterior of the capsids 19 ., Consequently , all HBV capsids that have undergone some degree of genome maturation or phosphorylation bind to importin α/β 18 and can be found in the nuclear basket 1 , 19 ., Early in infection when sufficient amounts of surface proteins have not yet been synthesized , nuclear entry of progeny genomes increases the number of nuclear HBV DNA copies 33 , which is generally low but can reach numbers of up to 491 per cell in HBV infected patient 34 ., Using Digitonin-permeabilized cells – a frequently used system for analysis of nuclear import 35 – it was however shown that Immat-C and in vitro phosphorylated capsids synthesized in E . coli ( P-rC ) failed to exit the nuclear basket and thus do not diffuse deeper into the karyoplasm ., In contrast , adding Mat-C to the cells led to the presence of intranuclear capsids and genome release 18 , 19 ., The import strategy of HBV capsids seems thus to follow an entirely different strategy than has been shown for other viruses ., In particular the arrest of Immat-C in the basket is unique ., To date the only examples for cargos with an aborted nuclear import reaction are some Nups that become incorporated into the NPC after cell division and the protein Ubc9 , which partly associates with Nup358 on the cytosolic fibers of the NPC ., Evidently , the viral capsids have to interact with the NPCs upon initial infection but during establishment of the infection as well , at which time the nuclear viral DNA becomes amplified ., Due to the incomplete understanding of HBV transport and disassembly , the unique strategy and the medical importance of HBV , we evaluated the molecular background of the capsid arrest in the NPC basket ., Using in vitro and in cellulo approaches we identified the cellular interaction partner and identified the underlying mechanism responsible for the selective release of mature viral genomes ., Collectively , these findings lead to a model of a multi-step , maturation-regulated nuclear entry of the HBV genome ., Abortion of a nuclear import reaction within the nuclear pore must be based on a direct or indirect interaction with proteins of the NPC ., The recently observed arrest of Immat-C and P-rC in the nuclear basket thus most likely involves an association with a Nup on the karyoplasmic face of the NPC ., According to the current view on the architecture of the NPC candidates would include Nup50 , 54 , 58 , 62 , 93 , 96 , 98 , 107 , 133 , 153 , 160 , Rae1 , Seh1 , Sec13 PBC68 , and Tpr ( summarized by 36 ) ., First , we wanted to identify the NPC proteins that are co-precipitated by Immat-C using a nuclear extract of rat liver nuclei ., The NPC is composed of tightly interacting mostly hydrophobic proteins requiring denaturating conditions or harsh detergents for separation ., As this treatment unfolds proteins we first tested whether refolding restored biological functions ., For this purpose , we investigated the importin β binding ability since it is one of the essential functions of Nup153 ., Affinity of Nup153 to importin β was shown to be stronger than to other nucleoporins ( Nup62 , Nup214 , Nup358 37 ) ., To test our experimental system , pull-down was performed by incubating recombinant functional importin β with the extract followed by binding of the nucleoporins to solid-phase bound antibody mAb414 ., This antibody binds preferentially to the FXFG-repeat containing nucleoporins of vertebrates as e . g . Nup62 , Nup153 , Nup214 and Nup358 38 , 39 with different efficiency ., Figure 1A depicts that importin β became co-precipitated ( lane 1 ) ., As the nuclear extract did not contain detectable amounts of importin β ( not shown ) this result implies that the importin β binding activity of the nucleoporin ( s ) was maintained or restored during extraction and renaturation ., Binding was specific as antibody-coated beads without the extracted proteins and uncoated beads failed to interact with importin β ., Binding occurred in the absence of an importin β-bound cargo ., This finding is in accordance with the observation that importin β after its dissociation from the cargo does not diffuse deeper into the karyoplasm but remains bound to the NPC before being exported into the cytoplasm 40 ., The nuclear extract was then subjected to co-immune precipitations using Immat-C which are arrested within the nuclear basket ., These capsids contain DNA replication intermediates of the viral genome; they interact with the NPCs 19 and can thus be responsible for nuclear genome amplification ., For visualization of co-precipitated proteins we used Sypro Red , staining all proteins after SDS PAGE ., As depicted in Figure 1B no co-precipitation could be observed in the absence of nuclear extract ., Faint bands were observed in the absence of capsids ., This indicates some unspecific binding of nuclear proteins to the carrier beads , probably due to nonspecific hydrophobic interactions ., Immat-C-driven pull-down however showed the precipitation of one dominant protein , strongly enriched in comparison to the control extract and the precipitation controls ., This protein showed an apparent molecular weight of a ∼180 kDa typical for Nup153 41 , 42 ., To confirm the identity of co-precipitated Nup153 we used the antibody mAb414 in Western blot ., Figure 1C confirms the Nup153 co-precipitation by Immat-C and provides evidence that neither Nup214 nor Nup358 were co-precipitated ., To further confirm specific Nup153 capsid-interaction we preincubated the nuclear extract with Immat-C prior to precipitation via mAb414 and detection of co-precipitated Immat-C ., Since antibodies against denaturated core proteins require very large amounts of protein to provide adequate signals we labeled Immat-C by 32P using the protein kinase activity inside the capsids ., Phosphoimaging showed that a single protein was labeled having the molecular weight of the core protein of 21 . 5 kDa ( Fig . 1D ) ., As depicted in Figure 1E Immat-C could be precipitated by mAb414-bound Nup ., No signal was obtained in the absence of nuclear extract or in the absence of mAb414 ., To determine whether the Nup153 binding was selective for Immat-C we included different capsid species in co-immune precipitations ., We used Mat-C , which enter the nucleus and comprise an rcDNA genome; P-rC , which contain E . coli-RNA and which is arrested in the basket like Immat-C , and capsids that were formed by C-terminally truncated core proteins ( ΔC-rC ) ., ΔC-rC do not contain RNA , cannot be phosphorylated , and cannot enter the basket as these capsids do not contain the NLS ., All capsids reacted equally well with the anti-capsid antibody used to precipitate them after preincubation with nuclear extract ( not shown ) ., Unexpectedly , all capsids were able to precipitate Nup153 as depicted by immune blot using mAb414 ( Fig . 2A ) ., The Sypro Red stain of co-precipitated proteins was restricted for the molecular weight of the interaction partners ., Proteins smaller ∼75 kDa could not be detected as this part of the SDS PAGE was heavily overloaded with high amounts of bead-bound antibodies and BSA used for saturation of unspecific binding sites ., We could have thus missed smaller adapter proteins connecting the capsids with Nup153 ., To obtain evidence of direct interaction we replaced the nuclear extract by E . coli-expressed Nup153 that was purified under native conditions ., The fusion protein was previously shown to be functional on nuclear import and export after integration into the NPCs of reconstituted Xenopus laevis oocyte nuclei 43 ., Figure 2B showed that all capsids precipitated GST-Nup153 ., As GST does not bind to the capsids ( not shown ) we conclude that Nup153 interacts directly with the surface of the capsid ., The signal was not derived from the capsid preparation since no signal was observed in the absence of GST-Nup153 ., Furthermore , GST-Nup153 precipitation was not observed in the absence of the capsids implying that GST-Nup153 did not bind directly to the antibody coated-bead ( Fig . 2C ) ., We next set out to determine to which Nup153 domain the capsids bind ., We expressed large parts of Nup153 in three fragments ( N: aa 53–272; Z: aa 272–543; C1: aa 618–999 ) as GST fusion proteins ., The Nup153 fragments were incubated with the different capsids and co-precipitated by biomagnetic bead-bound anti-capsid antibodies ., Co-precipitation of the Nup153 fragments was demonstrated by Western blot using anti-GST antibodies ., Figure 2D shows that Mat-C , Immat-C and P-rC failed to precipitate the 50 kDa N terminal part of Nup153 ., A strong band at 54 kDa was visible in all samples to which the biomagnetic beads were added ., This band was most likely the product of an unspecific binding of the secondary blot antibody to the heavy chain of the antibodies used in precipitation ., Using the zinc finger domain Z ( 56 kDa ) no precipitation could be observed ( Fig . 2D ) ., However all three capsid species co-precipitated the 68 kDa C1 fragment ., Binding was specific as no precipitates could be observed in control reactions without capsids or with capsids but without anti-capsid antibodies on the beads ., This Nup153 fragment comprises an importin β-binding domain 44 , 45 thus implying that importin β and capsids compete for Nup153 binding ., A fourth His-tagged fragment ( C2: aa 992–1219 ) containing most of the ∼30 FXFG repeats 46 was analyzed by co-precipitation but we observed an unspecific binding to the beads ., To circumvent this technical problem we performed retardation gels ., We incubated P-rC with His-Nup153 C2 and separated the complex by native agarose gel electrophoresis ., The Nup fragment was visualized by anti-His antibodies , the capsid by anti-capsid antibodies ., Control experiments were performed using the N and the C1 fragment ., Migration of the fragments was determined by anti-GST antibodies ., Figure 2E confirmed that the N fragment did not interact with the P-rC as no shift of capsid migration could be observed upon the presence of this Nup153 domain ., Nup153 C1 caused a retarded migration of the majority of capsids ( Fig . 2F ) and fragment C2 retarded the migration of all P-rC ( Fig . 2F ) ., Although fragment C2 is largely hydrophobic we assume that the capsid binding is not unspecific as the capsids are negatively charged and migrate on native agarose gels as ∼3000 bp linear double stranded DNA fragments ., As both fragments C1 and C2 barely overlap we conclude that there is more than one interaction site on Nup153 ., In vivo , numerous proteins and nuclear factors pass the NPC simultaneously requiring interaction with Nup153 ., As Immat-C and P-rC are arrested in vivo , one should expect that the affinity of the physiological transport complexes to Nup153 is weaker than that of the capsids ., We incubated capsids ( P-rC ) immobilized on biomagnetic beads with Nup153 and importin β in different ratios ., Figure 3A , lane 1 , shows that no GST-Nup153 bound to the beads in the absence of capsids thus demonstrating the specificity of Nup153 binding ., When incubating Nup153 in parallel to different amounts of importin β with the capsids only molar importin β excesses of more than 150-fold with regard to Nup153 prevented Nup153 binding to the capsids ., This result thus confirm that capsid and importin β binding sites on Nup153 overlap and show further that the capsid binding was much stronger than the importin β interaction ., We next asked whether bound capsids can be displaced from Nup153 by importin β ., We preincubated biomagnetic bead-bound P-rC with Nup153 and added importin β after removal of unbound Nup153 ., Figure 3B shows that even 6700-fold molar excesses of importin β did not remove Nup153 from the capsid ., In order to determine whether the capsid arrest in the basket is mediated by Nup153 we suppressed Nup153 expression in HeLa cells by siRNA ., We controlled Nup153 expression in parallel to the expression of Fibrillarin , which is a component of a nucleolar small nuclear ribonucleoprotein ( SnRNP ) involved in ‘house keeping’ for nucleolar integrity 47 ., Figure 4A shows that Fibrillarin expression was in fact unaltered but Nup153 reduced by 80% ., Suppression was limited because Nup153 is essential for cell viability ., After permeabilization of these cells by digitonin a nuclear import assay using P-rC was performed ., We visualized Nup153 and capsids using indirect immune fluorescence ., Nup153 staining was weaker in 80–90% of the siRNA-treated cells than in non-silenced cells ( Fig . 4B , lower panel ) ., The nuclei of these cells were larger than in mock-transfected cells most probably due to the role of Nup153 in mitotic progression 48 ., The nucleoporin Nup153 plays separate roles in both early mitotic progression and the resolution of mitosis ., In the nuclei of the control cells , P-rC accumulated at the nuclear envelope and did not enter the karyoplasm as demonstrated previously 18 ., In contrast , a proportion of capsids were found in the karyoplasm of Nup153 siRNA-transfected cells , ( still image: Fig . 4B upper panel; for videos see Videos S1 , S2 , S3 and S4 ) ., Quantification on 20 cells revealed that 18–19% of capsids entered the nucleus while ca ., 80% were still arrested at the NE ., We next asked how Mat-C proteins could enter the nucleus despite the strong binding to Nup153 ., We followed the hypothesis that capsids should remain intact while genome maturation continues but should liberate the genome from the basket once rcDNA is formed ., To test this hypothesis we cross-linked the Mat-C subunits , which were 32P-labelled by UV irradiation ( Mat-C UV ) and analyzed their integrity ., Figure 5A shows that the cross-linking caused a strong reduced migration of the core protein subunits in SDS PAGE ., Most of the protein was retained at the entry site of the SDS PAGE; only traces resulted in a smear >45 kDa ., To test whether cross-linking occurred between capsids , which would have caused unsuitable particle aggregates we separated Mat-C UV on native agarose gel and detected them by anti-capsid antibody used before ., We observed no difference in migration compared to the P-rC standard ( Fig . 5B ) , demonstrating that UV irradiation only induced bonds between subunits of individual capsids ., The result further shows that the UV treatment had not changed the surface charge ., We next investigated the transport competence of Mat-C UV in comparison to untreated Mat-C ., We injected 1×107 capsids into the cytosolic periphery of 6 Xenopus laevis oocytes and followed the fate of the capsids by electron microscopy ., We used Xenopus laevis oocytes sine more NPCs can be analyzed by EM in this system than by using permeabilized cell lines thus giving more reliable results ., Combining these two systems and comparing the results is however justified since , as far as is known , nuclear import is identical in both systems 49 , 50 ., As control , six Xenopus laevis oocytes were injected with Mat-C ., We restricted the incubation time after injection to 1 h , the earliest time point at which significant numbers of capsids arrive at the nuclear pore ( not shown ) ., Such a short time was chosen in order to detect possible differences in nuclear entry of the capsids ., As shown in Figure 6A both types of capsids entered the nuclear basket as intact particles ., We determined the number of capsids that arrived at 68 NPCs ( Mat-C ) and 74 NPCs ( Mat-C UV ) , and determined their location at the NPCs ., Figure 6B showed that a similar number of capsids arrived at the nuclear pore indicating that the cross-link neither affected the intracytoplasmic transport capacity nor the interaction with the NPCs ., We next analyzed the distribution of the capsids , showing that both capsid species exhibited the same distribution at the NPCs with a majority on the cytoplasmic face ., To our experience this dominantly cytoplasmic localization is related to the short incubation time ., However , the same proportion of Mat-C and Mat-C UV entered the pore and were found in the nuclear basket , implying that their transport competence was the same ., We next analyzed whether Mat-C UV diffuses deeper into the nucleus like untreated Mat-C or remained arrested at the nuclear envelope like Immat-C and P-rC ., These assays were performed with digitonin-permeabilized cells since the relatively low amounts of capsids would have been undetectable in Xenopus laevis oocytes ., Capsids were detected by indirect immune fluorescence using the same anti-capsid antibody that showed an unchanged reactivity after cross link ., Figure 6C shows that Mat-C entered the nucleus as it was previously reported for different permeabilized cells and in living cells 19 , 23 ., In contrast Mat-C UV failed to enter the karyoplasm and remained associated at the NPCs and some cytosolic structure ., We assume that the cytosolic binding sites are collapsed microtubules as such binding was demonstrated before for digitonin-permeabilized cells 23 ., Figures 2 and 3 showed that the capsids bound to a Nup153 domain that participates in importin β binding and that the interaction competes with importin β in vitro ., To obtain in situ evidence on Nup153 that is integrated into the NPCs we analyzed whether binding of capsids to Nup153 interfere with the importin β-dependent nuclear import pathway ., These experiments were performed with P-rC because the large numbers of capsids required for Nup153 saturation on the permeabilized cells were only available for this type of capsid ., We used the human hepatoma cell line HuH-7 cells which is able to synthesize hepatitis B virions after transfection with HBV DNA ., We incubated digitonin-permeabilized cells with different amounts of P-rC ., Nuclear import into the basket was facilitated by means of the nuclear transport receptors in rabbit reticulocyte lysate ., After incubation cells were washed ., In the last washing buffer no P-rC was detectable ., First we quantified the amount of NPCs and the number of bound capsids ., NPC number was calculated from the Nup153 signal obtained by Western blot , which was compared to a dilution series of E . coli-expressed Nup153 ( not shown ) ., We determined ∼6000 NPCs per HuH-7 nucleus , which is higher than the ∼2770 NPCs found in HeLa cells 9 ., It is however known that the NPC number varies strongly between different cell types ( 18 , 451+/−2 , 336 ( Purkinje cells ) to 402+/−67 ( oligodendrocytes ) 51 ., We used P-rC preloaded digitonin-permeabilized cells after washing and added new cytosol together with two fluorescent cargos – Alexa594 NLS-BSA and Alexa647 M9-BSA ., NLS-BSA is imported by importin β while M9-BSA uses transportin for nuclear entry ., Both cargos comprise the same number of nuclear localization signals , as determined previously 18 ., After import reaction , capsids and NPCs were stained by indirect immune fluorescence , and nuclear cargo concentrations were determined semi-quantitatively using confocal laser scan microscopy ., Quantification was performed in the equatorial region of the nuclei ., A positive control was performed on cells that were preincubated in the absence of capsids but to which the cargos were added in a second step and incubated at 37°C ., In the negative control no P-rC was added during the first incubation but the following import reaction with the fluorescent cargos was performed on ice , thus inhibiting active nuclear transport 52 ., Figure 7 shows no intranuclear fluorescence in the negative control ( Fig . 7A ) but strong signals for both cargos in the positive control ( Fig . 7B ) ., With increasing preload of the nuclei by capsids the import of both cargos became reduced ( Fig . 7C–H ) ., At the highest capsid concentration no intranuclear fluorescence could be observed ( Fig . 7H ) ., In this sample the average nuclear diameter increased from 200 µm2 in the negative control to 240 µm2 ., This observation is in accordance with a “plugging” of the NPCs by the capsids , which does not allow exchange of smaller molecules needed for equilibration of the osmotic pressure between nucleus and reaction mixture ., Figures 7B to G show further that the reduction of nuclear import appeared for NLS-BSA at lower preloaded numbers of capsids than for M9-BSA ., As no unbound P-rC was present after washing , we can exclude competition between capsid NLS and NLS-BSA for the transport receptors importin α and β ., We next quantified the import in 570 nuclei ., The mean intranuclear fluorescence in the positive control was taken as 100% ., Figure 8 shows scatter plots obtained for each capsid preload ., We observed a significant variation between the nuclei in one sample , for instance between 80% and 120% in the positive control ., This can be explained by variable nuclear permeability throughout the cell cycle 53 ., Moreover , this observation confirms that the nuclei remained intact during the transport reaction as a disruption would have resulted in equal concentrations ., The import via transportin and via importin β was correlated in all cells and followed a Gaussian normal distribution ., As indicated by the slopes of the regression lines no significant changes were observed at preloads from 0 to 0 . 2 capsids per NPC ( Fig . 8B–D ) ., With larger numbers of capsids , the importin-mediated import became greatly reduced ( Fig . 8E; 0 . 7 P-rC/NPC ) or undetectable ( Fig . 8F; 3 . 3 P-rC/NPC ) , while the transportin-mediated pathway remained significantly more active ( p≤10−9 ) ., Concentrations ≥3 . 3 capsids/NPC blocked both the NLS-and M9-mediated import , most likely the result of steric hindrance by capsids that got stuck in the channels of the NPCs ., Collectively the data however indicate that blocking the capsid binding sites on Nup153 interferes with the importin β-mediated nuclear entry but not with the transportin pathway ., Karyophilic cargos interact transiently with components of the NPC via transport receptors before they are released to the karyoplasm ., HBV capsids are an exception as they remain arrested in the nuclear basket ., We showed that HBV capsids bound solubilized and renatured rat Nup153 from a nuclear extract with importin binding activity ., This is consistent with the high degree of conservation of Nups among different species and the conserved mechanisms of nuclear translocation 54 ., It confirms further the observation that Immat-C is arrested at the NPCs of different cell lines 19 ., Neither co-purification of Nup98 or Nup160 , which connect Nup153 with the central frame work of the NPC 55 nor of Tpr ( 268 kDa ) , which is attached to the NPC via Nup153 7 , was observed ., This finding indicates that the NPC components became disassembled during purification and did not reassemble upon renaturation ., This interpretation is in accordance with previous findings that NPC reconstitution depends on the presence of RanGTP 56 ., RanGTP has a low molecular weight of 25 kDa and is removed from the nuclei prior to disintegration of the NPCs ., The absence of a 97 kDa band in the co-precipitation further confirms that the binding was not mediated by importin β , which is in accordance with the direct binding observed in co-precipitation of natively purified GST-Nup153 ., The latter finding provides further evidence that the binding did not require a protein linker , which may have been below the 75 kDa limit of the Sypro Red stained gel ., HBV capsids undergo complex modifications upon genome maturation which are not well understood ., Their changes comprise not only the reverse transcription of the encapsidated RNA to rcDNA but also protein phosphorylation 31 , 32 and eventually dephosphorylation ., We found that Nup153 was precipitated by Immat-C , containing replication intermediates and possibly empty capsids , and by Mat-C containing rcDNA ., This observation suggests that the type of nucleic acid within the capsid has no impact on Nup153 interaction ., The observation that E . coli-expressed RNA containing capsids that were phosphorylated in vitro by protein kinase C and empty E . coli-expressed capsids that lacked the RNA-binding and phosphorylated C-terminal domain interacted equally with Nup153 shows that the binding is mediated by the N-terminal assembly domain ., It further confirms that importin β , which interacts via importin α with the NLS on the C-terminal domain 18 do not interfer with capsid binding to Nup153 ., Search

Introduction, Results, Discussion, Materials and Methods

Virtually all DNA viruses including hepatitis B viruses ( HBV ) replicate their genome inside the nucleus ., In non-dividing cells , the genome has to pass through the nuclear pore complexes ( NPCs ) by the aid of nuclear transport receptors as e . g . importin β ( karyopherin ) ., Most viruses release their genome in the cytoplasm or at the cytosolic face of the NPC , as the diameter of their capsids exceeds the size of the NPC ., The DNA genome of HBV is derived from reverse transcription of an RNA pregenome ., Genome maturation occurs in cytosolic capsids and progeny capsids can deliver the genome into the nucleus causing nuclear genome amplification ., The karyophilic capsids are small enough to pass the NPC , but nuclear entry of capsids with an immature genome is halted in the nuclear basket on the nuclear side of the NPC , and the genome remains encapsidated ., In contrast , capsids with a mature genome enter the basket and consequently liberate the genome ., Investigating the difference between immature and mature capsids , we found that mature capsids had to disintegrate in order to leave the nuclear basket ., The arrest of a karyophilic cargo at the nuclear pore is a rare phenomenon , which has been described for only very few cellular proteins participating in nuclear entry ., We analyzed the interactions causing HBV capsid retention ., By pull-down assays and partial siRNA depletion , we showed that HBV capsids directly interact with nucleoporin 153 ( Nup153 ) , an essential protein of the nuclear basket which participates in nuclear transport via importin β ., The binding sites of importin β and capsids were shown to overlap but capsid binding was 150-fold stronger ., In cellulo experiments using digitonin-permeabilized cells confirmed the interference between capsid binding and nuclear import by importin β ., Collectively , our findings describe a unique nuclear import strategy not only for viruses but for all karyophilic cargos .

Viral capsids facilitate protection of the enclosed viral genome and participate in the intracellular transport of the genome ., At the site of replication capsids have to release the genome ., The particular factors triggering genome liberation are not well understood ., Like other karyophilic cargos , hepatitis B virus ( HBV ) capsids are transported through the nuclear pore using nuclear transport receptors of the importin ß superfamily ., Unlike physiological cargos , HBV capsids become arrested within the nuclear basket , which is a filamentous structure on the nuclear side of the nuclear pore ., Asking which interaction causes this unique strategy , we found that the capsids bind to a protein of the basket periphery , nucleoporin 153 ( Nup153 ) ., The findings were confirmed in situ using digitonin-permeabilized cells that support physiological genome delivery into the nucleus ., We observed that HBV capsids bound to Nup153 irrespective of the maturation of the encapsidated genome ., But while capsids with an immature genome remained in arrested state , capsids with a mature genome disassembled and released their DNA .

virology/host invasion and cell entry

journal.pgen.1006046

Characterization of Expression Quantitative Trait Loci in Pedigrees from Colombia and Costa Rica Ascertained for Bipolar Disorder

Dozens of investigations have now shown that the identification of local eQTL may play a crucial role in delineating the causal variant ( s ) contributing to genetic associations observed for complex disorders or quantitative traits 1–3 ., While it may be particularly informative to evaluate , for a given trait , eQTL specific to tissues implicated in the manifestation of that trait , this strategy may be infeasible on a large scale for human brain related traits , such as psychiatric disorders and their endophenotypes ., In this study we report the results of gene expression in lymphoblastoid cell lines ( LCL ) for 786 genotyped members of Costa Rican and Colombian pedigrees 4–5 ., While the subjects in this study were originally recruited as part of an investigation for severe bipolar disorder ( BP1 ) , we found no relationship between the observed gene expression data and BP1 ( S1 Text ) ., We selected LCL for ease of study and on the basis of the increasing evidence that a substantial proportion of local genetic regulation is conserved across tissues 6–8 ., While distal regulation has been found to have a higher degree of tissue-specificity vs . local regulation 9 , it is unclear to what extent this finding reflects the very limited power to detect distal associations in a given tissue and resulting underestimates for the extent of overlap across tissues ., Studying LCLs has enabled at least a partial reconstruction of the specific regulatory network ( i . e . the bipartite graph relating genetic variants to gene expression traits , the strength associated to each of these edges , and the overall impact of genetic variants on the variability of expression ) for these families , allowing us to identify those components that might show differences from the general population ., We study the genetic regulation of expression in these pedigrees at a multiscale level: we estimate heritability , evaluate the relative importance of local vs . distal genomic variation , identify variants with regulatory effects , and analyze the role of multiple associated SNPs in the same region ., By capitalizing on known pedigree structure , as well as extensive genotyping , we can compare different methodologies for heritability estimation ., The most interesting element of regulatory networks for our purpose is the localization of SNPs with regulatory effects ( eSNPs ) : these variants are candidates for future studies investigating association to the BP1 endophenotypes measured in our sample , and also provide insight into functional genetic variation in this unique population ., To control the rate of false discoveries of eSNPs , we adopt a novel hierarchical testing procedure that leads to the analysis of expression quantitative trait loci ( eQTL ) data in a stage-wise manner with increasing levels of detail ., The study subjects are members of 26 Costa Rican and Colombian pedigrees ascertained from local hospitals and clinics based on multiple individuals affected with BP1 ., Descriptions of pedigrees and ascertainment procedures are provided in 4 ., Written informed consent was obtained from each participant , and institutional review boards at participating sites approved all study procedures ( UCLA Medical Institutional Review Board 3 IRB # 11–000407; Scientific Ethics Committee of the University of Costa Rica Project No . 801-91-552; and the Bioethics Committee of the Institute of Medical Research , University of Antioquia Project Name “Genética de la enfermedad Bipolar . Endofenotipos bipolares en una población aislada genéticamente” ) ., Lymphoblastoid cell lines ( LCLs ) were established at two sites ., RNA was extracted from these cell lines and its expression quantified using Illumina Human HT-12 v4 . 0 Expression BeadChips ., Expression values were background corrected , quantile normalized , log2 transformed , and corrected for major known batch effects ., The outcome of these procedures is what we refer to as ‘probe expression’ for all subsequent analyses ., After quality control filters , the 34 , 030 probes included in the final set were aligned to at most 2 locations in hg19 , contained no common SNPs ( as defined in dbSNP 137 or 138 ) , their expression was detected in at least one individual , and queried 24 , 385 unique genes ., As discussed in S1 Text , both the choice of normalization procedure ( across all subjects rather than within pedigree ) and of detection threshold ( which is fairly generous ) affect downstream estimates including expression heritability ., For a detailed description of the processing steps used at each site and the RNA quantification , normalization , and quality control procedures , see S1 Text ., DNA was extracted from blood or LCLs using standard protocols ., Illumina Omni 2 . 5 chips were used for genotyping , in three batches ., A subset of samples was repeated in each batch to enable concordance checks ., A total of 2 , 026 , 257 SNPs were polymorphic and passed all QC procedures , including the evaluation of call rate , testing for Hardy Weinberg equilibrium , and Mendelian error ., A total of 1 , 024 , 051 autosomal SNPs with minor allele frequency ( MAF ) of at least 10% were selected for use in the subsequent analysis ., A threshold of 10% was chosen since we have very limited power to detect SNP-gene associations for SNPs with MAF < 10% given the size of our study population ., We would like to note , however , that this threshold may result in lower estimates for the local portion of gene expression heritability and possibly different numbers of independent local eSNPs vs . studies using a less stringent threshold ., After excluding married-ins with no descendants in the study and cases of possible contamination , the analyzed sample contains 786 individuals with both genotype and gene expression data ., ( See S1 Text for details . ), In order to adjust for both known and unknown factors affecting global gene expression , all association and heritability analyses include age , sex and batch as covariates , in addition to a set of PEER factors to adjust for latent determinants of global gene expression 10 ., We chose to include 20 PEER factors on the basis of the proportion of global gene expression explained , and found that these PEER factors were strongly correlated with batch , but not with family groupings , suggesting that they are in fact correcting for technical artifacts ., The eQTL literature documents a distinction between cis vs . trans regulation , although the precise definition of these is sometimes elusive ., Following the suggestion of 3 , we adopt the terminology “local” and “distal” regulation to distinguish the situations where genetic variants and the genes whose expression they regulate are nearby or far away in the genome , without any assumption on the mechanisms of this regulation ., Operationally , we define “local” associations as those between SNPs and probes where the SNP is located within 1Mb of either end of the probe , and “distal” as all other probe-SNP associations , including those across different chromosomes ., For each probe , we estimated the heritability of gene expression using two approaches: a variance components model relying on known family relationships as implemented in Mendel 11 and a variance decomposition based on observed genotypic similarities among individuals as implemented in GCTA 12 ., Both analyses included age , sex , batch and PEER factors as covariates ., In our primary GCTA analysis , we utilized a genetic relatedness matrix ( GRM ) based on the full set of genome-wide SNPs ., This allowed us to calculate the ratio of genetic variability over total phenotypic variability for each probe ., We then compared the estimates obtained using Mendel and GCTA ., To determine which probes were significantly heritable , we relied on the likelihood ratio test implemented in GCTA to obtain p-values for the significance of the genetic variance component ., To consider the effect of shared environment on the heritability of gene expression , we computed a version of the variance components model in Mendel with a variance component included for effects corresponding to pedigree membership ., We also examined correlations between spouses , siblings , and parent-child pairs using the function FCOR from the S . A . G . E . software package 13 , which allows the estimation of familial correlations and their asymptotic standard errors 14–15 ., Because FCOR does not allow the inclusion of covariates , expression was first regressed on all covariates and the residuals were used for correlation analysis ., To obtain an unbiased estimate for the mean heritability of gene expression , Price et al . 16 allow negative values for heritability ., Although Mendel requires heritability estimates to be constrained to 1 interval , GCTA allows this assumption to be relaxed; to assess the impact of this constraint , we compute both constrained and unconstrained estimates in GCTA for the ratio of genetic variability to total phenotypic variability ., As a secondary analysis , we used GCTA to refine the variance decomposition of probe expression to obtain estimates of the proportion of probe heritability due to local regulation ., Specifically , we utilized the multiple GRM option in GCTA ( which allows partitioning of the phenotypic variance into components explained by different SNP subsets; see for example 17 ) with two GRMs specified: one based on the set of SNPs within 1Mb of the probe of interest ( whenever a sufficient number of SNPs was present ) , and one based on all SNPs genome-wide ( a reasonable stand-in for relatedness based on distal SNPs ) ., This strategy allowed us to partition the heritability into local vs . global components and calculate the ratio of local genetic variability to total variability ., With regards to interpretation of the resulting estimates , we note that the goal of GCTA is to estimate the additive effects of the genotyped SNPs , rather than a true estimate of heritability ., Yang et al . 12 therefore recommend excluding related subjects since including these will bias the estimate of the proportion of variance explained by common variants upward due to factors such as shared environment or rare variants passed down within a family ., Since we include related subjects , our GCTA results will be inflated relative to those for unrelated subjects , and therefore are more similar to the family-based heritability estimates ., The relatedness of our subjects ( and the fact families share a number of environmental factors ) is also likely to affect the partitioning of heritability into local vs . distal components in GCTA: since the GRM from local SNPs will align less closely to the correlations due to family structure and shared environment than that of the GRM from genome-wide SNPs , the proportion of genetic variance to due local SNPs may be underestimated ., We computed association p-values for each SNP-probe pair using the pedigree GWAS option in Mendel including additive genetic and environmental variance components 11 , 18 ., The Mendel implementation relies on a score test to greatly increase the speed of computation of association p-values in mixed models ., For the most promising SNP-probe pairs , a standard likelihood ratio test ( LRT ) is conducted , and effect sizes are derived ., In our analysis , we included age , sex , batch and PEER factors as covariates ., We performed the LRT for the 100 most significant local and 100 most significant distal SNPs for each probe , with the score test used for the remaining SNP-probe pairs ., Our hierarchical testing approach is based on the selective procedure by Benjamini and Bogomolov 19 whose effectiveness in genetic association studies for multiple phenotypes in demonstrated through the simulations provided in 20 ., A version of this approach tailored to the eQTL context is implemented in the TreeQTL R package 21; the current work is the first application of the proposed methods to a real-world eQTL study ., The testing procedure is designed to take into account that local regulation is more common than distal ( the hypotheses in these two classes are tested separately ) and that SNPs with distal effects are likely to affect the expression of more than one probe ., While the possibility of identifying variants involved in the local regulation of each probe depends on the sample size and the signal strength , it is quite reasonable to expect that the expression of every gene could be affected by appropriate sequence variation in the genomic region surrounding it ., In contrast , one expects that only a small portion of the genotyped variants have any regulatory role ., Both to capitalize on this heterogeneity and because our ultimate interest is to identify genetic variants that have phenotypic effects , we apply a multiscale testing strategy to first identify SNPs that have regulatory effects ( eSNPs ) ., We control the FDR in these discoveries at a target level of 0 . 05 with the Benjamini-Yekutieli 22 procedure , a conservative approach which is robust to dependence among the test statistics and therefore appropriate given linkage disequilibrium among the SNPs ., In a second stage we investigate which specific probes are influenced by these eSNPs ., We control the expected average proportion of false SNP-probe associations across the selected SNPs at a target 0 . 05 level with the Benjamini Bogomolov ( BB ) method 19 , which has been shown to control the relevant error rates under the typical dependency structure of multi-trait GWAS 20 ., We adopt this hierarchical multiple testing strategy to improve the interpretability and relevance of our findings , as it controls error rates regarding the discovery of functional SNPs and the association of these SNPs to traits which are not controlled using standard non-hierarchical multiple testing corrections ., While our primary goal in adopting the hierarchical testing procedure is to control these important error rates , we are able to take advantage of the heterogeneity across genetic variants ( mentioned above ) to preserve power to the extent possible ., We studied the position of local eSNPs relative to the transcription start site ( TSS ) of the gene queried by the probe to which they were associated ., TSS information was derived from the UCSC Genome Browser ( http://genome . ucsc . edu/ ) ., We investigated the distal eSNPs by assessing their overlap with local eSNPs and by comparing their locations with the annotations derived by the Roadmap Epigenomics Project ( http://egg2 . wustl . edu/roadmap/web_portal/ ) for LCLs using ChIP-Seq and DNAse-Seq 23 ., Cross-study comparisons are hampered by many factors including changes in annotation resulting in different gene symbols , changes in SNP names , and the use of different versions of the human physical map ., We downloaded results from eQTL analysis of blood or LCL from the seeQTL database ( http://www . bios . unc . edu/research/genomic_software/seeQTL/ ) 24 , including results from 25–26 and a meta-analysis of HapMap LCLs , and also obtained results of 27 for associations with FDR less than 50% ., We used official gene symbols to compare results across studies ., For each probe associated to some of the discovered eSNPs , we constructed a multivariate linear mixed model relating expression to the genotypes at significant SNPs , local or distal ., Using the variance components model implemented in Mendel , a fixed effect was estimated for age , sex , batch , the PEER factors , and each of the genetic variants , while a random effect was used to capture family structure ., We then calculated the proportion of variance explained in this model by the collection of local eSNPs and distal eSNPs and compared it with the local and global heritability estimates obtained using the partitioning approach of GCTA ., To account for the fact that linkage disequilibrium may lead to the identification of a number of neighboring SNPs as associated to the same probe—even when the underlying association is effectively captured by one SNP alone—we performed model selection to determine the number of SNPs that might reasonably correspond to independent signals ., Specifically , after transforming the data to obtain independent observations ( using the appropriate variance covariance matrix determined from the mixed model analysis in Mendel ) , for each probe we carried out stepwise forward selection , relying on the BIC criteria , and using residual expression ( adjusted for covariates ) as the response and the eSNPs associated to the probe as the pool of predictors ., This procedure gave us an estimate of the number of independent eSNPs affecting each probe , as well as the value of the percentage of variance explained ( the adjusted r2 value ) for the resulting multivariate linear model ., For comparison , we also obtained the percentage of variance explained ( the r2 value ) for the univariate linear model using the most strongly associated eSNP ( local or distal ) as the only predictor ., We then computed the ratio of the r2 for each model to the heritability previously estimated using the variance components model in Mendel ., The distribution of heritability estimates across all 34 , 040 probes obtained using Mendel , shown at left in Fig 1 , had median 0 . 03 ( mean = 0 . 10 ) ., Estimates of the heritability of gene expression based on kinship obtained using Mendel correlated well with estimates of the proportion of phenotypic variation due to genome-wide SNPs obtained using GCTA ( r = 0 . 99 ) , suggesting agreement between the known pedigree structure and levels of genetic similarity in the subjects ( S1 Fig ) ; the estimates from Mendel tended to be slightly larger than those from GCTA , particularly for values closer to, 1 . The median proportion of variance explained by genome-wide SNPs as computed by GCTA was 0 . 04 ( mean = 0 . 10 ) when constrained to the 1 interval , while the unconstrained estimates had median 0 . 03 and mean 0 . 09 ( S1 Fig ) ., The likelihood ratio test for the significance of the genetic variance component in GCTA resulted in 12 , 631 rejections ( 37% ) at p<0 . 05; 10 , 630 rejections ( 31% ) at FDR threshold 0 . 05; and 4 , 496 rejections ( 13% ) applying the Bonferroni correction to target FWER 0 . 05 ., The median proportion of variance in gene expression explained by genome-wide SNPs among probes satisfying FDR<0 . 05 was 0 . 22 ( range 0 . 07–1 . 00 ) when constrained to the 1 interval , while the unconstrained estimates for these probes was 0 . 23 ( range 0 . 07–1 . 01 ) ., The inclusion of a family variance component in Mendel corresponding to pedigree membership reduced the heritability estimates from a median of 0 . 034 ( mean = 0 . 10 ) to a median of 0 . 026 ( mean = 0 . 09 ) ., The GCTA estimates of the proportion of phenotypic variability explained by genome-wide SNPs are most similar to the Mendel kinship-based estimates of heritability without the additional family variance component , suggesting that GCTA functions very similarly to kinship-based approaches in a family setting ( S1 Fig ) ., In our examination of familial correlations , we found very few spouse correlations ( which reflect shared environment but not kinship ) to be significant even at the 0 . 05 level ( 4% ) , whereas 23% and 27% of parent-offspring and sibling correlations were significant at this threshold ., The median ( mean ) correlations for the three relationship classes were 0 . 01 ( 0 . 009 ) for spouses , 0 . 04 ( 0 . 06 ) for sibling , and 0 . 03 ( 0 . 04 ) for parent-offspring pairs ., Sibling and parent-offspring correlations were well-correlated to the heritability estimates , whereas there was no correlation between heritability and spouse correlations ., Together , these two approaches suggest that shared environment accounts for a small proportion of our estimated heritabilities ., Among the 10 , 630 significantly heritable probes ( FDR<0 . 05 ) , 9 , 458 had a sufficient number of SNPs in the local region to obtain a GRM usable for partitioning; for these probes , a median of 30% of the total genetic variance was attributed to local genetic variation ( mean = 37% ) ., The distribution of the proportion of total genetic variance attributed to local genetic variation for these probes is shown at right in Fig, 1 . Probes with a low proportion of genetic variance attributed to local genetic variation ( <5% ) have a significantly smaller number of local SNPs than those with a larger proportion ( one-sided t-test p<2 . 2e-16 ) and are associated to a significantly higher number of distal eSNPs ( one-sided t-test p = 0 . 02 ) , suggesting that both a failure to measure relevant local SNPs and the effects of distal regulation may explain the fraction of heritable probes found to have a low local proportion of genetic variance ., QQ plots for the local and distal SNP-gene association analyses are shown in S2 Fig . Taken together , these plots demonstrate that the distribution of the test statistics under the null is as expected , and that there is strong evidence for a significant number of non-null hypotheses genomewide for both local and distal regulation ., Controlling the FDR of eSNP discoveries at a 5% level , we identify 139 , 668 local eSNPs and 11 , 016 distal eSNPs ., Controlling the expected value of the average proportion of false discoveries for probe-SNP association across the discovered eSNPs to 5% as well results in the identification of 305 , 635 local probe-SNP pair associations and 22 , 304 distal probe-SNP pair associations ., There are 10 , 065 distinct probes involved in these associations ( 9 , 645 in local regulation and 1 , 081 in distal , with an overlap of 661 ) ., We now consider some of the characteristics of the discovered eSNPs ., In keeping with current understanding of the mechanisms of local regulation , 72% of the local eSNPs are upstream from the gene they putatively regulate , and 15% of these are within 100kb upstream from the transcription start site ( TSS ) ., The distribution of local eSNPs by distance from the TSS , calculated as the TSS position of the queried gene minus the SNP position for each SNP-probe pair discovered , shows that the discoveries are most concentrated closest to the TSS ( at left in Fig 2 ) ., Among the discovered distal eSNPs , 50% also appear to act as local regulators , a phenomenon that has been noted before 27–28 ., On average , distal eSNPs affect 2 . 0 probes ( median = 1 . 0 ) , or 1 . 8 genes; the distribution of the number of genes controlled by distal eSNPs is shown at center in Fig, 2 . Utilizing the annotations from the Epigenomics Roadmap , we found that 27% of distal eSNPs fall within narrow peaks ( which reflect point sources such as transcription factors or chromatin marks associated with transcription start sites ) and 38% fall within broad domains ( which cover extended areas associated with many other types of histone modifications ) , indicating that a substantial portion of distal eSNPs are located within functional genomic regions ., The most strongly associated local eSNP to each probe with local associations had an average effect size ( absolute value of the estimated regression coefficient ) of magnitude 0 . 12; the comparable average for the distal setting was 0 . 21 ., The distributions of effect sizes for local and distal regulation are shown at right in Fig 2: the appreciable difference in effect sizes is likely due to the “winner’s curse” phenomenon given the large number of distal hypotheses ., Specifically , due to the more stringent selection criteria in the distal setting , there is effectively a higher threshold on the estimated effect sizes for distal associations , so larger eSNP effect sizes are to be expected and may not necessarily reflect a biological difference between distal and local regulation ., One possible way to explore the effect of this bias would be to compute false coverage rate confidence intervals for each estimated coefficients ( where wider intervals reflect stronger selection bias ) ; this is not completely straightforward given the hierarchical selection procedure , but is of interest in future work ., A more detailed investigation of the percentage of variance explained by local and distal eSNPs is given in a later section ., For a comparison of the number of discoveries under different error controlling strategies and their characteristics , see S1 Table and S3 Fig . To compare our discovered local eSNPs with those of other studies , we rely on the named genes they appear to regulate ., This allows us to implicitly account both for the effect of linkage disequilibrium and the different genotypes available ., Considering first local association and matching on gene name , our study and published studies had 14 , 174 gene names in common; 6 , 456 have significant local associations in our work , and 7 , 755 have local associations with p<0 . 0001 in the published studies ., Of the 6 , 456 genes we find significant and on which we have available data in other studies , 4 , 790 are significant in other studies ( 430 are significant in one other study , 1 , 354 are significant in two other studies , 1 , 182 are significant in 3 other studies and 1 , 194 are significant in all 4 studies examined ) ., Examination of distal associations in our work and published studies indicates 10 , 002 gene names in common; 409 have significant distal associations in our work , and 528 genes have distal associations with p<5e-08 in the published studies ., Of the 409 genes we find significant and on which we have data available in other studies , 63 are significant in other studies ( 17 in one other study , 24 in two studies , 16 in three studies , and 6 in all four studies examined ) ., Only 34 of these 63 genes identified as being significantly affected by distal variants in our study were also identified as having significant distal associations in published work to SNPs on the same chromosome as ours; and 23 of these 34 genes involved associations to SNPs <2Mb apart in our study compared to the published studies ( S2 Table ) ., We examined whether the same SNPs were involved in distal associations in multiple studies , without specifying that the associations were to the same genes ., We considered this question matching both on SNP name and on SNP position , requiring that the SNPs were selected as eSNPs in our work at FDR 5% and had associations in published studies at p<5e-08 ., We found 33 SNPs on six chromosomes to have distal associations to one or more genes in both our study and in published studies ( p<5e-08 ) ; however the distal associations were to different genes ( S3 Table ) ., There are only ten distal associations significant in our work ( controlling the expected average proportion of false associations involving the selected eSNPs to 5% ) and in published studies ( p<5e-08 ) that involve the same SNP and same gene: ( 1 ) LIMS1 on chromosome 2 at ~10 . 9Mb is associated to five SNPs on chromosome 6 , at 32 . 4–32 . 7Mb ( rs13192471 , rs3129934 , rs3763313 , rs9268877 , rs9272219 ) in our work and in 27; ( 2 ) three probes in DUSP22 on chromosome 6 at ~0 . 35Mb are associated to one SNP on chromosome 16 at ~35Mb ( rs12447240 ) and is also associated to this gene in 25; ( 3 ) OR2AG1 on chromosome 11 at ~6 . 8Mb is associated to one SNP on chromosome 21 at 34 . 6Mb ( rs1131964 ) in both our study and 25; ( 4 ) TSSC4 on chromosome 11 at ~2 . 4Mb is associated to one SNP on chromosome 6 at ~31 . 2Mb ( rs3131018 ) in both our study and 27; ( 5 ) NOMO1 on chromosome 16 at ~14 . 9Mb is associated to one SNP on chromosome 16 at ~16 . 3Mb ( rs4780600 ) in both our study and 25 ( 6 ) and lastly RTF1 on chromosome 15 at ~41 . 7Mb is associated to one SNP on chromosome 17 at 2 . 5Mb ( rs8081803 ) in both our work and 25 ., The sparser concordance of the inferred distal vs . local regulation in the cross-study comparison is not surprising: the power to detect distal effects is considerably smaller in all studies , while the impact of confounders stronger ., Two additional reasons might explain this difference ., On the one hand , our methodology to identify distal eSNPs has larger power to discover multiple genes regulated by the same variant ., On the other hand , some of our unique findings might be due to the ascertainment of the subjects , who are members of families carrying genes predisposing to BP1 and/or to extreme values of BP1-related quantitative traits ., To examine the explanatory power of the discovered eSNPs , we focus on the probes that they affect ., Of the 10 , 065 probes associated to any eSNPs , 7 , 280 were significantly heritable at an FDR of 5% ( 7 , 000 with local associations , 916 with distal associations , and 636 with both ) ., Among the non-heritable probes with eSNP associations , 94% had only local associations , suggesting that these discoveries reflect the less stringent multiplicity control for the discovery of local associations ., When the eSNPs for each of the 7 , 280 heritable probes were included as fixed effects in a variance components model of the probe expression , the genetic variance component was estimated to be 0 for 1 , 491 ( 20% ) of the probes , indicating that for these probes , the eSNPs capture essentially all of the genetic component of variation in probe expression ., The distribution of the proportion of genetic variance due to the selected eSNPs ( estimated as 1—the ratio of the genetic variance component when eSNPs are included as fixed effects to the genetic variance component when eSNPs are not included ) is shown in Fig 3 , assuming values less than 0 ( 12% ) are exactly 0 ., The median proportion of variance explained for the set of probes with only local , only distal , or both types of associations is 0 . 44 , 0 . 52 , and 0 . 97 , respectively , demonstrating that the eSNPs do explain a substantial proportion of the heritability of gene expression , particularly for probes with both significant local and significant distal associations ., The distributions of the local and total genetic proportions of variance under partitioning using GCTA for probes with only local , only distal , or both types of associations ( S4 Fig ) demonstrates that probes with local associations do in fact have larger proportions of variance due to local effects vs . probes not associated to any local SNPs ., To understand the number of independent signals represented by the eSNPs , we obtained the results from model selection using eSNPs as the pool of possible predictors , focusing again on the set of 7 , 280 significantly heritable probes with associations to any eSNPs ., For this set , the median number of eSNPs with significant marginal association was 23 ( mean 41 . 5 ) , with 521 probes ( 7 . 2% ) associated to only one eSNP ., The distribution of the number of eSNPs included in the best multivariate linear model had a median of 2 ( mean 3 . 1 ) , with 5 , 328 probes ( 73% ) associated to multiple eSNPs ., The large discrepancy in the number of associated SNPs underscores the fact that a substantial proportion of the pairwise SNP-probe associations is due to linkage disequilibrium among neighboring SNPs ., At the same time , it is interesting that the selected linear model includes multiple SNPs for 73% of the probes considered: this observation can be interpreted as the result of multiple variants with regulatory effects , but also as a sign that the causal variant is not typed and multiple typed SNPs allow a better reconstruction of the associated haplotype ., It is also possible , however , that probes with multiple regulatory SNPs are more likely to appear in the set of significantly heritable probes with any eSNP associations vs . those regulated by a single SNP ., To gain insight into the explanatory power of the univariate vs . multivariate models , we assessed the percentage of total phenotypic variance explained by the most significantly associated SNP and by the selected multivariate linear model ., The distribution of the percentage of variance explained for the most significantly associated SNP ( Fig, 4 ) has a median of 3 . 6% , around half th

Introduction, Methods, Results, Discussion

The observation that variants regulating gene expression ( expression quantitative trait loci , eQTL ) are at a high frequency among SNPs associated with complex traits has made the genome-wide characterization of gene expression an important tool in genetic mapping studies of such traits ., As part of a study to identify genetic loci contributing to bipolar disorder and other quantitative traits in members of 26 pedigrees from Costa Rica and Colombia , we measured gene expression in lymphoblastoid cell lines derived from 786 pedigree members ., The study design enabled us to comprehensively reconstruct the genetic regulatory network in these families , provide estimates of heritability , identify eQTL , evaluate missing heritability for the eQTL , and quantify the number of different alleles contributing to any given locus ., In the eQTL analysis , we utilize a recently proposed hierarchical multiple testing strategy which controls error rates regarding the discovery of functional variants ., Our results elucidate the heritability and regulation of gene expression in this unique Latin American study population and identify a set of regulatory SNPs which may be relevant in future investigations of complex disease in this population ., Since our subjects belong to extended families , we are able to compare traditional kinship-based estimates with those from more recent methods that depend only on genotype information .

We assess the heritability and genetic regulation of gene expression in a population of 786 individuals from Costa Rica and Colombia ., The subjects , originally recruited in a study of bipolar disorder , are related within 26 extended families ., This design allows us to compare estimates of the heritability of gene expression obtained using both traditional and genotype-based methods ., We address questions regarding the architecture of genetic regulation including the extent to which gene expression is influenced by variants located nearby vs . far away on the genome and how many variants affect the expression of a given gene ., In addition , we identify genetic variants which regulate gene expression; these serve as candidates for future studies to establish the genetic basis of complex traits , including those related to bipolar disorder , and also provide insight into the architecture of genetic regulation in this unique Latin American study population .

medicine and health sciences, evolutionary biology, gene regulation, population genetics, bipolar disorder, dna transcription, genome analysis, population biology, mood disorders, gene expression, mental health and psychiatry, phenotypes, gene regulatory networks, genetics, biology and life sciences, genomics, gene prediction, computational biology

journal.pgen.1004838

Genetic Control of Contagious Asexuality in the Pea Aphid

While sexuality is the dominant reproductive mode in metazoans , parthenogenesis - the development of an embryo from an unfertilized egg - occurs in most branches of the animal kingdom ( e . g . molluscs , insects , crustaceans , nematodes , fish , reptiles ) e . g . 1 , 2 , 3 ., Cyclical parthenogenesis ( CP ) represents a mixed reproductive mode with an alternation of sexual reproduction and parthenogenesis , and is reported in many animal species 4 ., The loss of the sexual phase in CP species - leading to permanently parthenogenetic taxa - have been shown to arise from diverse mechanisms , including microbial infection , hybridization , contagion via pre-existing parthenogenetic lineages or spontaneous mutations 5–9 ., Nevertheless , in case of contagious or mutational origin , the precise genomic regions responsible for the transitions to obligate parthenogenesis ( OP ) remain largely unknown , mostly because dissecting the genetic bases of that trait using recombination-based approaches is not possible in strictly asexual species ., However several species show coexisting CP and OP lineages , with OP lineages often retaining a residual production of males ., Such species offer ideal systems to decipher the heredity and therefore the genetic basis of the loss of sexual reproduction ., In the rare cases where it has been explored , genetic control of this trait has been shown to be rather simple , with the involvement of one to four loci , depending on the studied organisms 10–15 ., However , the precise location and underlying function of these genetic factors have not been elucidated ., The ancestral life-cycle of aphids is cyclical parthenogenesis 16 , which consists in an alternation of sexual and asexual generations ., In spring and summer , CP lineages produce asexual females through apomictic parthenogenesis ., In autumn , asexual females give birth to males and sexual females in response to photoperiodic cues ( note that CP lineages can also produce asexual females to some extent e . g . 17 ) ., Sexual females are strict clones of their asexual mothers , while one of the two X chromosomes is randomly lost to generate males 17 ., Eggs produced by sexual females are the only frost-resistant stage in the aphid cycle ., Hence , a CP life cycle is required to survive in regions with cold winters ., In addition , many aphid species also encompass OP lineages which are characterized by an altered response to sex-inducing environmental cues as they produce only asexual females ( although they often produce some males 18 , 19 ) ., These lineages are thus cold-sensitive because of their inability to lay eggs ., Yet , OP lineages are favoured in regions with mild winters where they have a major demographic advantage over CP lineages 20 , 21 ., Accordingly , CP lineages dominate in cold areas and OP lineages in warmer regions , and both coexist in regions with fluctuating winter temperatures 18–20 ., Because male production by OP lineages is difficult to prove in the wild , it has been demonstrated in a single study which also showed that these males actually contribute to sexual reproduction with CP lineages 22 , 23 ., While the switch from clonal to sexual reproduction in CP aphids is triggered by photoperiodic changes , the loss of sexual form production in OP aphids is genetically determined , changes in environmental conditions having little or no effect on their reproductive phenotype 10 , 19 ., Here , we combined QTL and genome scan approaches to decipher the genetic bases of reproductive mode variation in the pea aphid Acyrthosiphon pisum ., This species conveniently shows CP lineages ( here defined as those able to produce sexual females ) and OP lineages ( defined as those unable to produce sexual females ) , and these two types of lineages locally co-occur in regions with intermediate climate conditions 24 ., These independent QTL and genome scan approaches outlined the same genomic region as controlling obligate parthenogenesis , this trait being recessive and determined by an X-linked locus ., Our data also indicate that asexuality is transmitted in a contagious manner , leading to the conversion of sexual lineages into asexual ones ., We produced F1 crosses between males of an obligate parthenogenetic lineage ( L21V1 ) and sexual females of two cyclically parthenogenetic lineages ( JML06 and LSR1 ) ( Fig . 1 ) ., Five F2 crosses ( families 3 to 7 ) involving 6 F1 lineages were performed to obtain a genetic map and to locate QTL controlling the presence and proportion of sexual females by genotypes placed under sex-inducing conditions ., A total of 305 microsatellite markers ( out of 394 ) was successfully ordered on the genetic maps ., These loci clustered in four linkage groups that correspond to the four chromosomes of the pea aphid 25 ., 45 loci locate on the X chromosome ( LG1 following notation in 26 ) , and 85 , 135 , and 40 on LG2 , LG3 and LG4 , respectively ., Average map length ( over males and females ) was 113 , 95 , 79 and 59 cM for LG1 , LG2 , LG3 and LG4 , respectively ( Fig . 2 ) ., Of the 89 unmapped loci ( out of 394 ) , 51 were monomorphic in the 3-generation pedigree , five were homozygous in all F1 females , and 33 showed null alleles at high frequencies or inconsistent genotypes ( presumably due to difficulties to score alleles ) ., By contrast with the 61 F1 progeny which all produced sexual females ( hence were classified as CP ) segregation of reproductive phenotype was observed among the five F2 families ( Fig . 1 , families 3 to 7 ) ., All five families ( 203 F2 genotypes ) comprised a mixture of genotypes expressing either an OP ( no sexual females produced at all ) or a CP ( sexual female production ranged from 22% to 77% ) phenotype ., The percentage of OP F2 ranged from 7% to 35% depending on families ( Fig . 1 , see also S1 Figure ) ., Contrastingly , 97% of F2 lineages produced males: only 5 out of 35 OP lineages , and 2 out of 168 CP lineages did not produce males ( S1 Figure ) ., QTLs analyses on these five F2 families revealed one candidate genomic region located on the X chromosome ( LG1 ) for the control of reproductive mode variation ( measured as the proportion of sexual females or occurrence of sexual females ) , as evidenced by likelihood-ratio ( LR ) values for these traits above the LR thresholds corresponding to the null hypothesis of no QTL ., The QTL for the proportion of sexual females produced locates at 38 . 0 cM on LG1 based on highest LR values ., The 95% confidence interval CI for QTL position is 34 . 0–43 . 2 cM ( Fig . 2 ) ., The QTL for presence/absence of sexual females locates at 37 . 6 cM on LG1 and the 95% CI is 34 . 8–43 . 6 cM ( Fig . 2 ) ., We then accounted for the presence of a QTL at position ∼38 cM to test for a second QTL ( see Methods ) ., No significant support for a second QTL was found for either traits , as LR values along the four chromosomes were largely below the LR thresholds corresponding to 5% significance at the genome level ., We then focused on the genomic region pinpointed by the QTL analysis ( ∼38 cM on LG1 ) and looked at the alleles inherited by F2 individuals ., In three F2 families ( families 3 , 4 and 6 , Table 1 ) , the 29 F2 lineages that expressed an OP phenotype had the same genotype as the OP lineage L21V1 ( F0 ) at all markers located on the X-chromosome between T_128012_2_G ( 34 . 8 cM ) and T_126075_3_Y ( 49 cM ) ., For simplification , we refer to this multilocus genotype as “op1/op2” ( Table 1 , see also S1 Figure ) ., Contrastingly , the 89 CP F2 individuals from families 3 , 4 and 6 all possessed at least one allele inherited from their CP grandmothers in this genomic region ( the four possible alleles from the two CP grandmothers are collectively referred to as “CP” ) ., Hence these individuals were either op1/CP , op2/CP or CP/CP ( Table 1 , S1 Figure ) ., In these three F2 families , the op1 allele was transmitted through the F1 fathers from the OP grandfather ( L21V1 , of genotype op1/op2 ) ., Since chromosomes in male pea aphids do not recombine 27 , the entire X-chromosome of grandfather L21V1 that carries the op1 allele was transmitted to its grandchildren ., Conversely , the op2 allele was inherited from the F1 mothers , which themselves inherited the whole op2-bearing X-chromosome from their OP grandfather ., Recombination of the op2-bearing X-chromosome in the F1 mothers allowed reducing the region controlling reproductive modes between markers T_128012_2_G ( 34 . 8 cM ) and T_126075_3_Y ( 49 cM ) on the op2-bearing X chromosome ( S1 Figure ) ., Based on the results from QTL analyses , we performed two additional crosses ., Here only a subset of individuals per cross were phenotyped ( 24 and 27 , respectively ) , chosen according to their genotype at 8 microsatellite markers in the genomic region of interest ., A F3 cross ( cross 9 , see Fig . 1 and Table, 1 ) confirmed the location of the QTL and allowed further narrowing down its upper boundaries to marker 111865_3 48 . 5 cM ( see S1 Figure ) ., We finally crossed op1/CP2 females with op2/CP3 males in order to recombine the op1-bearing X-chromosome ( cross 8 , Table 1 ) ., All the 11 lineages that harboured the op1 allele in combination with the op2-bearing X-chromosome were OP , and recombination in the op1-bearing X-chromosome showed that the region controlling reproductive phenotype lies between markers 116879_10 ( 39 . 1 cM ) and D_111865_3 ( 48 . 5 cM ) on the op1-bearing X copy ( see S1 Figure ) ., These different crosses revealed that op1 and op2 alleles are recessive over CP alleles ( since the 76 op1/CP and the 41 op2/CP lineages are CP , and the 12 op2/op2 and 43 op1/op2 lineages are OP , Table 1 ) ., Noteworthy we observed that op1/op1 genotypes can have either a CP ( 11 lineages ) or an OP ( 6 lineages ) phenotype ( Table 1 ) , suggesting that other genetic or environmental factors mitigate the control of reproductive phenotype in lineages op1/op1 at the major candidate locus ., A 436-marker genome scan performed on 109 individuals from wild populations collected in environments selecting for different reproductive modes ( OP or CP ) revealed four loci having excessive genetic differentiation ( FCT ) at the α\u200a=\u200a0 . 01 threshold ( ARLEQUIN 3 . 5 analysis , Table 2 , S2 Figure ) ., FCT between populations under selection for different reproductive modes ranged from 0 . 14 to 0 . 31 at these four outlier loci , while the median FCT value estimated over the 436 markers was 0 . 014 ( average 0 . 025 ) ., Among these four outliers , T_111491_2 was also identified as outlier under balancing selection in the populations from CP-selecting environment ( FST among CP populations was 0 . 0003 , and He 0 . 56 ) when ARLEQUIN analyses were performed among populations assumed to share the same reproductive mode ( Table 2 ) ., This locus was not successfully genotyped in the families so its genomic location remains unknown ., Interestingly , the three other outliers co-locate on the X-chromosome and within the same genomic region identified with the experimental ( QTL ) approach ( Fig . 2 ) ., Accordingly , FCT values along the genetic map of the four chromosomes show a clear peak of genetic differentiation in the QTL region ( Fig . 2 ) ., In this region , expected heterozygosity in OP populations was lower than in CP populations ( S3 Figure ) , while heterozygosity values of the three CP populations and the three OP populations were similar along other regions of the chromosomes ., We have shown here that a key ecological trait – the variation in reproductive mode – was controlled by one main genomic region in the pea aphid ., This ∼9 cM-wide X-linked region was identified by two independent and complementary approaches: the co-segregation of molecular markers and phenotypes in experimental crosses and a large scale population genomic survey ( genome scans ) ., Interestingly , 100% of phenotypic variation was explained by the genotype at the candidate locus in five crosses ( crosses 3 , 4 , 6 , 8 , 9 , Table 1 ) ., In the two remaining families ( crosses 5 and 7 ) , this genomic region was also strongly associated with reproductive phenotype ( as all six OP F2 were op1/op1 at this candidate region ) but linkage was not absolute ( as 11 op1/op1 individuals are nevertheless CP ) ( Table 1 ) ., Two hypotheses can be invoked for this lack of association in some F2 genotypes ., First , an additional locus with minor effects might contribute to the control of reproductive mode variation , its contribution being only visible in individuals op1/op1 at the major locus ( all 68 individuals from crosses 5 and 7 that are not op1/op1 are CP ) ., A second hypothesis is that the production of sexual females depends on a threshold concentration of some unknown factor ( e . g . transcript , protein , hormone ) ., Under this assumption , minor environmental variation could have drastic effect on reproductive phenotype determination around the concentration threshold ., We tested for the presence of a second QTL ( first hypothesis ) , and found no statistical support for it ., Yet , power to detect such an additional QTL was low ( due to the small sample size of op1/op1 genotypes ) so we cannot at the moment disentangle these two hypotheses ., Nevertheless , the mostly single-locus recessive inheritance of obligate parthenogenesis in the pea aphid is in line with the few similar studies which showed that the transition from sexual to asexual reproduction is determined by a small number of loci 10–14 ., Transitions from cyclical parthenogenesis ( CP , i . e . the alternation of asexual and sexual generations ) to obligate parthenogenesis ( OP ) in aphids probably occur through loss-of-function mutations leading to an inability of lineages to produce females in response to the environmental cues that normally trigger the sexual phase ., Hence , any mutation ( i . e . point mutation , indel or rearrangements ) that disrupts the pathway leading to the production of sexual females might be responsible for this transition ., In theory , these loss-of-function mutations could occur repeatedly in the same gene , or on different genes involved in the same molecular cascade , these genes being either neighbours or scattered on the genome ., Herein , the OP grand-parent used for QTL mapping harbours two distinct alleles ( op1 and op2 ) at the identified QTL and the phenotype of homozygotes op2/op2 and op1/op1 significantly differs ( all 12 op2/op2 but only 6 of the 17 op1/op1 individuals are OP , test of proportion: p\u200a=\u200a0 . 0016 ) ., This indicates that at least two independent mutations in the same region are involved in the loss of sexual reproduction ., Remarkably , the genome scan demonstrates that the region identified by the QTL approach also shows signatures of divergent selection between populations under different selective regimes for reproductive mode ., This indicates that the QTL identified with three laboratory clones is also involved in the control of reproductive mode in wild populations originating from a large-scale geographic area ( populations were collected up to 700 km apart ) ., These population genomic data give further insights into the transitions from CP to OP ., In particular , the occurrence of outliers in the QTL region , combined with their low genetic diversity in OP- compared to CP-selecting environments , reveal that only one or a few mutations leading to the OP phenotype have reached high frequencies in OP-selecting environments ( otherwise this genomic region would not have been identified as FST-outlier ) ., Outside the candidate region , populations from CP- and OP-selecting environments are weakly differentiated and show highly correlated levels of genetic diversity along chromosomes , suggesting important gene flow ., The most likely scenario to explain these genomic patterns of differentiation involves bidirectional gene exchanges between CP and OP lineages: Let us consider that the rare males produced by OP lineages successfully mate with sexual females from CP ( as it is the case in laboratory conditions and presumably into the wild ) , producing CP offspring heterozygous at the candidate region ( op/CP ) ., These heterozygous CP lineages may produce OP progenies ( those homozygote for the op alleles ) that would survive if they encounter mild winter environments ., Some minimal amount of gene flow can maintain a low genetic differentiation between populations from OP- and CP-selecting environments at the scale of the genome since divergence for neutral loci at a migration drift equilibrium is prevented when Nm>1 , N being the effective population size and m the migration rate 28 , 29 ., Such bi-directional gene flow between OP and CP lineages may occur in the geographical regions with intermediate winter conditions where both CP and OP lineages coexist 22 , 30 ., Another scenario to consider relies on unidirectional gene flow from CP to OP ., Under the hypothesis that recessive op alleles are relatively frequent in CP populations , CP lineages will regularly produce new OP lineages ( those homozygous at the op alleles ) ., If such OP linages are generated frequently , low differentiation between populations from OP- and CP-selecting environments along the genome is expected , except at the candidate region ., This scenario is however less parsimonious than the former ., First , it requires very frequent production of OP lineages by CP ones in order to prevent genomic differentiation between these two compartments likely to result from the strong clonal fluctuations ( due to neutral factors and/or selection ) typical of asexual populations 31 , 32 ., Second , in absence of reciprocal gene flow from OP to CP lineages , positive selection on op alleles in CP populations should be invoked to maintain these alleles ., Yet , we know that op alleles are associated with a cost in CP selecting environments ( homozygous op/op individuals do not survive cold winters ) and therefore their frequencies are expected to decrease under these conditions ., Our data are thus best explained by bidirectional gene flow between populations of distinct reproductive modes and support the hypothesis of contagious asexuality in wild pea aphid populations ., Contagious asexuality has important consequences on the evolvability of the OP lineages ., Indeed , the bi-directional gene flow between CP and OP lineages allows genomes and alleles evolved under an asexual regime to enter the “sexual” pool via the few males produced by OP clones ., Once introgressed in a CP lineage , a genomic region evolved under an asexual regime will recombine , allowing the purging of deleterious alleles ., Later , if some of the CP individuals produce OP progeny ( those homozygous at the op/op genomic region ) , some of the alleles evolved under the asexual regime might then reintegrate an OP lineage ., Hence , contagious asexuality has the potential to combine the beneficial effects of sex ( purging of deleterious mutations and combination of beneficial mutations within the same genome 33 , 34 ) with the advantages of clonal reproduction that avoid the two-fold cost of sex 34 and can “freeze” a genome ( avoiding recombination load ) 35 ., This genetic system could thus favour the regular emergence of well fit OP lineages , which would be so fit because they would reuse alleles that competed and evolved under an OP-selecting environment ( during their long stay within OP lineages ) and that would have been separated from linked deleterious mutations during their sojourn in CP lineages ., The physical size of the ∼9 cM candidate region , that represents ∼2 . 6% of the whole genome in term of recombination units ( cM ) , is still unknown because scaffolds from the pea aphid genome sequence are not yet ordered on chromosomes 36 ., Hence the exact number and nature of the genes that are comprised within the candidate region are not known ., Nevertheless , already 66 genes encoding proteins have been identified in the three scaffolds covering part of the 9 cM genomic region of interest ( S1 Table ) ., It is too early to designate candidate genes responsible for the CP/OP phenotypes , mostly because half of them have no predicted functions ., However , recent works on the genetic programs involved in the seasonal switch from clonal to sexual reproduction in CP lineages allow highlighting in the candidate region three predicted genes putatively involved in photoperiod perception and brain signalling ( e . g . rhodopsin specific isomerase , insulin ) , two pathways identified as differentially expressed in aphids exposed to either clonal or sex induction regimes 37 ., Two genes putatively involved in the melavonate pathway ( farnesyl-pyrophosphate synthase like and hydroxymetharyglutaryl-CoA synthase ) upstream of the juvenile hormone synthesis , which is known as being a key regulator of reproductive orientation in aphids 38 , 39 , also locate within the candidate region ., To conclude , here we combined population genomics and quantitative genetics to identify the genetic bases of a key trait for aphid adaptation to climate - the loss or maintenance of sexual reproduction ., We found this trait to be controlled by one main genomic region located on the X chromosome ., The widespread geographical distribution of a few alleles associated with obligate asexuality suggests that these alleles might be particularly advantageous for OP lineages , and might have outcompeted previously established op alleles , a hypothesis that deserves further investigations ., We crossed individuals from three genotypes ( clones LSR1 , L21V1 , JML06 ) that present contrasted reproductive phenotypes ., These three F0 lineages were chosen based on their ability to produce or not sexual females under standard sex-inducing conditions ( i . e . short photoperiod with 12 h light ) 37 ., All aphids were reared on Vicia faba ( broad bean ) because it is a universal host for all known host races of the pea aphid species complex 40 , 41 and also because this plant is easier to grow compared to Medicago ssp ., LSR1 ( collected on Medicago sativa in New-York , USA in 1998 and used for complete genome sequencing 36 ) produces males ( 21% ) , sexual females ( 54% ) plus some parthenogenetic females ( 25% ) under standard sex-inducing conditions ., Under the same inducing conditions , JML06 ( sampled on Medicago lupulina in Jena , Germany in 2006 ) produces only sexual individuals ( 70% males and 30% sexual females ) ., Contrastingly , L21V1 ( sampled on M . sativa in Rennes , France in 2003 ) produces only parthenogenetic females ( 89% ) and a few males ( 11% ) ., LSR1 and JML06 are therefore classified as cyclical parthenogens ( CP ) and L21V1 as obligate parthenogen ( OP ) ., Crosses between males from the OP and ( sexual ) females from the CP lineages were performed ., For this , one L4 larva from each of the three grandparent clones was moved to a new broad bean plant and transferred to a climatic chamber with a 12 h photoperiod ( 18°C ) to trigger the production of sexual females and males in CP lineages ( and males in the OP lineage ) 37 ., Then , for each lineage , three larvae of the next clonal generation were isolated on three different broad bean plants ., Once the larvae reached adulthood and started to give birth to nymphs , groups of 10 larvae of the next generation were isolated on new broad bean plants until the asexual female stopped reproducing and died ., The morph of each individual of this second clonal generation ( i . e . male , sexual females , asexual females ) was determined at adult stage based on morphological characters ( males are slender than females , and the legs of sexual females are longer that those of asexuals ) ., The few individuals that died before reaching the adult stage ( hence before being sexed ) were also counted ., Then a total of 50 males from the clone L21V1 and 50 sexual females from the clone JML06 were put together on broad bean plants ( Vicia faba ) to generate a F1 family ( cross 1: JML06 ♀×L21V1 ♂ , Fig . 1 ) ., The 50 sexual females used in the cross are clonal ., However , males consist of two different genotypes because they inherit randomly one of the two X copies from their asexual mother ( approximately half of males are expected to bear the first X copy of their mother and the second half the other copy ) 17 ., A second F1 family was generated similarly by crossing 50 L21V1 males with 50 LSR1 sexual females ( cross 2 , Fig . 1 ) ., In Fig . 1 , dotted lines show lineages used as male and plain lines those used as female ., Eggs were kept at 4°C ( 80% humidity ) for 85 days and were then transferred at 18°C for hatching ., A few days after the first eggs hatched , 50 parthenogenetic larvae for each cross were isolated on new broad bean plants ., Each F1 lineage was kept for 7 to 9 months under conditions sustaining clonal reproduction ( 16 h light , 18°C ) ., Reproductive phenotype of the F1 lineages was then assessed similarly ., Six F1 lineages ( three per cross ) were then chosen to produce the next F2 generation ( Fig . 1 ) ., All F1 produced sexual females ( with from 27% to 71% and 28% to 64% sexual females for cross 1 and 2 , respectively ) , hence were CP ., The 6 F1 clones were thus chosen to cover the diversity in terms of the production of males ( that ranged from 0–73% and 0–55% males for cross 1 and 2 , respectively ) and asexual females ( that ranged from 0–42% and 1–53% for cross 1 and 2 , respectively ) ., Five F2 crosses ( crosses 3 to 7 in Fig . 1 ) were performed using the same protocol as for the F1 ., 44 to 47 F2 lineages per family ( hence 229 F2 lineages in total ) were then isolated and kept for subsequent assessment of reproductive mode phenotype ( same protocol as for the F0 and F1 ) ., Twenty-six F2 lineages ( out of 229 ) died before being phenotyped ., The three grand-parents , the six F1 parents and the 229 F2 individuals from families 3 to 7 were typed at 401 microsatellite loci ( see S2 Table for loci used and 42–44 for primer sequences ) ., We first checked for the presence of null alleles by looking at the inheritance of alleles in the 3-generation pedigree ., Homozygous individuals originating from parents displaying a null allele were transformed into missing data ., Loci located on the same chromosome were identified based on their complete linkage in males ( 2n\u200a=\u200a8 in the pea aphid and chromosomes in males do not recombine ) 27 , 45 ., Genetic maps were then constructed for each of the four chromosomes with Crimap 2 . 53 46 using Kosambi mapping function ., Linkage maps were drawn using MAPCHART v . 2 . 1 47 ., QTL detection was then performed with the interval mapping method implemented in QTLmap , using the LDLA approach 48 ., The phenotypic traits analysed for each F2 lineage ( from crosses 3 to 7 ) were the occurrence ( binary variable ) and the percentage ( quantitative variable ) of sexual females in the parthenogenetically produced offspring ., We focused on these traits because the production of sexual female is the most relevant variable to predict whether a population is able to reproduce sexually or not 10 , 49 ., In our analyses , we set parameter ndmin to 200 so that no information from males meioses was used to locate QTLs ( since chromosomes do not recombine in males , males are not informative to locate QTLs within chromosomes ) ., QTLs were detected based on likelihood-ratio ( LR ) along chromosomes ., LR values corresponding to a significance level of 0 . 05 for each chromosome were empirically determined from 1 , 000 simulations under the null hypothesis of the test ( i . e . no QTL ) ., Genome-wide significance levels ( i . e . LR values corresponding to adjusted p-values ) were computed to account for multiple testing ( i . e . four tests , corresponding to the four chromosomes ) ., The drop-off method implemented in QTLmap was applied to obtain 95% confidence intervals of the QTL location ., Similarly to the reduction of x-LOD when using LOD scores , the maximum LR value was reduced by 3 . 84 ( corresponding to a Chi2 distribution with one degree of freedom for p<0 . 05 ) to determine a threshold ., Region boundaries were then defined by the LR locations crossing this threshold upstream and downstream of the peak LR as described in 50 , 51 to identify the 95% CI of the QTL ., After identifying the first QTL ( see Results ) , we tested for the presence of a second QTL ., For this , genotype at locus T_121775_26 ( the closest marker from the peak of the QTL on LG1 ) was introduced as a fixed effect in the model ., This marker is highly discriminative as each of the three grandparents possesses different alleles ., LR for the presence of a QTLs against the hypothesis of no QTL was then compared to LR thresholds corresponding to a 5% significance determined by 1000 simulations in QTLmap ., The QTL approach led to the identification of a single genomic region , located on the X chromosome , which controls reproductive mode ( see Results ) ., Yet , in the three F2 crosses that were highly informative , all lineages used as mother inherited by chance the same X chromosome copy from their OP father ( remind that chromosomes do not recombine in male aphids ) ., From these crosses and from recombination events , we determined that the gene ( s ) that control ( s ) reproductive mode locate ( s ) between markers T_128012_2_G ( 34 . 8 cM ) and T_126075_3_Y ( 49 cM ) on this X copy ( see Results ) ., The segment of this X chromosome copy is referred to as “op2 allele” hereafter ., However , we had little power to test whether the same region also controls this trait on the second X copy from the OP grandparent ( that we refer to as “op1 allele” ) ., We therefore performed an additional F2 cross ( cross 8: X2_25♂×X1_3 ♀ ) to recombine the X-chromosome bearing the op1 allele ( the mother X1_3 ♀ bears one op1 allele ) ., We also crossed X6_2♂×X3_4♀ ( that each possesses an op2 allele , cross 9 ) to produce homozygous individuals at this X chromosome region ( i . e . op2/op2 ) to assess the dominance status of the different alleles in the candidate region ( i . e . op1 , op2 , and those inherited from CP clones , referred to as CP1 , CP2 , CP3 and CP4 ) ., Since these two crosses were performed after we had identified the genomic region controlling reproductive mode variation , only a subset of individuals were phenotyped ( 24 and 27 , respectively ) , chosen accordingly to their genotype at 8 microsatellite markers in the genomic region of interest ( see S1 Figure for markers used ) ., Pea aphid individuals were collected in alfalfa fields from six sampling sites ( S3 Table ) ., All A . pisum individuals were sampled from the same plant species ( Medicago sativa ) to prevent confounding effects of plant or reproductive mode specialization on genetic divergence 52 ., Three of the sites locate in north-east France or Switzerland and correspond to regions with cold winters ( “temperate continental climate” as defined in 53 ) ., In these areas , pea aphid populations consist mainly of CP lineages , because eggs are the only stage that survives cold winters 19 ( we thus consider these areas as CP-selecting environment ) ., Individuals were collected in spring 2008 , a few weeks after egg hatching to maximize the probability to sample locally overwintering populations ( these samples have been used in 43 , 44 ) ., The three other sampling sites locate in south-west France , and correspond to regions characterized by mild winters ( i . e . “temperate oceanic climate” as defined in 53 ) ., These areas are considered as OP-selecting environment ., Here , sampling took place in winter 2008–2009 because at this season , OP lineages can be discriminated from CP ones , since the former overwinter as parthenogenetic females while the latter spend winter as eggs ., Parthenogenetic females were collected from the six populations ( see 43 for further details ) ., To obtain sufficient amounts of DNA for genotyping hundreds of microsatelli

Introduction, Results, Discussion, Materials and Methods

Although evolutionary transitions from sexual to asexual reproduction are frequent in eukaryotes , the genetic bases of such shifts toward asexuality remain largely unknown ., We addressed this issue in an aphid species where both sexual and obligate asexual lineages coexist in natural populations ., These sexual and asexual lineages may occasionally interbreed because some asexual lineages maintain a residual production of males potentially able to mate with the females produced by sexual lineages ., Hence , this species is an ideal model to study the genetic basis of the loss of sexual reproduction with quantitative genetic and population genomic approaches ., Our analysis of the co-segregation of ∼300 molecular markers and reproductive phenotype in experimental crosses pinpointed an X-linked region controlling obligate asexuality , this state of character being recessive ., A population genetic analysis ( >400-marker genome scan ) on wild sexual and asexual genotypes from geographically distant populations under divergent selection for reproductive strategies detected a strong signature of divergent selection in the genomic region identified by the experimental crosses ., These population genetic data confirm the implication of the candidate region in the control of reproductive mode in wild populations originating from 700 km apart ., Patterns of genetic differentiation along chromosomes suggest bidirectional gene flow between populations with distinct reproductive modes , supporting contagious asexuality as a prevailing route to permanent parthenogenesis in pea aphids ., This genetic system provides new insights into the mechanisms of coexistence of sexual and asexual aphid lineages .

Asexual lineages occur in most groups of organisms and arise from loss of sex in sexual species ., Yet , the genomic bases of these transitions remain largely unknown ., Here , we combined quantitative genetic and population genomic approaches to unravel the genetic control of shifts towards permanent asexuality in the pea aphid , which conveniently shows coexisting sexual and asexual lineages ., We identified one main genomic region responsible for this transition located on the X chromosome ., Also , our population genetic data indicated substantial gene exchange between these reproductively distinct lineages , potentially leading to the conversion of some sexual lineages into asexual ones in a contagious manner ., This genetic system provides new insights into the mechanisms of coexistence of sexual and asexual lineages .

genomics, reproductive system, signatures of natural selection, anatomy, genome scans, natural selection, genome analysis, genetics, biology and life sciences, population genetics, computational biology, evolutionary biology, ecological selection, evolutionary processes, asexual reproduction

journal.pgen.1003387

TIP48/Reptin and H2A.Z Requirement for Initiating Chromatin Remodeling in Estrogen-Activated Transcription

Transcription activation relies on a choreography of local chromatin remodeling events that include posttranslational histone modifications and replacement of canonical histones by variants 1–4 ., Chromatin immunoprecipitation ( ChIP ) studies have provided extensive information on the recruitment of these complexes by the hormone bound estrogen receptor in ERα-positive breast cancer cells 5 ., The first complex to occupy promoter sequences is the ATP-dependent SWI/SNF chromatin remodeling complex and its catalytic subunit Brg1 ., Its activity enables subsequent binding of a plethora of histone and protein modifying assemblies which lead to transcription initiation by polymerase II 6–10 ., In contrast , it is less clear how local chromatin structure prepares for rapid and massive recruitment of the estrogen receptor itself in the presence of estrogen ., Incorporation of histone variants constitutes a means to alter nucleosome properties and positioning at specific genomic loci ., The histone H2A variant H2A . Z is frequently found within nucleosomes at regulatory sequences 11–14 ., In particular , H2A . Z occupancy characterizes inducible and constitutive DNAseI hypersensitive sites to which nuclear receptors bind 15 ., This variant is believed to induce a chromatin conformation that poises genes for transcription in human cells 16 ., In yeast , H2A . Z exchange is mediated by the SWR1 complex 17 , 18 ., However , in mammalian cells , the mechanisms of H2A . Z deposition are still poorly characterized and may require several distinct protein complexes depending on the cellular context ( for a review see 19 ) ., Among ATP-dependent chromatin remodeling complexes the TIP48/TIP49 containing SWR1/SRCAP 20 or TIP60/p400 21 , 22 complexes have been shown to play a role in H2A . Z deposition 17 , 18 , 23 ., p400 was also reported to be required for H2A . Z incorporation into the TFF1/pS2 gene concomitant to estrogen receptor binding 24 ., In an in vitro study Choi et al . demonstrated that the AAA+ family ( ATPases Associated with various cellular Activities ) members TIP48/TIP49 participate in the replacement of H2A by H2A . Z 25 ., This H2A . Z exchange was facilitated by TIP60-mediated H2A acetylation ., TIP48/TIP49 proteins ( also known as TIP49b and TIP49a , Rvb2 and Rvb1 , reptin and pontin ) are important for assembly and activity of the histone TIP60 acetyltransferase complex 26 ., To gain a better understanding of the early steps required in estrogen receptor mediated transcription activation and the coordination between remodeling complexes and chromatin structure , we analyzed transcription of the cyclin D1 gene ( CCND1 ) in ERα-positive MCF-7 breast cancer cells ., This oncogene is frequently overexpressed in human breast tumors 27 ., Its down-regulation increases migratory capacity and is linked to unfavorable prognosis 28 , 29 ., Cyclin D1 is a mitogenic sensor that modulates cell cycle progression ., CCND1 transcription is stimulated by 17β-estradiol ( E2 ) , inhibited by antiestrogens and cell cycle regulated in ERα-positive breast cancer cells 30 , 31 ., Here we show that TIP48 and H2A . Z associate with CCND1 promoter and enhancer sequences ., TIP48 is required for chromatin reorganization which is initiated by release of H2A . Z and opening of a repressive promoter-enhancer gene loop enabling TIP60 and the E2 bound estrogen receptor to be loaded to stimulate CCND1 transcription ., In ERα-positive MCF-7 cells grown in steroid stripped media only basal transcription levels of the Cyclin D1 gene ( CCND1 ) were measured ., Addition of E2 lead to a 2 . 5-fold increase in CCND1 mRNA levels in cells treated 6 h with 100 nM E2 ( Figure 1A ) ., H2A . Z has been reported to act in concert with ER to regulate the TFF1 gene 24 prompting us to examine H2A . Z association with CCND1 regulatory elements ., Analysis of ChIP-on-chip data revealed that H2A . Z was highly enriched at sequences 5′ and 3′ flanking the CCND1 gene , and largely absent from the open reading frame ( Figure 1B ) 32 ., By conventional ChIP , we found that the amount of H2A . Z present at the CCND1 promoter was reduced by 50% at the TSS ( Figure 1C ) ., Eeckhoute et al . identified an enhancer ( enh2 ) at the 3′ end of the CCND1 gene which acts as the primary site for ERα and cofactor binding during CCND1 transcriptional regulation 7 ., We thus analyzed the chromatin organization of enh2 ., Similar to the promoter , H2A . Z present at enh2 was removed during transcription activation ( Figure 1C ) ., Reduced binding was not due to a decrease in H2AFZ expression in the presence of E2 ( Figure 1A ) ., Chromatin modifications at the CCND1 promoter and enhancer appear to be coordinated ., Removal of H2A . Z from promoter sequences upon transcription activation correlates with observations in yeast and several mammalian cells and points to a mechanism of regulation distinct from the one of the ERα target gene TFF1 33 , 24 , 34 ., Replacement of nucleosomal H2A with H2A . Z has been shown to be catalyzed by the TIP48/49 complex in vitro 25 ., The TIP48/49 complex was thus a good candidate for regulating H2A . Z dynamics at CCND1 regulatory sequences ., TIP48 and TIP49 are ubiquitously expressed and are often part of the same complex ., In most cell types , and in particular in epithelial cancer cells such as MCF-7 cells , silencing one of the partners by interference RNA lead to degradation of the other partner 33 ., Better antibody specificity and efficiency prompted us to investigate TIP48 ., TIP48 was associated with the promoter and enh2 of CCND1 in non-induced ERα-positive MCF-7 cells ( Figure 1D ) ., Binding of TIP48 to the CCND1 TSS and enh2 decreased rapidly following addition of 100 nM estradiol ( E2 ) ., Expression of the gene coding for TIP48 was insensitive to E2 ( Figure S1A ) ., To examine the relationship between TIP48 and H2A . Z , we selectively depleted TIP48 by siRNA ( Figure S1A ) ., In siTIP48 transfected cells treated or not with E2 , H2A . Z binding to CCND1 was reduced compared to control , non–specific siRNA transfected cells ., Levels of H2A . Z binding in the absence of TIP48 were roughly equivalent to levels in E2 treated control cells ( Figure 2A ) ., Moreover , nucleosome density assessed by immunoprecipitating histone H3 was unchanged near the TSS ( Figure 2B ) ., Thus , eviction of H2A . Z upon initiation of E2 stimulated transcription was not due to general chromatin decondensation around the CCND1 gene and its promoter region in particular ., TIP48 appears to be necessary for recruiting H2A . Z to the CCND1 gene in MCF-7 mammary tumor cells ., We thus asked whether binding of H2A . Z or its release were important for regulating CCND1 transcription ., H2A . Z mRNA expression levels were reduced ∼5-fold 48 h post transfection with a smartpool siRNA directed against H2A . Z compared to control cells ( Figure S1B ) ., Reduced levels of H2A . Z did not alter basal CCND1 expression levels , but impeded activation by E2 ( Figure 2C ) ., Similarly , in the absence of TIP48 , basal transcription levels were conserved , while activation of CCND1 by E2 was compromised ( Figure 2C ) ., H2AFZ mRNA levels were not affected by selective knockdown of TIP48 ( Figure S2A ) ., Thus , Stimulation of CCND1 expression required release of H2A . Z concomitantly from both these DNA elements ., Absence of activation was likely due to failure of ERα fixation to the CCND1 promoter ., Under standard conditions , E2 stimulated ERα binding to both the promoter and enh2 of CCND1 ( Figure 3A ) 7 ., Selective knock down of TIP48 hindered ERα binding to these sites ( Figure 3A ) ., Reduced binding could not be attributed to altered or decreased ESR1 expression patterns in cells transfected with control or TIP48 siRNAs ( Figure 3B ) ., Therefore , TIP48 appears to be necessary to remodel CCND1 chromatin structure for productive ERα binding in the presence of hormone ., TIP48 and TIP60 have been found as part of the same complex 20 , 25 ., TIP60 also cooperates with ERα and other chromatin-remodeling enzymes during estrogen-induced transcription 34 , 35 ., We tested whether TIP48 and TIP60 binding to the CCND1 promoter was coordinated ., In the presence of E2 , TIP60 was recruited to the CCND1 promoter ( Figure 4A ) ., Upon depletion of TIP48 , TIP60 no longer associated with the CCND1 TSS ( Figure 4A ) ., Cooperation between TIP48 , ERα and TIP60 binding was likely to be necessary for transcription activation ., To unravel a functional link , we first over-expressed TIP60 in MCF-7 cells ( Figure S2B ) ., TIP60 overexpression stimulated E2 activated CCND1 transcription nearly 5-fold compared to control untreated cells , without affecting neither basal , non-induced mRNA levels ( Figure 4B ) nor the expression pattern of the H2AFZ gene ( Figure S2C ) ., In siTIP48 transfected cells , overexpression of TIP60 was no longer able to stimulate CCND1 transcription upon E2 stimulation ( Figure 4C ) , suggesting that TIP48 is required for TIP60 function ., TIP60 is found in protein complexes able to acetylate histones , with a preference for lysine 5 of H2A 36 ., Core histones are generally acetylated in the promoter region of transcribed genes ., Acetylation of the histone variant H2A . Z was shown to characterize active genes in yeast and recently also in prostate cancer cells 13 , 37 ., Using an antibody that specifically recognizes H2A . Z acetylated at 3 N-terminal lysines , we determined that a large fraction of H2A . Z bound to the CCND1 promoter and to the 3′ enh2 , was highly acetylated ( Figure 5A ) ., Acetylation levels of H2A . Z did not vary following E2 induced CCND1 gene activation in control samples ( Figure 5A ) ., However , because H2A . Z was released during transcription activation , the ratio of acetylated H2A . Z/total H2A . Z increased nearly 2-fold at these sites ( Figure 5B ) ., In siTIP48 transfected cells , we observed a decrease in acetylated H2A . Z present at the TSS and the enh2 ( Figure 5A ) ., The increased ratio of acetylated H2A . Z associated with the CCND1 gene following E2 was abolished in cells transfected with siTIP48 ( Figure 5B ) ., The reduced ratio of H2A . Z acetylation thus correlated with impeded transcription activation in siTIP48 transfected MCF-7 cells ( Figure 5B and Figure 2C ) ., In conclusion , failure of TIP60 to associate with CCND1 in the absence of TIP48 correlated with reduced binding of ERα ( Figure 3 ) , reduced levels of H2A . Z acetylation at the CCND1 gene ( Figure 5 ) and the inability to activate this gene by estrogen ( Figure 2C ) ., Long-range chromatin interactions between ERa recognition sequences and enhancers have been proposed to regulate ERa-target genes in breast cancer cells 38 , 39 ., The main enhancer regulating CCND1 is located at the 3′ end of the gene , 14 kb distant from the promoter 7 ., Gene looping via promoter-enhancer crosstalk is associated with repressed , low CCND1 expression in ERa-negative , MDA-MB231 cells 40 ., Thus we asked whether this loop also existed in MCF-7 cells and more importantly , whether looping was sensitive to hormone ., We used a chromatin conformation capture ( 3C ) assay ., The 3C method detects physical proximity between distal DNA sites by ligation of cross-linked restricted DNA fragments 41 , 42 ., Ligation products between enh2 and promoter , and between enh2 and a control fragment inside the CCND1 ORF were amplified and normalized to an amplified enh2 PCR product ( see Materials and Methods ) ( Figure 6A ) ., We measured significant interaction frequencies between enh2 and promoter sequences in MCF-7 cells grown in hormone-stripped media ( -E2 ) ., Interaction frequencies were reduced ∼10-fold 45 min after addition of E2 to the cell culture ( Figure 6B ) ., No significant amplification of ligation products between enh2 and the internal control fragment was detectable ., Hence , an extragenic loop mediated by specific promoter enhancer interactions was present when CCND1 expression is low ( Figure 6B and Figure 2D ) ., Upon transcription activation , gene looping is markedly reduced ., It was tempting to speculate that TIP48 plays a role in regulating looping ., We assessed the relative frequencies of interaction between enh2/promoter and enh2/internal control fragments in MCF-7 cells transfected or not by siTIP48 ., Depletion of TIP48 had no impact on enh2/promoter contacts in the absence of E2 ( Figure 6B ) ., This observation correlated with identical basal expression levels of CCND1 in control and siTIP48 transfected cells ( Figure 2C ) ., 45 min after addition of E2 to the cells , the frequency of enh2/promoter interaction was 5-fold greater in siTIP48 transfected cells compared to control cells ( Figure 6B ) ., Conservation of significant repressive gene looping could thus account for impeded E2 bound ERa binding to the CCND1 promoter and compromised transcription activation ., We propose a model ( Figure 7 ) in which TIP48 is required at early steps during transcription activation which is initiated by release of H2A . Z and subsequent dissociation of the enhancer from the promoter ., E2 bound estrogen receptor can then recognize the promoter and stimulate transcription of CCND1 ., We unraveled a role for TIP48 in initiating transcription activation of the CCND1 oncogene ., Recruitment of the histone acetyltransferase TIP60 is dependent on TIP48 and H2A . Z binding to the promoter and 3′ enhancer of the CCND1 gene ., We propose that low levels of CCND1 expression are regulated because the associated gene loop is transcription-dependent ., This regulation is brought about by the activity of TIP48 containing complexes which locally act upon chromatin structure to release a disabling loop ., Such a mechanism allows fine-tuning transcription regulation of genes pivotal for the cellular equilibrium in rapidly changing environments ., Our work describes early events implicated in E2 induction of CCND1 ., These events include dynamic exchange of a series of cofactors , namely the TIP48 complex and histone variant H2A . Z , recruitment of TIP60 and acetylation of H2A . Z enabling the main transcription factor , the estrogen receptor , to associate with its target sequences ., TIP60 can directly interact with ERa and its acetyltransferase activity is important during transcription initiation once ERa is bound to target gene promoters 34 ., TIP60 is a versatile enzyme that functions with a variety of partners in a gene and cell specific manner 34 , 43 ., Selective knock-down of TIP60 by siRNA compromises activation of some , but not all ERa target genes in MCF-7 cells , as well as nuclear receptor independent genes in several cell lines ( unpublished observations ) ., CCND1 was one of the genes found to be insensitive to siTIP60 34 ., This observation denotes that TIP60 can be replaced by other histone acetyltransferases in CCND1 transcription activation ., Thus , dependency of early chromatin remodeling steps on TIP48 and H2A . Z may be more generally applicable to allow cofactor recruitment for productive ERa binding in stimulated transcription ., We found that H2A . Z was removed from CCND1 regulatory elements while this variant had previously been shown to be recruited to the promoter of the TFF1 gene upon E2 treatment of MCF-7 cells 24 ., Differences in promoter structure are a plausible explanation for divergent remodeling mechanisms ., It is also likely that post-translational modifications of H2A . Z are important as shown in a recent genome wide study by Valdes-Mora et al . who found that H2A . Z acetylation at the TSS correlates with active transcription in prostate cancer cells 13 ., Indeed , the level of acetylation of H2A . Z near the TSS of CCND1 was equivalent at non-activated and E2 stimulated cells ., We propose that , in ERα-positive breast cancer cells , the ratio of acetylated H2A . Z/H2A . Z rather than the total amount of H2A . Z bound to the CCND1 promoter correlates with transcriptional activity ., Chromatin remodeling events are crucial for hormone stimulated activation of estrogen receptor target genes ., However , so far , all data available describe the recruitment of remodeling complexes and cofactors once the estrogen receptor is bound ., The Brg1 subunit of the SWI/SNF complex is one of the first proteins to associate with ERa and , although transcription is no longer activated in its absence , ERa remains bound in siBrg1 transfected cells 34 ., Here we demonstrate that chromatin remodeling events prior to ERa binding are essential for initiating transcription ., These events depend on TIP48 and H2A . Z specific nucleosome conformation ., Chromatin structure impedes ERa loading via intragenic looping ., Notably , interaction between promoter and enhancer sequences forms a repressive complex ., Reduced distances between 5′ and 3′ ends of gene loci have been attributed to greater chromatin density ., In this case , looping does not require changes in chromatin compaction ., Dynamic release of gene loops is consistent with rapid chromatin remodeling and transcription activation by hormone ., Finally , addition of hormone triggers large scale chromatin remodeling ., In breast cancer cells gene response to progestin is mediated by nucleosomes 44 and estradiol treatment leads to expansion of chromosome territories within minutes 45 ., This latter phenomenon was also observed in ERα-negative cells ( unpublished ) suggesting that chromatin decondensation is independent of the receptor and may prepare its binding in ERα-positive cells ., It is thus tempting to speculate that the signaling mechanism by which hormone addition primes chromatin triggers histone exchange and remodeling prior to ERα binding ., MCF-7 cells were purchased from ATCC and were maintained in DMEM/F12 without phenol red with Glutamax containing 50 mg/ml gentamicin , 1 mM sodium pyruvate and 10% heat-inactivated and steroid free fetal calf serum ( FCS ) ( Invitrogen ) ., MCF-7 cells were treated with 10−7 M estrogen E2 ( Sigma ) for the indicated times ., 5×106 MCF-7 cells were transfected with 20 nM of H2A . Z siRNA ON-TARGET plus SMARTpool , TIP48 siRNA ON-TARGET plus SMARTpool or scrambled ( scr ) siRNA ( Dharmacon Thermo Scientific ) using Interferine ( Ozyme ) ., Cells were mock-transfected ( pcDNA3 . 1 ) or transfected with 1 µg of pcDNA3 . 1/TIP60 ( gift from Dr . Didier Trouche ) using the Amaxa Cell line Nucleofactor Kit V program P-020 according to the manufacturers protocol ., TIP60 siRNA 43 was purchased from Eurogentec , and transfected using Interferine ( Ozyme ) ., 5×105 MCF-7 cells were seeded in 6 well plates ., 72 h following siRNA transfection , total cell extracts were isolated and protein levels of H2A . Z , TIP48 and TIP60 analyzed by immunoblotting on gel SDS-page 15% using antibodies against H2A . Z ( ABCAM , ab4174 ) , TIP48 ( gift of Dr . Mikhaïl Grigoriev ) TIP60 43 or GAPDH ( Millipore , mab374 ) ., Total RNA was extracted using an RNeasy mini-kit ( Qiagen ) and eluted with 35 µl of RNAase-free water ., First strand cDNA was generated using 2 µg of total RNA in a reaction containing random oligonucleotides as primers with the ThermoScript RT-PCR system ( Invitrogen ) ., Real-time PCR was performed on a Mastercycler ep realplex 4 ( Eppendorf ) using the platinum SYBR Green q-PCR SuperMix ( Invitrogen ) according to the manufacturers instructions ., Amplification conditions: 1 min at 50°C , 3 min at 95°C followed by 40 cycles ( 20 s at 95°C , 20 s at 60°C , 20 s at 72°C ) ., mRNA expression were normalized against expression levels of the RPLP0 ribosomal gene used as an internal control ., qRT-PCR primers: H2AFZ: 5′-CCTTTTCTCTGCCTTGCTTG-3′ and 5′-CGGTGAGGTACTCCAGGATG-3′ , CCND1: 5′-GCGTCCATGCGGAAGATC-3′ and 5′-ATGGCCAGCGGGAAGAC-3′ , RPLP0: 5′-TGGCAGCATCTACAACCCTGAA-3′ and 5′- CACTGGCAACATTG CGGACA-3′ , TIP48: 5′-TGAAGAGCACTACGAAGACGC-3′ and 5′-CCTTACTACCCAGCTC CTGA- 3′ ., ChIP analyses were performed as described previously 46 ., Samples were sonicated to generate DNA fragments <500 bp ., Chromatin fragments were immunoprecipitated using antibodies against H2A . Z ( ab4174 , ABCAM ) , acetyl H2A . Z ( ab18262 , ABCAM ) , TIP48 ( gift of Dr . Mikhaïl Grigoriev ) , ERα ( sc-543 , Santa Cruz ) , H3 ( ab1791 , ABCAM ) , TIP60 47 or an irrelevant HA antibody ( H6908 , Sigma ) ., The precipitated DNA was amplified by real-time PCR , with primer sets designed to amplify the promoter ( TSS ) and enh2 enhancer regions of the CCND1 gene ( Figure 1B ) ., qRT-PCR primers: CCND1 ( TSS ) : 5′-CGGGCTTTGATCTTTGCTTA-3′ and 5′-ACTCTGCTGCTCGCTGCTAC-3′ , distal CCND1 enhancer ( enh2 ) : 5′-CAGTTTGTCTTCCCGGGTTA-3′ and 5′- CATCCAGAGCAAACAGCAG-3′ ., All ChIP data are shown as percent input ., 3C assays were performed essentially as described 48 , 49 , with minor modifications ., MCF-7 cells were treated with E2 10−7 M for 45 mn or transfected with a scrambled control siRNA , with TIP48 SMARTpool siRNA ( Dharmacon Thermo Scientific ) , and cultured in phenol red-free DMEM containing 10% FBS-T for 72 h before cross-linking ., The culture medium was removed , and cells were fixed with 1 . 5% formaldehyde for 10 min at room temperature ., Cells were then washed twice with cold phosphate-buffered saline solution , and resuspended in ice-cold lysis buffer ( 10 mm Tris-HCl , pH 8 . 0 , 10 mm NaCl , 0 . 2% Nonidet P-40 , and protease inhibitor mixture ) ., Nuclei were resuspended in 1 ml of Buffer B 1 . 2× buffer ( MBI Fermentas ) supplemented with SDS 0 , 3% ., Triton X-100 1 , 8% was added to sequester the SDS and incubated for 1 h at 37°C ., The cross-linked DNA was digested overnight with 400 units of restriction enzyme Csp6I ( MBI Fermentas ) ., The restriction enzyme was inactivated by incubation at 65°C for 20 min ., The reactions were diluted with ligase buffer ( 50 mm Tris-HCl , pH 7 . 5 , 10 mm MgCl2 , 10 mm dithiothreitol , 1 mm ATP , and 25 µg/ml bovine serum albumin ) , supplemented with Triton X-100 ( 1% final concentration ) ., The DNA was ligated using T4 DNA ligase ( New England Biolabs , Ipswich , MA ) overnight at 16°C and an additional 100 units for 2 h at 37°C ., RNase was added for 30 min at 37°C , and samples were incubated with SDS overnight at 70°C to reverse the crosslink ., The following day , samples were incubated for 2 h at 45°C with proteinase K , and the DNA was purified by phenol-chloroform extractions and ethanol precipitation ., Interaction between chromatin domains was assessed by real-time-PCR amplification for each predicted ligation event 48 , 50 ., Primers have been designed on the digested BAC fragments , directly around the putative site of ligation for the four possibilities ., BAC clones RP11-300ID ( BACPAC Resources Center at Childrens Hospital Oakland Research Institute , Oakland , CA ) containing the CCND1 gene and downstream 160-kb region were used ., 40 ug of BAC was digested by Csp6I overnight and ligated ., This product was purified by phenol chloroform and precipitated in order to generate 3C control templates ., PCR primer efficiency was measured by amplifying 0 . 01 to 50 ng of digested BAC product and also tested on a fixed amount ( 50 ng ) of digested genomic DNA ., All primers have an annealing temperature between 65 to 70°C and a product size around 150–300 bp ., All primer combinations showed PCR efficiency between 90 and 100% ., 3C assay results are presented as the average from three independent preparations of 3C DNA , followed by qPCR analysis in triplicate ., qPCR for enh2 ( PCR primers design inside the Csp6I restriction fragment enh2 ) was used as an internal control to verify ligation events ., Non-digested sample and ligation between a control fragment and enh2 were also performed ( data not shown ) ., Primers used for one of the four ligation event tested: Enh2/Prom: 5′-CTGGGAGAGATGGAGCTGAG-3′ and 5′-GGTTTTGTTGGGGGTGTAGA-3′ , Enh2/ctrl: 5′-AAGCTCTCCCACAACCCATT-3′ and 5′-GTCAGCCCCACTGTTGACTC-3′ ., Other primers available upon request .

Introduction, Results, Discussion, Materials and Methods

Histone variants , including histone H2A . Z , are incorporated into specific genomic sites and participate in transcription regulation ., The role of H2A . Z at these sites remains poorly characterized ., Our study investigates changes in the chromatin environment at the Cyclin D1 gene ( CCND1 ) during transcriptional initiation in response to estradiol in estrogen receptor positive mammary tumour cells ., We show that H2A . Z is present at the transcription start-site and downstream enhancer sequences of CCND1 when the gene is poorly transcribed ., Stimulation of CCND1 expression required release of H2A . Z concomitantly from both these DNA elements ., The AAA+ family members TIP48/reptin and the histone variant H2A . Z are required to remodel the chromatin environment at CCND1 as a prerequisite for binding of the estrogen receptor ( ERα ) in the presence of hormone ., TIP48 promotes acetylation and exchange of H2A . Z , which triggers a dissociation of the CCND1 3′ enhancer from the promoter , thereby releasing a repressive intragenic loop ., This release then enables the estrogen receptor to bind to the CCND1 promoter ., Our findings provide new insight into the priming of chromatin required for transcription factor access to their target sequence ., Dynamic release of gene loops could be a rapid means to remodel chromatin and to stimulate transcription in response to hormones .

Our study investigates changes in the chromatin environment at the Cyclin D1 gene that are a prerequisite for transcriptional initiation in response to estradiol ., Gene expression is under control of chromatin structure ., Histone variants , including histone H2A . Z , are incorporated into specific genomic sites and participate in transcription regulation ., We show that H2A . Z is present at the transcription start-site and downstream enhancer sequences of CCND1 when the gene is poorly transcribed ., Stimulation of CCND1 expression required release of H2A . Z concomitantly from both these DNA elements ., The TIP48/reptin protein , which is part of several chromatin remodeling complexes , also associated with the CCND1 regulatory elements ., Here , TIP48 promotes exchange of H2A . Z , which triggers a dissociation of the CCND1 enhancer from the promoter , thereby releasing a repressive intragenic loop ., This release then enables estrogen receptor binding to the CCND1 promoter ., Acetylation of H2A . Z is required for these processes ., Our findings provide new insight into the priming of chromatin required for transcription factor access to their target sequence ., Hence , we propose a new model for early events in transcription activation that were not shown before ., Specifically , release of looping could be a rapid means to activate transcription efficiently in response to stimuli , in particular estrogen .

oncology, medicine, chromosome biology, cell growth, gene expression, genetics, epigenetics, biology, basic cancer research, molecular cell biology, gene function

journal.pgen.1006594

Genetic prediction of male pattern baldness

The UK Biobank study 19 ( http://www . ukbiobank . ac . uk ) is a large , population-based genetic epidemiology cohort ., At its baseline assessment ( 2006–2010 ) , around 500 , 000 individuals aged between 40 and 70 years and living in the UK completed health and lifestyle questionnaires and provided biological samples for research ., The present study reports a GWAS of male pattern baldness in the UK Biobank cohort , which is over four times the size of the previously-largest meta-analytic study 15 ., After completing the GWAS , we split the cohort into a large ‘discovery’ sample of 40 , 000 participants in which the GWAS was re-run ., The regression weights from this GWAS were used to perform a prediction analysis in the sub-sample of 12 , 000 participants who did not contribute to the GWAS ., We determined the accuracy of the polygenic profile score by discriminating between those with severe hair loss and those with no hair loss ., The mean age of the 52 , 874 men was 57 . 2 years ( SD 8 . 0 ) ., 16 , 724 ( 31 . 6% ) reported no hair loss , 12 , 135 ( 23 . 0% ) had slight hair loss , 14 , 234 ( 26 . 9% ) had moderate hair loss , and 9 , 781 ( 18 . 5% ) had severe hair loss ., The genome-wide association study of the four-category self-reported baldness measure in 52 , 874 White British men from UK Biobank yielded 13 , 029 autosomal hits from the imputed data ( P<5x10-8 ) , in addition to 117 hits ( out of 14 , 350 genotyped SNPs ) on the X chromosome ( Fig 1 ) ., The QQ plot for the autosomal GWAS is shown in S1 Fig . An LD clumping analysis indicated that these hits can be attributed to 247 independent autosomal regions ., All previously reported autosomal hits 10 , 13–16 that mapped to SNPs in our study ( 62 out of 68 SNPs ) replicated with a maximum P-value of 0 . 006 ( 54 out of 62 lookups had P<5x10-8 , S1 Table ) ., The previously reported X chromosome variant from Li et al . 15 and the variant from Richards et al . 10 also replicated with P-values that were effectively zero ( S1 Table ) ., The chromosome 6 hit ( rs4959410 ) from Liu et al . 13 , which was not supported by additional SNPs in the region , failed to replicate ( P = 0 . 37 ) ., All other hits from Liu et al . 13 had been previously reported in the literature ., A list of the top 20 independent autosomal hits are presented in Table 1 ., The top 10 independent X chromosome hits are presented in Table 2; rs140488081 and rs7061504 are intronic SNPs in the OPHN1 gene ., After conditioning on the top SNP ( rs73221556 ) , 47 SNPs ( including the two lead X chromosome SNPs from the literature: rs2497938 and rs6625163 ) remained significant at P<5x10-8 ., In the UK Biobank data , the two lead SNPs from the literature were in very high LD ( R2 = 0 . 98 ) ., Summary output for all of the SNPs is available at the following URL: http://www . ccace . ed . ac . uk/node/335 ., A list of the 287 independent loci are reported in S2 Table ., The gene-based analysis identified 112 autosomal genes and 13 X chromosome genes that were associated with baldness after a Bonferroni correction ( P<0 . 05/18 , 061 and P<0 . 05/567 , respectively ) ., The top gene-based hit was , as expected , the androgen receptor on the X chromosome ( P = 2 . 0x10-269 ) ., A full list of the autosomal significant gene-based hits is provided in S3 Table and significant genes on the X chromosome are shown in S4 Table ., A significant enrichment ( FDR <0 . 05 ) was found for 143 gene sets; the full results are presented in S5 Table ., Using common genetic variants with a minor allele frequency of at least 1% , GCTA-GREML analysis found that 47 . 3% ( SE 1 . 3% ) of the variance in baldness can be explained by common autosomal genetic variants , while 4 . 6% ( SE 0 . 3% ) can be explained by common X chromosome variants ., Genetic correlations were examined between male pattern baldness and 24 cognitive , health , and anthropometric traits using LD Score regression ., No significant associations were found; all estimates were close to zero ( S6 Table ) ., The GWAS for self-reported baldness was re-run on a sub-sample of 40 , 000 individuals—retaining an equal proportion of each of the four baldness patterns as observed in the full cohort—to allow a polygenic prediction score to be built and applied to the remaining , independent sample of 12 , 874 individuals ., The most powerful predictions from comparing the extreme phenotype groups were observed at the P<1x10-5 threshold for both the autosomal and X chromosome polygenic scores ( Table 3 ) ., The optimal autosomal polygenic score yielded an AUC of 0 . 75 for discriminating between those with no hair loss ( n = 4 , 123 ) and those with severe hair loss ( n = 2 , 456 ) ., The corresponding AUC for the optimal X chromosome polygenic score was 0 . 62 ., An additive combination of the autosomal and X chromosome polygenic scores gave an AUC of 0 . 78 ( sensitivity = 0 . 74 , specificity = 0 . 69 , PPV = 0 . 59 , NPV = 0 . 82 ) for severe hair loss , 0 . 68 ( sensitivity = 0 . 66 , specificity = 0 . 61 , PPV = 0 . 58 , NPV = 0 . 68 ) for moderate hair loss , and 0 . 61 ( sensitivity = 0 . 64 , specificity = 0 . 53 , PPV = 0 . 49 , NPV = 0 . 68 ) for slight hair loss ( Fig 2 ) ., Adding age as a covariate boosted the AUC to 0 . 79 for severe hair loss ( P<2x10-16 ) , 0 . 70 for moderate hair loss ( P<2x10-16 ) , and 0 . 61 for slight hair loss ( P = 0 . 019 ) ., Fig 3 shows the proportion of participants in the four baldness groups for each polygenic risk decile of male pattern baldness ., Of those with a baldness polygenic score below the median , 14% reported severe hair loss and 39% no hair loss ., By contrast , of those with a polygenic score in the top 10% , 58% reported moderate-to-severe hair loss ., The results of the partitioned heritability analysis indicated that 27 of the functional annotations from the baseline model were statistically significant ( S2 Fig and S7 Table ) ., These significant annotations included a broad array of functional elements including histone marks , enhancer regions , conserved regions , and DNaseI hypersensitivity sites ( DHS ) ., The ten tissue types were then tested for significance after controlling for the baseline model ., Following correction for multiple testing , all ten of the tissue groups showed significant enrichment ( S3 Fig and S7 Table ) ., We identified over two hundred independent , novel genetic correlates of male pattern baldness—an order of magnitude greater than the list of previous genome-wide hits ., Our top SNP and gene-based hits were in genes that have previously been associated with hair growth and development ., We also generated a polygenic predictor that discriminated between those with no hair loss and those with severe hair loss ., Whereas accurate predictions for an individual are still relatively crude , of those with a genetic score in the top 10% of the distribution , 58% reported moderate-to-severe hair loss ., The release of genetic data on the full UK Biobank cohort will further refine these predictions and increase our understanding of the genetic architecture of male pattern baldness ., Data came from the first release of genetic data of the UK Biobank study and analyses were performed under the data application 10279 ., Ethical approval for UK Biobank was granted by the Research Ethics Committee ( 11/NW/0382 ) ., Genotyping details including quality control steps have been reported previously 43 ., Briefly , from the sample with genetic data available as of June 2015 , 112 , 151 participants remained after the following exclusion criteria were applied: SNP missingness , relatedness , gender mismatch , non-British ancestry , and failed quality control for the UK BiLEVE study 43 ., For the current analysis , an imputed dataset was used for the autosomes ( reference set panel combination of the UK10K haplotype and 1000 Genomes Phase 3 panels: http://biobank . ctsu . ox . ac . uk/crystal/refer . cgi ? id=157020 ) ., Imputed data were not available for the X chromosome , hence only genotyped variants were considered ., X chromosome quality control steps included a minor allele frequency cut-off of 1% and a genotyping call rate cut-off of 98% 44 ., For the imputed autosomal data , we restricted the analyses to variants with a minor allele frequency >0 . 1% and an imputation quality score >0 . 1 ., From the sample of 112 , 151 unrelated White British participants with genetic data , we identified 52 , 874 men with a self-reported response to UK Biobank question 2395 , which was adapted from the Hamilton-Norwood scale 45 , 46 ., These men were asked to choose , from four patterns ( no loss; slight loss; moderate loss; severe loss ) , the one that matched their hair coverage most closely ., Fig 4 shows a screenshot of the four options ., A genome-wide association study was conducted using baldness pattern residuals as the dependent variable ., The residuals were obtained from a linear regression model of baldness pattern on age , assessment centre , genotyping batch and array , and 10 principal components to correct for population stratification ., The GWAS for the imputed autosomal dataset was performed in SNPTest v2 . 5 . 1 47 via an additive model , using genotype probability scores ., The GWAS for the X chromosome was performed in Plink 48 , 49 ., The number of independent signals from the GWAS was determined using LD-clumping 48 , 49 based on the LD structure annotated in the 1000 genomes project 50 ., Index SNPs were identified ( P<5x10-8 ) and clumps were formed for SNPs with P<1x10-5 that were in LD ( R2>0 . 1 ) and within 500kb of the index SNP ., SNPs were assigned to no more than one clump ., GWAS lookups were performed for the top hits reported in Richards et al . 10 , Li et al . 15 , Heilmann-Heimbach et al . 16 , Liu et al . 13 and Pickrell et al . 14 ., Gene-based analyses were performed using MAGMA 51 ., SNPs were mapped to genes according to their position in the NCBI 37 . 3 build map ., No additional boundary was added beyond the genes start and stop site ., For the autosomal genes the summary statistics from the imputed GWAS were used to derive gene-based statistics using the 1000 genomes ( phase 1 , release 3 ) to model linkage disequilibrium ., For genes on the X chromosome the genotype data from UK Biobank was used and the gene-based statistic was derived using each participant’s phenotype score ., Gene-set pathway analyses were carried out in DEPICT 52 using the genome-wide significant autosomal SNPs as input ., For the prediction analysis , the GWAS was re-run on a randomly selected cohort of 40 , 000 individuals to give regression weights for prediction , leaving an independent cohort of 12 , 874 in which to test the polygenic predictor ., The methods for the GWAS were identical to those reported for the full sample ., The regression weights from the GWAS on the 40 , 000 cohort were used to construct polygenic scores in the target dataset at P value thresholds of <1x10-20 , <5x10-8 , <1x10-5 , <0 . 01 , <0 . 05 , <0 . 1 , <0 . 2 , <0 . 5 , <1 using PRSice software 53 ., PRSice creates polygenic scores by calculating the sum of alleles associated with male pattern baldness across many genetic loci , weighted by their effect sizes estimated from the male pattern baldness GWAS ., Prior to calculating the scores , SNPs in the prediction dataset were clumped across 250kb sliding windows at an R2>0 . 25 ., Thereafter , each threshold was used to discriminate between those with no hair loss and those with severe hair loss via logistic regression with results being reported for the optimal predictor only ., A predictor for both the autosomes and X chromosome were built and assessed independently and additively ., Receiver operator characteristic ( ROC ) curves were plotted and areas under the curve ( AUC ) were calculated using the pROC package in R 54 , 55 ., SNP-based heritability of baldness was estimated using GCTA-GREML 56 after applying a relatedness cut-off of >0 . 025 in the generation of the autosomal ( but not X chromosome ) genetic relationship matrix ., Linkage disequilibrium score ( LDS ) regression analyses 57 were used to generate genetic correlations between baldness and 24 cognitive , anthropometric , and health outcomes , where phenotypic correlations or evidence of shared genetic architecture have been found ( S7 Table ) ., Due to the large effects in the APOE region for Alzheimers disease , 500kb was removed from around each side of this region and the analysis was repeated for the Alzheimers—male pattern baldness analysis ., The Alzheimer’s data set without this region is referred to as Alzheimer’s 500kb ., In total , we carried out 25 hypothesis tests ., Multiple testing was controlled for using a false discovery rate ( FDR ) correction 58 ., An overview of the GWAS summary data for the anthropometric and health outcomes is provided in S1 Appendix ., Stratified linkage disequilibrium score ( SLDS ) regression 59 was used to determine if a specific group of SNPs made a greater contribution to the heritability of male pattern baldness than would be expected by the size of the SNP set ., Firstly , a baseline model was derived using 52 overlapping , functional categories ., Secondly , a cell-specific model was constructed by adding each of the 10 cell-specific functional groups to the baseline model and the level of enrichment was obtained ., Multiple testing was controlled for using FDR correction 58 in both the functional category and cell-specific analysis .

Introduction, Results, Discussion, Methods

Male pattern baldness can have substantial psychosocial effects , and it has been phenotypically linked to adverse health outcomes such as prostate cancer and cardiovascular disease ., We explored the genetic architecture of the trait using data from over 52 , 000 male participants of UK Biobank , aged 40–69 years ., We identified over 250 independent genetic loci associated with severe hair loss ( P<5x10-8 ) ., By splitting the cohort into a discovery sample of 40 , 000 and target sample of 12 , 000 , we developed a prediction algorithm based entirely on common genetic variants that discriminated ( AUC = 0 . 78 , sensitivity = 0 . 74 , specificity = 0 . 69 , PPV = 59% , NPV = 82% ) those with no hair loss from those with severe hair loss ., The results of this study might help identify those at greatest risk of hair loss , and also potential genetic targets for intervention .

Living with male pattern baldness can be stressful and embarrassing ., Previous studies have shown baldness to have a complex genetic architecture , with particularly strong signals on the X chromosome ., However , these studies have been limited by small sample sizes ., Here , we present the largest genome-wide study of baldness to date , using data from over 52 , 000 male participants in the UK Biobank study ., We identify over 200 novel findings ., We also split our dataset in two to build and apply a genetic predictor of baldness ., Of those with a polygenic score below the median , 14% had severe hair loss and 39% no hair loss ., By contrast , of those with a polygenic score in the top 10% , 58% reported moderate-to-severe hair loss .

genome-wide association studies, medicine and health sciences, hair growth, integumentary system, physiological processes, genome analysis, genetic linkage, sex chromosomes, hair, chromosome biology, x chromosomes, genetic loci, cell biology, anatomy, heredity, physiology, genetics, biology and life sciences, genomics, genetics of disease, computational biology, chromosomes, human genetics

journal.ppat.1003061

Influenza Infection Rates, Measurement Errors and the Interpretation of Paired Serology

Each year , seasonal influenza is responsible for about three to five millions severe illnesses and about 250 , 000 to 500 , 000 deaths worldwide 1 ., These epidemics can generate important economic losses due to high levels of worker absenteeism as well as a saturation of emergency services at the peak of the epidemic 1 ., In addition , avian or swine influenza viruses occasionally adapt to humans and generate influenza pandemics like in 1918 , 1957 , 1968 and 2009 , sometimes with catastrophic consequences like in 1918 , when 20 to 50 million people died worldwide ., Appropriate assessment of the epidemiological characteristics of the influenza virus is important to guide control policies ., In particular , this requires being able to track the number of influenza cases with severe clinical outcomes ( i . e . the tip of the severity pyramid ) as well as the total number of people infected by an influenza virus ( i . e . the base of the severity pyramid ) ., For example , the case fatality ratio ( proportion of influenza cases who die ) is a key measure of severity that informs decision making during influenza pandemics , and takes the number of influenza related death as numerator and the number of influenza cases as denominator ., Estimates of infection attack rates are also essential for characterizing the spread of the virus in human populations in order to predict epidemic trajectory , the potential impact of control measures such as social distancing measures , and the likelihood and magnitude of subsequent epidemics arising from continued circulation of the same virus 2 , 3 ., Although it is usually possible to estimate the number of severe influenza cases from sentinel surveillance ( e . g . based on data collected at medical practices , clinics or hospitals ) , it is much harder to estimate the total number of people infected by an influenza virus ., First , a substantial proportion of influenza infections are asymptomatic 4 , 5 ., Second , among those with symptoms , only a proportion seek healthcare; and this proportion may vary from season to season or even during the course of an epidemic ., Last , Influenza-Like-Illness ( ILI ) symptoms are not specific to influenza ., So , a substantial proportion of patients consulting for ILI may not have been infected by an influenza virus ., Serological studies have become the gold standard approach for estimating influenza infection attack rates due to the difficulty of estimating infection rates by other means ., Although cross-sectional serological surveys can provide valuable and timely information , paired blood samples collected before and after an epidemic in a cohort of individuals is the optimal approach for precisely assessing infection rates ., The haemagglutination-inhibition ( HI ) assay remains the most commonly used approach for detecting serological evidence of recent influenza infection 6–12 ., The assay detects the presence of antibodies that prevent the haemagglutinin protein of the influenza virus from agglutinating red blood cells 13 , 14 ., For each serum sample , antibody titers are expressed as the reciprocal of the highest serum dilution that can still prevent a fixed concentration of virus from agglutinating red blood cells ., A rise in antibody titers between the first and second blood is taken as a marker of infection ., However , because the procedure is susceptible to measurement errors , a 2 fold rise ( that is a 1-dilution increase ) is usually considered as insufficient evidence for infection ., Seroconversion is therefore typically defined as a 4-fold rise ( i . e . a 2-dilutions increase ) or more in antibody titers ., This ad-hoc rule became established when these methods were first developed and is now widely adopted 15 , 16 ., In the meantime , however , statistical methods for addressing measurement errors have made substantial progress ., In particular , there is now an extensive body of literature on methods to ensure that the presence of measurement errors does not bias estimates of key parameters of interest ., Given these developments , it is timely to revisit the way serological data are interpreted ., Central to the traditional approach to analyzing serological data is the belief that data about 2-fold rises provide no information since such increases can be caused by frequent measurement errors ., This concern about measurement errors is certainly relevant when trying to make specific diagnoses for individual cases ., For example , one may be averse to the risk of false positives; but less so to the risk of false negatives ., However , estimating infection attack rates at the population level is a very different aim from setting up a specific diagnostic tool , and may benefit from a different use of the data ., First , it is important to note that estimating infection attack rates is not just a matter of specificity ( i . e . ensuring that subjects satisfying the diagnostic definition of infection were indeed infected by an influenza virus ) but also a matter of sensitivity ( i . e . ensuring that all subjects infected are diagnosed as such ) ., An approach that favours specificity over sensitivity may lead to underestimating infection attack rates ., A second important observation is that , even in a context of frequent 2-fold errors , data about 2-fold rises may still be informative ., Consider for example a situation where all individuals exhibit a 2-fold rise during the season: such a pattern cannot be explained by measurement error alone since measurement errors are made both at baseline and post-epidemic and should be about equally distributed provided the sample size is sufficiently large ., Here , we explore how modern statistics for the analysis of data with measurement errors can change and improve our interpretation of serology ., We present a new method to quantify errors in the measurement of antibody titers and to estimate the true distribution of paired serological measurements corrected for measurement errors ., The methodology is applied to data collected in a cohort study conducted in Vietnam between 2007 and 2009 ., We estimate that the 1-sided probability of a 2-fold error was 9 . 3% ( 95% CI: 3 . 3% , 17 . 6% ) when the true antibody titer was below detection levels , rising to 20 . 2% ( 95% CI: 15 . 9% , 24 . 0% ) otherwise ( posterior probability that latter larger than former: 98 . 7% ) ., There was a satisfying fit of the model to replicate measurement data ( Figure 1 ) ., The model where measurement errors were independent of true antibody titers failed to fit the data ( Figure S2 and Supplementary Material ) ., Figure 2 summarizes the distribution of paired serology , corrected for measurement errors for the different seasons ( 2008 , Spring 2009 , Autumn 2009 ) and subtypes ( H1N1 , H3N2 and B ) ., A range of observations can be made ., The first observation concerns 2-fold rises in antibody titers between baseline and post serology ( yellow bars ) ., Such increases are usually ignored in analyses because 2-fold errors are common ., In some instances , like for example subtypes H3N2 and B in 2008 and H1N1pdm09 in Autumn 2009 , 2-fold rises appeared negligible and at levels that could be generated by measurement errors alone , since 0 was within the 95% CI of the estimated proportion of subjects having a 2-fold rise ( Figures 2B , 2C , 2G ) ., In other instances , however , the proportion of individuals experiencing a 2-fold rise ranged from 20% to 33% with lower bounds of the 95% CIs above 0 ( range: 7%–23% ) , indicating that these rises cannot be solely explained by measurement errors ., Assuming that most of these 2-fold rises were due to infection , our estimate of infection attack rates for H1N1 in 2008 and H1N1 , H3N2 and B in Spring 2009 would be dramatically higher than traditional estimate based on 4-fold rises or more ( Figure 3A ) ., So , even if only a proportion of the 2-fold rises were due to influenza infections , the traditional estimate might still represent a substantial underestimate of the true infection attack rates The fact that and were very similar for H3N2 and B in 2008 and virtually identical for H1N1pdm09 in Autumn 2009 ( Figure 3A ) highlights important heterogeneities in the way antibody titers increase by season/subtype ( Figure 3B ) ., For example , for H1N1pdm09 in Autumn 2009 , almost all those experiencing a rise in antibody titers exhibited a 4-fold rise or more; but for H1N1 in 2008 , most of those experiencing a rise only had a 2-fold increase ., The absence of a simple linear relationship between and the proportion of 2-fold rises suggests that the standard approach of inflating by a fixed proportion ( generally equal to the proportion of PCR positive cases who do not seroconvert; around 10–20% ) to get corrected estimates of infection attack rates may be inappropriate ., Rather , corrections might have to be applied on a season-to-season and subtype-to-subtype basis ., The last notable observation is that decay in antibody titers is observed ., For example , 30% ( 95% CI: 22 , 36 ) of individuals exhibited a decay for subtype H3N2 in 2008 ., Figure 4 shows the observed rise in antibody titers for PCR positive cases ., Twenty seven percent of these cases experienced no rise or only a 2-fold rise in titer during the season ., This again suggests that the case definition of a 4-fold rise or more may underestimate attack rates by at least 27% ., PCR positive cases with low baseline titers experienced an average increase significantly larger than those with higher baseline titers ( p\u200a=\u200a0 . 026 ) ( Figure 4 ) 17 , 18 ., Simulations were run to test the hypothesis of an absence of cross-reactivity between subtypes H1N1 , H3N2 and B in 2008 and Spring 2009 ( see Supplementary Material ) ., We found that there was good adequacy between the data and patterns that would be obtained in the absence of cross-reactivity ., The hypothesis of an absence of cross-reactivity could therefore not be rejected ( Figure S3 ) ., Figure 5 compares the distribution of observed paired serology as observed in the data ( black point ) and as predicted by the model ., Model fit was satisfactory ., In a simulation study , we found that estimates of parameters characterizing measurement errors were unbiased ( Table 1 ) , as well as those characterizing the selection process ( Table S2 ) ., We also found that estimates of the proportion of subjects with an antibody titer increase ( empirical absolute bias: 0 . 1% ) , of the proportion of subjects with an antibody titer decay ( empirical absolute bias: 0 . 0% ) and of the probabilities characterizing jointly baseline antibody titers and the change in antibody titers during a season ( empirical absolute bias: 0 . 0% ) were unbiased ( Figure 6 ) ., Our statistical model describes the distribution of paired serology across all subjects ., However , since we infer true paired serology for each individual , it is possible to reconstruct a posteriori the distribution of true paired serology for the different age groups ., The age-specific distributions for true paired serology are presented in Figure S4 ., Interesting differences can be noticed between age groups ., For example and consistent with the literature , for H1N1pdm09 in Autumn 2009 , the proportion of 4-fold rises falls from 39% ( 95% CI: 37% , 39% ) in <18 y . o . to 15% ( 95% CI: 15% , 16% ) in 18–48 y . o . and 8% ( 95% CI: 7 , 9 ) in >48 y . o . For H3N2 in 2009 , the decay in antibody titers was more important among <18 y . o . ( 53%; 95% CI: 38% , 65% ) than among older age groups ( 25% , 95% CI 19% , 30% for 18–48 y . o . and 18% , 95% CI 12 , 22 for >48 y . o . ) ., For H3N2 in Spring 2009 , although the proportions of 4-fold rises were similar across age groups , our analysis suggests that the proportion of 2-fold rises may have been higher among <18 y . o ( 43% , 95% CI: 23 , 58 ) than in other age groups ( 30% , 95% CI 17% , 41% for 18–48 y . o . and 27% , 95% CI 13 , 38 for >48 y . o . ) ., We find that , for each age group , there is a satisfying adequacy between the observed distribution of paired serology and that predicted by the model ( Figure S5 ) ., In this paper , we have revisited the traditional interpretation of paired serological measurements of influenza antibody titers ., Until now , data on 2-fold rises have been largely ignored because of the belief that measurement errors made them unreliable ., Although this may be a valid concern if the aim is to get a specific diagnosis for individual cases , we argue that this is less so when the objective is to interpret antibody titer variations at the population level ., We have shown that it is possible to quantify measurement errors , and to reconstruct the distribution of paired serology corrected for measurement errors ., Our method gave unbiased estimates in a simulation study ., After correction for measurement errors for the Vietnamese data examined here , we found that for some seasons and subtypes the proportions of individuals with 2-fold rises in antibody titers was too large to be explained by measurement errors alone ., Estimates of infection attack rates varied greatly depending on whether or not 2-fold rises were included ., It is therefore important to determine the biological phenomenon that could cause such increases , in particular whether they are caused by exposure to influenza viruses ., A first hypothesis is that 2-fold titer increases are caused by infection by an influenza virus ., In support of this hypothesis , it is clear that a proportion of virologically- or RT-PCR- confirmed influenza cases do not achieve a 4-fold rise in HI titer ., This proportion was 27% in our dataset , similar to a large cohort of confirmed pandemic cases in the US 19 ., However , past work has shown this proportion to be as high as 77% in people who have high pre-existing antibody titers 17 , or as low as 10% in patients seeking medical care for pandemic H1N1 infection in 2009 20 ., It is clear that antibody titer changes following infection vary between individuals and are affected by factors including pre-existing titer and timing of serum collection ., In particular , since there is an upper limit to antibody concentrations , individuals with high pre-existing titers are limited in their ability to generate 4-fold rises and may produce only a 2-fold titer increase in response to infection 15 ., However , the analysis performed here shows that 2 fold titer changes are common even among individuals with low pre-existing titers ., Antibody concentrations reach a peak 4–7 weeks after infection and then decay over a period of around six months to a plateau that is maintained for several years 21 ., Although the profile of HA antibody decay is not well characterised , the probability of detecting 2- or 4- fold rises will vary with the interval following infection ., However , in our data the longest interval between the peak transmission period and blood sampling was in season 3 , when the proportion of 2-fold titer rises was lowest ., A second hypothesis is that 2-fold rises correspond to infection which is attenuated by mucosal or serological antibodies to homologous or heterologous strains , or by innate or cell mediated immunity ., Antibody responses to inactivated influenza vaccines clearly demonstrate the potential for antigenic stimulation without active infection and the phenomenon of boosting of immunity in exposed yet uninfected individuals is well documented for other viruses ( e . g . varicella zoster 22 ) ., A third hypothesis is that 2-fold rises are an artefact unrelated to influenza infection or exposure ., Seasonal variation in titres independent of infection might result from the presence of non-specific inhibitors of agglutinination ., For example , this could happen if the circulation of other viruses boosted the immune system , leading to small increases in all antibody titers ., In such a scenario , one might expect the effect to be similar on the different subtypes ., However , in 2007 , a large proportion of individuals exhibited 2-fold increases for H1N1 but not for H3N2 or B , suggesting that this hypothesis is not strongly supported by the data ., It is also important to understand why 2-fold titers changes were prominent during some seasonal influenza epidemics but not during the pandemic ., One possibility may be that there was greater antigenic mismatch for some seasonal strains because of unrecognised co-circulation of different influenza strains from those used as antigens in the HI assay ., In this situation , anti-HA antibodies generated by infection have lower avidity for the HA of the assay virus ., Conversely , original antigenic sin , where an infection results in an anamnestic response and the generation of antibodies directed towards an earlier infecting strain , might also explain 2-fold titer rises in response to infection 17 ., In all these scenarios however , 2-fold increases would still represent infection by an influenza virus ., It is unlikely that 2-fold increases represent cross-reactivity of HI antibodies to strains of one subtype with strains of other subtypes ., This is confirmed by our analysis that did not reject the hypothesis of an absence of cross-reactivity between subtypes ., It is therefore important for future work to determine if 2-fold titer increases represent infection , antigenic stimulation ( attenuated infection ) , or artefact ., If influenza infection rates are higher than currently recognised this might change our understanding of influenza transmission and of intra-host and inter-host immune mediated evolutionary pressures , and may have implications for the feasibility of control measures ., In the dataset examined here , 2-fold increases exceeded 4-fold increases for H1N1 in 2008 and H1N1 , H3N2 and B in Spring 2009 ., There was no clear pattern with respect to subtype or strain ., The seasonal H1N1 strain circulating in 2008 ( A/Brisbane/59/2007 ) was antigenically distinct from those circulating previously ( A/Solomon Islands/03/2006 and A/New Caledonia/20/1999-like ) , but this strain continued to circulate in Spring 2009 ., The seasonal H3N2 strain circulating in Spring 2009 ( A/Perth/16/2009 ) was antigenically distinct from the 2007/8 strain ( A/Brisbane/10/2007 ) ., H3N2 A/Perth/16/2009-like viruses have been difficult to propagate and we had difficulty propagating sufficient virus for the HI assays using A/Perth/16/2009-like viruses isolated from the cohort during the Spring 2009 season ., We therefore used a virus isolated from a patient in Hanoi by the National Influenza Center , and propagated in eggs followed by MDCK cells ( TX265M2E1 ) for undertaking HI testing of sera collected in Spring 2009 ., It is possible that the propagation in eggs this virus underwent might have resulted in some antigenic change , resulting in lower titers in the HI assay ., National influenza surveillance data indicates that both influenza B lineages - Yamagata and Victoria- co-circulated during the study period , with the Yamagata lineage dominating in 2007 and 2008 and the Victoria lineage in 2009 ., For all HI assays , we used the same influenza B virus , which was isolated in 2008 and was characterized antigenically as Yamagata lineage-like , as with all influenza B viruses isolated from the cohort in 2008 ., While Yamagata viruses dominated the influenza B samples we collected in 2007 and 2008 , the Victoria lineage was predominant in 2009 ., This may be a factor explaining the lower influenza B titer increases seen in that year ., If heterogeneities in the proportion of 2-fold titer rises are largely attributable to a poor match between assay antigen and infecting virus , future seroprevalence and seroincidence surveys will need to use a greater diversity of antigens than typically used currently ., There are often strong age-related patterns in influenza serology ., Ideally , we would therefore like to fit our statistical model independently for each age group ., However , simulation studies indicate that the relatively small number of observations per age group would lead to relatively inaccurate estimates ., We have therefore opted for an intermediate estimation strategy ., Our statistical model fits a single distribution of true paired serology to all subjects; but since we infer true paired serology for each individual , we can reconstruct a posteriori the distribution of true paired serology for the different age groups ., Even with such a conservative approach ( i . e . it favours scenarios where the different age groups exhibit similar distributions ) , we were able to detect clear age-related patterns ., In particular , it indicated that age may be another factor that influences the occurrence of a 2-fold rise ., Larger sample sizes will be needed to investigate this possibility further ., The presence of relatively large proportions of individuals experiencing a 2-fold increase in antibody titers is not a peculiarity of the Vietnamese data examined here ., Similar shifts were observed on data gathered by Cowling et al , with micro-neutralization assays for 2009 H1N1pdm09 influenza and on HI assays for seasonal influenza 23 ( Figure S6 ) ., It is well known that there may be substantial within- and between- laboratory variability in HI assays as well as in other serological assays such as virus neutralisation ( VN ) 24 ., The level of intra-laboratory variations may depend on both the laboratory and the type of assay used 24 ., Here , we have introduced an approach that allows controlling for within-laboratory variations ., The only additional data needed compared with standard serological surveys is that replicate measurements are performed for a subset of subjects ., These replicate measurements allow within-laboratory quantification of variation in assay performance ., With this information , it is then possible to reconstruct the distribution of paired serology that is corrected for the estimated level of within-laboratory variations ., Although our approach gives a better control on within-laboratory variation , it does not address the problem of between-laboratory variation ., The use of standards in bioassays is critical for minimising the impact of the latter problem 24 ., To conclude , while a 4-fold titer increase may be a highly specific diagnostic of infection by an influenza virus for individual cases , this criterion is less justifiable when the objective is to estimate community ARs ., Our work shows that requiring a 4-fold titer increase may lead to ARs being substantially underestimated ., More research is needed to determine what proportion of 2-fold rises are causally linked to exposure to influenza , and what proportion may be caused by other mechanisms ., It will be important to determine whether the high proportion of 2-fold titer increases seen in the settings of Vietnam and Hong Kong 23 are also observed in other ( e . g . temperate climate ) settings ., Samples were collected from a household-based cohort of 940 participants in 270 households in a single community in semi-rural northern Vietnam as previously described 5 ., None of the participants had ever received influenza immunisation ., Participants were under weekly active surveillance by village health workers for influenza-like-illness ( ILI ) and in the event of an ILI were asked to provide a nose and throat swab for detection of influenza RNA by reverse-transcription polymerase chain reaction ., Participants were also asked to provide serial blood samples at times when national influenza surveillance data indicated that influenza circulation was minimal ., The samples described here were collected over a period of three consecutive influenza seasons , from December 2007 through April 2010 ., The bleeding times were 1st–7th December 2007 ( bleed 1 ) , 9th–15th December 2008 ( bleed 2 ) , 2nd–4th June 2009 ( bleed 3 ) , and on the 3rd April 2010 ( bleed 4 ) ., This provided three sets of paired samples either side of an influenza transmission season: 548 paired samples for season 1 ( 2008 ) , 501 paired samples for season 2 ( Spring 2009 ) , and 540 paired samples for season 3 ( Autumn 2009 ) ., In season 1 , the influenza A virus strains detected in the cohort through ILI surveillance were A/H1N1/Brisbane/59/2007-like and A/H3N2/Brisbane/10/2007-like; in season 2 , they were A/H1N1/Brisbane/59/2007-like and A/H3N2/Perth/16/2009-like; and in season 3 , it was A/H1N1/California/7/2009-like ., There was co-circulation of influenza B Yamagata lineage and Victoria lineage in both season 1 and season 2 , with a predominance of Yamagata lineage in season 1 and Victoria lineage in season 2 ., Nasal and oropharangeal swabs were assessed by real-time reverse-transcriptase polymerase chain reaction ( RT-PCR ) , according to WHO/USCDC protocols 25 ., Influenza hemagglutination inhibition ( HI ) assays were performed according to standard protocols WHO 2011 manual ., The seasonal influenza A viruses used were isolated from participants swabs or from swabs taken from patients presenting in Ha Noi in the same season and propagated in embryonated hens eggs or in MDCK cells ., A reference antigen supplied by WHO ( A/H1N1/California/7/2009-like ) was used to assess season 3/pandemic sera ., A single influenza B virus isolated from a participant during 2008 was used to assess serum for both the first and second seasons ., The virus had a titer of 320 with B/Wisconsin/1/2010 ( Yamagata ) reference antisera and of <10 with B/Brisbane/60/2008 ( Victoria ) antisera ., Each virus was first assessed for haemagglutination of erythrocytes from chickens , guinea pigs and turkeys then titrated with optimal erythrocytes ., Serum was treated with receptor destroying enzyme ( Denka Seiken , Japan ) then heat inactivated and adsorbed against packed erythrocytes ., Eight 2-fold dilutions of serum were made starting from 1∶10 and incubated with 4 HA units/25 µl of virus ., Appropriate erythrocytes were added and plates read when control cells had settled ., Virus , serum and positive controls were included in each assay ., Pre- and post-season sera were tested in pairs ., Each serum was tested in a single dilution series ., The HI titre was read as the reciprocal of the highest serum dilution causing complete inhibition of RBC agglutination , partial agglutination was not scored as inhibition of agglutination ., If there was no inhibition of HI at the highest serum concentration ( 1∶10 dilution ) the titer was designated as 5 ., Only one sample had a titer >1280 and this was not adjusted ., Replicate HI assay measurements were performed on a subset of samples from patients that seroconverted ( i . e . 4-fold rise in titer ) as well as some others that had titers ≥20 in both pre and post-season sera ., A less technical description of statistical methods is given for non-specialists in Box 1 and Figure 7 ., The research was approved by the institutional review board of the National Institute of Hygiene and Epidemiology , Vietnam; the Oxford Tropical Research Ethics Committee , University of Oxford , UK; and the Ethics Committee of the London School of Hygiene and Tropical Medicine , UK ., All participants provided written informed consent .

Introduction, Results, Discussion, Materials and Methods, Ethics statement

Serological studies are the gold standard method to estimate influenza infection attack rates ( ARs ) in human populations ., In a common protocol , blood samples are collected before and after the epidemic in a cohort of individuals; and a rise in haemagglutination-inhibition ( HI ) antibody titers during the epidemic is considered as a marker of infection ., Because of inherent measurement errors , a 2-fold rise is usually considered as insufficient evidence for infection and seroconversion is therefore typically defined as a 4-fold rise or more ., Here , we revisit this widely accepted 70-year old criterion ., We develop a Markov chain Monte Carlo data augmentation model to quantify measurement errors and reconstruct the distribution of latent true serological status in a Vietnamese 3-year serological cohort , in which replicate measurements were available ., We estimate that the 1-sided probability of a 2-fold error is 9 . 3% ( 95% Credible Interval , CI: 3 . 3% , 17 . 6% ) when antibody titer is below 10 but is 20 . 2% ( 95% CI: 15 . 9% , 24 . 0% ) otherwise ., After correction for measurement errors , we find that the proportion of individuals with 2-fold rises in antibody titers was too large to be explained by measurement errors alone ., Estimates of ARs vary greatly depending on whether those individuals are included in the definition of the infected population ., A simulation study shows that our method is unbiased ., The 4-fold rise case definition is relevant when aiming at a specific diagnostic for individual cases , but the justification is less obvious when the objective is to estimate ARs ., In particular , it may lead to large underestimates of ARs ., Determining which biological phenomenon contributes most to 2-fold rises in antibody titers is essential to assess bias with the traditional case definition and offer improved estimates of influenza ARs .

Each year , seasonal influenza is responsible for about three to five million severe illnesses and about 250 , 000 to 500 , 000 deaths worldwide ., In order to assess the burden of disease and guide control policies , it is important to quantify the proportion of people infected by an influenza virus each year ., Since infection usually leaves a “signature” in the blood of infected individuals ( namely a rise in antibodies ) , a standard protocol consists in collecting blood samples in a cohort of subjects and determining the proportion of those who experienced such rise ., However , because of inherent measurement errors , only large rises are accounted for in the standard 4-fold rise case definition ., Here , we revisit this 70 year old and widely accepted and applied criterion ., We present innovative statistical techniques to better capture the impact of measurement errors and improve our interpretation of the data ., Our analysis suggests that the number of people infected by an influenza virus each year might be substantially larger than previously thought , with important implications for our understanding of the transmission and evolution of influenza – and the nature of infection .

medicine, public health and epidemiology, epidemiology, infectious disease epidemiology

journal.pcbi.1004254

Monte Carlo Planning Method Estimates Planning Horizons during Interactive Social Exchange

Successful social interactions require individuals to understand the consequences of their actions on the future actions and beliefs of those around them ., To map these processes is a complex challenge in at least three different ways ., The first is that other peoples’ preferences or utilities are not known exactly ., Even if the various components of the utility functions are held in common , the actual values of the parameters of partners , e . g . , their degrees of envy or guilt 1–6 , could well differ ., This ignorance decreases through experience , and can be modeled using the framework of a partially observable Markov decision process ( POMDP ) ., However , normal mechanisms for learning in POMDPs involve probing or running experiments , which has the potential cost of partners fooling each other ., The second complexity is represented by characterizing the form of the model agents have of others ., In principle , agent A’s model of agent B should include agent B’s model of agent A; and in turn , agent B’s model of agent A’s model of agent B , and so forth ., The beautiful theory of Nash equilibria 7 , extended to the case of incomplete information via so-called Bayes-Nash equilibria 8 dispenses with this so-called cognitive hierarchy 9–12 , looking instead for an equilibrium solution ., However , a wealth of work ( see for instance 13 ) has shown that people deviate from Nash behaviour ., It has instead been proposed that people model others to a strictly limited , yet non-negligible , degree 9 , 12 ., The final complexity arises when we consider that although it is common in experimental economics to create one-shot interactions , many of the most interesting and richest aspects of behaviour arise with multiple rounds of interactions ., Here , for concreteness , we consider the multi round trust task , which is a social exchange game that has been used with hundreds of pairs ( dyads ) of subjects , including both normal and clinical populations 14–18 ., This game has been used to show that characteristics that only arise in multi-round interactions such as defection ( agent A increases their cooperation between two rounds; agent B responds by decreasing theirs ) have observable neural consequences that can be measured using functional magnetic resonance imaging ( fMRI ) 16 , 19–22 ., The interactive POMDP ( IPOMDP ) 23 is a theoretical framework that formalizes many of these complexities ., It characterizes the uncertainties about the utility functions and planning over multiple rounds in terms of a POMDP , and constructs an explicit cognitive hierarchy of models about the other ( hence the moniker ‘interactive’ ) ., This framework has previously been used with data from the multi-round trust task 22 , 24 ., However , solving IPOMDPs is computationally extremely challenging , restricting those previous investigations to a rather minuscule degree of forward planning ( just two- out of what is actually a ten-round interaction ) ., Our main contribution is the adaptation of an efficient Monte Carlo tree search method , called partially observable Monte Carlo planning ( POMCP ) to IPOMDP problems ., Our second contribution is to illustrate this algorithm through examination of the multiround trust task ., We show characteristic patterns of behaviour to be expected for subjects with particular degrees of inequality aversion , other-modeling and planning capacities , and consider how to invert observed behaviour to make inferences about the nature of subjects’ reasoning capacities ., A Markov decision process ( MDP ) 25 is defined by sets 𝓢 of “states” and 𝓐 of “actions” , and several components that evaluate and link the two , including transition probabilities 𝓣 , and information ℛ about possible rewards ., States describe the position of the agent in the environment , and determine which actions can be taken , accounting for , at least probabilistically , the consequences for rewards and future states ., Transitions between states are described by means of a collection of transition probabilities 𝓣 , assigning to each possible state s ∈ 𝓢 and each possible action a ∈ 𝓐 from that state , a transition probability distribution or measure 𝓣 s s ^ a = 𝓣 ( s ^ , a , s ) : = ℙ s ^ ∣ s , a which encodes the likelihood of ending in state s ^ after taking action a from state, s . The Markov property requires that the transition ( and reward probabilities ) only depend on the current state ( and action ) , and are independent from the past events ., An illustration of these concepts can be found in Fig 1 . By contrast , in a partially observable MDP ( i . e . , a POMDP 26 ) , the agent can also be uncertain about its state, s . Instead , there is a set of observations o ∈ 𝓞 that incompletely pin down states , depending on the observation probabilities 𝓦 s ^ o a = 𝓦 ( o , a , s ^ ) : = ℙ o ∣ s ^ , a ., These report the probability of observing o when action a has occasioned a transition to state s ^ ., See Fig 2 for an illustration of the concept ., We use the notation st = s , at = a or ot = o to refer explicitly to the outcome state , action or observation at a given time ., The history h ∈ ℋ is the sequence of actions and observations , wherein each action from the point of view of the agent moves the time index ahead by 1 , ht: = {o0 , a0 , o1 , a1 , … , at−1 , ot} ., Here o0 may be trivial ( deterministic or empty ) ., The agent can perform Bayesian inference to turn its history at time t into a distribution ℙSt = st∣ht over its state at time t , where St denotes the random variable encoding the uncertainty about the current state at time, t . This distribution is called its belief state 𝓑 ( ht ) , with ℙ𝓑 ( ht ) St = st: = ℙSt = st∣ht ., Inference depends on knowing 𝓣 , 𝓦 and the distribution over the initial state S0 , which we write as 𝓑 ( h0 ) ., Information about rewards ℛ comprises a collection of utility functions r ∈ ℛ , r:𝓐 × 𝓢 × 𝓞 → ℝ , a discount function Γ ∈ ℛ , Γ:ℕ → 0 , 1 and a survival function H ∈ ℛ , H:ℕ × ℕ → 0 , 1 ., The utility functions determine the immediate gain associated with executing action a at state s and observing o ( sometimes writing rt for the reward following the tth action ) ., From the utilities , we define the reward function R: 𝓐 × 𝓢 → ℝ , as the expected gain for taking action a at state s as R ( a ,, s ) = 𝔼r ( a , s , o ) , where this expectation is taken over all possible observations o ., Since we usually operate on histories , rather than fixed states , we define the expected reward from a given history h as R ( a , h ) : = ∑s ∈ 𝓢 R ( a ,, s ) ℙs∣h ., The discount function weights the present impact of a future return , depending only on the separation between present and future ., We use exponential discounting with a fixed number γ ∈ 0 , 1 to define our discount function:, Γ ( τ - t ) = γ τ - t ∀ τ , t ∈ ℕ , τ ≥ t ., ( 1 ) Additionally , we define H such that H ( τ ,, t ) is 0 for τ > K and 1 otherwise ., K in general is a random stopping time ., We call the second component t the reference time of the survival function ., The survival function allows us to encode the planning horizon of an agent during decision making: If H ( τ ,, t ) is 0 for τ−t > P , we say that the local planning horizon at t is less than or equal to P . The policy π ∈ Π , π ( a , h ) : = ℙa∣h is defined as a mapping of histories to probabilities over possible actions ., Here Π is called the set of admissible policies ., For convenience , we sometimes write the distribution function as π ( h ) ., The value function of a fixed policy π starting from present history ht is, V π ( h t ) : = ∑ τ = t ∞ γ τ - t H ( τ , t ) 𝔼 r τ | π , h τ ( 2 ), i . e ., , a sum of the discounted future expected rewards ( note that hτ is a random variable here , not a fixed value ) ., Equally , the state-action value is, Q π ( a , h t ) : = R ( a , h t ) + ∑ τ = t + 1 ∞ γ τ - t H ( τ , t ) 𝔼 r τ | π , h τ ., ( 3 ) Definition 1 ( Formal Definition—POMDP ) ., Using the notation of this section , a POMDP is defined as a tuple ( 𝓢 , 𝓐 , 𝓞 , 𝓣 , 𝓦 , ℛ , Π , 𝓑0 ) of components as outlined above ., Convention 1 ( Softmax Decision Making ) ., A wealth of experimental work ( for instance 27–29 ) has found that the choices of humans ( and other animals ) can be well described by softmax policies based on the agent’s state-action values , to encompass the stochasticity of observed behaviour in real subject data ., See 30 , for a behavioural economics perspective and 10 for a neuroscience perspective ., In view of using our model primarily for experimental analysis , we will base our discussion on the decision making rule: π ( a , h ) = ℙ a | h = e β Q π ( a , h ) ∑ b ∈ 𝓐 e β Q π ( b , h ) ( 4 ) where β > 0 is called the inverse temperature parameter and controls how diffuse are the probabilities ., The policy π ( a , h ) ={ 1ifQπ ( a , h ) =max{Qπ ( b , h ) |b∈A} ( assumingthisisunique ) 0otherwise ( 5 ) can be obtained as a limiting case for β → ∞ ., Convention 2 . From now on , we shall denote by Q ( a , h ) , the state-action value Qπ ( a , h ) with respect to the softmax policy ., POMCP was introduced by 31 as an efficient approximation scheme for solving POMDPs ., Here , for completeness , we describe the algorithm; later , we adapt it to the case of an IPOMDP ., POMCP is a generative model-based sampling method for calculating history-action values ., That is , it builds a limited portion of the tree of future histories starting from the current ht , using a sample-based search algorithm ( called upper confidence bounds for trees ( UCT ) ; 32 ) which provides guarantees as to how far from optimal the resulting action can be , given a certain number of samples ( based on results in 33 and 34 ) ., Algorithm 1 provides pseudo code for the adapted POMCP algorithm ., The procedure is presented schematically in Fig 3 . The algorithm is based on a tree structure T , wherein nodes T ( h ) = ( N ( h ) , Q ˜ ( h ) , 𝓑 ( h ) ) represent possible future histories explored by the algorithm , and are characterized by the number N ( h ) of times history h was visited in the simulation , the estimated value Q ˜ ( h ) for visiting h and the approximate belief state 𝓑 ( h ) at, h . Each new node in T is initialized with initial action exploration counts N ( h , a ) = 0 for all possible actions a from h and an initial action value estimate Q ˜ ( h , a ) = 0 for all possible actions a from h and an empty belief state 𝓑 ( h ) = ∅ ., The value N ( h ) is then calculated from all actions counts from the node N ( h ) = ∑a ∈ 𝓐 N ( h , a ) ., Q ˜ ( h ) denotes the mean of obtained values , for simulations starting from node, h . 𝓑 ( h ) can either be calculated analytically , if it is computationally feasible to apply Bayes theorem , or be approximated by the so called root sampling procedure ( see below ) ., In terms of the algorithm , the generative model 𝓖 of the POMDP determines ( s′ , o , r ) ∼ 𝓖 ( s , a ) , the simulated reward , observation and subsequent state for taking a at s; s itself is sampled from the current history, h . Then , every ( future ) history of actions and observations h defines a node T ( h ) in the tree structure T , which is characterized by the available actions and their average simulated action values Q ˜ ( a , h ) under the policy SoftUCT at future states ., If the node has been visited for the N ( h ) th time; with action a being taken for the N ( h , a ) th time , then the average simulated value is updated ( starting from 0 ) using sampled simulated rewards R up to terminal time K , when the current simulation/tree traversal ends as:, Q ˜ new ( a , h ) = Q ˜ old ( a , h ) + 1 N ( h , a ) ( R - Q ˜ old ( a , h ) ) ., ( 6 ) The search algorithm has two decision rules , depending on whether a traversed node has already been visited or is a leaf of the search tree ., In the former case , a decision is reached using SoftUCT by defining, SoftUCT ( Q ( . ∣h ) ) Q ( a , h ) : = Q ˜ ( a , h ) + c log N ( h ) N ( h , a ) ℙ a | h = e β ( Q ( a , h ) ) ∑ b e β ( Q ( b , h ) ) ( 7 ), where c is a parameter that favors exploration ( analogous to an equivalent parameter in UCT ) ., If the node is new , a so-called “rollout” policy is used to provide a crude estimate of the value of the leaf ., This policy can be either very simple ( uniform or ε–greedy based on a very simple model ) or specifically adjusted to the search space , in order to optimize performance ., The rollout value estimate together with the SoftUCT exploration rule is the core mechanism for efficient tree exploration ., In this work , we only use an ε–greedy mechanism , as is described in the section on the multi round trust game ., Another innovation in POMCP that underlies its dramatically superior performance is called root sampling ., This procedure allows to form the belief state at later states , as long as the initial belief state 𝓑0 is known ., This means that , although it is necessary to perform inference to draw samples from the belief state at the root of the search tree , one can then use each sample as if it was ( temporarily ) true , without performing inference at states that are deeper in the search tree to work out the new transition probabilities that pertain to the new belief states associated with the histories at those points ., The reason for this is that the probabilities of getting to the nodes in the search tree represent exactly what is necessary to compensate for the apparent inferential infelicity 31–, i . e ., , the search tree performs as a probabilistic filter ., The technical details of the root sampling procedure can be found in 31 ., In the presence of analytically tractable updating rules ( or at least analytically tractable approximations ) , the belief state at a new node can instead be calculated by Bayes’ theorem ., This will also be the case for the multi round trust game below , where we follow the approximate updating rule in 22 ., An Interactive Partially Observable Markov Decision Process ( IPOMDP ) is a multi agent setting in which the actions of each agent may observably affect the distribution of expected rewards for the other agents ., Since IPOMDPs may be less familiar than POMDPs , we provide more detail about them; consult 23 for a complete reference formulation and 35 for an excellent discussion and extension ., We define the IPOMDP such that the decision making process of each agent becomes a standard ( albeit large ) POMDP , allowing the direct application of POMDP methods to IPOMDP problems ., Definition 2 ( Formal Definition—IPOMDP ) ., An IPOMDP is a collection of POMDPs such that the following holds: Agents are indexed by the finite set ℐ ., Each agent i ∈ ℐ is described by a single POMDP ( 𝓢i , 𝓐i , 𝓞i , 𝓣i , 𝓦i , ℛi , Πi , 𝓑 0 i ) denoting its actual decision making process ., We first define the physical state space 𝓢 phys i: an element 𝓢i ∈ 𝓢 phys i is a complete setting of all features of the environment that determine the action possibilities 𝓐i and obtainable rewards ℛi of i for the present and all possible following histories , from the point of view of, i . The physical state space 𝓢 phys i is augmented by the set 𝓓i of models of the partner agents θij ∈ 𝓓i , j ∈ ℐ\\{i} , called intentional models , which are themselves POMDPs θij = ( 𝓢ij , 𝓐ij , 𝓞ij , 𝓣ij , 𝓦ij , ℛij , Πij , 𝓑 0 i j ) ., These describe how agent i believes agent j perceives the world and reaches its decisions ., The possible state space of agent i can be written 𝓢 i = 𝓢 phys i × 𝓓 i and a given state can be written s ˜ i = ( s i , × j θ i j ) , where s i ∈ 𝓢 phys i is the physical state of the environment and θij are the models of the other agents ., Note that the intentional models θij contain themselves state spaces that encode the history of the game as observed by agent j from the point of view of agent, i . The elements of 𝓢i are called interactive states ., Agents themselves act according to the softmax function of history-action values , and assume that their interactive partner agents do the same ., The elements of the definition are summarized in Fig 4 . Convention 3 . We denote by S and S ˜ the random variables , that encode uncertainty about the physical state and the interactive state respectively ., When choosing the set of intentional models , we consider agents and their partners to engage in a cognitive hierarchy of successive mentalization steps 9 , 12 , depicted in Fig 5 . The simplest agent can try to infer what kind of partner it faces ( level 0 thinking ) ., The next simplest agent could additionally try to infer what the partner might be thinking of it ( level 1 ) ., Next , the agent might try to understand their partner’s inferences about the agent’s thinking about the partner ( level 2 ) ., Generally , this would enable a potentially unbounded chain of mentalization steps ., It is a tenet of cognitive hierarchy theory 9 that the hierarchy terminates finitely and for many tasks after only very few steps ( e . g . , Poisson , with a mean of around 1 . 5 ) ., We formalize this notion as follows ., Definition 3 ( A Hierarchy Of Intentional Models ) ., Since models of the partner agent may contain interactive states in which it in turn models the agent i , we can specify a hierarchical intentional structure 𝓓i , l , built from what we call the level l ≥ −1 intentional models 𝓓i , l ., 𝓓i , l is defined inductively from θ i j , - 1 ∈ 𝓓 i , - 1 ⇔ S i j , - 1 = 𝓢 phys i j × { ∅ } ., ( 8 ) This means that any level −1 intentional model reacts strictly to the environment , without holding any further intentional models ., The higher levels are obtained as θ i j , l ∈ 𝓓 i , l ⇔ 𝓢 i j , l = 𝓢 phys i j × 𝓓 i j , l - 1 ., ( 9 ) Here 𝓓ij , l−1 denotes the l−1 intentional models , that agent i thinks agent j might hold of the other players ., These level l−1 intentional models arise by the same procedure applied to the level −1 models that agent i thinks agent j might hold ., Definition 4 ( Theory of Mind ( ToM ) Level ) ., We follow a similar assumption as the so called k-level thinking ( see 12 ) , in that we assume that each agent operates at a particular level li ( called the agent’s theory of mind ( ToM ) level; and which it is assumed to know ) , and models all partners as being at level lj = li−1 ., We chose definition 4 for comparability with earlier work 22 , 24 ., Convention 4 . It is necessary to be able to calculate the belief state in every POMDP that is encountered ., An agent updates its belief state in a Bayesian manner , following an action a t i and an observation o t + 1, i . This leads to a sequential update rule operating over the belief state ℙ S ˜ t i ∣ h t i of a given agent i at a given time t: ℙ S ˜ t + 1 i = s ˜ 1 | { h t i , a t i , o t + 1 i } = η 𝓦 ( o t + 1 i , a t i , s ˜ 1 ) ∑ s ˜ ∈ 𝓢 i 𝓣 ( s ˜ 1 , a t i , s ˜ ) ℙ S ˜ t i = s ˜ | h t i ., ( 10 ), Here η is a normalization constant associated with the joint distribution of transition and observation probability , conditional on s ˜ , s ˜ 1 , o t + 1 i and a t, i . The observation o t + 1 i in particular incorporates any results of the actions of the other agents , before the next action of the given agent ., We note that the above rule applies recursively to every intentional model in the nested structure 𝓓i , as every POMDP has a separate belief state ., This is slightly different from 23 so that the above update is conventional for a POMDP ., Convention 5 . ( Expected Utility Maximisation ) ., The decision making rule in our IPOMDP treatment is based on expected utility as encoded in the reward function ., The explicit formula for the action value Q ( a t i , h t i ) under a softmax policy ( Eq ( 4 ) ) is: Q ( a t i , h t i ) = R ( a t i , h t i ) + ∑ o t + 1 i ∈ 𝓞 ℙ o t + 1 i | { h t i , a t i } ∑ w ∈ 𝓐 i γ ( i ) H ( t + 1 , t ) Q ( w , h t + 1 i | t ) ℙ b | h t + 1 i ., ( 11 ), Here h t + 1 = { h t i , a t i , o t + 1 i } and Q ( b , h t + 1 i ∣ t ) denotes the action value at t+1 with the survival function conditioned to reference time, t . γ ( i ) is the discount factor of agent i , rather than the i-th power ., This defines a recursive Bellman equation , with the value of taking action a t i given history h t i being the expected immediate reward R ( a t i , h t i ) plus the expected value of future actions conditional on a t i and its possible consequences o t + 1 i discounted by γi ., The belief state 𝓑 ( h t i ) allows us to link h t i to a distribution of interactive states and use 𝓦 to calculate ℙ o t + 1 i ∣ {h t i , a t i} , , in particular including the reactions of other agents to the actions of one agent ., We call the resulting policy the “solution” to the IPOMDP ., Our central interest is in the use of the IPOMDP to capture the interaction amongst human agents with limited cognitive resources and time for their exchanges ., It has been noted in 9 that the distribution of subject levels favours rather low values ( e . g . , Poisson , with a mean of around 1 . 5 ) ., In the opposite limit , sufficient conditions are known in which taking the cognitive hierarchy out to infinity for all involved agents allows for at least one Bayes-Nash equilibrium solution ( part II , theorem II , p . 322 of Harsanyi 8 ) and sufficient conditions have been shown in 36 , given which a solution to the infinite hierarchy model can be approximated by the sequence of finite hierarchy model solutions ., A discussion of a different condition can be found in 37; however , this condition does assume a infinite time horizon in the interaction ., In general , as 9 , p . 868 notes , it is not true that the infinite hierarchy solution will be a Nash equilibrium ., For the purposes of computational psychiatry , we find the very mismatches and limitations , that prevent subjects’ strategies to evolve to a ( Bayes ) -Nash equilibrium in the given time frame , to be of particular interest ., Therefore we restrict our attention to quantal response equilibrium like behaviours ( 30 ) , based on potentially inconsistent initial beliefs by the involved agents with ultimately very limited cognitive resources and finite time exchanges ., An IPOMDP is a collection of POMDPs , so POMCP is , in principle , applicable to each encountered POMDP ., However , unlike the examples in 31 , an IPOMDP contains the intentional model POMDPs θij as part of the state space , and these themselves contain a rich structure of beliefs ., So , the state is sampled from the belief state at the root for agent i is an I tuple ( s ^ i , θ ^ i 1 , … , θ ^ i ( ∣ ℐ ∣ − 1 ) ) of a physical state s ^ i and ( ∣ℐ∣−1 ) POMDPs , one for each partner ., ( This is also akin to the random instantiation of players in 8 ) ., Since the θ ^ i j still contain belief states in their own right , it is still necessary to do some explicit inference during the creation of each tree ., Indeed , explicit inference is hard to avoid altogether during simulation , as the interactive states require the partner to be able to learn 23 ., Nevertheless , a number of performance improvements that we detail below still allow us to apply the POMCP method involving substantial planning horizons ., Many social paradigms based upon game theory , including the iterated ultimatum game , prisoners’ dilemma , iterated “rock , paper , scissors” ( for 2 agents ) and the multi round trust game , involve repeated dyads ., In these , each interaction involves the same structure of physical states and actions ( 𝓢phys , 𝓐 ) ( see below ) , and all discount functions are 0 past a finite horizon ., Definition 5 ( Dyadic Repeated Exchange without state uncertainty ) ., Consider a two agent IPOMDP framework in which there is no physical state uncertainty: both agents fully observe each others’ actions and there is no uncertainty about environmental influence; and in which agents vary their play only based on intentional models ., Additionally , the framework is assumed to reset after each exchange ( i . e . , after both agents have acted once ) ., Formally this means: There is a fixed setting ( 𝓢phys , 𝓐 , 𝓣 ) , such that physical states , actions from these states , transitions in the physical state and hence also obtainable rewards , differ only by a changing time index and there is no observational uncertainty ., Then after each exchange the framework is assumed to reset to the same distribution of physical initial states 𝓢phys within this setting ( i . e . the game begins anew ) ., Games of this sort admit an immediate simplification: Theorem 1 ( Level 0 Recombining Tree ) ., In the situation of definition 5 , level 0 action values at any given time only depend on the total set of actions and observations so far and not the order in which those exchanges were observed ., Proof ., The level −1 partner model only acts on the physical state it encounters and the physical state space variable S is reset at the beginning of each round in the situation of 5 . Therefore , given a state s in the current round and an action a by a level 0 agent , the likelihood of each transition to some state s1 , 𝓣 ( s1 , a ,, s ) , and of making observation o , 𝓦 ( o , a , s1 ) , is the same at every round from the point of view of the level 0 agent ., It follows that the cumulative belief update from Eq 10 , from the initial beliefs 𝓑0 to the current beliefs , will not depend on the order in which the action observation pairs ( a , o ) were observed ., This means , that depending on the size of the state space and the depth of planning of interest , we may analytically calculate level 0 action values even online or use precalculated values for larger problems ., Furthermore , because their action values will only depend on past exchanges and not on the order in which they were observed , their decision making tree can be reformulated as a recombining tree ., Sometimes , an additional simplification can be made: Theorem 2 ( Trivialised Planning ) ., In the situation of definition 5 , if the two agents do not act simultaneously and the state transition of the second agent is entirely dependent on the action executed by the first agent ( as in the multi round trust task ) ; and additionally the intentional model of the partner can not be changed through the actions of the second agent , then a level 0 second agent can gain no advantage from planning ahead , since their actions will not change the action choices of the first agent ., Proof ., In the scenario described in theorem 2 the physical state variable S of the agent 2 is entirely dependent on the action of the other agent ., If the agent is level 0 , they model their partner as level −1 and by additional assumption the second agent does not believe that the partner can be made to transition between different intentional models by the second agent’s actions , hence their partner will not change their distribution of state transitions , depending on the agent’s actions and hence also their distribution of future obtainable rewards will not change ., Theorem 3 ( Trivialised Theory of Mind Levels ) ., In the situation of theorem 2 , we state that for the first to go agent , only the even theory of mind levels k ∈ {0}∪2ℕ show distinct behaviours , while the odd levels k ∈ 2ℕ−1 behave like one level below , meaning k−1 ., For the second to go partner equivalently , only the odd levels k ∈ {0}∪2ℕ−1 show distinct behaviours ., Proof ., In the scenario described in theorem 2 , the second to go level 0 agent behaves like a level −1 agent , as it does not benefit from modeling the partner ., This implies that the first to go agent , gains no additional information at the level 1 thinking , since the partner behaves like level −1 , which was modeled by the level 0 first to agent already ., In turn , the level 2 second to go agent gains no additional information over the level 1 second to go agent , as the their partner model does not change between modeling the partner at level 0 or level −1 ., By induction , we get the result ., Examples of the additional simplifications in theorems 2 and 3 can be seen in the ultimatum game and the multi round trust game ., The multi-round trust task , illustrated in Fig 6 is a paradigm social exchange game ., It involves two people , one playing the role of an ‘investor’ the other the one of a ‘trustee’ , over 10 sequential rounds , expressed by a time index t = 1 , 2 , … , 10 ., Both agents know all the rules of the game ., In each round , the investor receives an initial endowment of 20 monetary units ., The investor can send any of this amount to the trustee ., The experimenter trebles this quantity and then the trustee decides how much to send back to the investor , between 0 points and the whole amount that she receives ., The repayment by the trustee is not increased by the experimenter ., After the trustee’s action , the investor is informed , and the next round starts ., We consider the trust task as an IPOMDP with two agents ,, i . e ., , ℐ = {I , T} contains just I for the investor and T for the trustee ., We consider the state to contain two components; one physical and observable ( the endowment and investments ) , the other non-physical and non-observable ( in our case , parameters of the utility function ) ., It is the latter that leads to the partial observability in the IPOMDP ., Following 24 , we reduce complexity by quantizing the actions and the ( non-observable ) states of both investor and trustee—shown for one complete round in Fig 7 ., The actions are quantized into 5 fractional categories shown in Fig 7 ., For the investor , we consider aI ∈ {0 , 0 . 25 , 0 . 5 , 0 . 75 , 1} ( corresponding to an investment of $20 × aI , and encompassing even investment ranges ) ., For the trustee , we consider aT ∈ {0 , 0 . 167 , 0 . 333 , 0 . 5 , 0 . 67} ( corresponding to a return of $3 × 20 × aI × aT , and encompassing even return ranges ) ., Note that the trustee’s action is degenerate if the investor gives 0 ., The pure monetary payoffs for both agents in each round are, investor : χ I ( a I , a T ) = 20 - 20 × a I + 3 × 20 a I × a T ( 12 ) trustee : χ T ( a I , a T ) = 3 × 20 × a I - 3 × 20 a I × a T ., ( 13 ), The payoffs of all possible combinations and both partners are depicted in Fig 8 . In IPOMDP terms , the investor’s physical state is static , whereas the trustee’s state space is conditional on the previous action of the investor ., The investor’s possible observations are the trustees responses , with a likelihood that depends entirely on the investor’s intentional model of the trustee ., The trustee observes the investor’s action , which also determines the trustee’s new physical state , as shown in Fig 9 . Since all agents use their own planning horizon in modeling the partner and level k agents model their partner at level k−1 , inference in intentional models in this analysis is restricted to the guilt parameter α ., Using a categorical distribution on the guilt parameter and Dirichlet prior on the probabilities of the categorical distribution , we get a Dirichlet-Multinomial distribution for the probabilities of an agent having a given guilt type at some point during the exchange ., Hence 𝓑0 is a Dirichlet-Multinomial distribution ,, 𝓑 0 ∼ D i r M u l t ( a 0 ) a 0 = ( 1 , 1 , 1 ) ( 17 ), with the initial belief state, ℙ α partner = α i | h = ∅ = 1 3 ., ( 18 ), Keeping consistent with the model in 22 , our approximation of the posterior distribution is a Dirichlet-Multinomial distribution with the parameters of the Dirichlet prior being updated to, a t + 1 i = a t i + ℙ o t + 1 = observed action | α partner = α i ( 19 ), writing αpartner for the intentional models ., The level −1 models are obtained by having the level −1 agent always assume all partner types to be equally likely ( ℙ α partner = α i = 1 3 , ∀ i ) , setting the planning horizon to 0 , meaning the partner acts on immediate utilities only , and calculating the agent’s expected utilities after marginalizing over partner types and their respective response probabilities based on their immediate utilities ., In

Introduction, Results, Discussion, Materials and Methods

Reciprocating interactions represent a central feature of all human exchanges ., They have been the target of various recent experiments , with healthy participants and psychiatric populations engaging as dyads in multi-round exchanges such as a repeated trust task ., Behaviour in such exchanges involves complexities related to each agent’s preference for equity with their partner , beliefs about the partner’s appetite for equity , beliefs about the partner’s model of their partner , and so on ., Agents may also plan different numbers of steps into the future ., Providing a computationally precise account of the behaviour is an essential step towards understanding what underlies choices ., A natural framework for this is that of an interactive partially observable Markov decision process ( IPOMDP ) ., However , the various complexities make IPOMDPs inordinately computationally challenging ., Here , we show how to approximate the solution for the multi-round trust task using a variant of the Monte-Carlo tree search algorithm ., We demonstrate that the algorithm is efficient and effective , and therefore can be used to invert observations of behavioural choices ., We use generated behaviour to elucidate the richness and sophistication of interactive inference .

Agents interacting in games with multiple rounds must model their partner’s thought processes over extended time horizons ., This poses a substantial computational challenge that has restricted previous behavioural analyses ., By taking advantage of recent advances in algorithms for planning in the face of uncertainty , we demonstrate how these formal methods can be extended ., We use a well studied social exchange game called the trust task to illustrate the power of our method , showing how agents with particular cognitive and social characteristics can be expected to interact , and how to infer the properties of individuals from observing their behaviour .

journal.pbio.2004037

The neural system of metacognition accompanying decision-making in the prefrontal cortex

Decision-making is a process of evidence accumulation ., That evidence may come from sensory signals of external stimuli or from mental representations of internal cognitive operations ., Variations in evidence can create uncertainty in the person rendering a decision ., The decision maker is normally explicitly or implicitly aware of uncertainties about a decision and consequently confirms or revises a decision even prior to , or in the absence of , external feedback ., In the framework of cognitive control , the processes of decision uncertainty monitoring—and consequent decision adjustments—are termed metacognition , that is , ‘cognition about cognition’ 1–4 ., Although metacognition generally accompanies decision-making with uncertainty , the underlying neural system of the metacognitive processes in decision uncertainty monitoring and consequent decision adjustments remains less clear than that of the decision-making process per se 5 , 6 ., Much of the work on the neural bases of metacognition in humans has focused on metacognitive monitoring of internal states ( i . e . , confidence or uncertainty ) with regard to the cognitive processes such as episodic memory 7 , 8 and sensory perception 9 , 10 ., Behaviorally , confidence ratings , which reflect subjective accuracy beliefs regarding decisions , have often been found to deviate from the accuracy of an actual decision 11–13 ., These observations have suggested the existence of a separate neural processing system ( meta-level ) in the generation of decision confidence or uncertainty , independent of the decision-making process per se ( object-level ) ., We hereafter refer to this description of metacognition as separable from decision-making as “Theory 1” 11–18 ., The prefrontal cortex ( PFC ) has been proposed to play a critical role in metacognition 14 , and it has been demonstrated that interference with or lesions in PFC regions may impair metacognitive monitoring of perceptual decisions , but not decisions per se 15–18 , but see also 19 ., A contrary theory , which we will refer to as “Theory 2 , ” suggests that metacognition may be merely dependent on the decision-making process and therefore exclusively reliant on accumulated evidence 20–24 ., Specifically , this theory , based on bounded accumulation models , has interpreted divergence between decision accuracy and confidence reports as being caused by the accumulation of postdecisional evidence during the interval between decision-making and confidence reporting 20–24 ., Furthermore , it implies that decision adjustment naturally occurs as a part of this continuous postdecisional evidence accumulation and therefore is an integrated part of the initial decision-making process 21 , 24 ., Some proponents of this theory have argued that a separate neural system for metacognition to monitor and control decision-making should not be necessary because the processes are interdependent 24; however , not all work supporting this theory insists on this notion 20 ., Thus , one of the crucial issues in the debate between the two theories is whether a separate neural system for metacognition exists ., Single-decision paradigms ( depicted in Fig 1A and 1C ) are not sufficient to determine the existence or nonexistence of separable systems because the decision-making process and the metacognitive process are inevitably coupled in such tasks ., The purpose of retrospective metacognition is to confirm or revise foregone decisions ., Given an opportunity to make a decision on the same situation again ( i . e . , make a “redecision” ) ( as depicted in tasks shown in Fig 1B and 1D ) , a decision maker may revise an initial decision as well as confidence in the decision once s/he detects uncertainty regarding the initial decision 25 ., Thus , if a separate neural system for metacognition exists as proposed by Theory 1 , the metacognitive processes—in particular metacognitive control—should be more extensively involved in redecision , especially if the initial decision must be made quickly ., On the contrary , the neural system involved in redecision should be the same as those involved in an initial decision if there is not a separate neural system for metacognition ( as proposed by Theory 2 ) ., If a separate neural system for metacognition exists , the activity of this system should be manifest after an initial decision is reached , whereas Theory 2 suggests that they share the same underlying neural systems and that neural activity following either a single decision or a redecision should be the same ., Therefore , comparing behavioral and neural differences between the two phases of initial decision and redecision may allow us to test which theory better accounts for the neural processing of metacognition ., A specific perspective of metacognition derived from Theory 1 implies that decision uncertainty , rather than decision confidence , should be the key signal for metacognition ., If there is no uncertainty regarding a decision , it should not evoke the processes of metacognitive monitoring and control ., Therefore , the critical aim of this study was to elicit and analyze neural activity positively correlated with decision uncertainty , rather than that positively correlated with decision confidence ., In the present study , we employed a novel “decision–redecision” experimental paradigm to investigate neural activity associated with metacognition ., The participants were asked to make two consecutive decisions on the same situation using a perceptual decision-making task and a rule-based decision-making task ( Fig 2A ) ., We combined this novel paradigm with the functional magnetic resonance imaging ( fMRI ) technique to formally test the two theories and systematically investigate the underlying neural substrates of metacognitive processes accompanying decision-making ., Based on our previous study 25 , we expected that the frontoparietal control network would be associated with metacognitive processing ., In the current study , we focused on specific functions of the regions in the network believed to be involved in metacognition ., We found that dorsal anterior cingulate cortex ( dACC ) activity significantly correlated with metacognitive monitoring of decision uncertainty and that lateral frontopolar cortex ( lFPC ) activation correlated instead with metacognitive control ., These findings provide evidence for distinct neural processes involved in metacognition and decision-making ., We developed a novel decision–redecision paradigm for this study ( Fig 2A ) ., The participants made an initial decision ( decision phase ) , immediately followed by another decision on the same situation ( redecision phase ) ., This allowed the participant the opportunity to revise the initial decision and update their confidence rating , even without feedback ., The internal states of uncertainty regarding initial and final decisions were separately evaluated by confidence ratings ., Confidence was rated on a scale of 1 to 4 , immediately following the corresponding decisions ., Decision uncertainty was then the inverse of the confidence rating ( i . e . , a confidence rating of 4 corresponded to an uncertainty rating of 1 ) ., Critically , our task differed from previous paradigms used to analyze ‘change of mind’ 21 , 24 ., Previous task paradigms were only able to analyze the small portion of trials in which a participant happened to change their mind , while our paradigm allowed analysis of each trial ., We used two different types of decision-making tasks in the present study: one was a rule-based decision-making task ( Sudoku ) , the other a perceptual decision-making task ( random-dot motion RDM ) , which has commonly been used to investigate the neural process of decision-making 5 and more recently , metacognition 21 , 22 , 24 , 26 ., The decisions in the Sudoku task rely on internal informational operations , but decisions in the RDM task should be more dependent on accumulation of external information ., It is possible to continue accumulating evidence from external stimuli that may affect decision-making in the RDM task , but that is less likely in the Sudoku task because it is rule based ., For this reason , the two tasks should result in differential processing in metacognitive control to adjust initial decisions ., The sequences of both tasks were identical ( Fig 2A , illustrated for the main fMRI experiment fMRI1 ) ., After a Sudoku problem or RDM stimulus was presented for 2 s , the participant made a choice from 4 possible solutions within 2 s and then reported their confidence rating on that decision within 2 s ., A critical feature of our paradigm was that the same Sudoku problem or the same RDM stimulus was immediately repeated for 4 s , and the participant again made a choice within 2 s and again reported their confidence rating within 2 s ., To better distinguish the metacognitive process from the decision-making process , we intentionally set a short initial decision phase ( 2 s ) , to minimize metacognition during the initial decision-making phase , but set a longer duration in the redecision phase ( 4 s ) to allow enough time for metacognitive processing in redecision ., There was no explicit feedback or cue to indicate whether the decision was correct after either the initial decision or the redecision ., For both tasks , the task difficulty of each trial ( Fig 2B ) was adaptively adjusted by a staircase procedure 9 , 27 so that the average accuracy for the initial decision was converged to approximately 50% ( chance level was 25% ) ., For the control condition , the participant was shown a digital number in the target grid in the Sudoku task , and for the RDM task , s/he was shown an RDM stimulus with 100% coherence ., For the former , the participant only needed to press the button matching the number , and for the latter , the participant indicated the unambiguous RDM direction ., Prior to the experimental testing , the participant was trained to attain high-level proficiency in Sudoku problem-solving ., The current study was composed of 4 fMRI experiments: Twenty-one participants took part in fMRI1 ( see Materials and methods ) ., In both the Sudoku and RDM tasks , decision uncertainty levels were largely consistent with the percentage of incorrect initial decisions ( Fig 2C; Pearson’s r = 0 . 76 ± 0 . 12 mean ± SD , one-tailed t test , t21 = 7 . 3 , P = 1 . 7 × 10−7 in the Sudoku task; r = 0 . 71 ± 0 . 14 , t21 = 6 . 8 , P = 5 . 0 × 10−7 in the RDM task ) ., To examine the trial-by-trial consistency between objective erroneous decisions and subjective decision uncertainty levels in individual participants , a nonparametric approach was employed to construct the receiver operating characteristic ( ROC ) curve by using the decision uncertainty levels as thresholds to characterize the likelihood of erroneous decisions ., The area under curve ( AROC ) was then calculated to represent the individual uncertainty sensitivity , indicating how sensitive the participant was to the decision uncertainty 9 ., As observed in the previous studies , the uncertainty sensitivity of individual participants markedly deviated from decision accuracy in both tasks , which were controlled around 50% ( Fig 2D ) ., The response times ( RTs ) of option choices in the initial decision were positively correlated with the decision uncertainty levels ( Fig 2E; t21 = 6 . 9 , P = 4 . 0 × 10−7 in the Sudoku task; t21 = 4 . 3 , P = 1 . 6 × 10−4 in the RDM task ) ., The correlation coefficient between RT of option choices and the decision uncertainty level ( rRT-uncertainty ) in the initial decision was highly correlated with the uncertainty sensitivity ( AROC1 ) across the participants ( Pearson’s r = 0 . 61 , z test , z = 3 . 4 , P = 4 . 0 × 10−4 in the Sudoku task; r = 0 . 48 , z = 2 . 4 , P = 0 . 0085 in the RDM task ) ., Thus , the RT–uncertainty correlation also reflected individual uncertainty sensitivity ., The level of decision uncertainty was reduced by redecision ., The extent of decision uncertainty reduction via redecision was highly correlated with the decision uncertainty level in the initial decision phase ( Fig 2F; Goodman and Kruskal’s γ = 0 . 82 ± 0 . 11 , t21 = 8 . 8 , P = 2 . 1 × 10−8 in the Sudoku task; γ = 0 . 78 ± 0 . 14 , t21 = 7 . 7 , P = 8 . 2 × 10−8 in the RDM task ) ., Accordingly , accuracy also improved along with uncertainty reduction ( Fig 2G; Pearson’s r = 0 . 54 ± 0 . 13 , t21 = 4 . 2 , P = 2 . 3 × 10−4 in the Sudoku task; r = 0 . 39 ± 0 . 14 , t21 = 2 . 8 , P = 5 . 6 × 10−3 in the RDM task ) ., One could suspect that the improvement of uncertainty reduction and the change in accuracy in the redecision phase were caused by regression towards mean in the two separate decisions: higher uncertainty at the first measurement by chance would increase improvement at the second measurement ., However , the decision accuracy and decision uncertainty levels for the final decision-making phase remained significantly differential ( Pearson’s r = 0 . 35 ± 0 . 15 , t21 = 2 . 1 , P = 0 . 032 in the Sudoku task; r = 0 . 36 ± 0 . 14 , t21 = 2 . 6 , P = 8 . 9 × 10−3 in the RDM task in Fig 2C; Pearson’s r = 0 . 32 ± 0 . 14 , t21 = 2 . 0 , P = 0 . 042 in the Sudoku task; r = 0 . 32 ± 0 . 15 , t21 = 2 . 2 , P = 0 . 028 in the RDM task in Fig 2G ) , indicating that the participants’ performance in redecision reflected metacognitive processing ability rather than chance ., Despite the fact that both decision accuracy and decision uncertainty levels were improved in the redecision phase , the divergence between uncertainty sensitivity and decision accuracy remained significant ( Fig 2H ) ., Indeed , neither the individual uncertainty sensitivities nor those of individual differences were altered by redecision ( Fig 2I; t21 = 0 . 82 , P = 0 . 21 in the Sudoku task; t21 = 1 . 0 , P = 0 . 15 in the RDM task ) ., Similarly , neither the individual RT–uncertainty correlation coefficients nor those of individual differences were altered by redecision ( Fig 2E; t21 = −0 . 77 , P = 0 . 22 in the Sudoku task; t21 = 0 . 35 , P = 0 . 36 in the RDM task ) ., These results show that individual uncertainty sensitivity was stable , was intrinsic to individual metacognitive ability , and was independent of the accumulated evidence and the type of decision-making required ., Commonly across both tasks , brain activation during the initial decision phase was mainly restricted to brain areas posterior to the PFC , in particular the posterior portion of the PFC , the inferior frontal junction ( IFJ ) ( S1A Fig; Fig 3A ) ., In the redecision phase , a frontoparietal control network—consisting of the lFPC , dACC , anterior insular cortex ( AIC ) , middle dorsolateral PFC ( mDLPFC ) , and anterior inferior parietal lobule ( aIPL ) —was more extensively recruited ( Fig 3B; S1B and S2 Figs; S1 Table ) ., In contrast , the lFPC and mDLPFC regions of the frontoparietal control network were not activated when a new Sudoku problem or a new RDM stimulus was presented for the first time during the redecision phase , preceded by the control stimuli in the initial phase ( fMRI2 , n = 17; S1C and S3 Figs ) , while the dACC activity during the same phase became much weaker , and its response onset was much delayed from the onset of the stimulus presentation ( delay offset >3 s; S3 Fig ) ., Thus , the frontoparietal control network , in particular the regions of the lFPC , mDLPFC , and dACC in the anterior FPC , were more extensively involved in redecision than in the initial decision phase ., Trial-by-trial activity in the regions of the frontoparietal control network in redecision was positively correlated with decision uncertainty level for the initial decision ( Fig 3C and S2 Table ) ., Critically , these correlations remained significant even for the correct trials ( S1E Fig ) , indicating that these regions were encoding the decision uncertainty signal rather than the error signal ., Furthermore , task difficulty or RT could not explain their association with the decision uncertainty in these regions ., The residual fMRI signal changes after the components associated with the task difficulty and RT were regressed out remained highly correlated with decision uncertainty level , but residual fMRI signal changes after the components of the decision uncertainty level were regressed out were not further correlated with the task difficulty and RT ., Although the dACC and AIC regions were also partially activated during the initial decision phase ( S1A Fig and Fig 3A ) , this activity—as well as in other regions activated during the same phase—was neither positively nor negatively correlated with the decision uncertainty level ( S1D Fig ) ., Activity in the ventromedial PFC ( VMPFC ) and posterior cingulate cortex ( PCC ) regions of the default-mode network in redecision were negatively correlated with decision uncertainty level or positively correlated with its inverse—decision confidence ( S1F Fig ) ., Thus , the regional activity seen in the frontoparietal control network involved processes intrinsic to redecision but not the activity involved in decision-making for the initial phase ., In the third fMRI experiment ( fMRI3 , n = 25 ) , we confirmed that the strength of activity in the frontoparietal control network depended critically on whether redecision was required after the initial decision phase ., When decision uncertainty levels for initial decisions were matched in the two conditions ( two-tailed paired t test , t25 = 0 . 62 , P = 0 . 27 ) , activity in the frontoparietal control network was much stronger when redecision was required ( ‘redecision condition’ ) , in comparison with those when redecision was not required ( “non-redecision condition” ) ( Fig 3D ) , despite the fact that activation of the frontoparietal control network in the ‘non-redecision condition’ was also significant ( S1G Fig ) and was correlated with decision uncertainty level as well 25 ., Thus , the frontoparietal control network , more strongly activated in redecision , should not only be involved in metacognitive monitoring of decision uncertainty of the initial decision but also in metacognitive control in redecision ( Fig 1D ) ., We then putatively defined this frontoparietal control network as the metacognition network ., Because the duration of the redecision phase in fMRI1 was longer ( 4 s ) than that of the initial decision phase ( 2 s ) , it raised the question of whether the fMRI activity predominately observed during the redecision phase was induced by the longer exposure , specifically in the trials with more difficult decisions ., To address this , we scanned an independent group of participants ( fMRI4 , n = 20 ) while they underwent the same RDM task as fMRI1 except that the duration of redecision was set to 2 s ., The same behavioral and neural results were replicated as in fMRI1 ( S4 Fig ) ., Just as the extent of uncertainty reduction by redecision was found to be highly correlated with the decision uncertainty level of the initial decision ( Fig 2F ) , activity in the regions of the metacognition network were also found to be positively correlated with the extent of uncertainty reduction ( S1H Fig ) ., However , the strength of the correlations decreased somewhat after the components associated with the decision uncertainty level were regressed out ( S1I Fig ) ., Conversely , correlations with decision uncertainty level in the metacognition network remained significant after the components associated with the extent of uncertainty reduction were regressed out ( S1J Fig ) ., These partial correlations complementarily confirmed that the metacognition network in redecision was involved in both metacognitive monitoring and metacognitive control , indicating that the two processes interacted in redecision processing ., The two processes , although interactive , can be dissociated ., In the region involved in uncertainty monitoring , activity strength should dynamically represent decision uncertainty level ., As decision uncertainty levels were reduced by redecision , the strength of its activity should accordingly be reduced ., Therefore , the neural activity change should be negatively correlated with the extent of decision uncertainty reduction ., Alternatively , in the region that was critically involved in metacognitive control , its activity should become positively correlated with the extent of decision uncertainty reduction , representing the outcome or the extent of metacognitive control ., We found that the activity in the dACC and AIC regions at the late phase of redecision did in fact negatively correlate with the extent of decision uncertainty reduction after the components associated with the decision uncertainty level of the initial decision were regressed out ( Fig 4A , S1K and S1L Fig ) ., Conversely , the lFPC activity in the Sudoku task was positively correlated with the extent of decision uncertainty reduction after components associated with the decision uncertainty level were regressed out ( Fig 4B ) , but negatively in the RDM task ( Fig 4B and S1I Fig ) ., In addition , VMPFC activity was also positively correlated with the extent of decision uncertainty reduction in both tasks ( S1I Fig ) ., The regional activity in the default-mode network appeared intrinsically anticorrelated with the regional activity in the metacognition network ( further detail regarding activity in the default-mode network associated with metacognition will be discussed in another study ) ., Thus , the dACC and AIC regions were specifically involved in metacognitive monitoring ., In contrast , the lFPC was specifically involved in metacognitive control in redecision , particularly in the Sudoku task ., Therefore , their functional roles in metacognition appear to dissociate in redecision processing ., In the Sudoku task , whether the problem would be better solved should be conditioned to individual intrinsic motivation to engage metacognitive control because metacognitive control was effortful ., The ventral striatum ( VS ) activity during the redecision phase was positively correlated with the extent of decision uncertainty reduction in the Sudoku task , but not in the RDM task ( Fig 4C ) ., VS might encode the intrinsic motivation or the internal reward on reduction in uncertainty during the redecision phase in the Sudoku task ., Critically , the lFPC activity was significantly coupled with the interaction between the VS activity and the decision uncertainty level of the initial decision ( Fig 4D; see psycho–physiological interaction PPI analysis in Materials and methods ) ., Furthermore , the accuracy change of each participant by redecision was positively correlated with the coupling strength in the Sudoku task ( Fig 4E ) ., These results implied that the efficiency of lFPC involvement in metacognitive control in rule-based decision-making tasks ( i . e . , Sudoku ) might be facilitated by the VS activity ., The abilities of metacognitive monitoring and control are behaviorally embodied in two components: uncertainty sensitivity and accuracy change , respectively ., Throughout all sessions , including fMRI1 and the other repeated behavioral experiments , the individual uncertainty sensitivity was highly consistent across different sessions of the Sudoku task ( Cronbach’s α = 0 . 91; Fig 5A , left column , upper panel ) and the RDM task ( α = 0 . 89 , Fig 5A , left column , middle panel ) , as well as across the two tasks ( α = 0 . 85; Fig 5A , left column , lower panel ) ., In contrast , the individual accuracy change in redecision was not consistent across the two tasks ( α = 0 . 03; Fig 5A , right column , lower panel ) , although it was consistent between different sessions of the Sudoku task ( α = 0 . 80; Fig 5A , right column , upper panel ) and the RDM task ( α = 0 . 76; Fig 5A , right column , middle panel ) ., Thus , individual metacognitive abilities of uncertainty monitoring were reliably consistent , but individual metacognitive control was dissociable in the two tasks ., Accordingly , the individual uncertainty sensitivity ( AROC ) was positively correlated with the uncertainty-level regression β value of the fMRI signal changes ( i . e . , neural uncertainty sensitivity ) , primarily in the dACC and AIC regions ( Fig 5B , P < 0 . 001 , cluster size = 20; and Fig 5C upper , one-tailed t test , Pearson’s r = 0 . 79 , t19 = 5 . 6 , P = 6 . 0 × 10−6 in the Sudoku task; r = 0 . 55 , t19 = 2 . 9 , P = 0 . 0049 in the RDM task; S3 Table ) , but not in the lFPC region ( Fig 5B and 5C bottom; Pearson’s r = 0 . 17 , t19 = 0 . 8 , P = 0 . 22 in the Sudoku task; r = 0 . 21 , t19 = 1 . 0 , P = 0 . 17 in the RDM task ) , commonly in both tasks ., The differences of correlations were significant between the two regions ( t19 = 3 . 8 , P = 5 . 6 × 10−4 in the Sudoku task; t19 = 2 . 3 , P = 0 . 016 in the RDM task ) ., In contrast , the individual accuracy change was significantly correlated with the mean activity in the lFPC region ( Fig 5D , P < 0 . 001 , cluster size = 20; and Fig 5E bottom; Pearson’s r = 0 . 69 , t19 = 4 . 2 , P = 2 . 2 × 10−4 in the Sudoku task; r = −0 . 39 , t19 = 1 . 9 , P = 0 . 041 in the RDM task ) , but not in the dACC region ( Fig 5D and 5E upper; Pearson’s r = 0 . 18 , t19 = 0 . 8 , P = 0 . 21 in the Sudoku task; r = −0 . 02 , t19 = 0 . 09 , P = 0 . 47 in the RDM task ) ., When the lFPC activity was stronger , the accuracy change was more in the Sudoku task but became less in the RDM task ( Fig 5E ) ., The differences of correlations were significant between the two regions ( t19 = 2 . 7 , P = 0 . 007 in the Sudoku task; t19 = 1 . 8 , P = 0 . 045 in the RDM task ) ., Thus , the dACC activity ( AIC as well ) commonly represented individual metacognitive abilities in monitoring of decision uncertainty , whereas the lFPC differentially modulated individual metacognitive abilities in control of decision adjustment—in both the Sudoku and RDM tasks—consistent with their dissociated functional roles in metacognitive monitoring and metacognitive control , respectively ., The regions of the metacognition network were also activated in the trials of both tasks with confidence level 4 in comparison with their respective control conditions ( Fig 6B and S2 Fig ) ., These activity differences might be partially caused by differentially subjective uncertain states of the two conditions that were not reflected by the four-scale confidence ratings ( i . e . , the ceiling effect ) ., The averaged accuracy was about 80% in the certain trials of the tasks ( Fig 2C ) , but it was about 95% in the control conditions ., Nevertheless , the task baseline activity in the certain trials of the tasks could also predict the individual uncertainty monitoring bias and potential abilities of efficient metacognitive control of decision adjustment ., Individual uncertainty monitoring bias—as estimated by averaging the decision uncertainty levels of all trials in each session of the tasks , representing the individual’s overconfident or underconfident tendency—was consistent between different sessions in the Sudoku task ( α = 0 . 95; Fig 6A , left panel ) and in the RDM task ( α = 0 . 94 , Fig 6A , middle panel ) , as well as across the two tasks ( α = 0 . 91; Fig 6A , right panel ) ., Accordingly , individual uncertainty monitoring bias was positively correlated with the mean task baseline activity in the dACC region ( Fig 6C P < 0 . 001 , cluster size = 20; and Fig 6F left , Pearson’s r = 0 . 50 , t19 = 2 . 5 , P = 0 . 0096 in the Sudoku task; r = 0 . 44 , t19 = 2 . 1 , P = 0 . 022 in the RDM task ) but not in the lFPC region ( Fig 6C and 6F right; Pearson’s r = 0 . 18 , t19 = 0 . 80 , P = 0 . 22 in the Sudoku task; r = −0 . 04 , t19 = 0 . 17 , P = 0 . 43 in the RDM task ) , commonly in both tasks ., The differences of correlations were significant between the two regions ( t19 = 2 . 1 , P = 0 . 026 in the Sudoku task; t19 = 1 . 8 , P = 0 . 042 in the RDM task ) ., Meanwhile , the individual accuracy change in the Sudoku task was positively correlated with the mean task baseline activity in the lFPC region ( Fig 6D and 6G right; Pearson’s r = 0 . 45 , t19 = 2 . 2 , P = 0 . 020 ) but not with that in the dACC region ( Fig 6G left; one tailed t test , r = 0 . 14 , t19 = 0 . 62 , P = 0 . 27 ) ., In contrast , the individual accuracy change in the RDM task was negatively correlated with the mean task baseline activity in the lFPC region ( Fig 6E and 6G right; Pearson’s r = −0 . 40 , t19 = 1 . 9 , P = 0 . 035 ) but not with that in the dACC region ( Fig 6G left; r = −0 . 13 , t19 = 0 . 57 , P = 0 . 29 ) ., The differences of correlations were significant between the two regions ( t19 = 1 . 9 , P = 0 . 039 in the Sudoku task; t19 = 1 . 8 , P = 0 . 046 in the RDM task ) ., Furthermore , the differences of correlations with the individual uncertainty monitoring bias and the individual accuracy change in the dACC ( t19 = 2 . 2 , P = 0 . 020 in the Sudoku task; t19 = 2 . 8 , P = 0 . 0055 in the RDM task ) , as well as in the lFPC ( t19 = 1 . 8 , P = 0 . 046 in the Sudoku task; t19 = 2 . 0 , P = 0 . 030 in the RDM task ) , were significant ., Thus , the task baseline activity in the dACC region commonly reflected the individual uncertainty monitoring bias in both tasks , whereas that in the lFPC region could predict the individually differential potential abilities of metacognitive control for decision adjustment in both tasks ., Thus far , we have shown that the neural system of metacognition could be dissociated into at least two subsystems: the dACC and AIC regions involved in metacognitive monitoring of decision uncertainty , and the lFPC region involved in metacognitive control of decision adjustment ., To further elaborate the subsystems of the metacognition network , we performed analyses of interregional functional connectivity in the metacognition network ., By regressing out the mean activity and the modulations by the decision uncertainty level , the RT and the extent of uncertainty reduction , as well as their interactions , we calculated trial-by-trial correlations between each pair of regions in the metacognition network ( see Materials and methods ) ., The interregional functional connectivity patterns in both the task condition ( Fig 7A ) and the control condition ( Fig 7B ) were almost identical across the two types of tasks and were also similar to those at the resting state ( Fig 7C ) ., The interregional functional connectivity patterns consistently showed that the metacognition network might be divided into three subsystems: the lFPC region; the dACC and AIC regions; and the DLPFC and aIPL regions ., The interregional functional connectivity within each of the subsystems was systematically stronger than that across the subsystems ( paired t test , P < 0 . 05 in all comparisons ) ., So far , the functional roles of the subsystem consisting of the DLPFC and aIPL regions in metacognition remain unclear ., It is worth noting that the functional connectivity between the dACC and the regions of the other two subsystems in the task conditions was numerically stronger than the corresponding one at the resting state but was not statistically significant ., In the present study , we utilized a novel decision–redecision paradigm to examine the behavioral and neural associations of metacognitive processing in redecision , as compared to the processing in an initial decision ., The robust findings from our study showed that individual uncertainty sensitivity ( both AROC and rRT-uncertainty ) remained markedly stable over two consecutive decisions on the same situational task , between different sessions of the same tasks , and across the different tasks ., This indicates that individual uncertainty sensitivity was independent of evidence accumulation or the form of the decision-making process ., These findings provide evidence to contradict the theoretical prediction of Theory 2 ., If the processes of metacognitive processing and decision-making were integrated in one network , it should follow that , as more evidence is accumulated after redecision , the uncertainty sensitivity ( i . e . , AROC ) should be also improved 28 ., Our study did n

Introduction, Results, Discussion, Materials and methods

Decision-making is usually accompanied by metacognition , through which a decision maker monitors uncertainty regarding a decision and may then consequently revise the decision ., These metacognitive processes can occur prior to or in the absence of feedback ., However , the neural mechanisms of metacognition remain controversial ., One theory proposes an independent neural system for metacognition in the prefrontal cortex ( PFC ) ; the other , that metacognitive processes coincide and overlap with the systems used for the decision-making process per se ., In this study , we devised a novel “decision–redecision” paradigm to investigate the neural metacognitive processes involved in redecision as compared to the initial decision-making process ., The participants underwent a perceptual decision-making task and a rule-based decision-making task during functional magnetic resonance imaging ( fMRI ) ., We found that the anterior PFC , including the dorsal anterior cingulate cortex ( dACC ) and lateral frontopolar cortex ( lFPC ) , were more extensively activated after the initial decision ., The dACC activity in redecision positively scaled with decision uncertainty and correlated with individual metacognitive uncertainty monitoring abilities—commonly occurring in both tasks—indicating that the dACC was specifically involved in decision uncertainty monitoring ., In contrast , the lFPC activity seen in redecision processing was scaled with decision uncertainty reduction and correlated with individual accuracy changes—positively in the rule-based decision-making task and negatively in the perceptual decision-making task ., Our results show that the lFPC was specifically involved in metacognitive control of decision adjustment and was subject to different control demands of the tasks ., Therefore , our findings support that a separate neural system in the PFC is essentially involved in metacognition and further , that functions of the PFC in metacognition are dissociable .

Decision-making is often accompanied by a sense of uncertainty regarding the outcome ., In many situations , there is no explicit feedback or cue to indicate whether the decision is correct or not ., Fortunately , our brain can evaluate decision uncertainty using the internal signals and subsequently make appropriate adjustments to initial decisions ., The process of considering the outcome of a decision and whether a decision should be adjusted is called metacognition , and it tends to be automatically induced ., Thus , decision-making is usually accompanied by metacognition , and the two processes are inevitably coupled ., However , the neural systems supporting metacognitive processing remain unclear and have often been misattributed to the neural system of the decision-making process per se ., Here , we have analyzed this process in several volunteers by imaging the brain activity in specific regions while they performed Sudoku and random-dot motion ( RDM ) tasks ., Our results suggest the existence of a neural system located in the prefrontal cortex ( PFC ) mainly involved in metacognition and independent from the neural system of decision-making .

cingulate cortex, control theory, medicine and health sciences, decision theory, diagnostic radiology, functional magnetic resonance imaging, decision making, engineering and technology, prefrontal cortex, applied mathematics, brain, social sciences, neuroscience, magnetic resonance imaging, control engineering, cognitive psychology, mathematics, systems science, statistics (mathematics), cognition, brain mapping, neuroimaging, research and analysis methods, computer and information sciences, imaging techniques, metacognition, psychology, radiology and imaging, diagnostic medicine, anatomy, biology and life sciences, physical sciences, cognitive science

journal.ppat.0040042

Structural Insight into Epitopes in the Pregnancy-Associated Malaria Protein VAR2CSA

Adhesion of Plasmodium falciparum parasite-infected erythrocytes ( IE ) to the vascular bed is mediated by P . falciparum erythrocyte membrane protein 1 ( PfEMP1 ) , which interacts specifically with receptors on the vascular endothelium or placenta 1 , 2 ., The adhesion mechanism is thought to be developed by the parasite to avoid filtering through the spleen , where erythrocytes infected with late stage asexual parasites are removed from the circulation 3 ., Antibodies that target PfEMP1 and abrogate binding are believed to be important mediators of acquired malaria immunity ( reviewed in 4 ) ., Pregnancy-associated malaria ( PAM ) is caused by P . falciparum sequestering in the placenta by binding to chondroitin sulfate A ( CSA ) , which is a type of glycosaminoglycan attached on the surface of syncytiotrophoblasts 5 ., Women suffering from PAM develop antibodies which protect them and their offspring during subsequent pregnancies 6 ., These protective antibodies are thought to recognize a relatively conserved antigen as plasma and parasites from pregnant women from different malaria endemic areas cross-react 7 ., The PfEMP1 variant mediating placental binding was recently discovered and named VAR2CSA 2 , 8 ., The extracellular part of VAR2CSA consists of six Duffy-binding-like ( DBL ) domains , a large inter-domain ( ID2 ) and a C-terminal region predicted to be cytoplasmic ., Most PfEMP1 molecules , but not VAR2CSA , contain two cysteine-rich interdomain regions ( CIDR domains ) 9 , 10 ., Some CIDR domains bind to CD36 11 and they have been described as degenerated DBL domains 12 despite a very low sequence homology between DBL and CIDR domains ., The invasion of erythrocytes and the subsequent adhesion of IE to vascular endothelium or placenta are key events in the asexual life cycle of P . falciparum and thus of major importance for the virulence of this parasite ., Erythrocyte invasion is mediated by proteins belonging to the erythrocyte binding ligand family ( EBL ) and in P . falciparum the erythrocyte binding antigen ( EBA ) -175 is the best described EBL protein ., EBA-175 has some similarity to VAR2CSA: Firstly , EBA-175 contains two DBL domains ( called F1 and F2 ) ., Secondly , the EBA-175 DBL domains bind glycans on the sialylated glycophorin A on the erythrocyte surface 13 ., The monomeric structure of EBA-175 has been determined by X-ray crystallography and the primary features of the two DBL domains were found to be α-helices and an anti-parallel β-hairpin 14 ., EBA-175 also crystallized as a dimer , and the structure of this complex showed that the DBL domains of EBA-175 interacted in a reverse handshake orientation 14 ., The simian malaria parasite , Plasmodium knowlesi invades erythrocytes through the host receptor “Duffy antigen receptor for chemokines” ( DARC ) 15 ., This interaction is also mediated by a parasite-encoded DBL-containing protein , Pkα-DBL , and the crystal structure of Pkα-DBL has been shown to be very similar to PfEBA-175 despite extensive sequence variation 16 ., Based on the structure of Pkα-DBL , the DBL domain could be divided into three sub-domains named S1–S3 which are connected by short linkers ., Both the glycan binding site of PfEBA-175 and the DARC binding site of Pkα-DBL are predominantly located in S1 and S2 14 , 16 ., With the aim of making a vaccine that can reverse or inhibit parasite binding in the placenta , considerable effort has been put into defining the specific part/parts of VAR2CSA that bind to CSA ( reviewed in 17 ) ., The best way of determining this interaction would be to produce the extracellular part of VAR2CSA and co-crystallize this multidomain protein with CSA ., However , it is very difficult to express such a large protein and previous attempts to crystallize even single VAR2CSA DBL domains have failed ., Thus , novel methods are required to generate testable models and hypotheses on the overall 3D structure of PfEMP1 molecules ., The DBL domains of PfEMP1 are often illustrated as “pearls on a string” and vaccine development strategies are focusing on the VAR2CSA DBL domains as single entities in a larger protein ., We have recently published data showing a structural model of VAR2CSA DBL3X and mapped areas of DBL3X that are surface-exposed and reactive to naturally acquired antibodies on the native protein 18 ., In this previous study , we showed that most variable regions of DBL3X are located in flexible loops or surface-exposed parts of the model ., These findings were supported by an analysis of 106 VAR2CSA sequences which was published recently 19 , and showed that polymorphic sites in general are situated in flexible loop regions or other surface exposed areas of VAR2CSA DBL structure models ., In this study , we modeled the remaining five 3D7 VAR2CSA DBL domains , the VAR2CSA inter-domain 2 ( ID2 ) and a number of CIDR domains from different PfEMP1 molecules to get insight into the location of epitopes in the whole VAR2CSA molecule ., The models indicate that DBL domains contain features that are structurally conserved ., Furthermore it appears that there is homology between the ID2 , CIDR and part of the resolved structures of the DBL fold ., By absorbing antibodies on native VAR2CSA on the surface of IE and comparing antibody reactivity on a VAR2CSA peptide array before and after absorption , we define areas of the VAR2CSA molecule which may be accessible to antibodies in the native protein ., For all domains we find that relatively invariant parts are recognized and surface-exposed in the native VAR2CSA ., The surface-exposed epitopes on the six VAR2CSA domains are largely found within S1 and S2 , whereas S3 appears to be hidden in the complete VAR2CSA structure ., This finding is interesting because it leads to the first suggestions about the overall structure of VAR2CSA; based on these data we discuss the domain architecture of VAR2CSA and suggest a model where the protein is surface-exposed as a globular or multimerized structured protein stabilized by long α-helices in the S3 region ., The 3-D structures of the 3D7 VAR2CSA DBL domains ( Figure 1 ) were modeled using the HHpred server 20 developed for low-homology modeling ., HHpred template searches confirmed that the determined structures of the P . falciparum EBA-175 DBL domains F1 and F2 and the P . knowlesi DBL domain Pkα-DBL 14 , 16 could be used as templates for modeling ( HHpred probability scores were all 100% ) , although the sequence identity between the VAR2CSA DBL domains and templates was between 16–20% ( see modeling details in Table S1 ) ., The determined structures of the two EBA-175 DBL domains each contain a region where structural information is missing and the structure of Pkα-DBL has four such regions ( Table S1 and Figure 2 ) ., In crystallographic experiments , the local occurrence of missing structural information indicates regions of flexibility and loosely defined secondary structure ., For Pkα-DBL , the regions of missing structural information were used to divide the DBL fold into the subdomains S1–S3 14 and we adapted a similar classification for the VAR2CSA DBL domains ( Figures 1 and 2 ) ., VAR2CSA residues corresponding to regions of missing structural information in the templates were modeled as insertions relative to the template structure ., The structure of such inserted regions is difficult to predict correctly 21 ., Secondary structure predictions using the PSIPRED method 22 is part of the HHpred modeling protocol ., The prediction results are divided into helix , strand and coil , where the coil class consists of secondary structure types mostly found in loops ., Interestingly , the template regions of missing structural information aligned with VAR2CSA DBL sequences predicted to have coil secondary structure ., In general , the VAR2CSA sequence variation is high within these regions ( Figure 2 ) ., Taken together , this suggests that most of the variable regions in all VAR2CSA DBL domains form a variety of flexible loops with different conformations ., These results support the findings reported by Bockhorst et al . , who recently reported a similar tendency in structural models of VAR2CSA DBL2X and DBL3X 19 ., From a structural perspective these findings makes biological sense because the overall DBL fold could be preserved while surface-exposed parts and loop areas which are possibly less important for stabilization of the fold , would have more freedom to mutate ., The quality of a structure model obviously has a pronounced effect on the information that can be deduced from the model ., We used the automated structure analysis tool ANOLEA 23 for evaluation , and Z-scores ranging from 5 . 00 to 9 . 61 with 52%–68% high-energy residues were obtained ., The results indicated that the quality was lowest in the loop regions ., Analysis using Verify3d 24 resulted in a similar conclusion ( data not shown ) ., Since these results did not convince us that the models were correct , we decided to further investigate the quality of the models by inspecting them for conserved residues stabilizing the determined structures of EBA-175 and Pkα-DBL domains ., Structural alignment of the EBA-175 and Pkα-DBL domains identified a number of positions , which can be assumed to be important for the stabilization of the DBL fold ( Table S2 ) ., The positions of these residues were distributed throughout the domains in blocks ., We then made a multiple structural alignment including the six VAR2CSA DBL models and the EBA-175 and Pkα-DBL structure to identify VAR2CSA residues at the positions corresponding to the positions identified as conserved and stabilizing EBA-175 and Pkα-DBL ( Figure 2 ) ., This analysis showed that a high number of hydrophobic positions forming a hydrophobic core in the DBL structure were conserved in the models , together with positions of helix capping and positions of interacting buried polar residues ( Table S2 and Figure 2 ) ., This conservation of stabilizing positions indicates that the alignments of the model sequences to the template sequences are correct in regions surrounding these positions ., Additionally , we found that most template-stabilizing positions are located in semi-conserved blocks reported in an analysis of 106 VAR2CSA sequences 19 ( data not shown ) , which suggests that these stabilizing positions in the templates are conserved in VAR2CSA to stabilize the folding of the VAR2CSA DBL domains in general ., The structural alignment and sequence alignments used by HHpred for modeling were analyzed for conservation of cysteines forming disulfide bonds in the template DBL structures ., The models of the VAR2CSA DBL domains all contain conserved cysteine positions likely to form disulfide bonds ( Figures 1 and 2 ) ., The disulfide bonds were numbered according to the occurrence of the cysteines in the sequence ., Cysteines of disulfide bond 1 are conserved in all models except DBL6 ., Likewise , the cysteines of disulfide bond 5 are conserved in all models except DBL1 ., A number of cysteines in the models are in close proximity to a disulfide bond-forming template cysteine ., This suggests that the local alignments used for the modeling are sub-optimal or that alternative disulfide bonds are formed in the VAR2CSA DBL domains ., An interesting example is the disulfide bond 2 ( Figure 1 , number 2 and Figure 2 , positions 38 and 64 ) ., The disulfide bond is proximal to a region containing glycan-binding amino acids in the EBA-175 F1 and F2 domains ( Figure 2 , positions 44 , 48 , and 50–53 ) and it may play a role for the function of these domains ., Among the VAR2CSA DBL domains , only DBL3 has both cysteines conserved ., None of the other VAR2CSA DBL sequences have two cysteines in the proximity , and it is thus unlikely that the apparent variation stems from incorrect alignments ., The lack of the disulfide bonds in some regions of the VAR2CSA DBL domains may suggest higher flexibility and a more dynamic structure than in DBL domains stabilized by a higher number of disulfide bonds ., The analysis of different types of stabilizing characteristics shows that these to a large extent are conserved between the template and the VAR2CSA models ., Since our aim was to map experimental data onto the DBL models , rather than to analyze the structural conformations in detail , we concluded that the models were of sufficient quality for mapping of these data ., The extracellular part of VAR2CSA consists of six DBL domains and a sequence stretch consisting of 337 amino acids named inter-domain 2 ( ID2 ) 9 ., This part of the molecule has attracted little attention and has been viewed as an inter-domain spacer sequence ., We analyzed the ID2 sequence ( PFL0030c positions 879-1216 ) for homology to other proteins using the HHpred search and alignment tool ., Interestingly , the EBA-175 and Pkα-DBL domains were all identified as homologous to the ID2 domain with very high HHpred probability scores ( probabilities between 98 . 3% to 99 . 9% ) ., The similarity was pronounced in a region of 100 residues ( PFL0030c positions 1017–1116 ) , which aligned to the first two α-helices in S3 with a sequence identity of 19 . 8% ., Using the EBA-175 F2 DBL domain as template , we modeled the structure of ID2 positions 1017–1116 ( Figure 3A and 3B ) ., The secondary structure of the whole ID2 domain was then predicted using PSIPRED ( data not shown ) ., The topology of predicted secondary structure elements in the ID2 domain suggested that the C-terminal part of ID2 has a similar fold to DBL S3 , but did not support the notion that the N-terminal part of ID2 has a fold similar to DBL S1 and S2 ., In most PfEMP1 molecules the first DBL domains are separated by a cysteine-rich inter-domain region ( CIDR1 ) on which there is no structural data available ., Using the HHpred server homology between CIDRs and EBA-175 was identified ., To identify the significance of our results , and to make a more general analysis , we used a number of CIDR domains for the analysis ., Since VAR2CSA ID2 and CIDR1 are placed either after the first or the second DBL domain and can be divided into sub-groups , we examined CIDR1 sequences representative for the three subgroups CIDR-alpha , beta and gamma ., Similarly to the ID2 domain , the homology was detected in the first two helices of the DBL S3 and a structure model was made using the EBA-175 F2 DBL as template structure ( Figure 3B ) ., The homologous region was part of the CIDR M2 region defined by Smith et al . 10 ., Similarly to our results of ID2 , an analysis of topology in the predicted secondary structure suggested that C-terminal structures of CIDR domains are similar to the known structures of DBL S3 domains , but that the structure of the N-terminal may vary from DBL domains ., These predictions suggest that like ID2 , the C-terminal part of CIDR1 seems to form a structure similar to that of the DBL S3 ., In rational PAM vaccine design , it is important to establish which parts of native VAR2CSA are accessible to antibodies acquired by women who have developed immunity to pregnancy-associated malaria ., Additionally , the cross-reactivity of these antibodies is an important issue to investigate because of the variability in VAR2CSA sequences ., The latter has been addressed in several studies which have been investigating the cross-reactivity of naturally acquired human antibodies between different P . falciparum lines ., Both cross-reactivity 25 , 26 and isolate-specific recognition of antibodies have been reported 26 , 27 ., These studies were based on antibody reactivity to IE and the specificity as well as the target of the cross-reactive and isolate-specific antibodies are thus not known ., In our study , we have instead used VAR2CSA peptide arrays measuring binding to shorter peptides ., Anti-VAR2CSA IgG from pools of plasma was absorbed using VAR2CSA-expressing IE and the antibody reactivity of the pools in a VAR2CSA peptide array compared before and after absorption 18 ., Measuring the antibody reactivity in the peptide assays allowed qualitatively mapping of surface-exposed regions of the VAR2CSA DBL domains ., Plasma samples from 180 Tanzanian women , sampled at the time of delivery , were tested for reactivity towards CSA binding parasites ( 3D7CSA and FCR3CSA ) in flow cytometry and in ELISA towards recombinant VAR2CSA protein ., Two pools with high levels of antibodies towards IE and high levels of anti-VARCSA IgG were made using plasma from 32 and 10 Tanzanian pregnant women , respectively ., One of the plasma pools ( Human plasma pool 1 ) was depleted using erythrocytes infected with 3D7CSA parasites ., From the other pool ( Human plasma pool 2 ) , two depleted plasma pools were generated by depleting one part with erythrocytes infected with 3D7CSA , and depleting the other part on erythrocytes infected with the FCR3CSA parasite line ., As a control , we depleted a pool containing anti-VAR2CSA IgG ( Human plasma pool 3 ) on a 3D7 parasite expressing a non-VAR2CSA PfEMP1 variant ( VAR4 ) ., The antibody assays were performed on an array containing 442 overlapping 31mer peptides corresponding to the extracellular part of VAR2CSA based on the sequence of 3D7 ., Before absorption , we observed a number of intense peaks for most of the domains ( Figures S1–S6 ) ., The intensity varied between domains , for instance , the peptide reactivity in DBL1 resulted in highly intense peaks , whereas the peptide activity of DBL2 resulted in less intense peaks ., The antibody reactivity towards the majority of peptides was not affected by the depletion; nevertheless depletion on native VAR2CSA consistently removed antibody reactivity against some peptides in all domains ( Figures S1–S6 ) ., In general depletion on the non-CSA binding parasite did not reduce anti-VAR2CSA reactivity in the peptide array , but a small reduction of peptide-specific reactivity was detected in regions for DBL2 and DBL4 ., For DBL4 there was an overlap in the depletion on the non-CSA binding parasite and the CSA-binding parasite , indicating that some of the surface-exposed parts identified in DBL4 could be due to non-specific absorption ., In addition to the human plasma pool , a pool of sera from six rabbits each immunized with a different VAR2CSA DBL domain was absorbed and tested ., The pattern of reactivity in the peptide array with this pool before absorption was slightly different from the reactivity obtained with the human plasma pools and this difference was also reflected in the absorption experiments ., Taken together these results indicated that all domains including the N-terminal segment contained continuous peptide sequences accessible to antibodies when the VAR2CSA protein was expressed on the surface of CSA-binding IE ., It was difficult however , to detect a pattern for this reactivity between the domains when depicting the reactivity on a string of residues ., We therefore went on to visualize the reactivity on the DBL models ., To facilitate the mapping of depleted regions from the peptide array data we calculated depletion values ( DV ) by subtracting the depleted reactivity from the non-depleted reactivity ., DV were calculated for each of the four depletion experiments ( human pool 1 versus 3D7 or FCR3 , human pool 2 versus 3D7 , rabbit pool versus 3D7 ) and mapped onto the six structural DBL models ( Figures S7–S11 ) ., Figure 4 shows the results for DBL6 and it is apparent that the depletion experiments using the two human plasma pools identified essentially the same regions as target for surface reactive antibodies ( Figure 4A versus 4B ) ., The results obtained in the absorption experiment using FCR3-infected erythrocytes gave essentially the same results as the experiment using 3D7-infected cells ( Figure 4B versus 4C ) ., Since the peptide array was based on the 3D7 sequence this indicates that the surface-exposed epitopes on the 3D7 and FCR3 versions of VAR2CSA are cross-reactive or target-conserved epitopes ., The absorption experiment using the rabbit plasma pool only showed depletion of antibodies targeting peptides residing in S1 and S2 ( Figure 4D ) ., There was no indication of depletion of antibodies targeting S3 in any of the experiments ., To facilitate a comparison between the DBL domains , and to reduce the risk of false positive DV , we created consensus DVs for each domain by calculating a sum of normalized DV from each of the four absorption experiments and scoring the residue as positive if the value was above a fixed threshold ( Figure 5 ) ., Overall there was agreement between the models based on the individual absorption experiments and the consensus DV ( Figure 4A and 4B versus Figure 5 , DBL6 ) ., The results for DBL3 were also in agreement with the surface-exposed epitopes identified previously 18 ., When evaluating the consensus models it should be kept in mind that surface-exposed VAR2CSA regions can only be detected if the plasma pool contains antibodies against these regions and if the antibodies can be detected in the peptide array assay ., Thus , regions which are not targeted by antibodies , or regions targeted by antibodies that cannot be detected in the peptide array experiment ( because they either target non-linear sequences or polymorphic sequences not represented in 3D7 VAR2CSA , ) will not be scored as surface reactive ., As a consequence this method will only map a certain proportion of epitopes ., Likewise , residues buried in the native molecule but residing close to a region representing a surface-exposed epitope on the peptides will score as positive ., When comparing the consensus models for the six DBL domains , it was evident that the pattern of reactivity was comparable in DBL domains 1 , 2 , 3 , 5 , and 6 , whereas the pattern in DBL4 was unique ( Figure 5 ) ., For the former domains the targets of surface-reactive antibodies were mainly located in S1 and S2 , whereas little reactivity was found in S3 , which is located on the lower left side of the models in Figure, 5 . In S1 and S2 both loops and α-helices were targets of surface reactive antibodies ., The loop between S1 and S2 ( Figure 2 positions 81–87 and appearing most prominently in the upper left corner of the DBL2 model on Figure 5 ) is flexible in the Pkα-DBL structure and we observe a high sequence variation between the DBL domains in the region , suggesting that the corresponding loops in the VAR2CSA DBL domains are flexible , and reinforcing the possibility that VAR2CSA domains can be divided into sub-domains ., In DBL domains 2 , 3 , 5 , and 6 , a loop in S2 ( appearing most prominently in the lower right corner of DBL3 on Figure 5 ) was also targeted by surface reactive antibodies ., A loop region in S1 ( Figure 2 , positions 44–52 , appearing most prominently in the center of DBL6 on Figure 5 ) was recognized in DBL domains 1 , 2 , 3 , and, 6 . Interestingly , the corresponding regions in EBA-175 contain glycan-binding residues ., The S2 α-helix appearing in the upper right on all models was recognized in all domains but DBL4 , whereas some of the other α-helices in S1 and S2 were recognized to a varying degree ., The regions of DBL4 targeted by surface-reactive antibodies differed markedly from the other domains ., Relatively more reactivity was detected against S3 and the reactivity against S2 was mainly against regions on opposite side of the domain compared to the other domains ( Figures 5 and 6 ) ., The possibility that DBL4 is positioned different from the other domains in the quaternary VAR2CSA structure is in agreement with the finding that the DBL4-specific rabbit antibodies are not reacting with the native VAR2CSA on IE ( 28 and unpublished data ) ., The different pattern of DBL4 recognition could also be explained by the fact that some antibodies targeting DBL4 peptides were non-specifically absorbed on IE ( Figure S4 , lower panel ) ., There is no evidence that S3 is undergoing less diversifying selection , which might be expected considering the present experimental data ., To address whether epitopes in S3 in particular are conformationally arranged compared to S1+S2 and thus showing a bias in the results , we affinity purified rabbit antibodies on monomeric recombinant DBL2 and assessed these antibodies on the peptide array ., We found that epitopes in both S2 and S3 could be detected by the peptide array analysis ( data not shown ) , indicating that the system was functioning for epitopes in S3 as well ., Using a multiple sequence alignment of seven full-length VAR2CSA sequences , residues were classified as conserved if they were all identical and as polymorphic if any of the sequences showed variation in the particular position ., It is apparent from Figures 5 and 6 that all domains contained both conserved and polymorphic regions targeted by surface reactive antibodies , but the conserved regions were most prominent in DBL3 and DBL5 ., This is in agreement with data showing that antibodies raised against recombinant proteins representing DBL3 and DBL5 are more likely to cross-react with heterologous parasites , than antibodies raised against the other domains 28 ., Furthermore , these findings also support data showing that antibodies raised against one DBL3 or DBL5 variant are highly cross-reactive with a large panel of placental isolates ( Magistrado et al . , submitted ) ., Finally , the data is also in agreement with the reactivity of human monoclonal antibodies produced by immortalized B cells from malaria-exposed pregnant women which are directed predominantly against these two domains 29 ., To further investigate the presence of conserved immunogenic epitopes in VAR2CSA domains , we performed competition ELISA using a large panel of recombinant VAR2CSA DBL3 domains derived from placental isolates 18 ., One variant of VAR2CSA DBL3 was coated in ELISA plates and the antibody reactivity of a high-titered VAR2CSA plasma pool was compared before and after pre-incubation with a competing VAR2CSA DBL3 variant ., By this method it was possible to quantify the relative level of variant-specific IgG towards individual DBL3 variants ., As positive and negative controls we incubated the plasma pool with homologous VAR2CSA DBL domain or VAR2CSA DBL5 domain as competing antigen ( Figure 7 ) ., Pre-incubation with non-homologous DBL3 domain reduced the reactivity by 25%–100% ., No competition/absorption was seen with the control proteins ., These data strongly suggest that the insect-cell produced DBL3 domains share common and cross-reactive motifs ., To determine whether some of these conserved linear regions were exposed on the surface of the native parasite protein , we synthesized peptides corresponding to two regions in DBL3 and DBL5 suggested by the peptide array data to be surface-exposed ( Figure 5 ) ., The first peptide P62 , corresponded to aa position 1350–1370 in DBL3 ., This region is highly conserved having only one variant aa ( based on alignment of 43 VAR2CSA DBL3 sequences ) ., The second peptide , P63 , corresponds to aa 2045–2061 in DBL5 and is also relatively conserved having 2 variant aminoacids out of 17 ( based on alignment of 15 VAR2CSA DBL5 sequences ) ., P62 and P63 were tested in ELISA for reactivity with Tanzanian male and female plasma and both showed significantly higher reactivity with female plasma as compared to male plasma ( Mann-Whitney rank sum test , p < 0 . 05; data not shown ) ., The two peptides were screened in ELISA for reactivity against plasma from rabbits immunized with recombinant DBL3 or DBL5 protein and peptide-reactive rabbit sera was used to affinity purify antibodies on the peptides to create peptide-specific IgG reagents ., The affinity-purified antibodies were subsequently assayed in flow cytometry for reactivity with native VAR2CSA expressed on erythrocytes infected with CSA-binding 3D7 or FCR3 strain ( Figure 8 ) ., Both the DBL3 and DBL5 peptide purified antibodies reacted strongly with FCR3CSA strain and to lesser extend with the 3D7CSA parasite ., The DBL5 peptide antibodies were affinity-purified from a rabbit immunized with a FCR3 DBL5 domain , and the difference of two amino acids in the peptide sequence could explain the lower reactivity with 3D7CSA compared with FCR3CSA ., The DBL3 peptide-specific antibodies were affinity purified from a rabbit immunized with a placental variant of DBL3 protein and there is only one amino acid differing between this placental sequence and the FCR3 and 3D7 sequences ., It was intriguing to find that the affinity-purified antibodies recognized the parasites to varying degree ., One reason for this could be a difference in levels of VAR2CSA protein expressed on 3D7 and FCR3 , or the observed difference could be accounted for by the single polymorphisms in the peptide sequences ., A third explanation could be that polymorphic flexible loop regions , flanking the conserved surface exposed parts , influence the accessibility of the surface-exposed regions to varying degrees between 3D7 and FCR3 ., Conserved surface-exposed epitopes appear to be attractive vaccine targets ., However , protective immunity is acquired through successive pregnancies 6 , 30 , and is a function of transmission intensity 31 ., Naturally acquired protection could therefore depend on the ability to recognize several polymorphic VAR2CSA variants ., This is consistent with the finding that some targets of naturally acquired antibodies are under diversifying selection 18 ., The levels of antibodies against pregnancy-associated parasite-encoded antigens on the surface of the erythrocyte increase with the number of pregnancies , and are correlated to the adhesion inhibitory capacity 7 , 32 ., Therefore , our identification of antibody binding parts in more conserved regions of VAR2CSA may seem like a paradox ., However , there could be several explanations for this finding: Firstly , the identified regions may have importance for the function of VAR2CSA , and this could lead to conservation ., Second , the conserved regions may be less immuno-dominant than other more variable regions of VAR2CSA ., This would make the development of antibodies directed against these regions less frequent , and result in a delay in the development of immunity ., Finally , whole antibody binding surfaces may be composed of both conserved and variable regions , where the variance of a few residues would be enough to disrupt the binding ., Most antibody binding epitopes of globular proteins have been estimated to be discontinuous in nature 33 , and additionally it has been shown that most discontinuous epitopes are comprised by 14–19 amino acids , including a linear segment of 4–7 amino acids 34 ., Therefore , the complete binding surface of antibodies recognizing P62 or P63 could very likely be discontinuous , and comprised by both polymorphic and more conserved regions ., We have identified conserved VAR2CSA regions targeted by antibodies recognizing the surface of IE ., Furthermore we have been able to induce rabbit antibodies against these regions by immunization with recombinant DBL domains ., This is a promising finding for vaccine development and it will be important to establish whether antibodies against conserved surface-exposed VAR2CSA regions can inhibit the binding of parasites to CSA ., So far , little is known about the overall structure of VAR2CSA or any other PfEMP1 ., It has been suggested that the PfEMP1 protein architecture is comprised by a compact semi-conserved head structure which largely defines the binding affinity and a number of variable C terminal domains 12 ., Previous data have indicated that the general DBL fold is relatively conserved 14 , 16 , 18 , 35 and that DBL domains can interact with each other as building blocks to form binding sites 14 ., The data presented here indicate that DBL S3 is less surface-exposed than S1 and S2 ., Sub-domain 3 contains two long α-helices which are conserved in the template structures ., A number of multimeric protein complexes have been reported to be stabilized by interactions between long α-helices; a well-studied ex

Introduction, Results/Discussion, Materials and Methods

Pregnancy-associated malaria is caused by Plasmodium falciparum malaria parasites binding specifically to chondroitin sulfate A in the placenta ., This sequestration of parasites is a major cause of low birth weight in infants and anemia in the mothers ., VAR2CSA , a polymorphic multi-domain protein of the PfEMP1 family , is the main parasite ligand for CSA binding , and identification of protective antibody epitopes is essential for VAR2CSA vaccine development ., Attempts to determine the crystallographic structures of VAR2CSA or its domains have not been successful yet ., In this study , we propose 3D models for each of the VAR2CSA DBL domains and we show that regions in the fold of VAR2CSA inter-domain 2 and a PfEMP1 CIDR domain seem to be homologous to the EBA-175 and Pkα-DBL fold ., This suggests that ID2 could be a functional domain ., We also identify regions of VAR2CSA present on the surface of native VAR2CSA by comparing reactivity of plasma containing anti-VAR2CSA antibodies in peptide array experiments before and after incubation with native VAR2CSA ., By this method we identify conserved VAR2CSA regions targeted by antibodies that react with the native molecule expressed on infected erythrocytes ., By mapping the data onto the DBL models we present evidence suggesting that the S1+S2 DBL sub-domains are generally surface-exposed in most domains , whereas the S3 sub-domains are less exposed in native VAR2CSA ., These results comprise an important step towards understanding the structure of VAR2CSA on the surface of CSA-binding infected erythrocytes .

Individuals living in areas with high Plasmodium falciparum transmission acquire immunity to malaria over time and adults have markedly reduced risk of getting severe disease ., However , pregnant women constitute an important exception , and they become more susceptible to malaria during pregnancy ., This so called pregnancy-associated malaria ( PAM ) has severe consequences for both mother and child , and a vaccine would save hundreds of thousands of lives each year ., PAM is caused by P . falciparum–infected red blood cells that bind to receptors in the placenta ., By binding to the placental tissue , the parasites avoid being filtered though the spleen where they would have been killed ., The protein mediating this placental binding is a very large multidomain and variant protein named VAR2CSA ., Using structural modeling of VAR2CSA and antibody reagents from women who have had PAM , we show that antibodies tend to bind in similar regions , on one side of the individual VAR2CSA domains ., In addition , we show that highly conserved parts of this variant protein are accessible for antibodies ., This finding correlates with epidemiological data showing that woman acquire immunity towards PAM relatively fast , and the identification of these epitopes is thus a major step towards a protective vaccine .

immunology, plasmodium, microbiology, computational biology

journal.pcbi.1006759

RedCom: A strategy for reduced metabolic modeling of complex microbial communities and its application for analyzing experimental datasets from anaerobic digestion

Microbial communities are of major importance for human health 1 , 2 , geochemical cycles 3 , 4 and biotechnological processes 5–7 ., Despite of their importance , most microbial communities are still poorly understood due to their complex nature ., Mathematical modeling can help to uncover the interactions and dependencies of the members of these communities ., Different modeling formalisms have been used to simulate microbial communities including stoichiometric models , which can be analyzed by constraint-based methods 8–18 ., An increasing number of stoichiometric community models considers balanced growth as a key assumption stating that all organisms must grow with the same growth rate in a stable community 11 , 15 , 16 ., One central goal of these models is the characterization and prediction of possible community compositions and the analysis of the different modes of cross-feeding between the involved organisms ., Stoichiometric models of microbial communities with balanced growth usually result in bilinear models , where , in some equations , independent variables are multiplied with each other ., Thus , apart from their increased size , these models have a more complex nature than the linear metabolic models of single species ., To make bilinear models amenable to established constraint-based modeling approaches , they can be linearized by fixing either the community growth rate 16 or the community composition 11 , 15 ., In this study , we first provide a unified framework for setting-up , analyzing , and linearizing community models ., Even in linearized community models , the application of certain constraint-based techniques becomes quickly infeasible with an increasing number of organisms ., Furthermore , one shortcoming of existing methods for modeling of communities is that the solution space often contains unrealistic solutions ( where , for example , a species behaves unrealistically altruistic to produce substrates needed by other community members ) ., We therefore introduce a new approach , RedCom , to build reduced community models ., The main principle of RedCom is similar to what has been suggested by Taffs et al . 10 , namely to compute , in a first step , relevant net conversions of the single-species models which serve as reactions for the reduced model ., This reduced model can then be used to identify suitable combinations of single-species net conversions to obtain community-level conversions ., However , while Taffs et al . 10 used elementary modes to describe the single-species net conversions , RedCom is based on the more general concept of elementary flux vectors 19 , 20 ., This will be required to ensure balanced growth in the community model and to appropriately account for flux bounds and other ( e . g . proteome allocation ) constraints ., Reduced community models obtained with RedCom do not only focus on most relevant solutions but allow for a comprehensive characterization of solution spaces also for communities with more than only two or three species ., In the following , we apply the proposed techniques for different community models with increasing complexity from three up to nine species ., The investigated communities are capable of degrading different substrates to biogas , a renewable energy source ., Community models of the biogas process give insights on interdependencies and feasible community compositions and may contribute to increase productivity and stability of this process ., As one of the first studies , we also compare simulation results from the community models with experimental data of laboratory-scale biogas reactors for growth on ethanol and glucose-cellulose media ., Constraint-based ( stoichiometric ) modeling of metabolic networks 21 relies on the assumption of a steady-state for internal metabolite concentrations leading to the mass balance equation:, Nr=0, ( 1 ), The structure of the network is captured by the stoichiometric matrix N storing the stoichiometric coefficients of the metabolites ( rows ) in the metabolic reactions ( columns ) ., As consequence of eq ., ( 1 ) , steady-state flux vectors r fulfill the condition that no net accumulation or depletion of internal metabolites occurs ., Additionally to the steady-state condition , reversibility constraints ( 2 ) , flux bounds ( 3 ) and other types of inhomogeneous linear constraints ( 4 ) can be included:, rj≥0forj∈Irrev, ( 2 ), αj≤rj≤βj, ( 3 ), Ar≤b ., ( 4 ), The set Irrev contains the indices of irreversible reactions ., If only the steady-state ( 1 ) and the irreversibility constraints ( 2 ) are taken into account , the solution space forms a polyhedral ( flux ) cone; with any constraint of type ( 3 ) or ( 4 ) its shape becomes a ( flux ) polyhedron ., In order to create a community model combining all ( n ) single-species models , herein referred to as full model , a compartmented approach is usually employed 9 , 11 , 12 , 15 , 22 , 23 ., Each organism represents one compartment and an additional exchange compartment allows for exchange of metabolites ( substrates/products ) between organisms and with the medium ( Fig 1 ) ., With the new exchange compartment , the former external ( unbalanced ) metabolites become now internal ones and must be balanced in eq ., ( 1 ) ., Exchange metabolites used by several species are combined such that they exist only once in the community model ., As described in 15 the units of the ( specific ) single-species reaction rates must be adapted to refer to the total community ( instead of single-species ) biomass ., Accordingly , the units of all reaction rates change from mmol/gDWi/h to mmol/gDWc/h ., Exceptions are the n biomass synthesis ( growth ) reactions producing the species biomasses BMi from a ( species-specific ) set of precursors:, γi , 1pi , 1+γi , 2pi , 2+⋯+γi , qpi , q→1BMigDWi ( i=1…n ), ( 5 ), In the single-species models , the specific ( growth ) rates μi ( i = 1…n ) of these n reactions referred to unit 1/h , which is now changed to gDWi/gDWc/h ., We indicate the changed units of these reaction rates in the community model by the symbol μ˜i ( i=1…n ) ., Furthermore , n new pseudo-reactions are introduced in the community model to describe the integration of the n species biomasses into the community biomass BMc ( Fig 1 ) :, 1BMigDWi→1BMcgDWc ( rate:rBMi→BMcgDWi/gDWc/h ) ( i=1…n ), ( 6 ), Finally , a new community growth reaction is introduced “exporting” the synthesized community biomass to the medium ( Fig 1 ) ; the rate of this reaction is the community growth rate μc 1/h:, 1BMcgDWc→ ( rate:μc1/h ), ( 7 ), Note that , in steady state , μ˜i=rBMi→BMc and ∑i=1nμ˜i=∑i=1nrBMi→BMc=μc ., The obtained structure of the whole community network is captured in the community stoichiometric matrix Nc and the reaction rates in the community flux vector rc ( with units as described above ) ., As for the single-species models , we demand steady-state for the metabolites ( including all metabolites in the exchange compartment ) :, Ncrc=0 ., ( 8 ), In a stable continuous culture , the growth rate of microorganisms is typically equal to the dilution rate ., We assume that the same is true for a microbial community cultivated in a continuous process ., In that case , the growth rates μi of all organisms ( each normalized to the respective specific biomass ) must be identical and equal the community growth rate μc:, μ1=μ2=⋯=μn=μc ., ( 9 ), This concept of balanced growth of microbial communities has previously been proposed by Khandelwal et al . ( 2012 ) and is also an underlying principle of the OptDeg 15 and the SteadyCom 16 approach ., It has been argued that , even if there is no steady state in a continuous cultivation , the specific growth rates of the organisms need to be the same on average because otherwise the fastest organism would outgrow the others ., With constant growth rates , also the fractional biomass abundances, Fi=BMiBMc, ( 10 ), of each species i in the community biomass BMc must be constant ., The fractions Fi define the community composition F = ( F1 , … , Fn ) and sum up to unity:, ∑i=1nFi=1 ., ( 11 ), With balanced growth , the fraction Fi of species i is given by the ratio of the specific biomass production rate of species i ( normalized to the community biomass ) and the community growth rate:, Fi=BMiBMc=rBMi→BMcμc=μ˜iμc, ( 12 ) ), Note that the fractional contributions to the synthesis of the community biomass ( μ˜i=rBMi→BMc; normalized to BMc ) are not identical over the species , hence , the μ˜i need not fulfill ( 9 ) ., However , for the specific growth rates μi ( referring to BMi ) it holds that μi=μ˜i/Fi=μc and thus ( 9 ) is indeed satisfied ., For each species , we can rewrite ( 12 ) to the following constraint:, rBMi→BMc=Fiμc ., ( 13 ), ( Alternatively we could also use μ˜i instead of rBMi→BMc in this equation ) ., In the optimization problems considered below , constraints of type ( 13 ) need to be included only for n−1 species , because ( 6 ) , ( 7 ) , and ( 11 ) already imply ( 13 ) for the n-th species: rBMn→BMc=μc−∑i=1…n−1rBMi→BMc=μc−∑i=1…n−1Fiμc=μc− ( 1−Fn ) μc=Fnμc ., Due to the re-normalization of the reaction rates from specific to community biomass , as the last step in assembling the community model we also need to adjust the normalization of the original flux bounds ( 3 ) and other inhomogeneous conditions ( 4 ) by multiplying them with the fractional abundances:, Fiαij≤rijc≤Fiβij, ( 14 ), where αij and βij are the lower and upper bounds for reaction j in organism i and rijc is the reaction rate of reaction j in organism i in the community model ., Likewise , constraints ( 4 ) are adjusted for each organism to, Airic≤Fibi, ( 15 ), ( Ai , bi correspond to the respective variables in ( 4 ) for species i ) ., The irreversibility constraints for the reaction rates are kept from the single-species models:, rijc≥0forj∈Irrevi ., ( 16 ), To exclude solutions with non-zero fluxes rijc≠0 in organisms that are not present in the community ( Fi = 0 ) , we assume that every flux in species i is bounded ( i . e . , αij and βij are bounded ) ., With ( 14 ) , a non-zero flux rijc then implies Fi>0 ., In principle , with the chosen constraints , one can also consider the case where the community is not growing ( μc = 0 ) , i . e . , where dependencies arise exclusively from the maintenance metabolism of the participating species ., However , if the community is growing ( μc>0 ) , a non-zero flux rijc≠0 in species i implies again Fi>0 and , due to ( 13 ) , then also rBMi→BMc=μ˜i>0 ., In analogy to classical flux balance analysis ( FBA ) in single organisms , we may formulate a ( linear ) objective function maximizing certain ( combinations of ) reaction rates in the community model:, MaximizecTrcs . t ., ( 8 ) , ( 11 ) , ( 13 ) − ( 16 ), ( 17 ), Due to the multiplication of ( independent ) variables in constraints ( 13 ) , the community model and the associated optimization problem become bilinear ., While non-linear solvers can be employed to solve the optimization problem ( e . g . , to search for maximum community growth rates or to scan feasible ranges of fluxes or community compositions; see below ) , a linearization can be applied to enable application of standard linear programming solvers and methods routinely used in ( linear ) constraint-based modeling ., Two approaches have been used to linearize bilinear community models and to simplify its analysis ( Fig 2 ) ., In the first approach ( utilized in SteadyCom 16 ) , the community growth rate μc is fixed to a constant ( known ) value ., The constraints ( 13 ) become then linear and the optimization problem ( 17 ) thus treatable with standard linear programming ( LP ) solvers ., Linearization by fixing the community growth rate is useful , for example , to analyze which community compositions are feasible for a given community growth rate ., Repeating these analyses ( in discrete steps ) for the feasible range of community growth rates yields a more complete picture of the whole solution space ., An alternative linearization method was used in community FBA 11 and in the OptDeg approach 15 ., Here , instead of the community growth rate , the community composition , i . e . all the fractional abundances Fi , are fixed ., Eq ( 13 ) becomes then again linear allowing the utilization of LP solvers ., With given fractional abundances , constraint ( 11 ) can be removed from the optimization problem ( 17 ) ., This second linearization approach is useful to scan , for example , the feasible community flux space for a given community composition ., However , with a growing number of organisms , this scanning becomes very expensive in terms of the number of linear programs to be solved 16 ., In this study , we therefore linearize community models by fixing μc as proposed in the SteadyCom approach ., We used an iterative approach to find the maximum community growth rate μc , max in these linearized models ., First , we set μc to a value of 0 . 005 h-1 ., If a feasible flux distribution exists ( here , any ( including a zero ) objective function can be used in Eq ( 17 ) ) , we double μc and check again for a feasible flux distribution ., We repeat these steps until no feasible flux distribution is found ., We then take the average of this μc and the last feasible μc ( or zero if the first μc did not yield a flux distribution ) ., These steps are repeated ( check for feasibility , use average of latest feasible and infeasible μc as new constraint and check again for feasibility ) until the difference between the last feasible and infeasible μc is smaller than 0 . 00001 h-1 ., Generally , for both linearization variants , apart from the FBA-like optimization in ( 17 ) , other constraint-based methods like flux variability analysis ( FVA ) or metabolic pathway analysis based on elementary flux modes or elementary flux vectors can be carried out ( see below ) ., The described approaches for modeling communities under balanced growth can be used to define and analyze community solution spaces ., However , these solution spaces often include unrealistic solutions on the species-level ( e . g . , a species synthesizes , without any benefit for its own growth , products required by another species in the community 15 ) ., Consequently , the predicted ranges for community compositions or growth rates may become very large as they include many non-relevant phenotypes ., FBA could be used to find community compositions fulfilling certain optimality criteria , but the question of suitable objective function in communities arises ., In single-species models , a typical objective function is maximization of the growth rate ., In community models we can also maximize the community growth rate 11 ., But , again , even these optimal solutions may represent unrealistic community compositions in which some organisms waste substrate to ensure survival of the others 15 ., We therefore proposed previously an optimization approach to minimize a weighted sum of substrate uptake rates to find community compositions in which all organisms grow with their maximum biomass yields 15 ., This approach enabled us to narrow down the solution space to community compositions in which all organisms grow optimally with their maximum biomass yields at a given community growth rate ., When introducing our model reduction approach below , we will use a similar method to exclude unrealistic community flux distributions ., Elementary flux modes ( EFMs ) are non-decomposable flux vectors fulfilling Eqs ( 1 ) and ( 2 ) 24 ., EFMs represent balanced pathways or cycles and have become an important tool for exploring metabolic networks 20 , 25–28 ., However , one shortcoming of EFMs is that inhomogeneous constraints ( Eqs ( 3 ) and ( 4 ) in the single-species models and ( 14 ) and ( 15 ) in the community model ) , such as non-growth associated ATP maintenance demand and substrate-uptake limits , cannot be considered ., We therefore make use of the concept of elementary flux vectors ( EFVs ) , a generalization of EFMs which can account for inhomogeneous constraints 19 , 20 ., From the theory of EFVs , it is known that the flux polyhedron P resulting from a set of linear constraints is generated by convex combinations of bounded EFVs pk and conic linear combinations of unbounded EFVs xi and yj:, P={r∈ℜn|r=∑k∈Kγkpk+∑i∈Iαixi+∑j∈Jβjyj , γk≥0 , ∑k∈Kγk=1 , αi≥0}, ( 18 ), Due to combinatorial explosion , EFVs can usually only be calculated in medium-scale metabolic networks and , thus , only in smaller community models combining the central metabolism of two or three species ., We present RedCom , a new method to generate community models of reduced size and with reduced solution spaces excluding unrealistic community behaviors ., The main idea of the reduction approach , which has some similarities with but is not identical to an approach presented by Taffs et al . 10 , is to describe the metabolism of each organism by certain net conversions taken from the EFVs of the single-species models ( Fig 2 ) ., Since we are mainly interested in community compositions and metabolic interactions ( exchange reactions ) between the community members , it is often sufficient to focus only on net conversions of the respective species instead of taking its whole metabolic reaction network explicitly into account ., Furthermore , from the list of all net conversions of a species we select only those that obey certain optimality criteria avoiding unrealistic phenotypes in the community model ., The selected net conversions are used as reactions in the reduced community model to be built ., The construction of reduced community models with the RedCom approach is described in the following , a detailed example is given in S1 Text in the Supplements ., All models presented in the Results section were implemented and analyzed with CellNetAnalyzer version 2018 . 1 , a MATLAB package for structural and functional analysis of metabolic and signaling networks 29 , 30 ., CPLEX was used as a solver for linear optimizations and efmtool for computation of EFVs ., For solving bilinear problems , we used the fmincon solver for nonlinear optimization in MATLAB ., The solver needs an initial flux distribution that we retrieved from the linearized model ., Experimental data from a laboratory-scale biogas reactor on a defined glucose-cellulose medium were published earlier 31 and used for a comparison with predictions from the nine-species biogas producing community ( see Results ) ., The data were taken from steady-state conditions 31 ., We calculated the average methane and CO2 production rates over a course of 100 days ., To achieve steady-state conditions , the reactors were operated under similar conditions for 190 days prior to this time period ., Additionally to the data already published , we estimated biomass dry weights by measuring protein concentrations with the Lowry Assay 32 and dividing them by the factor 0 . 64 ( assumed fraction of protein of the total biomass in the model ) ., We used these data to calculate specific production and consumption rates for comparison with simulation results ., A detailed description of the procedures applied for inoculation , feeding , and sample analyses along with cultivation setup and parameters is given in the S6 Text ., Briefly , two 1 . 5 L bioreactor systems were inoculated with sludge from the aforementioned enrichment and fed with the same medium containing 14 . 6% ( v/v ) ethanol as main carbon source instead of glucose and cellulose ., After an adaption period , continuous cultivation mode was initiated using constant feeding rates and volume control ., In the following , different dilution rates were sampled at steady-state conditions , starting from 5 . 3∙10−4 h-1 further increasing until the biogas production collapsed ., Sampling and subsequent analyses comprised pH , biomass protein content , biogas composition and biogas volume produced ., In addition , samples were analyzed for residual ethanol and accumulated organic acids ., Finally , taxonomic analysis was carried out using an established MS-based metaproteomic workflow ( see S8 Text ) ., Using KEGG 33 and MetaCyc 34 as well as various literature references we manually constructed single-species models of the central metabolism of nine different organisms all being relevant for the biogas process: four primary fermenting bacteria ( Acetobacterium woodii , Escherichica coli , Clostridium acetobutylicum , Propionibacterium freudenreichii ) , three secondary fermenting bacteria ( Syntrophomonas wolfei , Syntrophobacter fumaroxidans , Desulfovibrio vulgaris ) and two methanogenic archaea ( Methanospirillum hungatei and Methanosarcina barkeri ) ., As suggested by Taffs et al . 10 , we consider each of these organisms as one functional guild in the biogas process with certain metabolic properties ., More specifically , under anaerobic conditions , E . coli produces ethanol as well as different organic acids like formate , lactate , acetate and succinate from glucose , glycerol and gluconate ., A . woodii is an homoacetogenic organism that can either ferment sugars like glucose and fructose but also lactate , formate or hydrogen and CO2 to produce acetate via the Wood-Ljungdahl pathway 35 , 36 ., P . freudenreichii can ferment glucose , glycerol and lactate to succinate and propionate ., The organism uses the methyl-malonyl-CoA pathway to produce propionate ., Some organisms using the methyl-malonyl-CoA pathway like Pelobacter propionicus are also capable of using ethanol as a substrate 37 ., Since we aimed to represent the functional guild of propionate producing bacteria using the methyl-malonyl-CoA pathway , we also added ethanol oxidation to propionate to the model ., C . acetobutylicum ferments glucose and glycerol to different organic acids and solvents like acetate , butyrate , ethanol , butanol and aceton ., The organism is known to grow in two different phases 38 ., In the first phase , the organism produces organic acids like acetate and butyrate ., These pathways have high ATP yields but the acids produced lower the pH in the medium ., In the second phase , acids are taken up and solvents like butanol and aceton are the main product ., C . acetobutylicum represents primary fermenting bacteria in our community model and we assumed that mainly production of formate , acetate , butyrate and ethanol is relevant in anaerobic digestion ., We therefore disabled production of the other solvents in the community model ., D . vulgaris is a sulfate-reducing bacterium that can grow on organic substrates like pyruvate , lactate and ethanol using sulfate or thiosulfate as an electron acceptor ., In the absence of electron acceptors , the organism can also grow in syntrophic associations with hydrogen utilizing organisms ., The products formed by D . vulgaris are either acetate and hydrogen plus CO2 or formate ( in syntrophic cultures ) or acetate plus hydrogen sulfide ( when sulfate is present ) ., Additionally , the organism can utilize hydrogen with acetate as a carbon source and sulfate as an electron acceptor ., S . fumaroxidans can grow on propionate in syntrophy or with sulfate as an electron acceptor 39 ., In pure culture the organism can grow on fumarate , fumarate plus propionate or succinate , formate or hydrogen plus sulfate 39 ., S . wolfei is a secondary fermenting bacterium that can degrade saturated fatty acids from butyrate through octanoate either to acetate and hydrogen ( even number of C-atoms ) or to acetate , propionate and hydrogen ( odd number off C-atoms ) in syntrophic cultures 40 ., Growth of S . wolfei is also possible on crotonate in monoculture 41 ., The methanogenic organism M . hungatei ( cytochrome-free ) produces methane from formate or from hydrogen plus CO2 while M . barkeri ( possesses cytochromes ) can use hydrogen plus CO2 , acetate , methanol and methylamines for methanogenesis ., In addition to different substrates utilized by the methanogens they also differ in ATP yields and substrate affinities ., M . barkeri has higher ATP yields but lower substrate affinity for hydrogenotrophic methanogens ., In our M . barkeri model we only implemented methanogenesis from acetate , methanol , and hydrogen with CO2 ., A summary of the single-species models with model dimensions ( number of metabolites and reactions ) and constraints is given in Table 1 ., The models of D . vulgaris , M . barkeri and M . hungatei were published before 15 ., We estimated flux bounds for substrate uptake and product formation from experimental data or existing models , partially also from closely related organisms ( see S3 Text ) ., Maintenance coefficients ( rATPmaint ) were taken from literature data but the reported values varied by more than one order of magnitude between the different species ( Table 1 , S3 Text ) ., Below we will therefore carry out a sensitivity analysis to investigate the influence of the maintenance coefficients on simulation results ., For model validation , we also compared model predictions with measured biomass yields reported in the literature ( see S4 Text ) ., All models are listed ( and also provided in SBML format ) in S1 Table in the Supplements ., For the simulations performed in this work , we focused on ethanol ( three and six-species community ) and glucose ( nine-species community ) as the only available substrates and switched the uptake of other substrates ( glycerol , gluconate , methanol , fructose , sulfate ) off to reflect the composition of media used in the experiments ., We investigated a three-species community model ( Table 2 ) consisting of D . vulgaris , M . hungatei and M . barkeri ., This community can convert ethanol to methane , CO2 , and acetate and thus covers the last two steps of anaerobic digestion ., A similar community was experimentally investigated by Tatton et al . 42 and simulated with FBA in a previous study 15 ., In analogy to the study of Tatton et al . 42 , the uptake of external CO2 was allowed to also include solutions in which the acetoclastic methanogen is non-essential ., We extended the three-species community model to a model with six of the nine model organisms by additionally integrating A . woodii , P . freudenreichii , and S . fumaroxidans ( Tables 1 and 2 ) ., The three additional organisms were chosen according to their potential of being part in an ethanol-degrading community; they represent functional guilds that extend the capability of the three-species community investigated above by additional pathways for homoacetogenesis and propionate fermentation ., Growth of the other ( remaining ) three organisms ( CA , SW , EC; see Table 1 ) is not supported with ethanol as substrate and they have therefore not been included yet ., Note that , at this initial point , no experimental data have been used yet to adjust the composition of the community model; this will later be done when including metaproteomic data from a concrete enrichment culture ., We finally simulated a community capable of growth on glucose ., Here , all of our nine guild organisms can potentially be involved in the process and are thus part of the community model ( Tables 1 and 2 ) ., In addition to the six-species community studied above , this model included E . coli , C . acetobutylicum and S . wolfei ., We first simulated the community with the bilinear model to predict the maximum community growth rate as well as ranges for substrate uptake , product excretion , biogas composition and methane yield ( Table 5 ) ., As already observed for the six-species model , a reliable prediction for μc , max was thus not possible with this model ( with the iterative approach in the linearized models we found that μc , max = 0 . 23 h-1 ) In contrast , the predicted ranges for reaction rates and yields seem reasonable ., We then compared predictions of the linearized full model and the reduced model with experimental data ( Table 5 ) from an enrichment culture grown on glucose-cellulose medium ( 31; see also Methods ) ., Data were available for two duplicate experiments with identical dilution rate ., Since hydrolysis of cellulose is not included in the model , we used glucose as a starting point and assumed that cellulose is converted to glucose by hydrolytic enzymes ., We set the community growth rate to 0 . 00067 h-1 , which corresponds to the dilution rate of the experiment and derived the corresponding linearized full community model and the reduced community model ., EFV computation was possible with the reduced model ( 213689 EFVs ) but not with the full model where we computed only ranges for biomass compositions , exchange rates , and methane yield via flux variability analysis ( Table 5 and Fig 8 ) ., Confirming findings from the three- and six-species models , we observed that the predicted ranges , especially of exchange rates and community compositions , are again considerably smaller in the reduced model compared to the linearized full model ., In fact , the calculated ranges of exchange rates of the linearized full model are almost identical to the ones from the bilinear model , although the latter did not consider a fixed growth rate ., The measured exchange rates were only slightly smaller than the minimum rates predicted by the models ., The predicted ranges of the reduced model lie on the lower end of the range of the linearized full model and are thus closer to the experimental data indicating that the organisms use their substrate efficiently as assumed by our model reduction approach ( Table 5 and Fig 8 ) ., The slight overestimation of the rates could again be a consequence of overestimating maintenance coefficients or an underestimation of ATP yields in the models ., Furthermore , we noticed a relatively high variance of the measurements for the exchange rates which may partially explain deviations between data and model predictions ., We also measured higher methane to CO2 ratios and lower methane yields than predicted by the models ., Typically , we would expect a ratio of 1 methane to one CO2 for carbohydrates like glucose ., However , some of the released CO2 might have been lost due to its better solubility in water ( compared to methane ) ., Microbial communities are of major importance for health , nature , and biotechnological applications ., Constraint-based stoichiometric modeling helps to obtain a better understanding of interrelationships in these communities and to make quantitative predictions ., However , compared to classical constraint-based modeling of single-species metabolic networks , analysis of community models based on the favored concept of balanced growth is hampered by four major technical difficulties: Our introduced RedCom approach , where reduced community models are constructed from net conversions of the linear single-species models , addresses three of the above four issues ( ( 2 ) - ( 4 ) ) ., Taffs et al . 10 also published an approach where EFMs ( instead of EFVs ) of single-species models were used as input for the community model ( “nested pathway consortium analysis approach” ) ., While the basic principle is the same , our RedCom approach uses EFVs instead of EFMs which is mandatory to guarantee balanced growth of the community and to allow the consideration of flux bounds , maintenance coefficients , and other inhomogeneous constraints ., A necessary pre-processing step is the calculation of EFVs in the single-species models for the fixed community growth rate followed by the selection of relevant EFVs projected onto their exchange fluxes ., Different optimization or selection criteria can be used for selecting the relevant single-species behaviors ., We decided to use all EFVs representing minimal conversions of exchange metabolites , which , as one particular advantage , ensures exclusion of unrealistic ( altruistic ) community behaviors of the respective species ( see point ( 3 ) ) ., Dependent on the application , other criteria could be used as well ., In the three- , six- , and

Introduction, Methods, Results, Discussion

Constraint-based modeling ( CBM ) is increasingly used to analyze the metabolism of complex microbial communities involved in ecology , biomedicine , and various biotechnological processes ., While CBM is an established framework for studying the metabolism of single species with linear stoichiometric models , CBM of communities with balanced growth is more complicated , not only due to the larger size of the multi-species metabolic network but also because of the bilinear nature of the resulting community models ., Moreover , the solution space of these community models often contains biologically unrealistic solutions , which , even with model linearization and under application of certain objective functions , cannot easily be excluded ., Here we present RedCom , a new approach to build reduced community models in which the metabolisms of the participating organisms are represented by net conversions computed from the respective single-species networks ., By discarding ( single-species ) net conversions that violate a minimality criterion in the exchange fluxes , it is ensured that unrealistic solutions in the community model are excluded where a species altruistically synthesizes large amounts of byproducts ( instead of biomass ) to fulfill the requirements of other species ., We employed the RedCom approach for modeling communities of up to nine organisms involved in typical degradation steps of anaerobic digestion in biogas plants ., Compared to full ( bilinear and linearized ) community models , we found that the reduced community models obtained with RedCom are not only much smaller but allow , also in the largest model with nine species , extensive calculations required to fully characterize the solution space and to reveal key properties of communities with maximum methane yield and production rates ., Furthermore , the predictive power of the reduced community models is significantly larger because they predict much smaller ranges of feasible community compositions and exchange fluxes still being consistent with measurements obtained from enrichment cultures ., For an enrichment culture for growth on ethanol , we also used metaproteomic data to further constrain the solution space of the community models ., Both model and proteomic data indicated a dominance of acetoclastic methanogens ( Methanosarcinales ) and Desulfovibrionales being the least abundant group in this microbial community .

Microbial communities are involved in many fundamental processes in nature , health and biotechnology ., The elucidation of interdependencies between the involved players of microbial communities and how the interactions shape the composition , behavior and characteristic features of the consortium has become an important branch of microbiology research ., Many communities are based on the exchange of metabolites between the species and constraint-based metabolic modeling has become an important approach for a formal description and quantitative analysis of these metabolic dependencies ., However , the complexity of the models rises quickly with a growing number of organisms and the space of predicted feasible behaviors often includes unrealistic solutions ., Here we present RedCom , a new approach to build reduced stoichiometric models of balanced microbial communities based on net conversions of the single-species models ., We demonstrate the applicability of our RedCom approach by modeling communities of up to nine organisms involved in degradation steps of anaerobic digestion in biogas plants ., As one of the first studies in this field , we compare simulation results from the community models with experimental data of laboratory-scale biogas reactors for growth on ethanol and glucose-cellulose media ., The results also demonstrate a higher predictive power of the RedCom vs . the full models .

ecology and environmental sciences, chemical compounds, enrichment culture, engineering and technology, atmospheric science, bioenergy, biological cultures, metabolic networks, organic compounds, biogas, metabolites, materials science, network analysis, alcohols, methane, research and analysis methods, propionates, computer and information sciences, chemistry, atmospheric chemistry, ethanol, fuels, biochemistry, hydrogen, chemical elements, environmental chemistry, organic chemistry, earth sciences, cell culturing techniques, biofuels, biology and life sciences, greenhouse gases, physical sciences, materials, metabolism, energy and power

journal.pbio.1000622

Clusters of Nucleotide Substitutions and Insertion/Deletion Mutations Are Associated with Repeat Sequences

A major challenge of evolutionary genetics is to determine the mechanisms underlying cryptic patterns of mutation rate variation and how they influence evolutionary outcomes 1 ., One of the most striking of these trends is the association between indel mutations and nucleotide substitutions 2–7 ., Inter-species genome comparisons have revealed this trend to be universal to all prokaryotic and eukaryotic genomes examined thus far 4–6 ., The prevailing explanation for this association is that indels , as “universal mutators” 4 , cause the accumulation of nucleotide substitutions in the hundreds of base pairs of sequence surrounding the indel 4 , 6 ., Although such studies have been unable to unequivocally determine if the clusters are due to a single multimutational event ( multiple mutation hypothesis ) , the indel per se ( the mutagenic indel hypothesis ) , or the region of sequence in which the indel is found ( the regional differences hypothesis ) , the mutagenic indel hypothesis has been adopted by workers in the field 8–12 ., The mechanism of indel mutagenicity proposed by Tian and co-workers is that indels , when heterozygous , cause paired chromosomes to form heteroduplex DNA during meiosis 4 ., This is posited to cause error-prone DNA repair systems to target indel-containing regions , leading to an increased likelihood of nucleotide substitution in the sequence surrounding the indel ., Over time , this increase in mutation rate is predicted to leave as its signature the clustering of nucleotide substitutions in the DNA surrounding indels , while corresponding non-indel-containing orthologous sequences should have a lower number of substitutions , in accordance with the background substitution rate ., In addition , because the proposed mutagenic effect of the indel is postulated to be dependent on its heterozygosity , the accumulation of substitutions should cease as soon as the indel becomes homozygous in the population ., These predictions contrast with the regional differences hypothesis; regional effects are predicted to cause both indel and non-indel haplotypes to accumulate substitutions whether the indel is heterozygous or not ., The multiple mutations hypothesis differs from both the regional and indel hypotheses in that clusters of mutations are due to a one-off mutation event ., Determining whether mutations have accumulated over time or are due to a single mutation event is difficult without the ability to examine indel divergence on a temporal scale ., Here we use a population genomics approach to tease apart the dynamics of indel divergence using the genomes of Escherichia coli , Saccharomyces paradoxus ( S . paradoxus ) , Drosophila , and humans ., We show that it is not the indel but rather the sequence region in which the indel occurs that is associated with the accumulation of nucleotide substitutions over evolutionary time scales ., We propose a mechanism whereby a DNA sequence that is prone to cause replication fork stalling causes the recurrent recruitment of error-prone DNA polymerases to certain DNA sites , resulting in an increased likelihood of nucleotide substitutions in the surrounding DNA sequence ., The detection of indel-associated mutation in bacterial species poses a dilemma for the mutagenic-indel hypothesis ., Prokaryotes are haploid; following the indel-causing event , the cell has only a brief heterogenote period during which , according to the mutagenic-when-heterozygous hypothesis , the indel is mutagenic ., After a few cell divisions , the daughter cell will produce only indel-containing copies of the genome and will not have a non-indel version to recognize that the indel is present ( Figure 2 ) ., The mutagenic-when-heterozygous theory then predicts ( at least in prokaryotes ) that nucleotide diversity does not accumulate over time ., To test this prediction , we generated pre-defined , non-overlapping sets of old and new indels in E . coli ., Old indels are those determined ( using an appropriate outgroup ) to have occurred before the divergence of the two strains under comparison; new indels are those that have occurred after their divergence ( Materials and Methods , Figure S2 ) ., As shown in Figures 3A and S3 , D values are significantly higher for old indels ( black lines ) than those for new indels ( grey lines ) ., This result demonstrates that , contrary to the mutagenic-when-heterozygous and multiple mutation hypotheses , mutations are accumulating at a higher rate in regions surrounding indels over time ., Background D ( Db ) is the average difference in the DNA sequences of two aligned orthologous regions ., An increase in the number of differences between the nucleotide sequences of two aligned orthologous regions above this average indicates an increase in the rate of the accumulation of substitutions ., The mutagenic indel hypothesis states that the indel per se is the cause of an increase in mutation rate and the accumulated nucleotide diversity in the surrounding sequence ., A consequence of this is that , of two aligned fragments of DNA , the indel-containing fragment should have a highly elevated D close to the indel and its corresponding non-indel-containing orthologous fragment should have a D equivalent to the background ., These predictions can be tested by choosing an orthologous sequence from a third E . coli genome as an outgroup to infer the ancestral state of the aligned sequence , thus allowing us to pinpoint in which of the two aligned genome fragments the indel event has occurred ., This is dependent on the assumption of parsimony—if indels are a convergent character , the indel haplotype could be mistakenly assigned ., D can be calculated for the sequence windows surrounding an indel-containing region ( the indel haplotype ) and the corresponding orthologous region without the indel ( the non-indel haplotype ) with which it is paired ., In order to minimize the bias caused by differences in the selective constraints upon aligned sequences , we employed stringent filters to ensure that the sequences compared are strictly orthologous ( see Materials and Methods ) ., Figure 3D shows that the values of D for both the indel- and non-indel-containing haplotypes , Di and Dni , are elevated in window 1 as compared to the background nucleotide diversity Db ., Although the values of Di in window 1 are often higher than Dni ( an average 14% difference in D ) , this was not significant ( two-sample Kolmogorov-Smirnov test , p>0 . 05 , Table S2 ) for any of the strains compared ., By contrast , when Di and Dni are compared to Db , in five out of six comparisons Di is significantly greater than Db ( an average 57% difference in D ) , while Dni is significantly greater than Db in four cases ( two-sample Kolmogorov-Smirnov test , p <0 . 05 , Table S2; average 40% difference in D ) ., Thus , for nearly as many instances as the indel haplotype , the non-indel haplotype has a D significantly higher than the background nucleotide divergence , confirming that the regional effect plays a role in the accumulation of nucleotide substitutions ., These results raise the possibility that the accumulation of mutations surrounding indels ( Figure 3C ) is mainly due to regional effects and not attributable to indels per se ., However , this conflicts with the inferences of previous studies 2 , 4 , 6 , that concluded that indels , not regions , are mutagenic ., In order to find the cause of this disagreement , we took a closer look at the results of those studies as well as our own data ., We noticed that the strains that are less diverged tended to have the largest difference between the indel and non-indel haplotypes ( Table S2 , Figure S4 ) ., Indels detected in the comparisons of two highly similar strains must have happened since their relatively recent divergence ., The fact that the more diverged strains differed less between the indel and non-indel haplotypes suggests that the indel-associated effect diminishes over time ., When we studied the results of 4 and 6 , we found the same trend ., For example , using data from 6 , when bacterial divergence was plotted against difference between Di and Dni , it showed that the difference between Di and Dni decreases with increasing divergence ( Figure S4 ) ., A further example is provided by Tian et al . s 4 analysis of heterozygote alleles at one-third and two-thirds frequencies in yeast ., The mutagenic-when-heterozygous mechanism predicts that indels occurring at a higher frequency in a population have been accumulating mutations for longer periods and should thus have a higher D value and a greater difference between Di and Dni ., Conversely , the indels at two-thirds frequency have a smaller Di/Dni ( 1 . 40 ) than the indels at one-third frequency ( 2 . 23 ) ., The fact that indels that have been segregating for longer time have a smaller difference between the indel and non-indel haplotypes indicates that spending more time as a heterozygote actually diminishes the indel-associated effect , contrary to the prediction of the mutagenic-when-heterozygous hypothesis ., The separation of D into Di and Dni allows us to calculate the proportion of D on the indel haplotype that can be attributed to the indel effect and to the regional effect , respectively ( see Materials and Methods ) ., Under the assumption that indel-causing events are uniformly distributed since the time of divergence , it follows that the level of divergence between two strains is correlated with the average age of the indels found during comparison ., If an indel constantly influences the accumulation of nucleotide substitutions in the surrounding sequence while polymorphic , we expect to see an increase in the difference between Di and Dni over time ., Conversely , if indels have a one-time-only effect on nucleotide diversity , we expect to find a decline in this difference over time ., We compared Di and Dni for alignments identifying new and old indels ( Materials and Methods , Table S3 ) ., Figure 4A shows that the difference between Di and Dni decreases with increasing divergence ( Pearsons correlation coefficient , r\u200a=\u200a−0 . 769 , p\u200a=\u200a0 . 0093 ) ., This negative correlation is striking when compared to the positive correlation between time since divergence and nucleotide diversity when the indel and region effects are not separated ( Figure 3C , Pearsons correlation coefficient r\u200a=\u200a+0 . 711 , p\u200a=\u200a0 . 00092 ) ., This result suggests that it is the region , but not the indel , that is constantly influencing the accumulation of substitutions over evolutionary time scales ., To test whether the aforementioned phenomenon is specific to prokaryotes , we carried out analogous indel analyses using the budding yeast Saccharomyces paradoxus ., This organism is suitable for analysis because genome sequences are now available for a variety of its strains 14 and because S . paradoxus , like many multicellular eukaryotes , spends most of its life as a diploid 15 ., The results of the analyses with S . paradoxus ( Figures 3B , 3E and 4B , Table S3 ) were in agreement with those obtained using E . coli sequences ., The S . paradoxus strains used here ( Table S1B ) cover a wider range of divergence than the E . coli strains 16; this allowed us to view the diminishing proportion of the indel-dependent component of D on a longer time scale ( Figure 4B , Pearsons correlation coefficient r\u200a=\u200a−0 . 963 , p\u200a=\u200a0 . 008 ) ., We then extended our analysis to Drosophila species ( Figure 4C ) ( see Materials and Methods ) ., Although few species diverged recently enough to be suitable for analysis , the results corroborate our prior findings that the proportion of D attributable to the indel decreases over time ( Pearsons correlation coefficient r\u200a=\u200a−0 . 980 , p\u200a=\u200a0 . 128 ) ., It should be noted that the ratio of ( Di − Db ) / ( Dni − Db ) was calculated for several yeast and fly alignments with greater divergence than shown in Figure 4; in all cases , this ratio was approximately one ( Table S3 ) ., All these results suggest that a difference between the indel and non-indel haplotype exists following the indel-causing event but that this difference decreases over time until stabilising with both haplotypes having the same amount of nucleotide diversity ., Because our study is able to track indel divergence within a species , this analysis provides unequivocal evidence that nucleotide diversity associated with indels decreases over time ., Mutations arise from inaccurate processing of DNA damage or errors incurred during DNA replication ., E . coli possesses five DNA polymerases of which two , Pol IV and Pol V , are error-prone ., These polymerases are recruited to stalled replication forks 17 , 18 and double-strand breaks 19 to restart DNA replication ., Errors made by DNA Pol IV are biased towards frameshifts 20 , and though genomes exhibit a bias towards transitions 16 , DNA Pol V most often causes transversion mutations 21–23 ., We analysed the ratio of transition to transversion changes for all aligned E . coli genomes and found that transversions are enriched close to indel and non-indel haplotypes ( two-sample Kolmogorov-Smirnov test , p <0 . 0001 ) ( Figure 5 ) ; this is also true for S . paradoxus and other eukaryotes 4 ., The accumulation of mutations at a specific site at a higher rate is uncharacteristic of mutations caused by a mutagenic chemical or another random event and is most likely due to the persistent recruitment of error-prone polymerases to that site over evolutionary time ., Impediments imposed by polynucleotide repeats or other repeat sequences are suggested to be common causes of DNA replication fork arrest 24 ., We performed a computational analysis on the 20 bp immediately flanking our collection of E . coli , S . paradoxus , and Drosophila indels to determine the distribution of repeats around indels ., We defined an indel as contiguous with a repeat if it occurred inside or immediately next to a repeat , and as repeat-proximal if some part of a repeat was positioned within 5 bp on either side of the indel ., For E . coli , 43% of indels were contiguous with a homopolymer , while 20% were proximal ., The corresponding numbers were 45% and 25% for yeast and 31% and 34% for flies , respectively ( Figure 6A ) ., The association between repeat sequences and indels is well understood: repeat sequences are prone to sustain strand slippage mutations 25 , 26 , which tend to cause indels 19 , 27 ., We propose a mechanism distinct from strand slippage for the regional increase in nucleotide substitutions , whereby repeat sequences and other polymerase-stalling motifs persistently cause the recruitment of error-prone DNA polymerases ., Each time DNA replication is restarted by an error-prone polymerase , DNA surrounding the region will be synthesized with a higher rate of error 17 , 18 , 28 , leading to an increased likelihood of nucleotide substitution ., The stalled fork also suffers a high rate of double-strand breaks , another route to error-prone repair 19 , 27 , 29 ., The 3R hypothesis predicts that regions of a genome with increased sequence diversity should be able to be identified by repeat sequence abundance ., We tested this prediction by using the recently sequenced genomes of three E . coli strains that we had previously not analysed ., We searched for repeat-rich regions by first generating pairwise alignments as for our indel analysis , dividing these into non-overlapping 100-bp windows , and then binning each window according to its number of 4-nucleotide homopolymer repeats ( see Materials and Methods ) ., We found that , even when indel-containing windows were excluded , windows with a higher number of repeat sequences had more nucleotide substitutions than those without ( 83% increase for SE11/REL606 and 71% increase for SE15/REL606 in windows with six repeats ) ., As for indel-based analyses , the more diverged two-strain comparison had a higher value of D , supporting that repeats cause the accumulation of substitutions over time ( Figure 6B ) ., We also found that the number of transversions relative to transitions was increased in repeat-rich regions ( 88% increase in windows with six repeats ) ( Wilcoxon Sum Rank , p<0 . 05 , Figure 6C , Table S5 ) ., The “bump” in nucleotide substitutions associated with the indel ( the difference between Di and Dni ) that we and others 4 , 6 often observe requires an explanation ., The declining ratio of Di/Dni shows that this bump is smoothened over evolutionary time ( Figure 4 ) ., One explanation for this is that indel mutagenicity is transient because the indel-containing allele is only mutagenic as a heterozygote and its mutagenic effect will vanish when it becomes homozygous ., The period for which bacteria exist as heterogenotes for an indel is orders of magnitude less than that for diploid eukaryotes ., However , a consistent decrease in Di/Dni is found across taxonomic kingdoms , an observation at odds with the proposal that heterzygosity/heterogenosity causes the indel “bump . ”, An alternative explanation is that the indel-associated bump in D may be due to the indel-causing event resulting in multiple nucleotide changes ., This possibility is not implausible considering the spectrum of mutations in bakers yeast ., Lang and Murray 30 found that in 63% of instances where two mutations occurred at the same time one was an indel and the other a nucleotide substitution; yet indels constituted only 6 . 67% of all mutations observed in that study ., Whichever explanation is correct , it is evident that the indel effect is transient and that it is the surrounding sequence that is associated with the accruement of substitutions over evolutionary time scales ., All the inferences made about indels , nucleotide substitutions , and repeat sequences have so far been drawn only from the comparisons of genomes ., In order to test predictions made by the 3R and mutagenic indel hypotheses , we utilized the comprehensive collection of spontaneous ura3 mutants gathered by Lang and Murray 30 ., This collection comprises 207 ura3 mutant alleles , each of which resulted from a single mutational event in a haploid ( and non-indel-containing ) gene ., The mutagenic indel hypothesis predicts that the clustering of mutations is caused by indels; thus , this set of independently occurring mutants should not cluster ., Conversely , the 3R hypothesis states that repeat sequences cause an increase in the likelihood of the surrounding sequence sustaining both indels and nucleotide substitutions; thus , according to this hypothesis , indels and substitution mutations collected from independent mutants should cluster around repeats ., Using a model based on a hyper-geometric distribution ( Materials and Methods ) , we first found that indels and substitutions cluster together ( p\u200a=\u200a0 . 019 ) , even though most substitutions occurred without a co-occurring indel ( 97% ) ., Next , we tested for the association of indel/nucleotide substitution mutations with any of the 264 four-nucleotide combinations of A , T , C , and G ( e . g . , ATCG , ATCA , ATCT , etc . ) ., It is expected by chance that 2 or 3 four-nucleotide combinations should be found to be significant; however , significant associations were found only with the repeat sequences TGTG ( p\u200a=\u200a0 . 00027 ) , AAAA ( p\u200a=\u200a0 . 0093 ) , and GTGT ( p\u200a=\u200a0 . 0098 ) ., These results confirm that indels , substitutions , and repeat sequences are associated independently of any initiating mutator indel ., We directly tested whether insertions of repeat sequences could increase the mutation rate of nearby regions in yeast ., We engineered a copy of the URA3 gene to contain either a poly ( A ) repeat , a poly ( G ) repeat , a poly ( TG ) repeat , or a random 12-mer sequence in the promoter , verified that these constructs did not abolish URA3 function , and then performed fluctuation tests using the maximum likelihood method to determine the mutation rate to URA3 inactivation ., We observed that ( G ) 11 and ( G ) 12 conferred a significant increase in the phenotypic mutation rate compared to the wild type ( paired t test , p<0 . 001 , Figure 7 ) ., Insertion of a shorter poly ( G ) sequence also conferred an increased rate , but the changes were less significant ., On the other hand , the insertion of a random 12-mer sequence , poly ( A ) , and poly ( TG ) showed no effect on the mutation rate ., The fact that poly ( G ) causes an increase in the mutation rate is interesting considering that tetranucleotides composed of G or C bases are absent in the URA3 gene and are 5–10-fold less common across E . coli , S . cerevisiae , and Drosophila genomes than A or T tetranucleotides ( unpublished data ) ., In order to determine if clusters of indels and substitutions influenced coding sequences in humans , we used alignments of recent segmental duplications ( <5% diverged ) 31 to detect indels in the human genome , restricting our analysis to those sequences that had been confirmed as expressed ( see Materials and Methods ) ., We found that indels and nucleotide substitutions occurring in human transcribed sequences follow the same patterns observed in other species , confirming that indel/region/repeat-associated mutation impacts genes expressed in humans ( Figure 8 ) ., Here we have provided evidence suggesting that regional effects have a strong influence on the accumulation of nucleotide substitutions over evolutionary time scales ., Although an indel effect is also observed , we have shown the proportion of D attributable to an indel effect diminishes over time ., In addition , it is not possible to formally exclude whether this effect is due to a mutagenic indel effect or a single multiple mutation causing event ., Although we found that many indels are associated with repeat sequences , many are not ., This finding may be explained by the existence of other non-repeat polymerase stalling sequence motifs; another possible explanation is that repeat sequences were destroyed by mutation , while the indel remained ., So what is the impact of the indel/region effect on phenotypic evolution ?, Most indels in E . coli are within 100 bp of the nearest gene ( Figure S5 ) ., In S . cerevisiae , 25% of promoters contain repeat sequences 32 and 600 seven-nucleotide homopolymer runs have been identified in essential genes 33 , putting cis-regulatory regions and coding sequences well within the range of the effect of indel/repeat-associated mutation ., The genomes and accession numbers used for E . coli/Shigella and S . paradoxus analyses are shown in Table S1 ., Genome sequences for alignments between Drosophila species were downloaded from the UCSC database ( http://www . biostat . wisc . edu/~cdewey/fly_CAF1/ ) , while those for melanogastor/melanogastor alignments were downloaded from http://www . dpgp . org ., The alignments of recent human segmental duplications were provided by 31 ., For pairwise comparisons , genome sequences were aligned using BLAST with default parameters and divided into orthologous regions of at least 3 kb in length and >80% nucleotide sequence identity ., Any region that could be aligned to multiple locations was not considered for analysis , ensuring that only orthologous sequences were used ., A program was written in Perl script to find indel mutations within orthologous regions; those regions not containing indels were discarded ., For three and four genome alignments , orthologous regions that were not common to all strains were discarded and those regions remaining were realigned using ClustalW ., In order to determine in which of two aligned fragments an indel has occurred , an appropriate outgroup was selected using the phylogenetic tree 34 and confirmed by our own approximations of relatedness ( Table S4 ) ., In addition to establishing in which of the fragments the indel had occurred , the number of nucleotide substitutions occurring in the indel containing haplotype ( Di ) and non-indel containing haplotype ( Dni ) was determined by comparison with an outgroup sequence ., For instance , when three genomes were aligned to determine indel and non-indel haplotypes , the number of mutations on the non-indel haplotype was counted by comparison of the non-indel fragment with the outgroup , and the number of substitutions on the indel haplotype was calculated by comparing the indel haplotype and the outgroup ., Statistical comparisons between indel- and non-indel-containing haplotypes were carried out using the non-parametric Kolmolgorov-Smirnov paired test ., See the statistical analysis plan below for more details ., An indel was designated as contiguous with a repeat for cases where the indel occurred inside the repeat ( A-AAA , AA-AA , or AAA-A ) , or immediately next to it ( −AAAA or AAAA− ) where − denotes the position of the indel ., It was defined as near a repeat if any part of a repeat was within five nucleotides on either side of the indel ( AAAANNN− , AAAAN− , etc . ) ., For the search for regions of high D on the basis of repeat sequence density , we used three E . coli strains not previously used in this study ( E . coli SE11 , E . coli SE15 , and E . coli B Str . REL606 ) ., We searched for repeat-rich regions by first generating pairwise alignments ( as described for the indel analysis above ) , followed by generating non-overlapping 100-bp windows and binning of windows according to the number of homopolymer repeats of at least 4 nt in length ., Repeat sequences interrupted by a substitution mutation so that the homopolymer was less than four continuous nucleotides in length were not included ., We then calculated total D for each window as well as the D for these classes of mutation: substitution , indel , transition , and transversion ., To test for statistically significant differences between different classes or 100-bp windows , we used the Wilcoxon Sum Rank test ., In order to extract indel-flanking sequences for analysis , the positions of indels were recorded in each orthologous region ., Next , the sequences ( 1 kb ) both up- and downstream were extracted and examined for additional indels ., If one of the flanking sequences was found to contain additional indels , that flanking region was discarded ., The sequence surrounding the indel was named and ordered into windows ( Figure S1 ) ., For every analysis in this study , the nucleotide divergence ( D ) was calculated for each window using the Jukes-Cantor method 35 ., Pairs of recently diverged strains were chosen based on a phylogenetic tree ( Figure 1B ) ., Each of these designations as highly related was supported by our own estimations of divergence provided by pairwise alignments ( Table S4 ) ., Two pairs of recently diverged strains were aligned by performing a new alignment of all four orthologous fragments in ClustalW , giving a total of four aligned genomes ., New indels were those that occurred within pairs of recently diverged strains; for indels to be detectable , they must have occurred since the recent divergence of these two strains ( see Figure S2 ) ., D for new indels was calculated using the alignment of two similar strains , of which one had been found to contain the indel ., Old indels were those sites which concurred within recently diverged pairs but were different between the two pairs ( see Figure S2 ) ., Such indels must have happened before the divergence of the highly similar strains yet after the divergence of the two sets of strains ., For calculating D , one from each of the sets of similar genomes was selected , so that two highly diverged genomes were compared and from this comparison D is calculated for old indels ., If there are double mutations ( sites where the two similar genomes are different from each other and the other diverged pair ) , these are scored as one substitution because the difference between the two similar strains must have happened since the divergence of the two diverged sets of strains and have already been scored in the new-indel analysis ., The background divergence ( Db ) used for the regression shown in Figure 3C was calculated as the average D from windows 3 to 10 for each E . coli pairwise alignment ( window 1 comprises the 50 bp closest to the indel; windows 3 to 10 were assessed as consistently outside the range of influence of the indel ) ( see Figure 1A ) ., The indel-associated divergence was calculated by subtracting the values obtained for Db from the value of D at window 1 ., For pairwise comparisons between indel and non-indel haplotypes , previous studies have used paired t tests , however we found that our data was not normally distributed ( Shapiro-Wilk test for normality , p<0 . 05 ) ., We used the two-sample Kolmogorov-Smirnov test to test for the appropriateness of the non-parametric Wilcoxon Sum Rank test for our samples ., If the samples were found to be different by the Kolmogorov-Smirnov test , the Kolmogorov-Smirnov test was named and p value given ( as was the case for the indel/non-indel analysis ) ., If the two-sample Kolmogorov-Smirnov test found the samples under comparison to be of the same shaped distribution , we carried out and presented the Wilcoxon Sum Rank test and p values ( this was the case for the repeat/window analysis ) ., A comparison of the amount of nucleotide substitutions attributable to the indel and regional effects for indels of different ages would provide for a test of the hypothesis that indel-associated mutations accumulate over time ., In principle , this could be achieved by using the sets of old and new indels used for the analysis presented in Figure 3A and 3B; however , the generation of the set of old indels required a four-genome alignment; a fifth genome needs to be added to determine the indel and non-indel haplotypes ., Because of our strict criteria for defining orthologous regions , the partitioning of the old and new indel sets into indel and non-indel haplotypes leaves prohibitively few orthologous regions for analysis ., An alternative is to consider pairwise sets of alignments ., The background nucleotide diversity for each pairwise comparison ( Figure 1 ) provides a measure of relatedness; the greater the average value of background D , the more diverged the two strains ., In order to gauge the range and degree of difference across these pairwise comparisons , the sets of background diversity values ( provided by the D values for windows 3 to 10 , which were chosen because they are outside the range of indel/region-associated influence ) were compared ., We found that most strains had distinct levels of sequence divergence from each other ( Tukeys HSD , p<0 . 05 , Table S4 ) , with an approximately 20-fold difference in D values between the most and least diverged strains ( see Table S4 for details ) ., In order to cover a range of pairwise comparisons of increasing divergence , we chose four strains and systematically compared them to strains from clades of increasing divergence ., The least divergent outgroup was always chosen ., Each value of D can be partitioned into composite fractions ( Figure 3D and 3E ) ., Di is attributable to the effect of the indel and the region together , whereas Dni is attributed to the region alone ., ( Di − Db ) / ( Dni − Db ) provides a measure of the total proportion of Di that is influenced by the indel ., If ( Di − Db ) / ( Dni − Db ) =\u200a1 , none of the increase in nucleotide diversity can be attributed to the indel ., As the value increasingly exceeds one , more of the nucleotide substitutions surrounding indels can be attributed to the indel effect ., The indels detected in pairwise comparisons of more diverged strains cannot be strictly called “old” indels; these pairwise alignments will also include indels that have occurred relatively recently ., However , increasingly divergent strains will be composed of a greater proportion of relatively old indels ., This method of comparing indels between less diverged and more diverged strains will therefore un

Introduction, Results and Discussion, Materials and Methods

The genome-sequencing gold rush has facilitated the use of comparative genomics to uncover patterns of genome evolution , although their causal mechanisms remain elusive ., One such trend , ubiquitous to prokarya and eukarya , is the association of insertion/deletion mutations ( indels ) with increases in the nucleotide substitution rate extending over hundreds of base pairs ., The prevailing hypothesis is that indels are themselves mutagenic agents ., Here , we employ population genomics data from Escherichia coli , Saccharomyces paradoxus , and Drosophila to provide evidence suggesting that it is not the indels per se but the sequence in which indels occur that causes the accumulation of nucleotide substitutions ., We found that about two-thirds of indels are closely associated with repeat sequences and that repeat sequence abundance could be used to identify regions of elevated sequence diversity , independently of indels ., Moreover , the mutational signature of indel-proximal nucleotide substitutions matches that of error-prone DNA polymerases ., We propose that repeat sequences promote an increased probability of replication fork arrest , causing the persistent recruitment of error-prone DNA polymerases to specific sequence regions over evolutionary time scales ., Experimental measures of the mutation rates of engineered DNA sequences and analyses of experimentally obtained collections of spontaneous mutations provide molecular evidence supporting our hypothesis ., This study uncovers a new role for repeat sequences in genome evolution and provides an explanation of how fine-scale sequence contextual effects influence mutation rates and thereby evolution .

An intriguing observation made during the comparison of genomes is that insertion and deletion mutations ( indels ) cluster together with nucleotide substitutions ., Two ( not mutually exclusive ) hypotheses have been proposed to explain this phenomenon ., The first postulates that an indel mutation causes an increase in the likelihood of the surrounding sequence incurring nucleotide substitutions , while the second claims that the region of DNA in which such a cluster is located is more likely to sustain both indels and substitutions ., Here , we present evidence suggesting that the region of DNA , and not the indel , is associated with the accumulation of clusters of mutations over evolutionary time scales ., We find that repeat sequences are closely associated with a large proportion of indels and that the abundance of repeat sequences is linked with regions of increased nucleotide diversity ., By analysing molecular data and measuring the mutation rates of genes engineered to contain repeats , we find that the mutation rate can be manipulated by the insertion of long repeat sequences ., On the basis of these results , we propose a model in which repeat sequences are prone to cause stalling of the high-fidelity DNA polymerase , leading to the recruitment of error-prone repair polymerases which then replicate the surrounding sequence with a higher-than-average error rate .

genetics and genomics/genomics, genetics and genomics/microbial evolution and genomics, evolutionary biology/microbial evolution and genomics, genetics and genomics/comparative genomics, evolutionary biology/evolutionary and comparative genetics, evolutionary biology/human evolution, molecular biology/molecular evolution, microbiology/microbial evolution and genomics, evolutionary biology/genomics, molecular biology/bioinformatics, genetics and genomics/bioinformatics, molecular biology/dna repair, genetics and genomics/population genetics

The authors propose that short repeat sequences may play an important role in causing the pervasive clustering of mutations across diverse genomes from prokaryotes to humans.

journal.pgen.1002563

Intracranial Aneurysm Risk Locus 5q23.2 Is Associated with Elevated Systolic Blood Pressure

Intracranial aneurysms ( IA ) are berry-shaped pouches at the branching sites of cerebral arteries ., 2–5% of the world population is estimated to harbor IA 1 ., Most IA go unnoticed during ones lifetime ., However , when they become symptomatic , it is usually due to rupture , causing subarachnoid hemorrhage ( SAH ) ., SAH is devastating intracranial bleeding , and half of those with SAH die within a year 2 , 3 ., SAH affects the working age population , with a median age of 55 4 ., Its incidence in Finland is 19/100 000/year 5 , 6 , triple than that of the rest of the world ., The reason for this higher than average incidence is unknown ., Aneurysmal SAH places a heavy burden on society both emotionally and financially ., The strongest known non-modifiable risk factor of SAH is family history of the disease , and the strongest modifiable risk factors are smoking , excessive alcohol intake , and hypertension 7 ., An important step in tackling SAH is to understand why IAs develop ., Our understanding of the environmental and genetic background of IA formation is limited ., Positive family history of IA or SAH , older age and female sex increase the risk of developing IA 1 ., Of the general cardiovascular risk factors , smoking has been shown to increase the risk of IA formation 8 , and high blood pressure has long been speculated to do so 9 ., The high , often undocumented , prevalence of high blood pressure in the control populations is likely the reason why it frequently fails to reach statistical significance as an IA risk factor 1 ., Chronic hypertension may contribute to IA formation by imposing constantly high shear stress on vascular walls 9 ., Multiple factors , such as familial aggregation of the disease , make a genetic contribution likely to the risk of IA ., A minority of IAs show familial aggregation ( under 10% ) 7 ., Linkage studies in IA families have highlighted numerous genetic regions and a recent exome sequencing study identified coding mutations in familial thoracic aortic aneurysm with intracranial aneurysm 10 ., However , the majority of IA is sporadic ., Sporadic IA is a complex disease and no gene with a certain role has been identified yet ., Recent genome-wide association studies ( GWAS ) 11 , 12 involving Finnish IA patients , have attempted to decipher the complex genetic background of IA ., From these studies , five loci emerged with strong association to IA ( p<5E-07 , posterior probability of association –PPA>0 . 5 ) , with the highest statistical significance at 9p21 . 3 , a risk locus of multiple cardiovascular diseases ., Further 14 loci exhibited suggestive association to IA ( 0 . 1≤PPA<0 . 5 ) ., Despite the success of GWAS in identifying IA susceptibility loci , the pathomechanism by which they contribute to IA formation remains elusive ., We hypothesize that hypertension , a strong modifiable risk factor of IA , may possess an overlapping genetic background with IA ., To test this hypothesis , we analyzed the IA loci so far identified , in well-characterized population-based cohorts consisting of more than 210 000 individuals with blood pressure measurements ., 41 SNPs from 19 independent IA loci 13 were first analyzed for association with blood pressure in the national Health 2000 survey ( H2000 ) 14 discovery cohort of 1581 individuals without blood pressure lowering medication ( Table S1 ) ., We adjusted the analysis for age and gender ( ROBUST model ) ., The most significant association ( p<0 . 1 ) were observed at 2q33 . 1 with diastolic blood pressure ( DBP ) and with mean arterial pressure ( MAP ) , at 4q31 . 23 and 19q13 . 12 with DBP , and at 5q23 . 2 with systolic blood pressure ( SBP ) , DBP , and MAP ., We did not detect association with pulse pressure ( PP ) ( Table S1 ) ., Next , we wanted to analyze the independence of the association signals observed ., We tested all 19 loci SNPs adjusting for further factors known to affect blood pressure , namely smoking habits , alcohol consumption , and body mass index ( BMI ) ( ADVANCED model ) ., There was no tendency of association with DBP at 4q31 . 23 and 19q13 . 12 with the ADVANCED model ., The strength of the association decreased for the four SNPs at the 2q33 . 1 locus for SBP , but increased marginally for DBP , and MAP ., At 5q23 . 2 the strength of association increased substantially for most blood pressure measurements , such as SBP , DBP , and MAP ( Table 1 and Table S2 ) for all three SNPs tested ., The IA risk alleles at 5q23 . 2 were associated with elevated blood pressure ., To confirm the initial association signals at the 2q33 . 1 and 5q23 . 2 loci observed in the H2000 discovery cohort , we tested them for association with blood pressure in three additional population-based cohorts from Finland ., SNPs at 2q33 . 1 failed to show significant association with DBP and MAP in any of the replication cohorts ( the Cardiovascular Risk in Young Finns Study-YFS 15 , 16 , the Northern Finland Birth Cohort 1966-NFBC1966 17 , and the Helsinki Birth Cohort Study-HBCS 18 ) ., When the results were combined from all cohorts in a fixed effect meta-analysis , they remained non-significant ( Table 2 ) ., At 5q23 . 2 SNPs showed significant association with SBP in YFS and NFBC1966 ( Table 1 ) ., In HBCS , although consistent in the direction of the effect , the association remained suggestive ., When the results were combined from all cohorts in a fixed effect meta-analysis , we detected significant association with SBP at 5q23 . 2 ( prs570682\u200a=\u200a4 . 80E-05 , prs2287696\u200a=\u200a6 . 81E-05 , prs335206\u200a=\u200a3 . 01E-05 ) ( Table 1 ) ., Comparing the mean SBP of the study participants stratified for their 5q . 23 . 2 genotypes indicated a positive correlation between the number of risk alleles and higher SBP for all three SNPs tested ( Figure 1 ) ., Study participants homozygous for the risk allele ( C , in the case of rs335206 ) , had on average 1 . 3 Hgmm higher SBP compared to those who were homozygous for the protective allele , and 0 . 9 Hgmm higher than those with the heterozygous genotype ., This effect size is comparable to those of most blood pressure loci identified by The International Consortium for Blood Pressure Genome-wide Association Studies ( ICBP-GWAS ) consortium 19 ., The observed linear effect of risk allele count is strongly suggestive of a true association ., Association at 5q23 . 2 with DBP ( prs570682\u200a=\u200a0 . 02 , prs2287696\u200a=\u200a0 . 04 , prs335206\u200a=\u200a0 . 03 ) and MAP ( prs570682\u200a=\u200a0 . 0007 , prs2287696\u200a=\u200a0 . 0010 , prs335206\u200a=\u200a0 . 0004 ) showed a reduction of significance when results were combined from all cohorts ( Table S2 ) ., To test whether the association at the 5q23 . 2 locus is unique to the Finnish cohorts , we attempted to replicate the association with the three SNPs in the multinational cohort ICBP-GWAS 19 ., All three SNPs showed significant association with SBP ( prs570682\u200a=\u200a0 . 0065 , prs2287696\u200a=\u200a0 . 00079 , prs335206\u200a=\u200a0 . 0014 ) in the ICBP-GWAS cohort of 200 000 individuals of European descent ., The risk allele for elevated SBP in the ICBP-GWAS cohort was the same as in our meta-analysis of four Finnish population-based cohorts ., When the results from the four Finnish cohorts were combined with the ICBP-GWAS results in a fixed effect meta-analysis , the strength of the association increased with all three SNPs tested ( Table 3 ) ., The strongest association was observed with rs2287696 ( pALL\u200a=\u200a8 . 13E-07 ) ., This suggests that the variant at 5q23 . 2 is a common risk factor present in multiple populations of European descent ., Further loci or results for DBP or MAP were not tested for association in ICBP-GWAS , since they failed to show significant association in our replication cohorts ., All three tested SNPs at 5q23 . 2 reside in intronic regions of the gene PR domain containing 6 ( short form: PRDM6 ) and showed comparable p-values ., To further explore the associated region in an attempt to pinpoint the causative variant , we examined all 1000 Genomes variants around PRDM6 in the four Finnish cohorts ( Figure 2 ) ., The strongest association was observed with rs163189 ( p\u200a=\u200a6 . 12E-06 ) near rs570682 and rs2287696 , in the second intron , where the most significantly associated SNPs clustered ., All five of the strongest associated SNPs are located within a 4 . 7 kb region at 122 . 4 MB ( Human genome build 36 ) , surrounding a Sterol regulatory element binding transcription factor 1 ( SREBP1 ) binding site ( Figure, 2 ) 20 ., Hypertension , a leading cardiovascular risk factor , is a strong modifiable risk factor for IA and its deadly rupture ., Our study establishes a genetic link between elevated SBP and IA formation ., Further , we demonstrate the benefits of using population isolates for mapping complex disease loci valid in multiple populations ., 5q23 . 2 was identified as a suggestive IA risk locus by Yasuno and colleagues 13 in a multinational GWAS including Finnish IA patients ., The strength of the association at 5q23 . 2 in their study mainly came from the Finnish cohort ( Figure S1 ) ., However , albeit weaker , association to IA at 5q23 . 2 was observable in all cohorts tested by Yasuno and colleagues ., In the two tier approach we applied , the suggestive aneurysmal locus at 5q23 . 2 showed robust association to blood pressure traits in three cohorts ( namely the discovery cohort H2000 , and the replication cohorts NFBC1966 and YFS ) ., The trend of the effect was the same while the association remained suggestive with blood pressure traits in the HBCS ., HBCS participants average age was higher ( 61 years ) than that of the rest ( 36 years ) ( Figure S2 ) ., With age , the relative contribution of genetic predisposition and lifestyle may change , potentially accounting for the less significant association in HBCS ., In our meta-analysis of candidate loci the most significant association was observed at 5q23 . 2 in PRDM6 ., Although an association can be observed throughout the whole gene , fine-mapping of the region with 1000 Genomes variants revealed the focus of association to be within a 4 . 7 kb region in the second intron ( Figure 2 ) ., PRDM6 encodes an epigenetic modulator of transcription with roles in endothelial 21 and vascular smooth muscle cells ( SMC ) 22 ., PRDM6 has a critical role in arterial wall SMC , where it is predominantly expressed ., PRDM6 participates in the phenotypic switch between proliferative and differentiating vascular SMC phenotypes 22; when active , PRDM6 inhibits differentiation and promotes proliferation ., Excess vascular SMC proliferation is an important pathomechanism in hypertension , and it exacerbates the vascular wall remodeling often seen in IA 23 , 24 ., When vascular SMCs re-enter the cell cycle to proliferate , they lose their contractile qualities ., Distinct from extracranial arteries , cerebral arteries lack an external elastic lamina and the adventitia is weakly developed , making them inflexible , and less resistant to stress 25 ., It is possible that when SMC proliferation further stiffens cerebral arteries , they become incapable of adjusting to shear stress , and give way to IA formation ., This is a plausible explanation to why the intracranial manifestation of a supposedly generalized vasculopathy can be so distinct ., Intriguingly , excessive vascular SMC proliferation is part of the pathomechanism of the strongest common IA risk locus at 9p21 . 3 26 ., However , to test possible causality , examination of whether the risk variant at 5q23 . 2 is associated with higher PRDM6 activity is necessary ., Although the causative variant remains elusive , we succeeded in narrowing down the associated region markedly ., The 4 . 7 kb region showing the strongest association harbors a SREBP1 binding site ., SREBP1 is a transcription factor governing cellular lipid biosynthesis ., Highlighting its biological significance in vascular traits , non-synonymous mutations in SREBP1 cause spontaneous hypertension in rats 27 ., It is possible that common variants facilitate SREBP1 binding , and thus , as shown by Zhou and colleagues 28 , cause vascular SMC proliferation ., We propose that this effect is conveyed by PRDM6 activation ., Although both the location and the function of the gene highlight PRDM6 as a likely candidate , it is not the only plausible gene near the association signal ., Centrosomal protein of 120 KD ( short form: Cep120 ) is just downstream from the region of association ( Figure 2 ) ., Cep120 is a centrosomal protein with preferentially high expression in neuronal progenitors during development 29 ., Cep120 could contribute to IA risk by causing perturbation in the neurovascular niche ., This is the first study establishing a shared genetic background at 5q23 . 2 for IA and its important risk factor , high blood pressure ., However , both IA 30 , 31 and hypertension 32 have shown linkage to 5q23 . 2 in previous studies ., Resequencing the genomic region in families that previously showed linkage to 5q23 . 2 might reveal penetrant variants causing familial IA or severe high blood pressure , or possibly both ., Notably , Vasan and colleagues 33 found that rs17470137 , less than 8 kb downstream from PRDM6 , is associated with aortic root size , a feasible proxy of blood pressure 34 ., GWAS are designed to identify associations , they do not prove causality ., Deep resequencing of the associated region may improve the fine mapping and guide closer to the causative variant , or even uncover it , although resequencing efforts of GWAS regions have had limited success 35 ., A further limitation of our study is that we were unable to address whether the identified risk variant at 5q23 . 2 increases the risk of developing IA as a consequence of elevated SBP ( causality between high SBP and IA ) or whether the variant modifies vessel wall structure in a way that elevates SBP and increases IA risk as a pleiotropic effect ( Figure S3 ) ., A study conducted in a cohort characterized both for IA and blood pressure would likely be a more suitable way of addressing this question ., Unfortunately , to the best of our knowledge , such a large-scale cohort does not currently exist ., The identified risk variant , however , is unlikely to confer its effect solely by increasing blood pressure , as leading hypertension risk loci fail to show association with IA ( data not shown ) ., Yet , the mechanical effect of elevated BP on the vessel wall , likely exacerbates IA formation ., The significance of the association identified in our study awaits confirmation in other ethnicities ., To further decipher the genetics of IA , it is important to test if genetic links can be established between IA and other strong risk factors , such as smoking and alcohol consumption ., In conclusion , our results highlight the link between IA and blood pressure ., Four Finnish population-based cohorts were included in our study ( Table 4 ) ., These cohorts were not characterized for IA ., We utilized genome-wide genotyped participants with available blood pressure data , excluding those on blood pressure medication and those for whom blood pressure medication data was not available ( nexcluded\u200a=\u200a1373 ) ., In our two tier approach , the discovery cohort ( ndiscovery\u200a=\u200a1581 ) was a subsample of the H2000 14 ., The H2000 study was carried out in several regions of Finland from fall 2000 to spring 2001 , and was designed to provide information on the health of the Finnish population ., A subset of this cohort , consisting of metabolic syndrome cases and matched controls , was genotyped and utilized in this analysis ., The replication cohort ( nreplication\u200a=\u200a8312 ) consisted of the YFS ( n\u200a=\u200a1874 ) 15 , 16 , the NFBC1966 ( n\u200a=\u200a5361 ) 17 , and the HBCS ( n\u200a=\u200a1077 ) 18 ., YFS participants were recruited from all around Finland for a large follow-up study on cardiovascular risk factors in young individuals in 1980 ., Clinical data are from the follow-up at age 27 performed in 2007 ., NFBC1966 comprises individuals born in 1966 in the two northernmost provinces of Finland ( Oulu and Lapland ) ., Clinical examinations took place at the follow-up at age 31 in 1997 ., The HBCS participants were recruited from the Helsinki region ., The study examines the impact of fetal environmental factors on childhood and adult life ., Clinical examinations took place during 2001–2004 ., HBCS participants had the highest average age ( Figure S2 ) ., The ICBP-GWAS represents a union of numerous prior blood pressure GWAS consortia to create a discovery meta-analysis of over 200 000 individuals of European ancestry ., NFBC1966 is part of the ICBP-GWAS; however , this overlap does not represent a significant risk for bias , due to the small relative contribution of NFBC1966 to the ICBP-GWAS results ., All Finnish cohorts were genotyped using Illumina arrays ( Illumina Inc . San Diego , CA , USA ) : Illumina Infinium HD Human610-Quad BeadChip for H2000 , Illumina HumanCNV370-Duo BeadChip for NFBC1966 , and Illumina Human670K custom BeadChip for YFS and HBCS ., For SNPs to be successfully genotyped , a per individual and per marker success rate minimum of 95% was defined as default ., 36 out of 41 candidate SNPs were successfully genotyped in all cohorts ., For SNPs with no directly genotyped data available , we imputed genotypes with MACH 36 using HapMap CEU from Phase II as the reference panel ( further referred to as HM2 imputed data ) ., If a SNP was not present in HM2 imputed data , we used genotypes imputed with IMPUTEv2 using the 1000 Genomes pilot data CEU panel ( August 2009 haplotypes ) combined with HapMap Phase 3 ( Public Release #2 ) haplotypes as the reference panel 37 , extended with Finnish specific HapMap Phase 3 haplotypes 38 ( further referred to as 1000G+HM3 imputed data ) ., All missing genotypes were imputed , so the number of individuals included in the analyses for each SNP is the same and equals the final number ( Table 4 ) ., Candidate loci were selected based on IA GWAS results 13 ., Loci associated with IA with PPA≥0 . 1 were included ( Table 2 ) ., PPA was calculated as described by Yasuno et al 12 ., Briefly , a uniform prior probability of association of 1/10 000 was assumed for all SNPs and used to provide a probabilistic measure of evidence ., We tested 41 SNPs from 19 independent loci ., We defined SBP , DBP , MAP , and PP as quantitative outcome variables ., MAP was counted as the average of SBP and DBP ( ( SBP+DBP ) /2 ) and PP as the difference of the two ( SBP-DBP ) ., We tested all 19 loci in the discovery cohort ( H2000 ) , and those showing suggestive association ( uncorrected p<0 . 1 ) with any outcome variable were tested in the replication cohorts ( YFS , NFBC1966 , and HBCS ) ., Association analyses with an additive genetic model were performed with ProbABEL 39 for HM2 imputed data , and with SNPTESTv2 40 , 41 for 1000G+HM3 imputed data ., The analyses were adjusted for age and gender ( ROBUST model ) , or for age , gender , smoking habits , alcohol consumption , and BMI ( ADVANCED model ) ., Additionally , in the metabolic syndrome case-control subset of the H2000 cohort we corrected for case-control status in both models ., Population stratification was corrected for by calculating principal components from genome-wide SNP data and including significant principal components in the association models as covariates ., Association results were combined in a fixed effect meta-analysis with MetABEL 39 for HM2 imputed data , and with METAv1 . 2 42 for 1000G+HM3 imputed data ., The best result at 5q23 . 2 in PRDM6 was tested for association in the ICBP-GWAS 19 cohort of 200 000 individuals of European descent ., In the ICBP-GWAS association with SBP was tested by linear regression assuming an additive model and correcting for age , age-squared and BMI ., To test the per-allele effect size of risk alleles on blood pressure , we calculated the mean blood pressure for the three genotypic states for the three 5q23 . 2 SNPs using Plink v1 . 07 43 ., Results were plotted using the Microsoft Excel charts function ., To further investigate the strongest associated locus , we analyzed all 1000 Genomes variants , with minor allele frequency greater than 1% , in and around PRDM6 ., We took uncertainty of imputation into account by using the maximum likelihood estimates of the reference allele counts as genotypes ( these estimates may be fractional and range from 0 to 2 ) ., Fine mapping of the 5q23 . 2 region was performed with 1000G+HM3 imputed data ., Results were plotted with LocusZoom 44 .

Introduction, Results, Discussion, Materials and Methods

Although genome-wide association studies ( GWAS ) have identified hundreds of complex trait loci , the pathomechanisms of most remain elusive ., Studying the genetics of risk factors predisposing to disease is an attractive approach to identify targets for functional studies ., Intracranial aneurysms ( IA ) are rupture-prone pouches at cerebral artery branching sites ., IA is a complex disease for which GWAS have identified five loci with strong association and a further 14 loci with suggestive association ., To decipher potential underlying disease mechanisms , we tested whether there are IA loci that convey their effect through elevating blood pressure ( BP ) , a strong risk factor of IA ., We performed a meta-analysis of four population-based Finnish cohorts ( nFIN\u200a=\u200a11 266 ) not selected for IA , to assess the association of previously identified IA candidate loci ( n\u200a=\u200a19 ) with BP ., We defined systolic BP ( SBP ) , diastolic BP , mean arterial pressure , and pulse pressure as quantitative outcome variables ., The most significant result was further tested for association in the ICBP-GWAS cohort of 200 000 individuals ., We found that the suggestive IA locus at 5q23 . 2 in PRDM6 was significantly associated with SBP in individuals of European descent ( pFIN\u200a=\u200a3 . 01E-05 , pICBP-GWAS\u200a=\u200a0 . 0007 , pALL\u200a=\u200a8 . 13E-07 ) ., The risk allele of IA was associated with higher SBP ., PRDM6 encodes a protein predominantly expressed in vascular smooth muscle cells ., Our study connects a complex disease ( IA ) locus with a common risk factor for the disease ( SBP ) ., We hypothesize that common variants in PRDM6 can contribute to altered vascular wall structure , hence increasing SBP and predisposing to IA ., True positive associations often fail to reach genome-wide significance in GWAS ., Our findings show that analysis of traditional risk factors as intermediate phenotypes is an effective tool for deciphering hidden heritability ., Further , we demonstrate that common disease loci identified in a population isolate may bear wider significance .

When multiple genes or genetic regions contribute to the inherited risk of a disease , it is referred to as a complex disease ., Genome-wide association studies ( GWAS ) aim to detect common genetic variations that associate with complex traits or diseases ., Although GWAS have been successful in identifying strongly associated genetic loci , they lack the means to point out true , but less strong , associations ., Studying conditions that are related to the disease of interest can help sort out less strong associations ., Intracranial aneurysms ( IA ) are berry-like dilations in cerebral arteries ., Most IAs do not give symptoms until they bleed , causing a highly fatal form of stroke ., Half of the people who suffer bleeding of an IA die ., IA is a complex disease ., Both inherited risk and environmental factors contribute to the risk of developing IA ., Women , smokers , those with high alcohol intake or high blood pressure are more prone to develop IA and bleeding ., GWAS found 19 genetic regions increasing the risk of IA ., Here we show that one of these loci , on the long arm of chromosome 5 , in addition to raising IA risk also increases systolic blood pressure ., We speculate that the cause is modified vascular wall structure .

medicine, neurological disorders, neurology, genetics, biology, genetics and genomics, cardiovascular

journal.pgen.1000157

Selective Constraints in Experimentally Defined Primate Regulatory Regions

Changes in gene regulation are likely to play an important role in evolution 1 , 2 ., However , compared to protein-coding sequences , the fitness effects of regulatory mutations remain poorly understood ., Furthermore , the relationship between changes in gene regulatory regions and the expression phenotype of the regulated gene are unclear ., Both of these issues are partly a result of poor annotation of the sites that control gene regulation , the vast majority of which are likely to be noncoding ., For example , transcription factor binding sites ( TFBSs ) , a major component of regulatory architecture , are small ( 6–15 bp ) , laborious to identify experimentally and potentially degenerate ., Furthermore , due to their small size , genuine TFBS are difficult to differentiate from similar , randomly-occurring sequences that are present in large numbers in mammalian genomes ., In an attempt to address the problem of annotation , evolutionary conservation has become popular as a metric for identifying putative regulatory regions 3 ., However estimates of the true level of selective constraint ( defined as the proportion of mutations which are strongly deleterious ) in regulatory DNA will , by definition , be biased upwards in datasets predicted using evolutionary conservation as a criterion ., One way to address this problem is to focus solely on regulatory regions which have been defined primarily by experimental rather than evolutionary criteria ., In this study , we estimated levels of selective constraint in mammalian regulatory noncoding DNA using two complementary datasets , both of which draw upon experimental data ., The first was derived from the literature collected in the TRANSFAC database 4 , such that every TFBS is supported by at least a single refereed publication ., The advantages of this dataset are twofold ., First , our dataset consists of individual TFBSs for which experimental support exists and which , according to an analysis by publication date ( see Discussion ) , appear to be subject to relatively little ascertainment bias with respect to evolutionary conservation ., Second , the literature in TRANSFAC also provides substantial information on the gene regulated and the transcription factor bound for each TFBS ., Thus , the TFBSs in our dataset can be assigned to a specific gene reliably , and we can also determine at least some of the transcription factors ( TFs ) which regulate a specific genes expression ., Our second dataset comprises sequences which have been identified as potentially transcription-factor-binding using chromatin immunoprecipitation combined with genomic microarray ( ChIP-chip ) analyses ., Specifically , we combine the sequences annotated in refs 5–11 ., While the resolution at which regulatory sites are identified is undoubtedly lower in the ChIP-chip dataset than in our TFBS dataset , our ChIP-chip sequences will still be highly enriched for functional regulatory DNA ., Using these two datasets , we addressed the following questions:, ( i ) what fraction of regulatory mutations in primates are strongly deleterious ,, ( ii ) does the fraction of strongly deleterious mutations at TFBSs vary between primates ,, ( iii ) what fraction of substitutions in human regulatory regions have been driven to fixation by positive selection ,, ( iv ) how does the selective constraint of human regulatory noncoding regions relate to the expression profile of the gene they regulate and, ( v ) does the rate of protein evolution of a TFs also relate to the expression profile of the regulated gene ?, We next estimated the level of selective constraint at TFBSs , pCRMs and in ChIP-chip sequences ( Figure 4 ) ., TFBSs appear to be reasonably highly constrained , approximately equivalent to a 2-fold degenerate synonymous site ., This result is in good agreement with previous studies which have suggested that a reasonable proportion of TRANSFAC binding sites are conserved between human and a variety of mammalian species 12–16 ., Estimates of constraint in our putative cis-regulatory modules and sequences annotated by ChIP-chip experiments are somewhat similar ( 0 . 14 and 0 . 11 , respectively ) suggesting that ChIP-chip studies can serve as reliable guides to functional regulatory regions in humans when compared with more traditional methods of identification ., It has been suggested that regulatory DNA in primates is under relaxed selective constraint relative to rodents 17 ., This has been attributed to the reduction of effective population size in primates facilitating the fixation of slightly deleterious mutations in gene control regions ., Effective population size is likely to vary between humans , chimpanzees and macaques and we therefore investigated whether any significant difference in constraint of regulatory noncoding regions existed between humans and their close relatives ., It is clear from Figure 5 that selective constraints in regulatory noncoding DNA vary significantly between primate species ( 1-way ANOVA , P<10−16 ) , and that this is primarily a result of a reduction in constraint in hominins ( post-hoc Tukey test , P<10−5 human vs rhesus , chimp vs rhesus ) when compared with rhesus macaques ., Summing over both coding and noncoding sites , we also find that mean selective constraint is also reduced somewhat in humans , compared with chimpanzees ( 0 . 254 versus 0 . 279 ) , although this difference is only marginally significant ( Bootstrap t-test , P<0 . 07 ) ., Two possible explanations for the reduction in constraint are that humans and chimpanzees have accumulated substantially greater numbers of deleterious mutations or are experiencing higher rates of adaptive evolution in their regulatory regions ., In order to investigate whether the reduced constraints we observed in human regulatory noncoding DNA were the result of adaptive evolution we estimated the proportion of substitutions ( α ) which were driven to fixation by positive selection in our pCRMs and ChIP-chip sequences using the McDonald-Kreitman framework 18 , 19 ., We were able to map 232 of our pCRMs and ChIP-chip sequences onto regions sequenced by the NIEHS Environmental Genome Project ( EGP; http://egp . gs . washington . edu ) ., Polymorphism data was taken from the EGP as this dataset is free of ascertainment bias , relative to other large polymorphism datasets , such as HapMap 20 ., McDonald-Kreitman analyses assumes that all mutations can be divided into strongly selected ( positively or negatively ) or strictly neutral classes ., One problem with this is that a non-negligible fraction of new mutations in species with small effective population sizes , such as primates , may be weakly negatively selected ., To account for this possibility we estimated α using both all segregating sites and excluding those sites where the minor allele frequency MAF ranged from 0 . 01 to 0 . 30 , many of which are likely to be slightly deleterious 21 ., We find no evidence of adaptive evolution in human regulatory regions and our estimate of α is not significantly different from zero across the entire range of excluded , low frequency polymorphisms ( Figure 6 ) ., We next investigated whether constraint in our regulatory sites covaried with the expression breadth of the gene regulated ., Expression profile of the genes inferred to be regulated by our pCRMs and ChIP-chip sequences was estimated from the human microarray data of Su et al 22 ., A gene was defined as expressed in a specific tissue based on the Affymetrix MAS5 presence/absence calls ., Our results were qualitatively unchanged when gene expression was designated using a cutoff probe intensity value ( data not shown ) ., For comparison , we also estimated constraint at the nonsynonymous ( 0-fold degenerate ) and synonymous sites of the genes adjacent to our regulatory regions ., The results of this analysis are presented in Figure, 7 . There is a clear relationship between selective constraint in regulatory regions and breadth of expression of the regulated gene ., pCRM and ChIP-chip selective constraint is significantly negatively correlated with the number of tissues in which a gene is expressed ( P<0 . 005 , P<5 . 07×10−7 , respectively ) ., This is not a function of the number of annotated TFBSs in our pCRMs , which is uncorrelated ( Pearson r\u200a=\u200a0 . 015;P<0 . 738 ) with pCRM constraint ., A similar relationship appears to exist between TFBS selective constraint and expression breadth ., In particular , the TFBSs of tissue-specific genes are more highly constrained than those of intermediate and broadly expressed genes ( 2-sided t-test; P<0 . 012 ) ., However , the equivalent regression is not significant at least in part due to the high error involved in estimating selective constraint from a small number of sites between closely related species ., The relationship between constraint and expression profile is reversed in protein-coding sequence where constraint increases with increasing expression breadth ( P<1 . 11×10−15 and P<1 . 70×10−6 , nonsynonymous and 2-fold degenerate , respectively ) , a result supported by previous work 23 , 24 , 25 ., Interestingly , constraint at 4-fold degenerate synonymous sites is also positively correlated with expression breadth ( P<2 . 60×10−6 ) , suggesting that constraints on mRNA stability and/or splicing efficiency reflect those on protein structure , with respect to expression breadth ., These results are not a product of different rates of nucleotide substitution in the intronic control regions of genes with differing expression breadth; we find that divergence in all controls used in our study was uncorrelated with breadth of expression of the gene in which they reside ( Pearson r\u200a=\u200a−0 . 004 P>0 . 83; Figure S2 ) ., It has previously been shown that mammalian promoters can be divided into two classes , CpG-rich and CpG-poor , based on the distribution of %CpG in human promoter regions 26 and these two classes of promoter region are associated with expression breadth ., Following ref 26 , we divided our pCRM and ChIP-chip sequences into CpG-rich and CpG-poor classes , to investigate whether this could explain the relationship we find between expression breadth and conservation ., The majority ( 95% ) of our pCRMs and ChIP-chip sequences are CpG-rich by the definition in ref 26 i . e . they have a normalized CpG content of >0 . 35 ., Within this CpG-rich class , constraint of regulatory regions is still significantly negatively correlated with expression breadth ( Pearson r\u200a=\u200a−0 . 103 , P<3 . 68×10−8 ) ., We also tested the influence of CpG content by regressing pCRM and ChIP-chip constraint on their %CpG ., The slope of this regression is negative and significantly different from zero ( simple linear regression b\u200a=\u200a−0 . 058 , P<0 . 028 ) ., However , the residuals of this regression are still negatively correlated with expression breadth ( Pearson r\u200a=\u200a−0 . 086 , P<5 . 99×10−7 ) ., These results suggest that , while CpG content is indeed correlated with constraint of regulatory DNA this does not explain the majority of the relationship we see between regulatory constraint and expression profile ., One advantage of our TFBS dataset is that we can identify which TF ( s ) control the expression of a specific gene , and that this relationship is also supported by experimental evidence ., We therefore investigated whether the rate of protein evolution ( estimated as Dn/Ds , the ratio of nonsynonymous to synonymous substitution ) in a TF bore any relationship to the expression breadth of the regulated gene ., Dn/Ds was estimated summing over all sites of all TFs which were known to regulate a specific gene ., We obtained Dn/Ds estimates for 185 TFs which regulate 349 genes ., The results of this analysis are presented in Figure, 8 . Interestingly we find that TF Dn/Ds ratio is significantly positively correlated with gene expression breadth ( Pearson r\u200a=\u200a0 . 15;P<0 . 005 ) ., We tested whether this result was an artifact of summing across multiple TFs by restricting our analysis to the 99 genes which were regulated by a single TF ., Despite this reduced dataset TF Dn/Ds is still marginally significantly correlated with gene expression breadth ( P<0 . 076 ) ., One major factor which influences the rate of protein evolution is their structure ., We therefore tested whether the relationship between transcription factor Dn/Ds and gene expression profile was influenced by the transcription factor structural class ., We divided the regulating TFs into four protein “superclasses” based on the transcription factor protein classification tree in TRANSFAC ., Of our 185 TFs we were able to assign 141 to either leucine zipper factors ( LZ; 27 TFs ) , zinc-coordinating DNA-binding domains ( ZC; 50 TFs ) , helix-turn-helix proteins ( HTH; 42 TFs ) and β-scaffold factors with minor groove contacts ( BSF; 22 TFs ) ., As expected , we find clear differences in the Dn/Ds ratio of each the four classes ( 1-way ANOVA P≪1 . 69×10−5; Figure S3 ) ., However , we find no relationship between protein structural class and expression breadth of the regulated gene ( 1-way ANOVA P<0 . 464; Figure S4 ) ., This suggests that the relationship we observe between the rate of protein evolution in a TF and the expression breadth profile of the regulated gene are independent of the protein structure of the TF ., We have presented a study of the fitness effects of mutations in primate regulatory noncoding DNA ., The regulatory regions included in our study are supported by a variety of experimental sources , both based on the extensive experimental biology literature , and inferred from more recent , high-throughput studies ., Our study confirms that experimentally validated regulatory noncoding regions are selectively constrained , a result supported by other previous studies of datasets of TRANSFAC TFBSs in mammals 12–16 ., Our estimates imply that ∼37% of new spontaneous mutations in primate TFBSs have a strongly deleterious effect and are removed by purifying selection ., We find that the proportion of strongly deleterious noncoding regulatory mutations varies significantly even between closely-related primate species , reflecting a similar trend in coding DNA ., We find no evidence for adaptive evolution in human regulatory regions , suggesting that these differences in selective constraint between primate taxa are likely to primarily reflect variations in effective population size ., Our study also clearly shows that the level of selective constraint in primate regulatory DNA depends upon the expression profile of the gene regulated ., Intriguingly , we also find higher constraint in the regulatory regions of tissue-specific genes is reflected in the rate of protein evolution of the TFs that interact with them ., Our study suggests that at least some fraction of human regulatory DNA is accumulating slightly deleterious mutations at an accelerated rate relative to other , closely-related primate species ., We find no evidence of adaptive evolution in our regulatory regions ., Nonetheless , a number of recent reports have suggested accelerated evolution in human noncoding DNA 27 , 28 , 29 ., There may be a number of reasons that we do not observe such an effect ., Firstly , we restrict our analysis to experimentally-supported regulatory noncoding DNA and exclude CpG prone sites entirely from our analysis and may therefore lack sufficient power to detect all but very strong selection ., Secondly our analysis is based upon the McDonald-Kreitman test which assumes that all adaptive mutations are strongly selected ., However , recent work has suggested that at least some fraction of adaptive mutations may be weakly selected 30 ., Although our degree of confidence in our estimates of α is small , the increasing numbers of high quality ChIP-chip datasets combined with larger resequencing studies will improve the accuracy of estimates of this important parameter ., The results we have presented also shed light on the relationship between gene expression and selective constraint of both the TF and TFBSs which ultimately control this expression ., A straightforward interpretation of our results is that selective constraint of regulatory DNA parallels the “complexity” of expression of the gene it regulates i . e . genes that are required to be “switched on” ubiquitously have a simpler , more degenerate regulatory architecture than those genes which require delicate control of the location and timing of expression ., This interpretation is supported by a recent study of human-mouse promoter regions 31 ., Furthermore , this hypothesis is intuitively appealing when we consider that tissue-specific genes may require regulatory sites both to up-regulate expression in the correct tissue , but also to suppress expression in an inappropriate tissue , a function that is presumably absent from the regulatory region of a broadly-expressed gene ., Taken together with estimates of constraint in protein-coding sequence our study suggests the following: broadly expressed genes produce a protein whose structure is tightly maintained by purifying selection but whose regulatory architecture is degenerate ., Tissue-specific genes on the other hand require a more elaborate and specific regulatory apparatus , but the protein produced by such genes is less rigorously maintained by selection ., It has been suggested that mutations affecting the regulation of tissue-specific genes are less likely to be strongly deleterious than those in broadly-expressed genes , given that they are expressed in a subset of tissues 32 ., However , our results support the opposite interpretation ., Although the correlations we observe between regulatory constraint and expression breadth are weak , we note that the experimental methods of annotation of regulatory sites are imperfect , and the numbers of sites which we have used in this study are relatively small , by genomic standards ., In addition , estimates of selective constraint are essentially a ratio of ratios , making them inherently noisy ., In the light of this , the strength of our correlations is perhaps less surprising ., It is also likely that our results to a certain extent reflect the variation in constraint of the noncoding DNA surrounding different functional “classes” of genes , as demonstrated previously ( e . g . ref 33 ) ., We note , however , that the relationship between gene expression profile and gene functional class as assigned by ontological classification is uncertain ., In addition , without a complete annotation of functional noncoding sites , we cannot distinguish whether between-gene variation in constraint of surrounding noncoding regions reflects variation in the number of constrained sites or in the intensity of purifying selection at these sites ., One advantage of the approach we have employed here is that we can at least partially disentangle these two factors; our results suggest that the intensity of purifying selection at primate TFBSs is indeed greater in tissue-specific genes ( Figure 7 ) ., We note that our estimates of constraint may also be biased upwards for two reasons ., The UCSC whole genome alignments are assembled with reference to the human genome and it is therefore possible that their use in our study could exclude weakly conserved unalignable TFBSs ., This could potentially lead to an overestimate of the true level of constraint ., However , we suggest that the impact of this is likely to be small given that such a bias will affect our control regions also , and thus will cancel in the estimation of constraint ., In addition , although ascertainment bias in the TRANSFAC annotations is reduced , compared to some computationally-predicted regions , it is unlikely to be zero , as phylogenetic footprinting has become more frequently used over time as a means of selecting candidate regulatory regions for experimental testing ., Unfortunately it is difficult to quantify this bias ., However , if phylogenetic footprinting has had a significant effect upon our TFBS dataset we might predict that , on average , those TFBSs that were annotated relatively recently would be less diverged than those annotated in the more distant past , given the dramatic increase in the use of comparative genomics in recent years ., We find that divergence is not significantly correlated with year of appearence of the supporting publication ( Figure S5 ) ., We do find that TFBSs published before 1996 ( the median age of publication of human TRANSFAC TFBSs ) are marginally ( ∼8% ) more diverged than those published during or after 1996 , although this difference is not significant ( Bootstrap t-test , P<0 . 19 ) ., Thus , although our estimates of TFBS constraint may be upwardly biased , this bias is likely to be small ., One straightforward implication of our results is that deleterious regulatory mutations are more likely to disrupt genes with tissue-specific expression , as a result of higher levels of constraint in both their regulatory regions and the protein-coding sequence of the TFs that bind to these regions ., We estimate that deleterious mutations will occur on average 1 . 6-fold more often in regulatory regions of tissue-specific ( ≤3 tissues ) than housekeeping genes ( >35 tissues ) ., This conclusion has interesting implications when we consider recent evidence suggesting that there are substantially more tissue-specific genes in primates compared with rodents 34 ., Our data imply that the penalty for an increase in expression “complexity” is a concurrent increase in the genomic deleterious mutation rate ., This penalty may , however , be offset by a corresponding decrease in the proportion of deleterious protein-coding mutations ., The data used in this study were collected from two sources ., We first used the literature in TRANSFAC release 10 . 2 4 to compile a dataset of known , experimentally-supported TFBSs ., For those TFBSs which were linked to a specific EMBL accession , we BLASTed the binding site and up to 400 bp of flanking sequence against the human genome ( assembly 18 ) ., Query sequences which matched a single , unique region in the human genome with a BLAST e-value of <10−5 were accepted ., Those regions which matched more than a single location were resolved manually by comparison with any existing annotation in TRANSFAC , or excluded ., For those TFBSs which were not linked to an existing EMBL record , we BLASTed the binding site sequence against the transcript of the RefSeq gene regulated , as recorded in TRANSFAC , with 20 kb flanking sequence ., We accepted any binding site which matched a single unique location in this sequence , with <99% identity , for the full length of the binding site ., We hereafter refer to these data as “TFBS” sequences ., All binding sites were checked to be in the appropriate chromosomal location with respect to the gene they regulate ., Our second dataset was derived from DNA sequences bound by a variety of TFs in 7 recent chromatin immunoprecipitation-coupled DNA microarray ( ChIP-on-chip ) analyses 5–11 ., The locations of these sequences were extracted from the “fragment” table of TRANSFAC 10 . 2 and updated to the latest assembly of the human genome ., We hereafter refer to these data as “ChIP-chip” sequences ., To estimate the level of selective constraint , we needed to compare substitution rates in our TFBS and ChIP-chip datasets with those in an appropriate neutrally-evolving control region , which has a mutation rate equal to that of the region of interest ., Previous analyses 33 , 35 have suggested that , in mammals , intronic regions outside the first intron and the splice sites are the fastest evolving in the genome and among the best candidates for neutrally-evolving sequence ., Because sites in both datasets were highly nonrandomly distributed across the genome , we sought to define a single control region for a “case” region of binding sites , rather than for each individual annotated sequence ., A “case” was defined as a group of TFBSs or ChIP-chip sequences in which the maximum distance between each cluster member and its nearest neighbour was 100 kb ., A control region for each “case” was defined as the window which extended up to 250 kb either side of the midpoint of cluster ., All non-first intronic sequence , excluding the first and last 100 bp , within this 500 kb window were denoted as control regions for the “case” region ., Given that mutation rates in mammals appear to vary across megabase scales 36 , 37 it is likely that the mutation rate in our control sites will not differ significantly from that in our “case” sites ., In a minority of cases ( <5% of TFBSs and ChIP-chip sequences ) , suitable intronic controls were unavailable ., In this case , we used nearby intergenic sequence which was greater than 1 kb from an annotated coding sequence ., All exon locations were taken from RefSeq annotations ., The selection of an arbitrary between-site distance of 100 kb allowed us to define 473 unique , nonoverlapping binding site “cases” , each with a unique set of intronic controls ., Likewise , we defined 6712 ChIP-chip “cases” from our 10104 unique ChIP-chip regions ., For all TFBSs , ChIP-chip sequences and their corresponding control regions , aligned sequence data from the human , chimpanzee ( assembly 2 ) and macaque ( assembly 2 ) genomes was extracted from the 28-way vertebrate alignments available in the UCSC genome browser database 38 ., In order to minimize the effects of poor sequence quality in the chimp and macaque genomes we masked all sites which were assigned a base quality of less than 20 in either species ., Lineage-specific substitution rates were estimated using parsimony ., Estimates of substitution rates were not corrected for multiple hits , given that this will make little difference between closely related species ., In all cases , selective constraint , C , was estimated as:where O is the number of substitutions observed in the TFBS or other region of interest and E is the number of substitutions expected under neutral evolution:where n is the length of the TFBS or other region of interest and K is the substitution rate estimated from the control region ., Unless stated otherwise , all confidence intervals were estimated by bootstrapping the data by binding site “case” , 1000 times ., The method of estimation of selective constraint employed here explicitly accounts for local mutational variation ., Previous studies of experimentally validated mammalian regulatory DNA ( e . g . refs 14–16 ) , have not accounted for such variation ., This is particularly important in our study for two reasons ., Firstly , substantial within-genome mutational variation is known to occur in mammals 37 , 39 meaning that the expectation of conservation under neutrality will vary from one genomic region to the next ., This can substantially impact estimates of conservation between very closely related species , such as humans and chimpanzees ., Secondly , regulatory regions frequently reside in CpG islands , where the level of CpG hypermutability is known to differ from other , more heavily methylated regions of the genome ., Given that CpG mutations make up a disproportionately large number of all mutations in mammals , it is important to correct for variations in the level of CpG hypermutability to avoid overestimating constraint in regions of lowered CpG hypermutability such as CpG islands ., Here , we account for variation in the frequency and mutability of CpG dinucleotides by excluding non CpG-prone sites ( not preceded by ‘C’ or followed by ‘G’ ) ., TF Dn/Ds ratios were estimated from human-macaque alignments in the Cornell orthologues dataset using PAML 40 ., In the case where multiple factors were known to regulate a gene x , Dn/Ds ( ωx ) was estimated summing over all TFs , Tx = t1 , … , tn as:where KA ( ti ) and Ks ( ti ) are the number of pairwise nonsynonymous and synonymous substitutions in TFi , and NA ( ti ) and Ns ( ti ) are the number of pairwise nonsynonymous and synonymous sites in TFi , respectively .

Introduction, Results, Discussion, Materials and Methods

Changes in gene regulation may be important in evolution ., However , the evolutionary properties of regulatory mutations are currently poorly understood ., This is partly the result of an incomplete annotation of functional regulatory DNA in many species ., For example , transcription factor binding sites ( TFBSs ) , a major component of eukaryotic regulatory architecture , are typically short , degenerate , and therefore difficult to differentiate from randomly occurring , nonfunctional sequences ., Furthermore , although sites such as TFBSs can be computationally predicted using evolutionary conservation as a criterion , estimates of the true level of selective constraint ( defined as the fraction of strongly deleterious mutations occurring at a locus ) in regulatory regions will , by definition , be upwardly biased in datasets that are a priori evolutionarily conserved ., Here we investigate the fitness effects of regulatory mutations using two complementary datasets of human TFBSs that are likely to be relatively free of ascertainment bias with respect to evolutionary conservation but , importantly , are supported by experimental data ., The first is a collection of almost >2 , 100 human TFBSs drawn from the literature in the TRANSFAC database , and the second is derived from several recent high-throughput chromatin immunoprecipitation coupled with genomic microarray ( ChIP-chip ) analyses ., We also define a set of putative cis-regulatory modules ( pCRMs ) by spatially clustering multiple TFBSs that regulate the same gene ., We find that a relatively high proportion ( ∼37% ) of mutations at TFBSs are strongly deleterious , similar to that at a 2-fold degenerate protein-coding site ., However , constraint is significantly reduced in human and chimpanzee pCRMS and ChIP-chip sequences , relative to macaques ., We estimate that the fraction of regulatory mutations that have been driven to fixation by positive selection in humans is not significantly different from zero ., We also find that the level of selective constraint in our TFBSs , pCRMs , and ChIP-chip sequences is negatively correlated with the expression breadth of the regulated gene , whereas the opposite relationship holds at that genes nonsynonymous and synonymous sites ., Finally , we find that the rate of protein evolution in a transcription factor appears to be positively correlated with the breadth of expression of the gene it regulates ., Our study suggests that strongly deleterious regulatory mutations are considerably more likely ( 1 . 6-fold ) to occur in tissue-specific than in housekeeping genes , implying that there is a fitness cost to increasing “complexity” of gene expression .

Changes in gene expression have been suggested to play a major role in mammalian evolution ., In eukaryotes , gene expression is primarily controlled by sites , such as transcription factor binding sites ( TFBSs ) , located in the noncoding region of the genome ., The majority of these TFBSs remain unannotated , however , because they are typically short , degenerate , and laborious to identify experimentally ., As a result , the effects of mutations in TFBSs on organism fitness remain poorly understood ., We collected a dataset of TFBSs derived from the experimental biology literature and recent high-throughput studies to estimate the proportions of new mutations in TFBSs that have strongly deleterious and strongly beneficial effects upon organism fitness ., We find that a relatively high proportion of new mutations in TFBSs are strongly deleterious , although it appears that relatively few are adaptive ., We also demonstrate that the fraction of strongly deleterious regulatory mutations is correlated with the breadth of expression of the regulated gene ., Thus , ubiquitously expressed genes are likely to experience fewer deleterious regulatory mutations than those expressed in a small number of tissues .

genetics and genomics/gene expression, evolutionary biology/human evolution, evolutionary biology/evolutionary and comparative genetics, evolutionary biology, genetics and genomics/population genetics

journal.pgen.1003814

dTULP, the Drosophila melanogaster Homolog of Tubby, Regulates Transient Receptor Potential Channel Localization in Cilia

The auditory system allows animals to communicate and obtain information about their environment ., The hearing organs transform sound into an electrical signal through a process called mechanotransduction , the conversion of a mechanical force impinging on a cell into an intracellular signal 1 ., Although the recent discovery of several molecules involved in mechanotransduction allows interpretation of the biophysical properties of the mechanotransduction process for hearing 2 , many additional molecular players in auditory development and function are waiting to be unveiled ., Drosophila melanogaster has been suggested as a model organism to study the fundamental process of hearing 3 , 4 ., Hearing in the fly is necessary for the detection of courtship songs 5–7 ., Male-generated courtship song causes females to reduce locomotion and enhances female receptivity , whereas it causes males to chase each other 8 ., The ability to hear courtship songs is ascribed to Johnstons organ ( JO ) in the second antennal segment ., Near-field sounds rotate the sound receiver; the third antennal segment and the arista and this rotation of the antennal receiver transmits mechanical forces to the JO in the second antennal segment , which is connected to the third antennal segment by a thin stalk 9 ., Each JO sensilla consists of two or three chordotonal neurons and several supporting cells ., The outer dendritic segments of the JO neurons are compartmentalized cilia which are directly connected to the antennal sound receiver via extracellular caps ., The distortion of the junction between the second and third segment stretches the cilia and stimulates the JO neurons ., Several transient receptor potential ( TRP ) channels have been shown to be required for Drosophila hearing transduction and amplification 4 , 10–14 ., Mutation in nompC , the Drosophila TRPN channel , resulted in substantial reduction of sound-evoked potentials 4 ., Reports showing that NompC and TRP-4 ( the C . elegans ortholog of NompC ) are bona fide mechanotransduction channels support the idea that NompC is the Drosophila hearing transducer 15 , 16 ., Two Drosophila TRPV channel , inactive ( iav ) and nanchung ( nan ) , mutants showed complete loss of sound-evoked action potentials 11 ., However , they have not been considered to be the hearing transduction complex per se; rather they are thought to be required to amplify the electric signal generated by the hearing transduction complex , since Iav and Nan reside in the proximal cilia which are distant from the distal cilia where NompC is localized and mechanical force is directly transmitted 17 , 18 ., A recent study which employed a new method to measure subthreshold signals from the JO neurons suggested the opposite possibility that the TRPV ( Iav and Nan ) complex is the hearing transduction complex modulated by TRPN ( NompC ) 14 ., Although the exact roles of each TRP in Drosophila hearing are still controversial , it is clear that TRPN and TRPV have essential and distinct roles in Drosophila hearing ., Several attempts have been made to identify molecular players regulating the function of the ciliated mechanoreceptor neurons ., Gene expression profiling identified chordotonal organ-enriched genes from campaniform mechanoreceptors , developing embryo chordotonal neurons , and the second antennal segment 19–21 ., Alternatively , chordotonal neuron-specific genes were identified by searching for regulatory factor X ( RFX ) -binding sites , because ciliogenesis of the chordotonal neurons mainly depends on the RFX transcription factor 22 ., However , so far only a limited number of genes involved in TRP channel localization in the JO neuron cilia have been identified and characterized , including axonemal components and intraflagellar transports ( IFTs ) 17 , 23 ., IFTs are indispensable for the formation and maintenance of cilia as well as for the transport of proteins along the microtubules in and out of the cilia 24–26 ., Therefore , mutation of many of the characterized genes results in not only delocalization of the TRPs but also profound structural abnormality in cilia , rendering it difficult to delineate the gene functions specific to TRP localization ., Tubby is the founding member of Tubby-like proteins ( TULPs ) 27 ., Loss-of-function of the Tubby gene exhibits adult-onset obesity , retinal degeneration , and hearing loss in mice ., The Drosophila genome encodes one Tubby homolog called King tubby ( hereafter designated dTULP ) , which shares approximately 43% amino acid identity with mouse Tubby ( Figure S1A ) 28 ., At the embryonic stage , dTULP is expressed in various types of neuronal cells including the chordotonal neurons ., Although previous expression analyses and bioinformatic approaches detected dTulp in the chordotonal organs , its presence did not attract much interest because of its distribution in various neuronal cell types 22 ., In this study , we aimed to investigate the novel molecular function of dTULP in Drosophila hearing ., dTULP is localized to the well-defined ciliary structure of Drosophila auditory organs ., Loss of dTULP has no effect on the ciliary structure of the JO neurons , but NompC and Iav localization in cilia was severely altered ., These data demonstrate a new role of dTULP as a regulator of TRP localization in the hearing organs ., To test whether dTULP plays a role in Drosophila hearing , we generated two dTulp mutant alleles by ends-out homologous recombination 29 ., The first allele was dTulp1 , which harbours a deleted C-terminal containing the conserved “tubby domain” ( residues 220 to 460; Figure 1A ) ., The second allele , dTulpG , was generated by replacing an N-terminal portion of the dTULP coding region ( residues 18 to 261; Figure S1B ) with GAL4 coding sequences at the site corresponding to the initiation codon of the short splicing variant of dTulp ., Genomic PCR analyses showed that the dTulp genomic locus was deleted in dTulp1 and dTulpG flies ( Figure 1B and Figure S1B ) ., We raised antibodies to dTULP , which recognized a 51 kDa protein as predicted in wild-type fly extracts on a Western blot , and confirmed that dTULP was not detected in dTulp1 and dTulpG fly extracts ( Figure 1C ) ., Both alleles are homozygous viable and fertile ., Since both dTulp1and dTulpG mutant alleles showed postural problems and uncoordinated movement , we performed a climbing assay ., Flies were banged down to the bottom of a vertical tube and the percentage of the flies climbing above half of the height of the vertical tube within 10 seconds was recorded as the climbing index ., dTulp1 , dTulpG , and transheterozygote flies exhibited a decreased climbing index compared to control flies ( Figure 1D ) ., Introduction of a Pacman clone containing the dTULP coding region ( CH321-59C17 ) in the dTulp1 mutant background rescued this phenotype 30 ., These data suggested that dTulp mutants may have functional defects in the JO neurons 13 ., To check for hearing defects in dTulp mutant flies , we recorded extracellular sound-evoked potentials in wild-type and dTULP-deficient flies ., Sound-evoked potentials were completely abolished in dTulp1 , dTulpG , and dTulp1 in trans with a deletion that completely removed dTulp , Df ( 2R ) BSC462 ., Genomic rescue using the Pacman clone produced sound-evoked potentials similar to those in the wild-type , suggesting that the hearing defect was specifically due to dTulp ablation ( Figure 1E and 1F ) ., To test whether dTULP is expressed in the JO neurons , we first attempted to take advantage of the GAL4/UAS system using the dTulpG allele ., However , the GAL4 reporter inserted in dTulpG was not expressed ., This may be caused by inserting GAL4 at the site corresponding to the initiation codon of the short splicing variant of dTulp rather than the long splicing variant ., Therefore , we performed immunohistochemistry with dTULP antibodies ., We found that dTULP was expressed in the cilia as well as the cell body of the chordotonal neurons ( Figure 2A , left ) ., We did not detect dTULP immunoreactivity in the JO neurons in dTulp1flies , indicating that the immunosignal is specific for dTULP ( Figure 2A , right ) ., To further characterize the ciliary localization of dTULP , we compared the localization of dTULP with that of Iav and NompC ., The subcellular localization of Iav and NompC are in the proximal and distal cilia , respectively , in a mutually exclusive manner ( Figure 2B ) 11 , 17 , 18 , 31 ., dTULP staining extended from the proximal to distal cilia with a much weaker signal observed in the distal portion ( Figure 2C and 2D ) ., The mouse Tubby protein has been reported to shuttle from the plasma membrane to the nucleus upon Gq-coupled G protein-coupled receptor ( GPCR ) activation 32 ., dTULP was also detected in the cell body as well as the nucleus in the JO neurons ( Figure 2A and Figure S7B ) ., We also found that dTULP was expressed in other types of sensory neurons with cilia ( Figure S2 ) ., To examine whether the dTulp mutants have developmental defects in the JO neuron structure , we observed the expression of a membrane-targeted GFP ( UAS-mCD8:GFP ) driven by the pan-neuronal promoter ( elav-GAL4 ) in the JO neurons ., We found no gross structural abnormalities in dTulp1flies ( Figure 3A ) ., Electron microscopy of the JO revealed that most dTulp mutants had normal ciliary ultrastructure ( Figure 3B ) ., Approximately 9 . 3% of chordotonal scolopidia appeared abnormal in terms of cilia number or cap-cilia connections ( Figure S3 ) ., In addition , we did not observe any discernible changes in the expression of the dendritic cap protein NompA , which transmits mechanical stimuli to the distal segment of chordotonal neurons in dTulp mutants ( Figure 3C ) 33 ., These observations suggest that structural changes in the JO cannot account for severe hearing impairment in dTulp mutants ., Mutations of trps , including iav and nompC , cause hearing defects in Drosophila 4 , 11 ., To investigate the possibility that dTULP controls the expression of TRPs and other genes which are indispensable for Drosophila hearing , we performed quantitative PCR analysis of such genes and no significant differences in expression levels were present between wild-type and dTulp1 antennae ( Figure S4 ) ., This suggested that dTULP plays other roles in Drosophila chordotonal neurons rather than as a transcription factor that controls transcription of known hearing related genes , although we cannot exclude the possibility that dTULP regulates the expression of hearing related genes we did not survey ., Next we examined the ciliary localization of Iav and NompC in the dTulp mutants ., Surprisingly , Iav was not localized to the proximal cilia in dTulp1 flies ( Figure 4A ) ., Furthermore , NompC , characteristically localized to the distal cilia ( Figure 2B ) , was redistributed toward the proximal cilia ( Figure 4B ) ., Spacemaker ( Spam ) is an extracellular protein which protects cells from massive osmotic stress 34 ., Localization of Spam was also altered in dTulp mutants from its two typical locations: the luminal space adjacent to the cilia dilation and the scolopidium base ( Figure 4C ) 35 ., Introduction of the dTulp+ transgene rescued the localization of Iav , NompC , and Spam ( Figure 4 ) ., IFTs are involved in the localization of Iav , NompC , and Spam 17 , 23 ., Because IFT mutants show similar phenotypes to the dTulp mutant , we investigated the localization of IFT proteins in dTULP-deficient flies ., Ciliary localization of the two IFTs , NompB ( the ortholog of human IFT-B , IFT88 ) and RempA ( the ortholog of human IFT-A , IFT140 ) , was unaffected in dTulp1 mutants ( Figure S5 ) ., To further address the functional relationship between dTULP and IFTs , we examined distribution of dTULP in three IFT ( nompB , rempA , and oseg1 ) mutants and a retrograde motor dynein heavy chain ( beethoven ) mutant ., Although the rempA , oseg1 , and beethoven mutants show different degrees of defective cilia structure , dTULP is localized to the deteriorated cilia of each mutant , suggesting that rempA , Oseg1 , and beethoven are not required for the transport of dTULP into the cilia ( Figure S6A–S6C ) ., Since the nompB mutant does not develop cilia structure , dTULP was present in the inner segment at a high level ( Figure S6D ) 36 ., However , it is possible that other IFTs may play a role for dTULP ciliary localization even though the IFTs we examined are not involved in ciliary localization of dTULP ., Mammalian Tubby have two distinct domains: nuclear localization signal ( NLS ) and phosphoinositide ( PIP ) -binding domain ., An NLS , which allows Tubby to translocate into the nucleus , resides in the N-terminal region of Tubby 32 ., Recently , a short stretch of amino acids including the NLS in TULP3 , a mammalian member of the Tubby-like protein family , has been reported as an IFT-A binding domain 37 ., A PIP-binding domain in the C-terminal tubby domain allows Tubby to be localized under the inner leaflet of the plasma membrane through binding to specific phosphoinositides ., These domains are also conserved in dTULP ( Figure 5A ) ., In order to investigate the mechanism by which dTULP regulates the ciliary localization of Iav and NompC , we introduced mutations into the putative NLS/IFT-binding ( dTULPmutA ) , PIP-binding domain ( dTULPmutB ) , or both domains ( dTULPmutAB ) of dTulp cDNA and generated UAS-wild-type dTulp ( UAS-dTulpwt ) , UAS-dTulpmutA , UAS-dTulpmutB , and UAS-dTulpmutAB transgenic flies , respectively ., To eliminate positional effects , all transgenes were integrated into the same loci using site-specific recombination with an attP landing site on the third chromosome 38 ., To test the effect of each mutation on the subcellular localization of dTULP , we examined the subcellular localization of dTULPwt , dTULPmutA , and dTULPmutB in Drosophila salivary glands ., dTULPwt was detected mainly in the plasma membrane and nucleus ( Figure S7A ) ., Mutations in the NLS/IFT-binding domain or PIP-binding domain of dTULP resulted in significant exclusion from the nucleus or accumulation in the nucleus , respectively , which suggested that the NLS/IFT-binding and PIP-binding properties of mouse Tubby are conserved in dTULP in Drosophila salivary glands ( Figure S7A ) ., However , the localization of dTULPwt , dTULPmutA , and dTULPmutB in the JO neurons in terms of the cell body and nuclear distribution was virtually the same ( Figure S7B ) ., These data suggested that dTULP is not shuttled between the plasma membrane and the nucleus in the JO neurons and these domains may have other functions in the JO neurons rather than controlling the translocation of dTULP from the plasma membrane to the nucleus ., To evaluate the functional consequences of each mutation , we expressed dTULPwt , dTULPmutA , dTULPmutB , or dTULPmutAB in the JO neurons of dTulp1 flies ., The expression of dTULPwt in the dTulp mutant background restored the distribution and the expression level of Iav and NompC similar to those of wild type ( Figure 5B and 5C ) ., The expression of dTULPmutA or dTULPmutB rescued the Iav trafficking defect of the dTulp mutant , but the expression levels of Iav in the proximal cilia in dTULPmutA- or dTULPmutB-expressing flies were reduced compared to those of dTULPwt-expressing flies ( Figure 5B and 5E ) ., NompC localization to the distal cilia in dTULPmutA- or dTULPmutB-expressing flies was similar to that in dTULPwt-expressing flies ( Figure 5C ) ., dTULPmutAB could not rescue the Iav or NompC localization defects of the dTulp mutant ., This difference was not due to the expression levels of the mutant dTulp transgene since the expression levels of mutant forms of dTULP were similar to those of wild-type dTULP ( Figure S8 ) ., Next , we examined whether the different degrees of rescue of Iav and NompC localization was due to differential ciliary trafficking of variant forms of dTULP ., The ciliary expression level of dTULPmutB was similar to that of dTULPwt , whereas the ciliary expression levels of dTULPmutA and dTULPmutAB were reduced compared with those of dTULPwt ( Figure 5D and 5F ) ., These data suggested that the putative NLS/IFT-binding domain of dTULP has a regulatory function to control the trafficking of dTULP into the cilia ., Consistent with immunohistochemical analyses , dTULPwt fully rescued the hearing defect of the dTulp mutant ., dTULPmutA and dTULPmutB restored a partial function and dTULPmutAB had no such activity ( Figure 5G and 5H ) ., In the current study , we demonstrate that dTULP is a cilia trafficking regulator in the Drosophila hearing system ., Mutation of dTulp results in hearing loss due to the mislocalization of two TRP channels , Iav and NompC , which are ciliary membrane proteins ., In addition , Spam , whose localization is dependent on the IFT machinery , is also mislocalized in dTulp mutants ., How does dTULP regulate the ciliary distribution of TRPs in the JO neurons ?, Several studies have shown that mutations in IFT machinery or cilia components result in mislocalization of Iav , NompC , and Spam , along with abnormal axonemal structure 17 , 23 ., It is notable that , in contrast to IFT or cilia component mutants , ciliogenesis and maintenance appear normal in dTULP-missing flies ., Furthermore , the altered distribution of Iav , NompC , and Spam in dTulp mutants was not due to the mislocalization of IFT proteins , since the localization of two IFTs ( NompB and RempA ) was normal in dTulp mutants ( Figure S6 ) ., These data suggest that dTULP acts downstream of the IFTs to regulate TRP localization ., Even though the mutation of dTulp affected the trafficking of both Iav and NompC , the compartmentalized ciliary localization of Iav and NompC is differentially regulated by dTULP ., An individual mutation in either the putative IFT- or PIP-binding domain reduced Iav expression levels in cilia , whereas NompC localization was not altered until both domains were mutated ., Even after the double mutations in both domains of dTULP , NompC is still situated inside the cilia , but in abnormal locations ., These findings demonstrate that ciliary entry of NompC is not dependent on dTULP while the distal ciliary localization of NompC is dependent on dTULP ., One possibility is that dTULP allows NompC to disengage from the IFT complex at the distal cilia so that NompC is enriched in the distal cilia through the mechanism that required both IFT- and PIP-binding domains ., It is also possible that the distal ciliary localization of NompC is regulated by an unidentified factor ( s ) whose ciliary localization is dTULP-dependent as is Iav ., Both the putative IFT- and PIP-binding domains play important roles in the proper Iav distribution in cilia , but they appear to have different roles ., Even though the IFT- or PIP-binding mutant forms of dTULP could only partially rescue the ciliary levels of Iav to the similar extent , the mutation of the IFT-binding domain reduced the ciliary levels of dTULP while disruption of the PIP-binding domain had no effect on the ciliary levels of dTULP ., These findings suggest that two domains play distinct roles in the regulation of the ciliary localization of Iav ., The IFT-binding domain is the motif required for the ciliary entry for dTULP , and the PIP-binding domain is not related to dTULP ciliary entry itself , rather it affects recruitment of Iav-containing preciliary vesicles to dTULP ., By these two linked steps , Iav localization to cilia would be facilitated by dTULP ., In mammals , IFT-A directs the ciliary localization of TULP3 through physical interaction between TULP3 and the IFT-A core complex ( WDR19 , IFT122 , and IFT140 ) , and in turn , promotes trafficking of GPCR to the cilia ., Indeed , the depletion of individual IFT-A core complex components affects the ciliary localization of TULP3 , which results in the inhibition of GPCR trafficking to the cilia 37 ., It appears that dTULP and TULP3 have the similar molecular mechanisms to regulate ciliary membrane proteins ., However , unlike TULP3 , dTULP ciliary access is not dependent on IFT-A ., dTULP ciliary trafficking was not affected by the mutation of Oseg1 ( an ortholog of human IFT-A , IFT122 ) or rempA ( an ortholog of human IFT-A , IFT140 ) ., Furthermore , the presence of dTULP in cilia did not determine the normal localization of Iav ., For example , in the rempA mutant , even when dTULP was localized to the cilia ( Figure S6B ) , Iav was not found in cilia 23 ., Taken together , dTULP facilitates the relay of preciliary vesicles to the IFT complex at the base of cilia rather than moving together with ciliary membrane proteins into the cilia as an adaptor between IFT and cargo ., dTULP may have other additional roles in cilia , which needs to be explored in the future ., Based on our finding that dTULP but not Iav could be found in cilia of IFT mutants , it is also possible that recruitment of Iav-containing preciliary vesicles requires dTULP and additional unknown factors , whose function is altered in IFT mutants ., Thus , Iav-containing preciliary vesicles may not be able to form stable interactions with dTULP and IFTs ., After the cloning of the Tubby gene two decades ago , one promising hypothesis has been that Tubby is a transcription factor , since Tubby translocates to the nucleus upon GPCR activation and the N-terminal region of Tubby has transactivation potentials 32 , 39 ., However , candidate target genes for Tubby have not been identified ., Tubby is thought to have additional functions including vesicular trafficking , insulin signaling , endocytosis , or phagocytosis 40–43 ., It is still not clear how these molecular functions lead to the in vivo phenotypes observed in the tubby mouse ., Meanwhile , several studies have hinted at possible connections between the phenotypes of tubby mutant mice and ciliary dysfunction ., Tubby mice phenotypes comprise syndromic manifestations that are commonly observed in ciliopathies such as Bardet-Biedle syndrome 44 and Usher syndrome 45 , 46 ., Recently , GPCR trafficking into neuronal cilia was reported to be misregulated in tubby mice 47 ., Mutation of Tulp1 , a member of the TULPs , in human and mice , exhibits retinal degeneration due to the mislocalization of rhodopsin 48 ., TULP3 represses Hedgehog signalling , which is a crucial signalling cascade in cilia , via the regulation of the ciliary localization of GPCRs 49 ., Our current study provides additional supports for the idea that TULPs play an important role in ciliary signalling and that the tubby mouse syndrome might be due to the ciliary defects ., In contrast to mammalian cells , only specialized cell types have the ciliary structure in Drosophila , and the expression of dTULP is not restricted to organs with the ciliary structure , which suggested that dTULP may have other roles not related to the ciliary function 28 ., For example , dTULP mediates rhodopsin endocytosis in Drosophila photoreceptor cells which do not have cilium in contrast to its mammalian counterpart 50 ., In summary , we demonstrate an intriguing role of dTULP in governing the ciliary localization of TRP proteins ., This is the first in vivo evidence showing that dTULP may have important roles in the maintenance of ciliary functions by regulating the localization of ciliary proteins , thereby maintaining sensory functions ., All fly stocks were maintained in regular laboratory conditions ( conventional cornmeal agar molasses medium , 12-h light/12-h dark cycle at 25°C , 60% humidity ) ., Iav-GFP and NompA-GFP were reported previously 13 , 33 ., RempA-YFP and NompB-GFP were from M . Kernan ., Y . Jan and M . Noll provided UAS-NompC:GFP and Poxn-GAL4 , respectively ., Df ( 2R ) BSC462 , elav-GAL4 , UAS-mCD8:GFP , AB1-GAL4 , F-GAL4 , and Orco-GAL4 were from the Bloomington Stock Center ( Bloomington , IN ) ., We employed ends-out homologous recombination to generate dTulp mutant alleles ., To make the dTulp1 allele , 3 kb genomic DNA at the 5′ and 3′ ends of the tubby domain ( 220 to 460 residues ) coding sequence was PCR amplified from w1118 and subcloned into the pw35 vector ., The primer sequences for the 5′ homologous arm of the pw35 vector are 5′-AAAGCGGCCGCCACCGGTGACATCCTCATGTTC-3′ and 5′-AAAGCGGCCGCGTTGCATCACGAACTGGTCGATATTG-3′ ., The primer sequences for the 3′ homologous arm of the pw35 vector are 5′-TGAGCTGGCTGGGATCCTCGGGTTGG-3′ and 5′-GTGGATCCTTCCTGGTTGGCATCACGTTGAC-3′ ., To generate the dTulpG allele , we used the pw35GAL4loxP vector in which GAL4 and white are flanked by loxP sequences so the cassette can be removed by introducing Cre recombinase ., We subcloned the 3 kb of genomic DNA from each of the 5′ and 3′ ends of the dTULP coding region ( 18 to 261 residues ) into the pw35GAL4loxP vector ., The primer sequences for the 5′ homologous arm of the pw35GAL4loxP vector are5′-ACAGATCTCACCGTCGCCTGGCTCAGTGCCC-3′ and 5′-GTGGTACCCAGCTGGCGCTGCAAAGCAGTTAAATC-3′ ., The primer sequences for the 3′ homologous arm of the pw35GAL4loxP vector are5′-AAAGCGGCCGCGTGGGTTATTGATAGTGATCCTCTA-3′ and 5′-AACCGCGGCGTACAGAATACTCCCTGTTCATGTCT-3′ ., We generated transgenic flies by germ line transformation ( BestGene Inc . , Chino Hills , CA ) and screened for the targeted alleles as described previously 51 ., Targeted alleles were subjected to outcross for five generations into a w1118 genetic background ., We amplified dTulp cDNAs from cDNA clones ( RE38560 ) with PCR and subcloned the fragments into the pUASTattB vector ., These constructs were subjected to further modification ., We generated the dTulpmutA and dTulpmutB mutant constructs using site-directed mutagenesis to change the sequence encoding R23QKR to L23AAA , and K292LR to A292LA , respectively ., The dTulpmutAB construct was generated by introducing the mutation corresponding to dTulpmutB into the dTulpmutA construct ., To generate genomic rescue transgenic flies , we obtained the BAC clones CH321-59C17 from the BACPAC Resource Center ( Oakland , CA ) and used these as genomic rescue constructs ., Transgenic flies were generated using PhiC31 integrase-mediated transgenesis on the third chromosome to minimize position effect ( Bloomington stock number 24749 ) ., Sound-evoked potentials were recorded as described by Eberl et al 4 ., Briefly , the flys antennal sound receivers were stimulated by computer-generated pulse songs ., Neuronal responses were detected using a recording electrode inserted in the junction between the first and second antennal segment and a reference electrode was inserted in the dorsal head cuticle ., The signals were subtracted with a DAM50 differential amplifier ( World Precision Instruments , Sarasota , FL ) and digitized using Superscope 3 . 0 software ( GW Instruments , Somerville , MA ) ., Each trace represents the average responses to 10 stimuli ., For whole-mount staining , antennae were dissected at the pupa stage and the labellum and legs were prepared at the adult stage ., Salivary glands were dissected from third instar larvae ., For antenna sections , fly heads were embedded in OCT medium and 14 µm frozen cryostat sections were collected ., Dissected tissues and sections were fixed for 15 min with 4% paraformaldehyde in 1× PBS containing 0 . 2% TritonX-100 ( PBS-T ) and washed three times with PBS-T ., The fixed samples were blocked for 30 min with 5% heat-inactivated goat serum in PBS-T and incubated overnight at 4°C in primary antibodies diluted in the same blocking solution ., The tissues were washed three times for 10 min with PBS-T and incubated for 1 h at room temperature in secondary antibodies diluted 1∶500 in blocking solution ., Following three washes with PBS-T , the samples were mounted with Vectashield ( Vector Laboratories , Burlingame , CA ) and examined using a Zeiss LSM710 confocal microscope ( Jena , Germany ) ., To quantify Iav-GFP and dTULP expression levels in cilia , all samples were prepared at the same time and all confocal images were obtained under the same conditions ., The pixel intensity of each protein was measured using Zen Software ( Jena , Germany ) ., Iav-GFP intensity was measured without immunostaining ., Rabbit dTULP antibodies were raised by injecting animals with a purified His-tagged dTULP fusion protein ( residue 95–339 ) , followed by affinity purification ., The primary antibodies were used in immunohistochemistry at the following dilutions: rabbit anti-dTULP , 1∶400; 22C10 , 1∶200 ( Hybridoma Bank , University of Iowa ) ; 21A6 , 1∶200 ( Hybridoma Bank ) ; rabbit anti-Orco , 1∶1 , 000 ( gift from L . Vosshall ) ; rabbit anti-NompC , 1∶20; rabbit anti-GFP , 1∶1 , 000 ( Molecular Probes , Eugene , OR ) ; mouse anti-GFP , 1∶500 ( Molecular Probes ) ., The secondary antibodies used were Alexa 488- , Alexa 568- , and Alexa 633-conjugated anti-mouse or anti-rabbit IgG ( Molecular Probes; 1∶500 ) ., DNA and actin were visualized by DAPI and Alexa Fluor 633 Phalloidin ( Molecular Probes ) staining , respectively ., Fly head or antennae lysates from each genotype were subjected to electrophoresis on SDS-polyacrylamide gels and transferred onto polyvinylidene fluoride membranes ., The membranes were blocked for 1 h with 5% nonfat milk plus 0 . 1% Tween-20 ., Membrane-bound proteins were analyzed by immunoblotting with primary antibodies against dTULP ( 1∶1 , 000 ) and tubulin ( Hybridoma Bank , 1∶2 , 000 ) ., Fly heads were dissected and fixed in 2% paraformaldehyde , 2 . 5% glutaraldehyde , 0 . 1 M cacodylate , and 2 mM CaCl2 , pH 7 . 4 ., The tissue was embedded in LR white resin ., Thin sections were cut , mounted on formvar-coated single slot nickel grids , counterstained with uranyl acetate and lead citrate , and examined on a Hitachi H-7500 electron microscope ( Hitachi , Tokyo , Japan ) ., Total RNA was extracted from adult antennae using Trizol reagent ( Invitrogen , Carlsbad , CA ) ., cDNA was generated from 0 . 5 µg of RNA from each genotype using the SuperScript III First Strand Synthesis System ( Invitrogen ) ., Quantitative PCR was performed using an ABI7500 real-time PCR machine ( Applied Biosystems , Foster City , CA ) and the ABI SYBR green system ., Transcript levels were normalized to rp49 as an internal control and the ΔCT ( CT\u200a=\u200athreshold cycle ) method was used to calculate the relative amount of mRNAs . The primers used for qRT-PCR are listed in Table S1 ., Fifteen 3- to 6-day-old flies were placed in an empty fly food vial ., The climbing index is the fraction of flies that climb halfway up the vials in 10 s after being tapped down to the bottom of the tube ., We performed each experiment twice and used the average of the two trials to calculate the climbing index ., Data shown are the mean ± SEM ., To compare two sets of data , unpaired Students t-tests were used ., ANOVA with the Tukey post-hoc test was used to compare multiple sets of data ., Asterisks indicate statistical significance .

Introduction, Results, Discussion, Materials and Methods

Mechanically gated ion channels convert sound into an electrical signal for the sense of hearing ., In Drosophila melanogaster , several transient receptor potential ( TRP ) channels have been implicated to be involved in this process ., TRPN ( NompC ) and TRPV ( Inactive ) channels are localized in the distal and proximal ciliary zones of auditory receptor neurons , respectively ., This segregated ciliary localization suggests distinct roles in auditory transduction ., However , the regulation of this localization is not fully understood ., Here we show that the Drosophila Tubby homolog , King tubby ( hereafter called dTULP ) regulates ciliary localization of TRPs ., dTULP-deficient flies show uncoordinated movement and complete loss of sound-evoked action potentials ., Inactive and NompC are mislocalized in the cilia of auditory receptor neurons in the dTulp mutants , indicating that dTULP is required for proper cilia membrane protein localization ., This is the first demonstration that dTULP regulates TRP channel localization in cilia , and suggests that dTULP is a protein that regulates ciliary neurosensory functions .

Tubby is a member of the Tubby-like protein ( TULP ) family ., Tubby mutations in mice ( tubby mice ) cause late-onset obesity and neurosensory deficits such as retinal degeneration and hearing loss ., However , the exact molecular mechanism of Tubby has not been determined ., Here we show that Drosophila Tubby homolog , King tubby ( dTULP ) , regulates ciliary localization of transient receptor potential protein ( TRP ) ., dTULP-deficient flies showed uncoordinated movement and complete loss of sound-evoked action potentials ., dTULP was localized in the cilia of chordotonal neurons of Johnstons organ ., Two TRP channels essential for auditory transduction , Inactive and NompC , were mislocalized in the cilia of chordotonal neurons in the dTulp mutants , indicating that dTULP is required for proper cilia membrane protein localization ., This is the first demonstration that dTULP regulates TRP channel localization in cilia , and thus provides novel insights into the pathogenic mechanism of tubby mice .

journal.pcbi.1003759

Bursts and Heavy Tails in Temporal and Sequential Dynamics of Foraging Decisions

Rather than regular , like a metronome , or homogenous ( i . e . , a constant overall rate of activity ) , timing of behavior and/or events in humans , non-human animals , and natural phenomena is often non-homogeneous , with periods or bursts of high activity separated by long inactive periods 1 , 2 ., Examples in humans include e-mail 1 , 3–7 or mail communication 8 , library loans 3 , financial trading 9 , 10 , on-line movie watching 11 , internet browsing 3 , 12 , printing requests 13 , and mobile communication 14 , 15; in non-human animals , locomotion 16–21 , and flying patterns 22; and in natural phenomena , rainfall 23 , tsunamis 24 , and earthquakes 2 , 25 ., A telltale diagnostic feature used to characterize non-homogeneous temporal processes is a heavy tail in the distribution of the inter-event intervals ( i . e . , the time interval between consecutive events ) 1 ., A heavy tail reflects a larger number of longer inter-event intervals than occurs with homogeneous Poisson processes ( i . e . , those in which the events occur at an overall constant rate , but are independent of one another ) ., Although a non-homogeneous process has been suggested as a universal feature of natural dynamical systems 2 , different specific underlying mechanisms can lead to a heavy-tailed distribution of the inter-event intervals 26 ., For example , it has been suggested that the bursty nature of human interactions results from the combined effects of different periodicities at different timescales: e . g . , a circadian rhythm , as well as weekly , monthly , etc . cycles; and , in fact , bursty behavior can derive from a cascading non-homogeneous Poisson process model that combines multiple Poisson processes with different timescales 6 , 27 , 28 ., At the same time , the bursty behavior of human interactions can also be induced by intrinsic correlations between actions 6 , 27–31 ., Indeed , bursty behavior might also derive from a combination of such processes , which we explore in the current study ., Here , we focus on foraging , a fundamental and frequent behavior for survival ., Foraging mechanisms underlie the daily energy budget allocation across activities 32–34 ., Unlike nature phenomena , feeding , and more generally , foraging behavior is influenced by both internal biological and external environmental factors: internal factors include preference , nutrition , memory , hunger and satiety; external factors include the daily light-dark cycle ( leading to a circadian rhythm ) , seasonal and social/societal effects 32 , 35 ., Thus , the study of foraging behavior provides the opportunity to examine decision mechanisms that result from the interaction of important internal and external influences ., Feeding behavior has been studied in large data sets of farm animals , pets , and captive wild animals , including cattle , pigs , chickens , ducks , turkeys , rats , and dolphins 33 , 35–39 ., The temporal structure of feeding behavior consists of high frequency feeding events that are separated by relatively long non-feeding periods: i . e . , it is bursty 33 , 38 ., In the current study , our first objective was to test the hypothesis that foraging timing is based on bursty behavior that is influenced by both the level of satiety ( internal ) and by the daily light- dark cycle ( external ) ., Indeed , we found a heavy-tailed distribution of the inter-choice intervals ( ICI , the time interval between two choices ) , reflecting a non-homogenous process ., Moreover , the ICI distribution exhibited bimodality , reflecting distinctive processes for short and longer timescales: bursty behavior for short ICIs and circadian rhythmic activity for longer ICIs ., To explain this bimodality in foraging behavior , we propose a dual-state model consisting of active and inactive states , with correlated behavior producing bursty activity in the active state , and relatively uncorrelated behavior influenced by a circadian rhythm in the inactive state ., Once activity timing is characterized , the decision dynamics of which option to select and whether to continue selecting it over repeated choices must be specified 40–43 ., Although progress has been made on characterizing outcome-driven behavior as governed by the goal-directed system 44 , 45 , and stimulus-driven behavior as governed by the habit system 45–48 , it nonetheless remains difficult to predict an individuals preference and choice responses over a long period of time ., For example , an individuals preference for different foods or music seems to fluctuate over time even when they have experienced the available options extensively and thus know all options well: e . g . , even if ones favorite food is a hamburger , it typically is not eaten every single day ., Thus , the underlying mechanisms that lead to dynamically changing preference-based choice behavior remains unclear , especially with qualitatively different rewards in stable environments , in which an agent ‘knows’ the reward contingencies and thus does not require further learning ., Therefore , the second objective of the current study was to help specify the mechanisms underlying seemingly unpredictable preference-based choices with, ( a ) multiple qualitatively different options; and, ( b ) repeated choices over an extended period in a stable environment that reflects real-world choice behavior ., Here we extracted two distinctive features from an individuals dynamic choice sequence: ( 1 ) preference bias ( i . e . , the skew of the choice distribution based on the individuals rank order of choice options ) , and ( 2 ) choice persistence ( i . e . , the degree to which choices are repeated ) , which capture distinct underlying control processes that determine what to choose and whether to continue choosing it , respectively ., We found individual differences in preferences that nonetheless could be characterized by choice option rank , reflecting a value-based process , as well as some persistent choice behavior , in which choices tended to be repeated , with an increasing likelihood of repeating a choice as a run of identical choices increased , reflecting a preferential-attachment process ., We then developed a dual-control model incorporating a combination of goal-directed and habitual control to describe the dynamical patterns of the choice sequences ., We investigated the continuous choice behavior of 12 rats over the course of two weeks using a four-armed bandit task with four differently flavored pellets: chocolate , banana , coffee , and cinnamon ., Each rat lived in an operant chamber for the entire two-week duration as a “closed economy” 49 with continuous access to water and the food pellets in the environment ., Each trial was initiated by nose-poking in a lighted opening , after which four levers would extend from the opposite wall of the chamber ( Figure S1 ) ., The rat then obtained one of the flavored pellets by pressing the corresponding lever ., To examine when and what the animals chose , timing and choice sequences of lever-pressing activity for all rats were recorded for the entire experiment ., With respect to when they chose , the animals actively foraged during the dark cycle and sporadically so during the light cycle as shown in Figure 1A ., With respect to what they chose , we found dynamic changes in the animals food choices , indicating that the rats did not commit themselves to a specific option but rather intermittently explored alternatives ., To assess the degree of the animals exploration or exploitation , we first computed entropy of choice sequences every hundred trials 50 , which is a measure of the uncertainty in choices , with zero being deterministic and solely exploitative and high entropy indicating a high degree of exploration ( Figure 1B ) ., We found that the entropy of choice sequences fluctuated to some degree throughout the experimental period ., Although entropy changes varied slightly across subjects , overall , there was no significant tendency of entropy to decrease at the group level , indicating that the animals maintained some level of exploring alternatives throughout the experiment rather than converging toward a particular option ., Next , we compared the entropy of empirical choice sequences with randomly shuffled ones , which removes any dependency on past choices , to determine whether the degree of exploration or exploitation depended on previous choice history ( Figure 1B ) ., We found that the levels of entropy in the empirical choice sequences were significantly lower than in randomly shuffled ones for all subjects ( paired t-test , p<0 . 001 ) ., Thus , this result shows that previous choices influenced the current choice , consistent with other reports 40 , 42 , 51–54 ., We next examined the amount of consumed pellets with respect to flavor , location , and rank ., Rank was defined as the order of overall consumption of each food type for an individual , which would reflect the order of an individuals subjective values for the qualitatively different rewards ., The percentages of mean choice for the four different locations – left ( LL ) , middle left ( ML ) , middle right ( MR ) and right ( RR ) – were not significantly different ( one-way ANOVA , F ( 3 , 44 ) =\u200a0 . 781 , p\u200a=\u200a0 . 511 ) ( Figure 1C ) , reflecting the counterbalancing of flavor and position across subjects , and demonstrating that there was no preferred location overall ., In addition , to test whether there were differences in effort to reach each lever location from the initial nose poke position , we compared the response latencies between nose-poke and lever pressing for each location ., The response latency medians across locations were not significantly different ( one-way ANOVA , F ( 3 , 44 ) =\u200a0 . 009 , p\u200a=\u200a0 . 998 ) , suggesting that the animals response vigor for each location was similar 55 ., The consumption rates for each flavor were significantly different ( one-way ANOVA , F ( 3 , 44 ) =\u200a5 . 043 , p<0 . 01 ) : the chocolate flavor was statistically more consumed than the coffee flavor at the group level ( Dunnett-T3 post hoc test , p\u200a=\u200a0 . 021 ) ( Figure 1D ) , although this was not the case for all subjects ( e . g . , Figure 1A ) ; nonetheless , all rats showed distinct individual preferences among the different flavors ., Since the rats exhibited individual differences in preference , and since quality has no obvious natural corresponding number to represent its value ( especially when quality was essentially flavor ) , we analyzed choice behavior based on rank , which should be driven by an individuals subjective values of the options , and which provides a common scale to compare individuals ., Comparing the percentages of mean choice for rank , there was a clear difference between food pellets of different ranks as shown in Figure 1E ( one-way ANOVA , F ( 3 , 44 ) =\u200a74 . 897 , p<0 . 001; Dunnett-T3 post hoc test ) ., Interestingly , choice rate appeared to decrease by nearly half as rank increased ., To confirm this tendency , we transformed the percentage of food choice by rank to a log-linear scale ., We found that the mean distribution of the choice percentage p as a function of rank r was well described by the log-linear distribution ( Figure 1E ) , where the slope of p versus log ( r ) was −70 . 7±4 . 95 ( mean ± standard error of the mean s . e . m . , adj . R2\u200a=\u200a0 . 994 ) , indicating that preference was highly skewed toward the higher rank ., Examining the timing characteristics of the choice behavior in more detail , we found periodic changes in food consumption ., First , the animals consumed more pellets during the dark than the light cycle ( Figure 2A ) ., To investigate the relationship between the foraging pattern and the daily light-dark cycle ( i . e . , a potential circadian rhythm effect ) , we measured the periodicity of the foraging pattern by calculating the time interval between peaks in the average autocorrelogram ., The rats foraging pattern period was approximately 24 hours , consistent with their circadian rhythm ( Figure 2B ) , indicating that it was one of the key factors that determined foraging timing in general ., The remaining issue was how the specific timing of foraging was determined at a short timescale ., We characterized the underlying action dynamics by analyzing the features of the inter-choice interval ( ICI ) distribution ., We found that the majority of ICIs were short , but very long ICIs also sporadically occurred , indicating that there were bursts of activity separated by relatively long inactive periods ( Figure 2C ) ., To measure this burstiness in the timing of foraging behavior , we used a burstiness index B , defined as , where and are the mean and the standard deviation of the ICI distribution , respectively 31 ., B ranges between −1 and 1: B\u200a=\u200a1 is the most bursty signal , B\u200a=\u200a0 is neutral , and B\u200a=\u200a−1 is a completely periodic signal ., We found that B of the foraging behavior was 0 . 794±0 . 008 ( mean ± s . e . m ) , indicating that the majority of activity was densely concentrated in short durations ., Next , to characterize a memory effect , we calculated the correlation coefficient of consecutive inter-choice intervals , which is defined as , where is the number of ICIs measured from the timestamps , and m1 ( m2 ) and are the mean and standard deviations of the ICIs s ( s ) , respectively 31 ., M ranges between −1 and 1: M is positive when the length of the current ICI is positively proportional to the length of the previous ICI; whereas , M is negative when the length of the current ICI is inversely proportional to the length of the previous ICI; M\u200a=\u200a0 is neutral; and M\u200a=\u200a−1 is a completely periodic signal ., We found that M of the foraging behavior was 0 . 046±0 . 006 ( mean ± s . e . m ) , indicating that the foraging activity had a relatively low correlation between consecutive ICIs ., The bursty nature of the foraging behavior was reflected in the heavy-tailed ICI distributions ., The cumulative distribution of ICIs , which is the probability of ICIs longer than a given ICI ( i . e . , the survival function ) , exhibited a heavy tail that was clearly seen in a log-log scale , representing a deviation from an exponential distribution resulting from a simple homogeneous Poisson process ( Figure 2D ) ., This indicates that the time interval between spontaneous behaviors is not simply governed by a random process , but is modulated in a more sophisticated way by other processes at a longer timescale ., In addition , heavy tails were also observed in the distributions of ICIs in both the light and dark cycles ( Figure 2D ) ., Interestingly , the empirical ICI distribution exhibited bimodality ( Figure 2E ) ., For short ICIs , the probability density function of the ICIs was highly left-skewed; whereas for longer ICIs , the probability density function did not appear to reflect the same left-skewed characteristic ., The highly left-skewed component of the distribution for short ICIs was well fit by the power-law ( p\u200a=\u200a0 . 68±0 . 09 for the fit to the power-law distribution—i . e . , the empirical and power-law distributions were not significantly different; see “Estimation of parameters in the inter-choice interval ( ICI ) distribution” in Material and Methods ) ( Figure 2E inset ) ., The second component of the distribution for longer ICIs appeared to follow the Weibull distribution , exhibiting a stretched exponential decay; however , with combined light and dark cycles , the empirical and Weibull distributions were significantly different ., When we decomposed the overall ICI distribution into the component light and dark cycles , however , the distributions of the short ICIs for both cycles followed the power-law distribution , and the distributions of the longer ICIs for both cycles followed the Weibull distribution ( Table 1 and Figure 2F ) ., Thus , the cumulative bimodal ICI distributions for both the light and dark cycles could be described as the following:where is the lowest time boundary , is a time constant used to separate activities into independent bursts , µ is the power-law exponent , λ is a scale parameter , and γ is the shape parameter of the distribution ., We calculated the value of τ0 as the local minimum of the bimodal distribution of ICIs , which separated the short and longer ICIs in the distributions ., The estimated parameters of the bimodal ICI distributions are shown in Table 1 ( see “Estimation of parameters in the inter-choice interval ( ICI ) distribution” in Material and Methods for details ) ., This bimodality in the ICI distributions suggests, ( a ) different underlying processes at different timescales of ICIs , and, ( b ) similar underlying processes in both the light and dark cycles leading to the power-law and Weibull distributions ., We take up these implications in the discussion ., When comparing the fitted parameters in the light and dark cycles , we found that the distributions for longer ICIs between the light and dark cycles exhibited different exponential decays reflected in the scale parameter λ ( light: 1 . 20±0 . 15 ×104 , dark: 2 . 74±0 . 2 ×103 , Sign test , p<0 . 001 ) , whereas the power-law distributions for the short ICIs in both cycles appeared to have similar slopes ( light: 2 . 21±0 . 07 , dark: 2 . 07±0 . 05 , Sign test , p\u200a=\u200a0 . 146 ) ( Table 1 and Figure 2F ) ., This finding comparing the light and dark cycles implies that the underlying mechanism governing longer ICIs was influenced by the circadian rhythm; whereas , the mechanism governing short ICIs may have been more weakly influenced by the circadian rhythm ., We next analyzed the choice patterns to examine the sequential dynamics governing what is chosen over trials ., First , we determined how long the rats continued to make the same choice ., We defined a “run” as a series of consecutive identical choices ., A trial-dependent change in a distribution of runs was then calculated , as shown in Figure 3A ., The cumulative distribution of runs , defined as the probability of runs longer than a given length of run ( i . e . , the survival function ) , revealed a heavy tail in a log-log scale ( Figure 3B ) , indicating that the choice pattern consisted of a large number of short runs and a few extremely long runs ., To test for a sequential dependency of previous choices , we compared the run distributions of the empirical sequences with those of randomly shuffled sequences of the same data for each rat ., The randomly shuffled sequence has no dependency on previous choices yet maintains the same choice frequency as the empirical data ., The cumulative run distribution of the empirical data was significantly different from that of the randomly shuffled choice sequences for all subjects ( Monte Carlo hypothesis testing , p<0 . 001 ) 6 ., This result indicates that the choice sequences were highly influenced by the choice histories 40 , 42 , 52 , 54 ., In addition , we examined whether there was an effect of choice history regardless of rank by comparing the run distribution of empirical data for each rank with that of randomly shuffled data ( Figure 3C ) ., Although the lower ranking flavors had fewer long runs than the higher ranking ones , the run distribution of the empirical data for all ranks was significantly different from those of the randomly shuffled choice sequences for all subjects , with the exception of the fourth rank for two of the twelve subjects ( Monte Carlo hypothesis testing , p<0 . 001 ) 6 ., The shared heavy-tailed feature of the run distribution for every rank suggests that the underlying processes determining whether a run would continue were relatively insensitive to reward outcome ., Conducting a simple calculation with the cumulative distribution of runs , we obtained the hazard rate for ending a run as a function of the number of preceding choices in a run for each rank , i . e . , the conditional probability of ending a run at a given length of a run ( Figure 3D ) ., We found that the hazard rate for ending a run decreased logarithmically and converged relatively quickly to approximately zero in all ranks ., This indicates that a run was more likely to be terminated when the length of the preceding choices in a run was short; and the run was more likely to continue when the length of the preceding choices in a run was increased ., In addition , the hazard rate converging to zero resulted in extremely long runs regardless of rank; indeed , there was no significant difference in the decreasing rate of the hazard rate between ranks ( one-way ANOVA , F ( 3 , 44 ) =\u200a0 . 666 , p\u200a=\u200a0 . 577 ) ., Thus , in general , the rats were more likely to choose what they had chosen previously , irrespective of outcome , reflecting a status quo bias or preferential-attachment process that tends to continue a run until switching ones choice finally becomes more compelling ., A bimodal distribution has been suggested as a mixture of distinct distributions formed by different underlying processes 14 , 25 , 38 ., We found that the empirical ICI distribution underlying the foraging behavior under free conditions exhibited bimodality with the power-law and Weibull distributions for short ICIs and longer ICIs , respectively ., To characterize the bimodal temporal dynamics , we propose a dual-state model that can provide an integrative account of both the bursty and periodic features of the foraging behavior ., The model consists of an active state and an inactive state , which executes correlated actions in bursts in the active state , and elicits intermittent uncorrelated actions in the inactive state ( Figure 4A ) ., We consider an animal to be in an active state when the animal exhibits a high frequency of activity , with short ICIs that are less than a certain time period , and we assume that the events within the active state are correlated due to the influence of the motivational drive 2 ., In our case , the motivational drive for feeding is to appease hunger ( i . e . , reach satiation ) ., A known physiological mechanism underlying short-term regulation of feeding ( within a meal ) is that feeding is governed by a feedback mechanism from the delayed gastrointestinal aftereffects of eating 36; the digestion of food inhibits eating , but the inhibitory effect is delayed ., Here , we focus on the delay between the swallowing of food and the digestion of food , resulting in the delayed satiety signal as feedback ., And this characteristic of feeding leads us to propose a satiation-attainment process , i . e . , an active waiting process based on feedback for upcoming satiation within each active state ., In this process for the active state , we assume that whenever animals eat , they wait for the feedback signal by which they determine whether to eat more or stop ., In other words , animals initiate eating and wait until they receive the satiety signal , which informs them that satiation is attained ., If the satiety signal is lower than the satiation threshold , they would continue to eat and wait for the next feedback signal ., Thus , the waiting time between eating and the feedback signal is important to determine time intervals between actions in an active state ., Instead of a constant time delay of feedback , we assume that there is a non-linear relationship in the waiting time between eating and the feedback signal ., A number of studies on human dynamics have suggested that the waiting time based on feedback in human communication patterns follows a power-law distribution 1 , 7 , 8 , 14 ., Considering a similarity in the waiting process for feedback between feeding and human communication , we assume that the waiting time between eating and the feedback signal follows a power-law distribution; in active states , the probability density function of the time interval between choices is for where 1<µ<3 ., In addition , an animal is considered to be in an inactive state when there is a period of inactivity longer than ; and thus the inactive state is defined as the time between the last event in a given active state and the first event in the next active state , which by definition , is longer than ., We model timing in the inactive period with a non-homogeneous Poisson process with the inactivity rate , i . e . , the reciprocal of the mean inactive duration as a function of time ., To capture the strong influence of the circadian rhythm on the longer ICIs , two temporal properties of the inactivity rate are further specified ., First , the inactivity rate depends on time in a periodic manner , as expressed by the equation , where T is the period of the process ., Since the animals periodic activity is modulated by a circadian rhythm , we set the period T as 1 day ., Second , the inactivity rate is proportional to the daily distribution of choice behavior in the inactive state , , where is the average rate in the inactive period , is the probability of beginning an active state during a particular hour of the day 6 , and b is the shape parameter ., To quantify the transition between active and inactive states , we assume that a state transits from the active state to the inactive state with a probability ξ after each choice and remains in the active state with probability 1 – ξ ., With the computational processes that determine when choices are made specified , we next delineate those that determine what choices are made ., Here , we propose a simple heuristic model that accounts for two key sequential features of decision-making: ( 1 ) the heavy-tailed nature of the run distribution , reflecting choice persistence as habitual behavior , and ( 2 ) the biased rank distribution , reflecting goal-directed outcome valuation ., First , to account for persistence in choice behavior , we assume an underlying preferential-attachment process , which has been proposed as the mechanism underlying heavy-tailed distributions 42 , 56 , 57 ., In this process , the probability of continuing a run increases as a run proceeds ( thus , it also has been called the “rich get richer” process ) ., We suggest that the same mechanism underlies choice behavior , in which the probability of choosing a particular option is proportional to the number of times the option was chosen previously ., The process may underlie response persistence found in choice behavior in humans and nonhuman primates 40 , 42 , 58 , 59 ., In addition , the preferential-attachment process occurs regardless of outcome type , reflecting its property of insensitivity to outcome , which is a defining feature of habitual behavior ( Figure 3D ) ., Thus , this process may underlie the acquisition and maintenance of habits ., We therefore more generally call this mechanism , habitual control ., In the habit system , in addition to the preferential-attachment process , we apply a leaky integrator to the dynamic trial-by-trial model of habitual behavior , in which the integrated choice frequency over previous trials is discounted as a function of the distance passed from a given trial 52 , 57 , 60 , 61 ., Thus , this integrator includes the effect of past choices 42 ., Because the preferential-attachment process is insensitive to outcome , we assume that the discount rate is identical for all options regardless of rank ., In habitual control , the action value of a particular option i at trial t , , is determined by the local choice history of that option with leakage: where is a weighting coefficient for choices occurring trials ago with an exponential decreasing profile , equal to , where is a free parameter for the decay constant , and is a binary vector denoting a chosen option i on trial t ., The choice vector is 1 if option i was chosen on trial t and 0 if the option was not chosen on that trial ., Second , for goal-directed control , we use a temporal difference ( TD ) reinforcement learning algorithm that updates the action-value on each trial according to its prediction error 62–66 ., The TD learning algorithm provides a theoretical framework for instrumental reward learning in which actions must be chosen to optimize long-term rewards 63 , 67 ., In addition , we incorporate a decay factor , which updates the chosen option and decays unchosen options 66 , 68 , 69 ., Thus , at each trial t , the action value for the chosen option c and for the unchosen option u are updated according to:where and are learning rates and and are the reward prediction errors at given trial t for the chosen and unchosen options , respectively ., The reward prediction errors , i . e . , the difference between the expected and received reward values , for the chosen and unchosen options are as follows: where is the reward value for the chosen option ., We deductively estimated the reward value based on the mean choice rate across days , R: where is a parameter of sensitivity of behavior to differences in reward values among alternatives 70 ., We refer to this outcome-driven process as “goal-directed . ”, The goal-directed process plays an important role in determining the initial choice for a new run on the basis of value , which in turn generates a certain degree of bias toward a more valued option ., Finally , for action selection , to capture the effects of both the habit and goal-directed systems on choice behavior , the goal-directed value and habit value are derived in parallel 71 ., We then assume that the probability to choose an option i at trial t , , is determined according to a softmax choice function 63:where the softmax inverse temperature parameters and represent the degree to which a choice is focused on the highest-valued option in goal-directed value and habit value , respectively ., Note that , together , the combination of goal-directed and habit systems create two key features of sequential dynamics: a bias among choice options and a bursting property in which very long runs are interspersed among a majority of short runs ., Regarding when choices were made , we found bursts of rapidly occurring actions separated by time-varying inactive periods , partially based on a circadian rhythm ., These characteristics of foraging behavior were reflected in a bimodal inter-choice interval ( ICI ) distribution comprised of a power-law for the short timescale ( i . e . , short ICIs ) and the Weibull distribution for the longer timescale ( i . e . , longer ICIs ) ., Although the specific mechanisms of the bimodal inter-event times could vary across different systems 9 , 10 , 14 , 24 , 25 , 74 , 75 , a common dynamical feature of the underlying mechanisms appears to be the combination of distinct processes at different timescales 14 , 25 , 37 ., To capture the temporal dynamics underlying foraging behavior , we propose a dual-state model consisting of active and inactive states for short and longer timescales based on a satiation-attainment process for bursty activity in the active states , and a non-homogeneous Poisson process for longer inactivity between bursts in the inactive states ., For the short timescale , we found an inverse square power-law distribution for short ICIs with exponent ., Interestingly , a recent study in human short message correspondence , which requires feedback between individuals , suggests that the waiting time of the bursty communication follows the power-law distribution with exponent ., Analogously , a satiation-attainment process could govern the timing of feeding activity by waiting for satiation feedback ., In fact , it is well known that short-term feeding is regulated by feedback from the delayed gastrointestinal aftereffects of eating and satiety signals: based on this feedback , meal termination is determined 36 , 76 ., For the longer timescale , we found that longer ICIs follow the Weibull distribution in both the light and dark cycles ., At the same time , the cumulative distributions of the longer ICIs in the light and dark cycles exh

Introduction, Results, Models, Discussion, Conclusions, Materials and Methods

A fundamental understanding of behavior requires predicting when and what an individual will choose ., However , the actual temporal and sequential dynamics of successive choices made among multiple alternatives remain unclear ., In the current study , we tested the hypothesis that there is a general bursting property in both the timing and sequential patterns of foraging decisions ., We conducted a foraging experiment in which rats chose among four different foods over a continuous two-week time period ., Regarding when choices were made , we found bursts of rapidly occurring actions , separated by time-varying inactive periods , partially based on a circadian rhythm ., Regarding what was chosen , we found sequential dynamics in affective choices characterized by two key features:, ( a ) a highly biased choice distribution; and, ( b ) preferential attachment , in which the animals were more likely to choose what they had previously chosen ., To capture the temporal dynamics , we propose a dual-state model consisting of active and inactive states ., We also introduce a satiation-attainment process for bursty activity , and a non-homogeneous Poisson process for longer inactivity between bursts ., For the sequential dynamics , we propose a dual-control model consisting of goal-directed and habit systems , based on outcome valuation and choice history , respectively ., This study provides insights into how the bursty nature of behavior emerges from the interaction of different underlying systems , leading to heavy tails in the distribution of behavior over time and choices .

To understand spontaneous animal behavior , two key elements must be explained: when an action is made and what is chosen ., Here , we conducted a foraging experiment in which rats chose among four different foods over a continuous two-week time period ., With respect to when , we found bursts of rapidly occurring responses separated by long inactive periods ., With respect to what , we found biased choice behavior toward the favorite items as well as repetitive behavior , reflecting goal-directed and habitual responding , respectively ., We account for the when and what components with two distinct computational mechanisms , each composed of two processes:, ( a ) active and inactive states for the temporal dynamics , and, ( b ) goal-directed and habitual control for the sequential dynamics ., This study provides behavioral and computational insights into the dynamical properties of decision-making that determine both when an animal will act and what the animal will choose ., Our findings provide an integrated framework for describing the temporal and sequential structure of everyday choices among , for example , food , music , books , brands , web-browsing and social interaction .

behavioral neuroscience, computational neuroscience, biology and life sciences, computational biology, neuroscience

journal.pcbi.1004579

Temporal Gillespie Algorithm: Fast Simulation of Contagion Processes on Time-Varying Networks

Networks have emerged as a natural description of complex systems and their dynamics 1 , notably in the case of spreading phenomena , such as social contagion , rumor and information spreading , or epidemics 1–3 ., The dynamics of contagion processes occurring on a network are usually complex , and analytical results are attainable only in special cases 3 , 4 ., Furthermore , such results almost systematically involve approximations 3 , 4 ., Numerical studies based on stochastic simulations are therefore necessary , both to verify analytical approximations , and to study the majority of cases for which no analytical results exist ., The development of fast algorithms is thus important for the characterization of contagion phenomena , and for large-scale applications such as simulations of world-wide epidemics 2 , 5 ., The Doob-Gillespie algorithm 6–11 ( also known as Gillespie’s Stochastic Simulation Algorithm—SSA or Gillespie’s direct method ) , originally proposed by David Kendall in 1950 for simulating birth-death processes and made popular by Daniel Gillespie in 1976 for the simulation of coupled chemical reactions , offers an elegant way to speed up such simulations by doing away with the many rejected trials of traditional Monte Carlo methods ., Instead of checking at each time-step if each possible reaction takes place , as rejection sampling algorithms do , the Gillespie algorithm draws directly the time elapsed until the next reaction takes place and what reaction takes place at that time ., It is readily adapted to the simulation of Poisson processes on static networks 12–16 and can be generalized to non-Markovian processes 17 ., Systems in which spreading processes take place , e . g . , social , technological , infrastructural , or ecological systems , are not static though ., Individuals create and break contacts at time-scales comparable to the time-scales of such processes 18–20 , and the dynamics of the networks themselves thus profoundly affect dynamical processes taking place on top of them 21–27 ., This means that one needs to take the network’s dynamics into account , e . g . , by representing it as a time-varying network ( also known as a time-varying graph , temporal network , or dynamical network ) 28 ., The dynamical nature of time-varying networks makes the adaptation of the Gillespie algorithm to such systems non-trivial ., The main difficulty in adapting the Gillespie algorithm to time-varying networks is taking into account the variation of the set of possible transitions and of their rates at each time step ., We show that by normalizing time by the instantaneous cumulative transition rate , we can construct a temporal Gillespie algorithm that is applicable to Poisson ( constant rate ) processes on time-varying networks ., We give pseudocode and C++ implementations for its application to simulate the paradigmatic Susceptible-Infected-Susceptible ( SIS ) and Susceptible-Infected-Recovered ( SIR ) models of epidemic spreading , for both homogeneous and heterogeneous 29 populations ., We verify the accuracy of the temporal Gillespie algorithm numerically by comparison with a classical rejection sampling algorithm , and we show that it is up to ∼ 500 times faster for the processes and the parameter ranges investigated here ., While Poissonian models are of widespread use , real contagion phenomena show memory effects , i . e . , they are non-Markovian ., Notably , for realistic infectious diseases , the rate at which an infected individual recovers is not constant over time 30 , 31 ., Social contagion may also show memory effects , e . g . , one may be more ( or less ) prone to adopt an idea the more times one has been exposed to it ., To treat this larger class of models , we show how the temporal Gillespie algorithm can be extended to non-Markovian processes ., We give in particular an algorithm for simulating SIR epidemic models with non-constant recovery rates ., We define in this section the type of stochastic processes for which the temporal Gillespie algorithm can be applied ., At the time of writing , the main domain of application of the algorithm is the class of compartmental models for contagion processes on time-varying networks , which we introduce below ., For definiteness , algorithms detailing the application of the temporal Gillespie algorithm will concern this class of stochastic processes ., In general , we consider a system whose dynamics can be described by a set of stochastic transition events ., We assume that the system can be modeled at any point in time by a set , Ω ( t ) , of M ( t ) independent stochastic processes m , which we term transition processes; the rate at which the transition m takes place is denoted λm ., The set Ω ( t ) thus defines the possible transition events at time t and in general changes over time , depending on both external factors and the evolution of the system itself; the number of possible transitions , M ( t ) , thus also generally changes over time , while λm may or may not vary over time ., For the classic “static” Gillespie algorithm to be applicable , Ω ( t ) is allowed to change only when a transition ( or chemical reaction in the context of Gillespie’s original article ) takes place ., For processes taking place on time-varying networks , the medium of the process—the network—also changes with time ., As these changes may allow or forbid transitions , Ω ( t ) is not only modified by every reaction , but also by every change in the network ., This is the domain of the temporal Gillespie algorithm , which only requires that the number of points in which Ω ( t ) changes be finite over a finite time-interval 32 ., The assumption that the transition processes are independent is essential to the validity of the Gillespie algorithm , as it allows the calculation of the distribution of waiting times between consecutive transitions ., This assumption is not overly restrictive , as it only requires a transition process to be independent of the evolution of the other simultaneous transition processes ., A transition process may depend on all earlier transitions , and the current and past states of all nodes ., As such , Gillespie algorithms can notably be applied to models of cooperative infections and other non-linear processes such as threshold models 17 , and has even been applied to model the dynamics of ant battles 33 ., A straightforward way to simulate a stochastic process is to use a rejection sampling algorithm , akin to the classical Metropolis algorithm ., Here one divides the time-axis in small time-steps Δt , where Δt should be chosen sufficiently small such that this discretization does not influence the outcome of the process significantly; on time-varying networks , the choice of Δt often comes naturally as the time-resolution at which the network is measured or simulated ( Fig 1A ) ., At each time-step t = 0 , Δt , 2Δt , … , we test whether each possible transition m ∈ Ω ( t ) takes place or not ., In practice this is done by drawing a random number rm that is uniformly distributed on 0 , 1 ) for each m and comparing it to λmΔt: if rm < λmΔt the reaction takes place , if rm ≥ λmΔt nothing happens Fig 2 ( Transitions ) ., ( Note that one should technically compare rm to 1 − exp ( λmΔt ) to ensure that λm defines a proper transition rate for finite Δt ., However , the two procedures are equivalent in the limit Δt → 0 . ) From the design of the rejection sampling algorithm we see that the proportion of trials that are rejected is equal to a weighted average over {1 − λmΔt}m ., Thus , since we require λmΔt ≪ 1 in order to avoid discretization errors , the vast majority of trials are rejected and the rejection sampling algorithm is computationally inefficient ., The Gillespie algorithm lets us perform stochastically exact Monte Carlo simulations without having to reject trials ., For Poisson processes on static networks , it works by recognizing that the waiting time between two consecutive transitions is exponentially distributed , and that each transition happens with a probability that is proportional to its rate ., Specifically , the ( survival ) probability that the transition m has not taken place after a time τ since the last transition event is, S m ( τ ) = e - λ m τ ., ( 1 ), Since each transition takes place independently , the probability that no event takes place during the interval τ since the last event is, S ( τ ) = ∏ m S m ( τ ) = e - Λ τ , ( 2 ), where Λ = ∑ m = 1 M λ m is the cumulative transition rate ., The above result is obtained by using the fact that while Ω and M do depend on t , they only change when an event takes place and not in-between ., They can thus be treated as constant for the purpose of calculating the waiting time between events ., The distribution of the waiting times τ is then given by the probability density p ( τ ) = Λe − Λτ , while the probability density for the reaction m being the next reaction that takes place and that it takes place after exactly time τ is equal to pm ( τ ) = λm e − Λτ The static Gillespie algorithm thus consists in drawing the waiting time τ∼ Exp ( Λ ) until the next transition and then drawing which transition m takes place with probability πm = λm/Λ ., Here τ∼ Exp ( Λ ) is short for: τ is exponentially distributed with rate Λ . For processes taking place on time-varying networks however , the set of transition process , Ω ( t ) , changes with time independently of the transition events , e . g . , for the case of an SIR process nodes may become infected only when in contact with an infected individual ( Fig 1A ) ., This means that the survival probability does not reduce to a simple exponential as in Eq ( 1 ) ; it is instead given by, S m ( τ ; t * ) = exp ( - ∫ t * t * * I m ( t ) λ m d t ) , ( 3 ), where t* is the time at which the last transition took place , t** = t* + τ is the time when the next transition takes place , and Im ( t ) is an indicator function that is equal to one when the process m may take place , e . g . , when two given nodes are in contact , and zero when m may not take place ., The meaning of Im is exemplified in Fig 1A: the node i may be infected by the infectious node j only when the two nodes are in contact; if we let m denote this transition process , Im ( t ) is then one for t = Δt , 3Δt , 4Δt and zero for t = 0 , 2Δt ., Note that for processes taking place on adaptive time-varying networks , whose changes only depend on the process itself , Im ( t ) only changes when a transition takes place and Eq ( 3 ) reduces to Eq ( 1 ) ., This means that from the point of view of the algorithm , such networks are effectively static and the classic “static” Gillespie algorithm may simply be used there 14 , 16 ., We now consider the general case where Ω ( t ) may change independently of the processes evolving on the network ( described in Sec . 1: “Stochastic processes on time-varying networks” ) ., Using , as in the previous section , that transition processes are independent , we can write the probability that no event takes place during an interval τ ( the waiting time survival function ) :, S ( τ ; t * ) = ∏ m ∈ Ω S m ( τ ; t * ) = exp ( - ∑ m ∈ Ω ∫ t * t * * I m ( t ) λ m d t ) , ( 4 ), where Ω denotes the set of all possible transitions ( transition processes ) on the interval between two transition events , ( t* , t** , i . e . , Ω is the union over Ω ( t ) for t ∈ ( t* , t** , and M is the total number of transition processes on the same interval ( the size of Ω ) ., We switch the sum and the integral in Eq ( 4 ) to obtain, S ( τ ; t * ) = exp ( - ∫ t * t * * ∑ m ∈ Ω I m ( t ) λ m d t ) ., ( 5 ), Finally , using that Im ( t ) = 0 for all m ∉ Ω ( t ) , we may write, S ( τ ; t * ) = exp ( - ∫ t * t * * Λ ( t ) d t ) , ( 6 ), where, Λ ( t ) = ∑ m ∈ Ω ( t ) λ m ( 7 ), is the cumulative transition rate at time t ., The dynamics of empirical time-varying networks is highly intermittent and we cannot describe Ω ( t ) analytically ., This means that we cannot perform the integral of Eq ( 6 ) to find the waiting time distribution directly ., We may instead normalize time by the instantaneous cumulative transition rate , Λ ( t ) : We define a unitless normalized waiting time between two consecutive transitions , τ′ , as, τ ′ = L ( t * * ; t * ) = ∫ t * t * * Λ ( t ) d t , ( 8 ), i . e . , equal to the cumulative transition rate integrated over ( t* , t** ., The survival function of τ′ has the following simple form:, S ( τ ′ ) = exp ( - τ ′ ) ., ( 9 ), The time t** when a new transition takes place is given implicitly by L ( t * * ; t * ) = τ ′ , while the probability that m is the transition that takes place at time t = t** is given by:, π m ( t ) = I m ( t ) λ m / Λ ( t ) ., ( 10 ), This lets us define a Gillespie-type algorithm for time-varying networks by first drawing a normalized waiting time τ′ until the next event from a standard exponential distribution i . e . with unit rate , τ′ ∼ Exp ( 1 ) , and second , solving L ( t ; t * ) = τ ′ numerically to find t** ., In practice , since Λ ( t ) only changes when a transition takes place or at tn = nΔt with n ∈ N , we need only compare τ′ to, L ( t n + 1 ; t * ) = ( t n * + 1 - t * ) Λ ( t * ) + Δ t ∑ i = n * + 1 n Λ ( t i ) , ( 11 ), for each time-step n ( Fig 2A–2C ) ., Here n* is the time-step during which the last transition took place , and Λ ( t* ) is the cumulative transition rate at t* , immediately after the last transition has taken place ., The first term of Eq ( 11 ) is the cumulative transition rate integrated over the remainder of the n*th time-step left after the last transition; the second term is equal to L ( t n + 1 ; t n * + 1 ) ., A new transition takes place during the time-step n** where L ( t n * * + 1 ; t * ) ≥ τ ′ ( Fig 2D ) ; the precise time of this new transition is, t * * = t n * * + τ ′ - L ( t n * * ; t * ) Λ ( t n * * ) ; ( 12 ), the reaction m that takes place is drawn with probability given by Eq ( 10 ) ( Fig 2D ) ., We then update Ω and Λ to Ω ( t** ) and Λ ( t** ) ( Fig 2E ) , draw a new waiting time , τ′ ∼ Exp ( 1 ) , and reiterate the above procedure ( Fig 2F ) ., The algorithm can be implemented for contagion processes on time-varying networks as follows ( see Methods for pseudocode for specific contagion models and S1 Files for implementation in C++ ) : By construction , the above procedure produces realizations of a stochastic process for which the waiting times for each transition follow exactly their correct distributions ., The temporal Gillespie algorithm is thus what we term stochastically exact: all distributions and moments of a stochastic process evolving on a time-varying network obtained through Monte Carlo simulations converge to their exact values ., Rejection based sampling algorithms are stochastically exact only in the limit λmΔt → 0 . A large literature exists on the related problem of simulating coupled chemical reactions under externally changing conditions ( e . g . , time-varying temperature or volume ) 35–40 ., Most of these methods consider only external perturbations that can be described by an analytical expression ., In this case the problem reduces to that of defining a static , yet non-Markovian , algorithm ., Some methods , and notably the modified next reaction method developed by Anderson 37 , can be adapted to a completely general form of the external driving and thus , in principle , to simulate dynamical processes taking place on time-varying networks ., These methods are based on a scheme that is conceptually similar to Gillespie’s direct algorithm , the next reaction method , proposed by Gibson and Bruck 35 ., The next reaction method draws a waiting time for each reaction individually and chooses the next reaction that happens as the one with the shortest corresponding waiting time ., It then updates the remaining waiting times , draws new waiting times ( if applicable ) , and reiterates ., To generalize the next reaction method to processes with non-exponential waiting times , Anderson introduced the concept of the internal time for each transition process 37 ., In the notation used in the present article it is defined as T m ( t ) = ∫ 0 t I m ( t ) λ m d t and is thus equivalent to the normalized time , L ( t , 0 ) , only for an individual transition process ., By construction , the next reaction method needs to draw only one random number per transition event , where the Gillespie algorithm draws two ., However , this reduction in the number of required random variables comes at a price: one must draw a random number for each individual transition process and keep track of , compare , and update each of the individual waiting times ., For chemical reactions , where the number of different chemical reactions is small ( it scales with the number of chemical species ) , this tradeoff favors the next reaction method ., However , for contagion processes on networks , each individual is unique ( if not intrinsically , at least due to its position in the network ) ., The number transition processes thus scales with the number of nodes and contacts , which favors the Gillespie algorithm as it does not need to keep track of each of them individually 17 ., On time-varying networks ( or for time-varying external driving ) one must furthermore update relevant internal times each time the network structure ( external conditions ) changes in the next reaction method ., Chemically reacting systems are usually close to being adiabatic , i . e . , the external driving changes slowly compared to the time-scales of chemical reactions ., Thus , the additional overhead related to updating individual internal times is practically negligible ., However , the dynamics of temporal networks is highly intermittent and the time-scale of network change is typically smaller than the time-scales of relevant dynamical processes ., Here one must thus update the internal times many times between each transition event , inducing a substantial overhead ., Since the temporal Gillespie algorithm operates with a single global normalized waiting time , it handles these updates more efficiently ., Finally , the modified next reaction method may in principle be extended to non-Markovian processes taking place on time-varying networks ( as treated in Sec . 6: “Non-Markovian processes” using the temporal Gillespie algorithm ) ., However , such an approach would , for each single transition , require solving numerically Eq ., ( 13 ) of 37 for the internal waiting time of each individual transition process , taking into account the time-varying network structure , finding the shortest corresponding waiting time in real time , and then updating the internal waiting times of all the other reactions , rendering the next reaction method even more inefficient in this general case ., For real-world contagion processes , transition rates are typically not constant but in general depend on the history of the process 30 , 31 ., Such processes are termed non-Markovian ., The survival probability for a single non-Markovian transition process taking place on a time-varying network is given by:, S m ( τ ; F t ( m ) ) = exp ( - ∫ t * t * * I m ( t ) λ m ( t ; F t ( m ) ) d t ) ., ( 15 ), Here F t ( m ) is a filtration for the process m , i . e . , all information relevant to the transition process available up to and including time t; typically , F t ( m ) will be its starting time and relevant contacts that have taken place since ., As above , t* is the time of the last transition and t** = t* + τ is the time of the next ., Note that since λm now depends explicitly on t , we may absorb Im in λm; however , to underscore the analogy with the Poissonian case , we keep the factor Im explicitly in Eq ( 15 ) . We use again that the transition processes are independent , to write the waiting time survival probability:, S ( τ ; F t ) = exp ( - ∫ t * t * * Λ ( t ; F t ) d t ) , ( 16 ), with, Λ ( t ; F t ) = ∑ m ∈ Ω ( t ) λ m ( t ; F t ( m ) ) , ( 17 ), and where F t is the union over F t ( m ) for m ∈ Ω ., For a static network , Eq ( 6 ) reduces to the result found in 17 ., This can be seen by noting that M ( t ) = M and Ω ( t ) = Ω are then constant , and thus that λ m ( t ; F t ( m ) ) = - d S m ( t ; F t ( m ) ) / d t / S m ( t ; F t ( m ) ) = d { ln 1 / S m ( t ; F t ( m ) ) } / d t and S m ( t ; F t ( m ) ) = S m ( t + t m ; F t ( m ) ) / S m ( t m ; F t ( m ) ) , yielding directly Eq ., ( 7 ) of 17 ., As in the Poissonian case ( Sec . 4: “Temporal Gillespie algorithm” ) we define the normalized waiting time , τ′ , as, τ ′ = L ( t * * ; t * , F t ) = ∫ t * t * * Λ ( t ; F t ) d t ., ( 18 ), This gives us the same simple form as above for the survival function of the normalized waiting time , τ′ ,, S ( τ ′ ) = exp ( - τ ′ ) , ( 19 ), and the probability that m is the transition that takes place at t = t** ,, π m ( t ; F t ) = I m ( t ) λ m ( t ; F t ( m ) ) Λ ( t ; F t ) ., ( 20 ), Until now our approach and results are entirely equivalent to the Poissonian case considered above ., However , since λm ( t ) in general depend continuously on time , the transition time t** is not simply found by linear interpolation as in Eq ( 12 ) ., Instead , one would need to solve the implicit equation L ( t * * ; t * ) = τ ′ numerically to find t** exactly ., To keep things simple and speed up calculations , we may approximate Λ ( t ) as constant over a time-step ., This assumes that ΔΛ ( t ) Δt ≪ 1 , where ΔΛ ( t ) is the change of Λ ( t ) during a single time-step ., It is a more lenient assumption than the assumption that Λ ( t ) Δt ≪ 1 which rejection sampling relies on , as can be seen by noting that in general ΔΛ ( t ) /Λ ( t ) ≪ 1 . The same assumption also lets us calculate L ( t n + 1 ; t * ) as in the Poissonian case:, L ( t n + 1 ; t * , F t ) = ( t n * + 1 - t * ) Λ ( t * ) + Δ t ∑ i = n * + 1 n Λ ( t i ; F t ) , ( 21 ), and the time , t** , at which the next transition takes place:, t * * = t n * * + τ ′ - L ( t n * * ; t * , F t ) Λ ( t n * * ; F t ) ., ( 22 ), Using the above equations , we can now construct a temporal Gillespie algorithm for non-Markovian processes ., This algorithm updates all λm ( t ) that depend on time at each time-step , where for the Poissonian case we only had to initialize new processes , i . e . , contact-dependent processes ( type b and c , Sec . 1: “Stochastic processes on time-varying networks” ) ., This means the algorithm is only roughly a factor two faster than rejection sampling compare dotted lines ( ϵ = 0 ) in Fig 6 ., To speed up the algorithm , we may employ a first-order cumulant expansion of S ( τ ; F t ) around τ = 0 , as proposed in 17 , 38 for static non-Markovian Gillespie algorithms ., It consists in approximating λ m ( t ; F t ( m ) ) by the constant λ m ( t * ; F t ( m ) ) for t* < t < t** and gives a considerable speed increase of the algorithm full ( ϵ → ∞ ) in Fig 6 ., However , the approximation is only valid when M ( t ) ≫ 1 43 , which is not always the case for contagion processes ., Notably , at the beginning and end of an SIR process , and near the epidemic threshold for an SIS process , M is small and the approximation breaks down; the approximate algorithm for example overestimates the peak number of infected nodes in a SIR process with recovery rates that increase over time compare full black line ( ϵ → ∞ ) to the quasi-exact full red line ( ϵ = 0 ) in Fig 7A ., An intermediate approach , which works when the number of transition processes is small , but is not too slow to be of practical relevance , is needed ., We propose one such approach below 44 ., We have presented a fast temporal Gillespie algorithm for simulating stochastic processes on time-varying networks ., The temporal Gillespie algorithm is up to multiple orders of magnitude faster than current algorithms for simulating stochastic processes on time-varying networks ., For Poisson ( constant-rate ) processes , where it is stochastically exact , its application is particularly simple ., The algorithm is also applicable to non-Markovian processes , where a control parameter lets one choose the desired accuracy and performance in terms of simulation speed ., We have shown how to apply it to compartmental models of contagion in human contact networks ., The scope of the temporal Gillespie algorithm is more general than this , however , and it may be applied e . g . to diffusion-like processes or systems for which a network description is not appropriate ., Tables 2 and 3 list the notation used in the manuscript ., Table 2 gives notation pertaining to the temporal Gillespie algorithm , and Table 3 lists notation pertaining to time-varying networks and compartmental contagion processes ., All simulations for comparing the speed of algorithms were performed sequentially on a HP EliteBook Folio 9470m with a dual-core ( 4 threads ) Intel Core i7-3687U CPU @ 2 . 10 GHz ., The system had 8 GB 1 600 MHz DDR3 SDRAM and a 256 GB mSATA Solid State Drive ., Code was compiled with TDM GCC 64 bit using g++ with the optimization option -O2 ., Speedtests were also performed using -O3 and -Ofast , but -O2 gave the fastest results , both for rejection sampling and temporal Gillespie algorithms ., For SIR processes simulations were run until I = 0; for SIS processes simulations were run for 20/ ( μΔt ) time-steps ( as in Fig 3 ) or until I = 0 , whichever happened first ., Between 100 and 10 000 independent realizations were performed for each data point depending on μΔt ( 100 for low μΔt and 10 000 for high μΔt ) ., For simulations on empirical contact data , data sets were looped if necessary ., We here give pseudocode for the application of the temporal Gillespie algorithm to three specific models: the first subsection treats the Poissonian SIR process , the second treats the Poissonian SIS process , and the third treats a non-Markovian SIR process with recovery times following a general distribution ., We assume in the following that the time-varying network is represented by a list of lists of individual contacts taking place during each time-step ., An individual contact , termed contact , is represented by a tuple of nodes , i and, j . The list contactListst gives the contacts taking place during a single time-step , t , for t = 0 , 1 , … , T_simulation-1 , where T_simulation is the desired number of time-steps to simulate ., The state of each node is given by the vector x , where the entry xi ∈ {S , I , R} gives the state of node, i . As one may always normalize time by the duration of a time-step , Δt , we have in the following set Δt = 1 ., Note that beta and mu in the code then corresponds to βΔt and μΔt , respectively ., When simulations are carried out on data which are looped due to their finite length , the speed of the temporal Gillespie algorithm may be further increased for processes with an absorbing state , such as the SIR process , by removing obsolete contacts to nodes that have entered such a state ., Pseudocode 3 shows pseudocode for removing obsolete contacts; its replaces lines 11 to 19 of Pseudocode 1 ., Pseudocode 3: Pseudocode for counting possible S → I transitions with removal of outdated contacts ., C++ code is given in S1 Files ., 01 CLEAR m_SI //S nodes in contact with I nodes 02 FOR contact in contactListst 03 ( i , j ) = contact 04 IF xi==S 05 IF xj==I 06 APPEND i to m_SI 07 ELSE IF xj==R //remove if xj==R 08 REMOVE contact from contactListst 09 ENDIF 10 ELSE IF xi==I 11 IF xj==S 12 APPEND j to m_SI 13 ELSE //remove if ( xi , xj ) ==I or xi==R 14 REMOVE contact from contactListst 15 ENDIF 16 ELSE //remove if xi==R 17 REMOVE contact from contactListst 18 ENDIF 19 ENDFOR

Introduction, Results, Discussion, Methods

Stochastic simulations are one of the cornerstones of the analysis of dynamical processes on complex networks , and are often the only accessible way to explore their behavior ., The development of fast algorithms is paramount to allow large-scale simulations ., The Gillespie algorithm can be used for fast simulation of stochastic processes , and variants of it have been applied to simulate dynamical processes on static networks ., However , its adaptation to temporal networks remains non-trivial ., We here present a temporal Gillespie algorithm that solves this problem ., Our method is applicable to general Poisson ( constant-rate ) processes on temporal networks , stochastically exact , and up to multiple orders of magnitude faster than traditional simulation schemes based on rejection sampling ., We also show how it can be extended to simulate non-Markovian processes ., The algorithm is easily applicable in practice , and as an illustration we detail how to simulate both Poissonian and non-Markovian models of epidemic spreading ., Namely , we provide pseudocode and its implementation in C++ for simulating the paradigmatic Susceptible-Infected-Susceptible and Susceptible-Infected-Recovered models and a Susceptible-Infected-Recovered model with non-constant recovery rates ., For empirical networks , the temporal Gillespie algorithm is here typically from 10 to 100 times faster than rejection sampling .

When studying how e . g . diseases spread in a population , intermittent contacts taking place between individuals—through which the infection spreads—are best described by a time-varying network ., This object captures both their complex structure and dynamics , which crucially affect spreading in the population ., The dynamical process in question is then usually studied by simulating it on the time-varying network representing the population ., Such simulations are usually time-consuming , especially when they require exploration of different parameter values ., We here show how to adapt an algorithm originally proposed in 1976 to simulate chemical reactions—the Gillespie algorithm—to speed up such simulations ., Instead of checking at each time-step if each possible reaction takes place , as traditional rejection sampling algorithms do , the Gillespie algorithm determines what reaction takes place next and at what time ., This offers a substantial speed gain by doing away with the many rejected trials of the traditional methods , with the added benefit of giving stochastically exact results ., In practice this new temporal Gillespie algorithm is tens to hundreds of times faster than the current state-of-the-art , opening up for thorough characterization of spreading phenomena and fast large-scale applications such as the simulation of city- or world-wide epidemics .

journal.pgen.1003112

A Genome-Wide RNAi Screen Reveals MAP Kinase Phosphatases as Key ERK Pathway Regulators during Embryonic Stem Cell Differentiation

Embryonic stem cells and induced pluripotent stem cells ( iPS cells ) are currently generating intense interest due to their potential therapeutic roles in regenerative medicine ( reviewed in 1 ) ., We are beginning to understand the rules governing the establishment and maintenance of the pluripotent state and , in particular , the signaling and transcriptional networks which define this state ( reviewed in 2–3 ) ., A number of genome-wide si/shRNA screens have been instrumental in deciphering these networks 4–6 ., In contrast , less attention has been directed towards understanding how embryonic stem cells lose their pluripotency and begin to differentiate ., Mouse embryonic stem cells can be maintained in a pluripotent state by culturing under a variety of defined conditions ( reviewed in 7 ) ., Traditionally , these cells are cultured in medium containing serum and the cytokine leukaemia inhibitory factor ( LIF ) 8–9 ., However , more recently , it was demonstrated that mouse embryonic stem cells can be maintained in a pluripotent ground state by using two specific protein kinase inhibitors ( known as “2i” conditions ) which target the ERK pathway component MEK and glycogen synthase kinase ( GSK3 ) ( 10; reviewed in 11 ) ., Removal of these two inhibitors promotes exit from the naïve ground state ., These studies therefore revealed an important role for the ERK and GSK3 pathways to enter into lineage commitment ( reviewed in 12 ) ., Moreover , the suppression of ERK signalling in the mouse embryo is sufficient to expand the pluripotent compartment in the early mouse embryo 13 and can enhance the efficiency of iPS cell generation by promoting completion of reprogramming 14–15 ., Importantly , the same pathways may operate in a functionally analogous manner in human pluripotent stem cells that have been genetically manipulated 16–17 ., The ERK pathway has previously been shown to trigger mouse ES cell differentiation 18–19 and is implicated in numerous developmental processes ( reviewed in 20 ) in addition to playing an important role in a variety of different stem cell types ( reviewed in 21 ) ., Less is known about GSK3 function in development and stem cell biology and the role for GSK3 is usually attributed to its ability to regulate β-catenin stability and hence limit the responses to Wnt pathway signalling ( reviewed in 11 , 22 ) ., Recently , a β-catenin-dependent mode of action has been demonstrated for GSK3 in the context of mouse embryonic stem cells , although this mode of action is not sufficient to explain all the effects of GSK3 signalling in this context ( 23–24; reviewed in 25 ) ., One major function of ERK MAP kinase signalling , is to orchestrate gene expression programmes in the cell ., In particular , this pathway directly targets a number of transcription and chromatin regulators and thereby controls their activities ( reviewed in 26–27 ) ., However , which of the ERK targets are important in embryonic stem cell differentiation are unknown ., It is also unclear how the canonical ERK pathway is controlled in these cells ., In this study , we took advantage of the fact that the combinatorial use of ERK pathway and GSK3 inhibitors maintains mouse embryonic stem cell pluripotency 10 and carried out a genome-wide siRNA screen to identify regulators and mediators of these pathways that influence the exit from pluripotency ., This has led to the identification of over 400 genes whose functions are required for efficient embryonic stem cell differentiation away from the pluripotent ground state ., The vast majority of these genes have not previously been implicated in this process; therefore our study provides an important new resource for the community ., Moreover , further downstream analysis has partitioned these genes into classes that functionally interact with the ERK and/or GSK3 pathways and has revealed an important role for MAP kinase phosphatases in controlling embryonic stem cell fate ., To identify the programme of genes involved in the loss of pluripotency and subsequent differentiation of embryonic stem cells , a genome-wide RNAi screen was performed using E14Tg2a mouse ES cells which are engineered to express an unstable version of GFP from the endogenous rex1 ( also known as zfp42 ) locus ., This reporter gene is regulated in an analogous manner to endogenous rex1 23 and provides a convenient readout for the loss of a naieve pluripotent stem cell marker Rex1 28 ( reviewed in 11 ) ., Rex1GFPd2 ES cells were maintained in media containing MEK and GSK inhibitors ( 2i ) to maintain their ES cell status and treated with siRNAs pools targeting ∼17 , 000 individual genes ., After 24 hrs , cells were exchanged into fresh media lacking these inhibitors and the levels of GFP in each cell were assessed over time ( Figure 1A ) ., A gradual loss of GFP expression occurred upon inhibitor withdrawal over a ∼2 day time period , with conversion of the majority of cells to low expression ( 1A ) ., We wanted to conduct the screen at the earliest possible time point to maximise the chances of detecting genes directly involved in the exit from pluripotency rather than secondary effectors ., The control siRNAs for fgf4 and gsk3β both significantly reduced GFP loss at 27–30 hrs ( Figure S1B ) ., Therefore we monitored the ratio of cells expressing high and low levels of GFP at this time point ., siRNAs were scored as positive hits when this ratio increased by more than two standard deviations ( SD ) above the mean of all siRNAs on each plate ., A conservative threshold was selected at this stage to be more inclusive before further downstream validation was performed ., This led to the identification of 792 siRNAs that delayed the loss of GFP expression , and hence target genes potentially involved in promoting pluripotency loss and/or cell differentiation ( Figure 1B; Table S1A ) ., Examples , include 2400001e08rik , raf1 and jarid2 ( Figure 1C; Figure S2A–S2D , left panels ) ., Importantly , this primary screen identified RNAi pools targeting nras , raf1 and gsk3β , as would be expected due to their known roles in the ERK and GSK3 pathways ., Moreover , further validation of the efficacy of our screen was demonstrated by the identification of a large number of siRNAs targeting genes encoding proteosomal proteins , as would be expected due to the subsequent increased half-life of the unstable GFP protein used as a readout in these assays ., In addition , this primary screen also revealed 130 siRNAs that accelerate the loss of GFP expression and hence target genes that function to maintain pluripotency and/or inhibit cell differentiation including known effectors such as esrrb , stat3 , and ctr9 4 , 29–30 ( Figure 1B; Figure S2E; Table S2 ) ., Furthermore , several of genes identified in our screen in this category were also identified in other screens designed to identify genes required for pluripotency 4–5 , 31–33 , including stat3 and smc1a ( both identified in 2 and 3 additional screens , respectively ) ( Table S3 ) ., As our primary interest was on the mechanisms of escape from the pluripotent ground state rather than the maintenance of pluripotency , we subsequently focussed on genes that were required for modulating the onset of differentiation ., Two secondary screens were performed with a different set of siRNA pools targeting the genes identified in the primary screen and either the same reporter cells ( ie Rex1GFPd2 ) or ES cells containing an alternative reporter gene , where GFP is instead driven by the oct4 ( also known as pou5fl ) promoter , thereby providing an independent readout for the loss of pluripotency ( Figure 1A; Figure S1C ) ., These screens gave rise to 398 and 420 positive hits respectively , and 316 of these siRNAs scored positive in both secondary screens ( Figure 1A , 1D and 1E; Figure S2A–S2D , Table S1B and S1C ) ., These 316 siRNAs therefore define a high confidence dataset of genes that are required for the efficient loss of pluripotency and/or promoting the onset of differentiation of ES cells ., A number of these genes have already been implicated in embryonic stem cell differentiation control including tcf7l1 ( tcf3 ) , jarid2 , and dpy30 34–37 ( Table S4 ) further supporting the quality of our dataset ., Moreover , comparisons to other RNAi and overexpression screens performed on mouse ES cells 4 , 31–33 , 38 identified several genes in common , including jun and mbd3 which were both identified in two of these screens in addition to our own ( Table S3 ) ., However , the vast majority of genes we have identified here , have not been previously implicated in controlling the escape from the pluripotent ground state ., To assess the types of biological processes and potential mechanisms of actions of these 316 genes , gene ontology ( GO ) analysis was performed and prominent terms identified included a number of signalling pathways and also genes encoding transcriptional regulators ( Figure 1F and 1G; Figure S3 ) ., Thus cellular signalling events and subsequent gene expression control appear to play prominent roles in the early events associated with ES cell differentiation ., Having established the core network of genes working in concert with the GSK3 and ERK pathways we wanted to discover the relative contributions of these genes to the actions of the individual pathways ., First we performed a counter screen in the presence of both pathway inhibitors ( “+2i” ) to eliminate siRNAs which promoted accumulation of GFP in the cells irrespective of the activity of the ERK and GSK3 pathways ( Figure 2A ) ., This eliminated a further 42 siRNAs , including 14 that targeted proteosomal components and hence stabilised the GFP ( Table S5 ) ., This left 274 siRNAs which define genes required for efficient signal-dependent loss of pluripotency and the onset of differentiation ., The differentiation of ES cells away from pluripotency is maximally promoted by removing inhibitors of both GSK3 and the ERK pathway ., However , the removal of a single inhibitor permits ES cell differentiation and loss of Rex1-GFP signal , albeit with delayed kinetics ( Figure S4 ) ., We took advantage of this to partition our dataset and identify genes whose functions are specifically required for differentiation driven by either the ERK pathway or the GSK3 pathway alone ., siRNAs targeting the genes constituting the high confidence data set from the “2i” withdrawal screens were tested for their effect on Rex1-GFP loss upon single inhibitor withdrawal ( ie “1i” withdrawal screens; Figure 2A ) ., Of the 274 siRNAs tested , 133 delayed GFP loss upon withdrawal of the MEK inhibitor and 168 upon withdrawal of the GSK3 inhibitor ., Amongst these , 106 were in common ., A further 79 siRNAs had no effect on Rex1-GFP expression under either condition ( Figure 2A and 2B; Figure S5 ) ., Thus there are four functionally distinct classes of hits identified that are involved in promoting the onset of differentiation:, ( i ) in the context of the ERK pathway ( “ERK only hits” eg nras , Figure S2A; identified upon MEK inhibitor withdrawal only ) ;, ( ii ) in the context of the GSK3 pathway ( “GSK only hits” eg dmbx1 , Figure S2B; identified upon GSK3 inhibitor withdrawal only ) ;, ( iii ) in the context of either pathway ( “ERK/GSK hits” eg jun , Figure S2C; identified upon GSK3 or MEK inhibitor withdrawal ) ; and, ( iv ) in the context of both pathways together ( “ERK and GSK hits” eg gli3 , Figure S2D; no effect when either inhibitor is withdrawn ) ., Next to gain an insight into how the ERK and GSK3 pathways might function in the context of embryonic stem cells , we used gene ontology analysis to determine whether different groups of genes identified from the single inhibitor ( “1i” ) screens are associated with different biological processes ., Generally , the enriched GO terms for the genes from the initial 2i screen closely resemble those enriched in the “GSK” dataset ( Figure S6 ) ., However , closer inspection of the data revealed enriched GO terms that are more specific for genes which were associated with either the ERK or the GSK3 pathway , thereby revealing functionally distinct contributions of these pathways to the exit from pluripotency ( Figure 2C and 2D; Figure S7A–S7D ) ., For example , genes associated with either the ERK or GSK3 pathways are enriched in different signalling pathways ( Figure 2C ) and a number of terms associated with mitochondrial function are preferentially enriched in the genes associated with the GSK pathway ( Figure S7C ) ., However , other groups of GO terms were identified with generally high enrichment for genes associated with both the GSK and the ERK pathways ., This is typified by a large number of GO terms associated with transcriptional control ( Figure 2D ) ., Weaker enrichment of specific terms could be discerned for genes functionally associated with either the ERK or the GSK pathways ( Figure 2D ) ., We then created a network out of the genes from the high confidence dataset identified in the “2i” screen based on previous knowledge of physical and functional interactions ., Functionally related subnetworks could be identified , two of the most prominent of which are composed of genes encoding proteins associated with regulating chromatin modifications and sequence-specific DNA binding transcription factors ( Figure 2E; Figure S8A ) ., These genes showed strong interconnectivities with the rest of the network as might be expected from their regulatory functions ., Although only a limited number of connections between ERK and GSK3 signalling pathway components identified in the screen were revealed during network construction , these connections are made to transcription and chromatin regulators associated with the correct respective pathways ( eg Jun is connected to the Ras pathway and Gli3 is connected to Gsk3β; Figure 2E; Figure S8B ) ., In summary , by comparing single inhibitor assays , we have been able to subcategorise the genes required for embryonic stem cell differentiation and tentatively assign them to mediating or regulating the effects of either the ERK pathway , or the GSK3 pathway or both ., Each pathway appears to require genes associated with overlapping and yet distinct biological processes ., Our RNAi screen identified genes belonging to many functionally related categories and they are potentially involved in many biological processes ., However , to begin to understand the roles of the genes we have identified in controlling the loss of pluripotency and subsequent differentiation , we decided to focus mainly on the genes which were required for ERK-mediated differentiation as this pathway has a well established role in triggering mouse ES cell differentiation 18–19 ., The majority of “ERK only” genes and a subset of “ERK/GSK” genes were taken for further investigation alongside several control genes from the “GSK only” hits ( Figure 3A ) ., The relative strength of the effect of the knockdown of each gene in the context of the “1i” screens is illustrated in Figure 3A ., First we validated the roles of these genes by using RT-qPCR to monitor the loss of the pluripotency markers rex1 and nanog and the appearance of the early differentiation marker fgf5 ., The majority of the siRNAs tested showed increased rex1 and nanog expression relative to control siRNAs upon “2i” withdrawal ( Figure 3B; Figure S9B ) ., Importantly , an excellent correlation was observed between effects on rex1 and nanog expression ( Figure 3B; R2\u200a=\u200a0 . 81 ) ., Conversely , more than half of the siRNAs tested reduced the accumulation of fgf5 mRNA ( Figure S9C ) ., However , there was generally reduced concordance between the severity of the effects on fgf5 and rex1 ( Figure 3C ) or fgf5 and nanog ( Figure S9D ) expression ., For example depletion of jarid2 and pabpc1 causes some of the largest effects in maintaining rex1 expression but has no effect on reducing fgf5 accumulation ., Conversely , reductions in ets1 and dmbx1 limit fgf5 expression while having only a small effect on rex1 expression ., Nevertheless , a group of siRNAs can be identified that limit the loss of rex1 expression and show reduced accumulation of fgf5 ( Figure 3C; quadrant 1 ) and hence have effects on both loss of naive pluripotency and the onset of differentiation ., In contrast , there is another large group of genes that appear to affect pluripotency status but have little effect on the onset of early differentiation ( Figure 3C , quadrant 2 ) ., It is unclear why this occurs but it might reflect that although individual siRNAs promote retention of pluripotency , they might also trigger the activation of subsets differentiation markers , thus the two processes need not be tightly linked ., To extend the analysis of differentiation events , we focused on the two of the top hits attributed to ERK signalling , gmnn and 3830406c13rik , and also asked whether the appearance of markers of the three embryonic cell lineages was affected ., First we determined whether pluripotent cells remained in the population by alkaline phosphatase staining ., Increased numbers of alkaline phosphatase stained cells were identified 5 days after “2i” withdrawal upon depletion of either gene , confirming their importance for escape from the pluripotent ground state ( Figure 3D ) ., Depletion of gmnn caused reductions in the expression of all three lineage markers at both 3 and 5 days following “2i” withdrawal , consistent with a general role in regulating the escape from the pluripotent ground state ( Figure 3E ) ., Similarly , depletion of 3830406c13rik , caused reduced expression of all three markers at day 3 ( albeit only marginally for tbx6 ) , and reduced levels of nestin after 5 days ( Figure 3E ) ., However , increased expression of gata4 and tbx6 was observed at this later timepoint , suggesting a lineage specific role for this gene ., Thus , the contributions of individual genes identified in our screen towards individual lineage commitment are likely complex ., In summary , the use of marker genes allows us to further validate the hits in our screen , although the effects of depleting individual genes on the loss of naive pluripotency and/or differentiation vary according to the gene involved ., Next , to further investigate the function of the hits identified in our screen , we investigated how this subset of genes impacted on ERK pathway regulation and function ., In theory , genes might act to control ERK pathway activity or alternatively might mediate the effects of ERK pathway signaling ., Therefore as a first step to partition genes as acting up or downstream of ERK , we used western blotting to monitor the active phosphorylated form of ERK ( Figure 4A , Figure S10 ) ., Using this assay , siRNAs targeting 21 different genes were identified as upstream regulators of ERK ., Importantly , none of the “GSK3 only” hits affected ERK activation , further validating our partitioning of the data ( Figure 4A; Figure S10 ) ., Furthermore , while “ERK only” hits are partitioned evenly as acting up and downstream of ERK activation , the “ERK/GSK” hits are more prominent downstream of ERK ( Figure 4B ) , as might be expected for genes which are important for GSK-mediated differentiation when ERK signaling is inhibited ., To further delineate their point of action , we then tested the subset of siRNAs which acted upstream of ERK for their effects on Ras activation by an ELISA-based assay ( Figure 4C ) ., Eleven genes were identified whose point of action is upstream of both Ras and ERK ( Figure 4C; Figure S11 ) ., Importantly , one of these genes was nras itself ., These assays therefore enabled us to position genes from the “ERK only” , and “ERK/GSK” datasets at different points in the ERK pathway , either acting upstream of Ras eg plekh1 or on the core pathway downstream from Ras ( Figure 4D ) ., The rest of the genes analysed appear to act downstream from ERK and hence are likely mediators of ERK pathway function ., Interestingly , transcription factors are over-represented in the subgroup of genes which act downstream of ERK ( Figure 4E ) , in keeping with the known major role of ERK signalling in controlling gene expression programmes ( reviewed in 26–27 ) ., Together , these findings indicate that we have identified groups of genes which affect either signalling through the ERK pathway and/or the downstream consequences of ERK activation ., While in this study we have focussed on studying genes which affect escape from pluripotency , and are associated with the ERK and GSK3 pathways , it is likely that many of the genes we have identified might also play a more general role in controlling stem cell pluripotency ., Indeed , several genes identified in our study were also identified previously in other siRNA screens conducted in cells maintained in the presence of serum and LIF rather than the “2i” conditions we used ( Table S3 ) ., To investigate this further , we tested 8 genes for their role in escape from pluripotency in Rex1GFPd2 ES cells maintained in serum and LIF and induced to differentiate by withdrawal of LIF ., Depletion of three of these genes , otx2 , etv5 and mbd3 , caused an increased retention of rex1 promoter-driven GFP expression , consistent with a disruption in escape from pluripotency ( Figure S12 ) ., Thus , it is likely that many of the genes we have identified in this screen will play a more general role in controlling cell fate decisions in ES cells maintained in serum plus LIF or “2i” conditions ., A group of 10 genes was identified which acted downstream of Ras but affected ERK phosphorylation levels and hence ERK activity ( Figure 4D ) ., To further probe the point of action of these genes , we tested MEK activation levels following their depletion but saw little difference ( data not shown ) ., Next , we therefore focussed on MAP kinase phosphatases ( also known as dual specificity phosphatases DUSPs ) , and hypothesised that increases in the levels and/or activity of these enzymes might be responsible for the reduced ERK activation that we observed and consequent effects on embryonic stem cell differentiation ., First we examined the set of genes we identified which accelerated differentiation in our primary siRNA screen for candidate dusp genes as we expected the loss of DUSPs would be predicted to enhance ERK phosphorylation and promote exit from pluripotency ., Dusp1 , dusp3 and dusp15 were amongst this category of genes ( Table S2 ) ., We therefore determined the expression of these genes and a range of additional phosphatases in embryonic stem cells before and after “2i” removal ., Amongst the genes tested , dusp1 , dusp5 and dusp6 levels all increased following “2i” withdrawal while dusp14 levels were fairly constant ( Figure 5A; Figure S13A ) ., The increased expression of all these phosphatases was dependent on active ERK pathway signalling as expected from other cellular systems ( reviewed in 39 ) but in the case of dusp1 combinatorial inhibition of ERK and GSK signalling was required for maximal inhibition ( Figure 5B; Figure S13B ) ., However , at the protein level , Dusp1 levels gradually declined following “2i” withdrawal while Dusp6 levels increased in line with the increases in their mRNA levels ( Figure S13C ) ., Due to their dynamic expression , we focussed on Dusp1 , Dusp5 and Dusp6 as these have the potential for controlling ERK pathway activity during embryonic stem cell differentiation ., We therefore asked whether depletion of any of the genes identified in our screen would affect Dusp1 , Dusp5 and Dusp6 expression at the mRNA or protein levels ., Almost all the siRNAs tested ( 9/10 ) caused an increase in basal dusp1 mRNA levels and the same was observed on dusp6 levels for 4/10 genes ( Figure 5C ) ., In contrast , none of the siRNAs caused increases in dusp5 levels under these conditions ( Figure S13D ) ., Similarly , the levels of these dusps followed a similar pattern in response to siRNA treatment after release from “2i” for 40 mins ( Figure S13E ) ., Importantly , increases in Dusp1 and Dusp6 at the protein level were also observed which generally correlated with the effects of these siRNAs on mRNA levels ( Figure S13F ) although there were exceptions typified by Rab24 whose depletion does not affect dusp1 mRNA levels but instead appears to act post-transcriptionally to cause increased levels of Dusp1 protein ., An increase in the basal levels of MAP kinase phosphatases would likely lead to changes in the ERK activation kinetics , leading to the decreases in phosphorylated ERK levels we observed previously ( Figure 4A ) ., Indeed all of the siRNAs tested which promote increases in Dusp levels also cause a delay in peak activation of ERK and a subsequent reduction in the magnitude of this activation ( Figure 5D; Figure S13G ) ., Importantly other control siRNAs do not elicit this effect ( Figure S14A ) ., This suggests a causative link between the genes we identified in our screen , their effects on dusp gene expression and subsequent changes in ERK activity and downstream differentiation ., Two key predictions of this model are that reductions in Dusp levels should first increase the rate and level of ERK pathway activation , and secondly , promote differentiation of embryonic stem cells ., Indeed , depletion of Dusp6 and Dusp1 levels caused premature and higher amplitude activation of ERK whereas depletion of Dusp3 and a range of other Dusps had little effect on ERK activity levels ( Figure 5E; Figure S14A ) ., Importantly , while depletion of dusp1 and dusp6 caused increased levels of ERK activation , no increases could be detected on the low levels of Jnk and p38 phosphorylation , demonstrating a specific effect on the ERK pathway ( Figure S14B ) ., In our primary siRNA screen , we found that dusp1 depletion enhanced the loss of rex1 promoter-driven GFP expression ( Figure 5F ) ., We therefore depleted other Dusps to examine whether might function in an analogous manner and found that amongst these , only reductions in dusp6 levels triggered more efficient inactivation of the rex1-GFP reporter gene ( Figure 5G ) ., Similarly , dusp1 and dusp6 depletion caused increased loss of mRNA expression of the pluripotency marker nanog whereas dusp3 depletion had little effect ( Figure 5H ) ., Thus Dusp1 and Dusp6 appear to play an important role in maintaining pluripotency ., We extended this analysis to examine whether depletion of dusp1 or dusp6 affected lineage commitment by examining the expression of different marker genes 5 days after “2i” withdrawal ., The depletion of dusp1 caused increased expression of all three lineage markers , consistent with a general role for this gene in inhibiting loss of pluripotency ( Figure 5I ) ., In contrast , depletion of dusp6 only caused increased levels of the ectoderm marker nestin , suggesting a more specific role in controlling differentiation into this lineage ( Figure 5I ) ., Together , these results therefore demonstrate that our RNAi screen has enabled us to identify an important role for a subset of MAP kinase phosphatases in determining the rate and efficiency of ERK pathway activation in embryonic stem cells , and hence influence their ability to escape from pluripotency and begin to differentiate ., The derivation of pluripotent iPS cells and the controlled differentiation of embryonic stem cells into defined cell fates are two of the most important areas of research in the area of regenerative medicine ., Numerous studies have helped build up a view of the complex signaling and transcriptional networks involved in maintaining the pluripotent state of embryonic stem cells ( reviewed in 2–3 ) but in contrast , much less is known about the pathways leading to the loss of pluripotency ., Here we have conducted a genome-wide siRNA screen and identified over 400 genes which play a role in the onset of differentiation which allows ES cells to initiate escape from pluripotency ., The vast majority of these genes have not previously been implicated in this process ., This dataset therefore provides an important resource for the community and is a rich source of information for further investigating this phenomenon and also for a more basic understanding of the mechanisms governing the regulation and action of the core ERK and GSK3 signaling pathways ., Due to the controlled conditions used in our screen , we were able to link the genes which we identified to either the ERK and/or the GSK3 pathways as potential regulators or mediators of pathway functions ., Importantly , it appears likely that many genes we have identified might also be important in the context of different culture conditions such as the commonly used serum and LIF-containing media ( see Figure S12 ) ., However , further analysis on a case by case basis is required to substantiate a role for individual genes under these conditions ., It is important to emphasise that ES cells grown in LIF and “2i” conditions exhibit very different epigenetic landscapes , so only a partial overlap in regulatory factors is expected when comparing these conditions 40 ., Indeed , this is not unexpected considering that RNAi screens , including our own , commonly identify chromatin and transcriptional regulators as major important functionally enriched categories ( see Figure 2E ) ., Here we focused on genetic interactions with the ERK pathway , and we were able to place a large number of genes as acting upstream or downstream from ERK ( Figure 4 ) ., Further subpartitioning of the dataset enabled us to identify genes which functioned upstream of Ras or between Ras and ERK ( Figure 4D; Table S6 ) ., A surprising finding was that all of the genes which acted downstream of Ras , controlled ERK activation levels through controlling the levels of the MAP kinase phosphatases Dusp1 and/or Dusp6 ., The major point of control was at the transcriptional level ., MAP kinase phosphatases are known regulators of MAP kinases activity in different cellular contexts , and Dusp6 in particular operates as part of a feedback loop in response to ERK activation ( reviewed in 39 ) ., While Dusp6 is able to specifically dephosphorylate and inactivate ERK in vitro , Dusp1 can also target the stress activated MAP kinases , JNK and p38 ( reviewed in 39 ) ., However , we saw no evidence for elevated levels of phosphorylated Jnk and p38 in mouse embryonic stem cells upon depletion of Dusp1 , indicating its effects are likely via ERK ., Fluctuations in both Dusp1 and Dusp6 levels occur upon ERK pathway activation in ES cells , suggesting that they play an important feedback regulatory role in this system ., It appears likely that the combined amounts of these phosphatases helps set the threshold for ERK activation and hence ERK-mediated loss of pluripotency ( Figure 5J ) ., Indeed , tampering with this threshold control switch , either by depleting genes that control Dusp levels , or by directly depleting dusp1 or dusp6 , alters this threshold and changes the activation kinetics of the ERK pathway ., This in turn accelerates the loss of pluripotency and increases the expression of lineage-specific markers , indicating that Dusps help control the equilibrium between pluripotency and differentiation by maintaining the correct levels of ERK activity ., Our demonstration of a key role for Dusps in early ES cell differentiation , adds to the literature demonstrating the role of these enzymes in controlling developmental processes ( reviewed in 41 ) and illustrates the importance of establishing signaling thresholds by balancing activating and inactivating mechanisms which converge on ERK pathway signaling ., Indeed , a recent study demonstrated a role for a different phosphatase , Dusp9 , in maintaining pluripotency in mouse embryonic stem cells maintained in the presence of LIF and BMP4 42 ., In this study , BMP4 was implicated in upregulating dusp9 expression through Smad pathway activation and hence leading to a dampening down of ERK activity ., Importantly , they demonstrated that Dusp9 was not relevant to ERK control in stem cells maintained in 2i conditions , and

Introduction, Results, Discussion, Materials and Methods

Embryonic stem cells and induced pluripotent stem cells represent potentially important therapeutic agents in regenerative medicine ., Complex interlinked transcriptional and signaling networks control the fate of these cells towards maintenance of pluripotency or differentiation ., In this study we have focused on how mouse embryonic stem cells begin to differentiate and lose pluripotency and , in particular , the role that the ERK MAP kinase and GSK3 signaling pathways play in this process ., Through a genome-wide siRNA screen we have identified more than 400 genes involved in loss of pluripotency and promoting the onset of differentiation ., These genes were functionally associated with the ERK and/or GSK3 pathways , providing an important resource for studying the roles of these pathways in controlling escape from the pluripotent ground state ., More detailed analysis identified MAP kinase phosphatases as a focal point of regulation and demonstrated an important role for these enzymes in controlling ERK activation kinetics and subsequently determining early embryonic stem cell fate decisions .

Embryonic stem cells and induced pluripotent stem cells represent potentially important therapeutic agents in regenerative medicine ., Manipulation of these cell types could allow us to replace dead or diseased cells in our bodies and hence potentially provide a solution to a wide range of medical problems ., However , before we can perform such manipulations , we need to understand how the stem cells are wired so that we are able to re-wire them in a logical way to produce the desired cell types ., Here we have attempted to understand this wiring by using an RNAi screen in which each individual component of the cell is systematically removed and the consequences on cellular fate determined ., We have identified hundreds of genes that are required for efficient loss of stem cell characteristics and hence conversion into other cell types ., By studying a subset of these genes , we have been able to show that many converge on two related negative regulators of one of the key pathways that act to promote loss of stem cell identity ., These negative regulators , Dusps , normally limit the ability of stem cells to change their function and hence be converted to different cell types .

mechanisms of signal transduction, cell differentiation, gene function, developmental biology, stem cells, signaling pathways, mapk signaling cascades, erk signaling cascade, feedback regulation, embryonic stem cells, biology, molecular biology, signal transduction, genetic screens, genetics, wnt signaling cascade, molecular cell biology, gene networks, genetics and genomics, cell fate determination, signaling cascades

journal.ppat.1003204

Mutualistic Co-evolution of Type III Effector Genes in Sinorhizobium fredii and Bradyrhizobium japonicum

Eukaryotes universally encounter bacteria that inhabit , infect , and often provide them with significant fitness benefits ., In many cases , bacterial mutualist lineages exhibit intimate interactions with hosts , giving each partner opportunity to shape the phenotype of the other ., Two diametric paradigms remain unresolved for the co-evolution of bacterial mutualists with their eukaryotic hosts 1 ., One common paradigm models mutualist-host interactions as an antagonistic arms race , as is the case for co-evolution of pathogens and their hosts ., Under this model , natural selection is predicted to shape partners to rapidly evolve traits to maximize their own selfish gains from the interaction and minimize costs invoked by the other 1 ., This paradigm predicts that there is constant conflict over the fitness gain that each partner receives from the interaction even though both partners can attain net fitness benefits ., The alternative framework assumes that conflicts between microbe and host are largely resolved 2 , 3 ., It is predicted that the common genotypes are more likely to find compatible partners than the rare genotypes ., As a consequence , the interaction is predicted to exhibit evolutionary stability , with lower rates of evolutionary change ., Testing these competing frameworks by comparing the genetic patterns of known host-association genes between mutualists and pathogens will help to examine whether bacteria-eukaryotic mutualisms represent reciprocally exploitative interactions , as they have often been characterized , or alternatively , if these interactions exhibit a “mutualistic environment” in which evolutionary stasis is maintained 1 , 3 ., A striking and well-studied example of arms race co-evolution occurs between proteobacterial pathogens and plant hosts ., Plants have multiple defense systems to recognize and respond to bacterial infection ., One key plant defense is pattern-triggered immunity ( aka PAMP-triggered immunity; PTI ) , in which pattern recognition receptors detect conserved microbe-associated molecular patterns and trigger defenses 4 ., To counteract host defenses , many phytopathogenic bacteria use type III secretion systems ( T3SS ) to deliver collections of type III effector proteins ( T3Es ) to dampen host defenses , thereby allowing the bacteria to proliferate within host tissues and cause disease ., A second line of host defense is effector-triggered immunity ( ETI ) in which resistance ( R ) proteins surveil for corresponding microbial effectors to trigger a robust defense often associated with a localized programmed cell death ( hypersensitive response; HR ) ., Plant pathogen T3Es exhibit patterns of genetic variation that reflect rapid evolution , as predicted by the antagonistic arms race model 1 , 5–8 ., In Pseudomonas syringae , the phytopathogenic species with the most extensive experimentally-validated set of T3Es , strains vary dramatically in T3E gene content , both in terms of the total number and sequence of effector genes 1 , 7 ., Even highly related strains exhibit T3E presence/absence polymorphisms and insertion/deletion mutations that affect their coding sequences 2 , 3 , 9 ., An important aspect of pathogen T3E collections is that their robustness is ensured via T3E redundancy so that any individual T3E gene is dispensable 1 , 3 , 10 ., Hence , under the arms race scenario , rapid evolution of T3Es is advantageous to phytopathogens as novel collections of T3Es are more likely to avoid recognition while balancing sufficiency in subverting host defenses ., Functional T3SS orthologs have been uncovered in diverse mutualistic species of proteobacteria , including nitrogen-fixing rhizobial species Sinorhizobium fredii ( Ensifer fredii ) , Bradyrhizobium japonicum , and Mesorhizobium loti 4 , 11 , 12 ., Analyses of T3SS and T3E ( Nops; Nodulation Outer Proteins ) of rhizobia reveal many parallels to those of phytopathogens , pointing to the possibility that rhizobial nop genes are also under selection to maximize rhizobial fitness at potential expense of the fitness of the host ., For instance , multiple studies have shown that T3SS and Nops of rhizobia are necessary for the establishment of mutualist infections and can modulate host PTI 13–20 ., Moreover , T3Es of rhizobia also risk detection by host defense surveillance systems ., In fact , legume loci responsible for “nodulation restriction” are R genes that restrict rhizobia in a T3SS-dependent manner and are linked to loci associated with resistance against phytopathogens 21–24 ., This is consistent with the repeated observations that rhizobial strains deleted of genes encoding T3SS-secreted proteins gain new hosts that were once incompatible 19 , 24 , 25 ., Since no study has examined the molecular evolution of T3Es in the context of mutualism , it is presently unknown whether these lineages exhibit patterns of genetic variation that would reflect arms race evolution with their hosts 5–8 ., To this end , we investigated the molecular evolution of T3E genes in two lineages of mutualistic rhizobia and tested the arms race versus mutualistic environment paradigms ., We used an experimentally validated set and compared their genetic patterns against the patterns of T3Es from five monophyletic strains of P . syringae ( group I strains ) and four that infect legumes ( legume pathovars ) to test the null hypothesis that collections of T3Es of mutualists evolve in a manner similar to those in proteobacterial phytopathogens ., We selected three S . fredii and five B . japonicum strains based on the criterion of demonstrable reliance on T3SS for host infection 11 ., For B . japonicum , we also chose strains based on the genetic diversity inferred from their phylogenetic relationship 26 ., At the initiation of this study , the only available finished genome sequences were for S . fredii NGR234 and B . japonicum USDA110 27 , 28 ., We used paired-end Illumina sequencing to generate draft genome sequences for S . fredii USDA207 , USDA257 , and B . japonicum USDA6 , USDA122 , USDA123 , and USDA124 ( Table S1 ) ., Initial and post hoc analyses based on comparisons to reference and corresponding finished genome sequences completed subsequent to our efforts , respectively , indicated that the assemblies and annotations are of sufficient quality and covered the majority of the genomes for whole-genome characterization and comprehensive genome mining ( Figure S1 ) ., We next used multiple measures to compare the within-group diversity for the rhizobial groups to that of the group I and legume P . syringae pathovars to determine the suitability of the latter two for genomic comparisons ( Figure 1; 7 ) ., Quantitative measures of phylogenetic diversity ( PD ) fell within a narrow range with the two rhizobial groups having the higher PD values 29 ., We also compared bacterial group PD values to those derived from equally sized groups of strains randomly assigned from the 17 used in this study ., The within-group diversity of S . fredii , B . japonicum , and P . syringae , are similar , marginally , and significantly lower , respectively , relative to expectations due to chance ., Additional measures based on average reliable single nucleotide polymorphisms ( SNPs ) per kilobase ( kb ) and average percent of orthologous pairs of genes were also consistent ( Figures 1 , S1 , and S2 ) ., In total , the data demonstrate that the levels of genome-wide , within-group genetic diversity are higher in the S . fredii and B . japonicum groups , respectively , relative to either of the P . syringae groups ., Candidate T3E genes were identified based on their association with a tts-box , a cis element proposed to be recognized by TtsI , a regulator of T3SS genes in rhizobia 13 , 16 ., We identified a total of 305 putative tts-boxes ( Table 1 ) ., In S . fredii NGR234 , we identified two additional tts-box sequences that were not previously reported 13 ., In the finished genome sequence of B . japonicum USDA110 , we found 52 tts-boxes , of which 29 were previously identified ( Table 1; 30 ) ., Fourteen of these tts-boxes are located upstream of 13 genes ( bll1862 has two upstream tts-boxes ) that encode proteins that are secreted in a T3SS-dependent manner 30 ., We searched up to 10 kb downstream of the 305 tts-boxes and identified a total of 268 candidate T3E genes that clustered into 92 different families ( Table 1 ) ., We adopted the Δ79AvrRpt2 reporter in the γ-proteobacterium P . syringae pv ., tomato DC3000 ( PtoDC3000 ) for high throughput testing of candidate rhizobial T3E for T3SS-dependent translocation into plant cells , the most important criterion for defining a T3E 31 ., We first selected NopB and NopJ from S . fredii NGR234 as likely T3E candidates for validation of heterologous T3SS-dependent translocation ., NopB is secreted in vitro in a flavonoid- and T3SS-dependent manner from S . fredii NGR234 , and NopJ is a member of the YopJ/HopZ T3E family 32 , 33 ., PtoDC3000 carrying either the nopB::Δ79avrRpt2 or nopJ::Δ79avrRpt2 fusions elicited HRs within the same time frame ( ∼20 hours post inoculation; hpi ) and to the same degree as the positive control , a fusion between the full-length avrRpm1 P . syringae T3E gene and Δ79avrRpt2 ( Figure 2A ) ., Although Arabidopsis ecotype Col-0 can elicit ETI in response to both AvrRpm1 and AvrRpt2 , the observed HR is known to be a consequence of perception of the latter by RPS2 34 ., Each of the tested nopB::Δ79avrRpt2 gene fusions were sufficient for PtoDC3000 to trigger an HR at 20hpi , confirming that this family encodes bona fide T3Es ( Figure 2B ) ., The NopB family is polymorphic with NopBNGR234 sharing ≥98% amino acid identity with NopBUSDA207 , but only 32% with NopBUSDA110 ., In contrast , PtoDC3000 lacking fusions to Δ79avrRpt2 failed to elicit an HR but eventually showed tissue collapse approximately 28 hpi , indicative of PtoDC3000-caused disease symptoms ( data not shown ) ., The T3SS-deficient mutant of PtoDC3000 ( ΔhrcC ) , regardless of the gene it carried , failed to elicit any phenotype throughout the course of the study , thereby demonstrating the T3SS-dependent delivery of T3Es ( Figure 2A ) ., The demonstration that members of a polymorphic T3E family behaved identically in the heterologous delivery assay allowed us to test just a subset of 127 genes that represent the diversity present in the 268 candidates ., From these , 87 T3Es belonging to 47 families between the two rhizobial lineages were confirmed for T3SS-dependent translocation ( Table 1; Figure 3 ) ., We also used the sequences of members of confirmed T3E families to re-survey all draft genome sequences and identified an additional 21 homologs that were interrupted by physical and sequence gaps ., Nine CDSs were amplified using PCR and sequenced and all were classified as functional based on the absence of premature termination codons ., The remaining 12 genes belonged to 10 families with four homologs having upstream sequences similar to a tts-box but with bit-scores below our threshold ., Seven had no discernible upstream tts-box , and one ( nopM2 ) potentially represents a subgroup of the nopM family since two copies are present in B . japonicum USDA123 ., Of the 24 candidate and confirmed T3E families that were identified prior to this study , our computational method identified 21 of which 19 were experimentally validated as T3Es ( Table S2; 35 , 36 ) ., Of the five that we failed to confirm , NopA may in fact be a secreted structural component of the T3SS 37 ., NopT , in contrast , is likely a bona fide T3E but its cytotoxic effects in Arabidopsis could have caused misleading conclusions in the translocation assay 18 ., NopC , NopH and NopD lacked a detectable tts-box or failed to meet the requirement of being >100 amino acids in length ., The T3Es were assigned to families according to guidelines developed for T3Es of pathogenic bacteria 38 ., Newly identified T3E families were assigned NopY through NopBT whereas 16 previously named families , that were confirmed in this study as representing T3Es , remain unchanged 11 ., A relational table of the validated T3E genes is provided ( Table S2 ) ., Other than those previously identified , none of the translated sequences of the T3E genes identified in this study have detectable homology to proteins of known function ., We compared the genetic patterns of the rhizobial T3Es to:, 1 ) those of the group I strains of pathogenic P . syringae , and, 2 ) the core genome of the respective bacterial groups ( i . e . , genes ubiquitous to all strains within a group ) to test the null hypothesis that rhizobial T3Es exhibit signatures of arms race evolution similar to what has been characterized in pathogenic P . syringae lineages 5–8 ., The T3Es of rhizobia were predominantly core , unlike the T3Es of the group I strains of P . syringae ( Figure 4A ) ., In fact , the representation of T3Es among the four categories of core , singletons ( present in only a single strain of a group ) , pseudogenes ( premature termination codon relative to a full-length family member ) , and other ( polymorphic in regards to presence/absence ) , was significantly different ( Figure 4B ) ., Next , we compared the proportion of core and accessory T3E genes in the S . fredii , B . japonicum , and P . syringae group I strains to the proportion of genes that are core and accessory to each group ( Figure 4C ) ., Analysis indicated that the proportions of core T3E genes were significantly more than core genes for both groups of rhizobia ., In contrast , the proportion of core T3E genes for the group I strains of P . syringae was significantly less ., Thus , the collections of T3E genes of rhizobia are significantly more conserved than the collection of T3E genes of P . syringae and relative to their core genomes ., The sequences within T3E families of S . fredii and B . japonicum are also highly conserved , as more than 75% of the within-family pairwise comparisons had ≥90% amino acid identity ( Figure 5A ) ., Strikingly , twenty of the T3E families had all members with ≥99% identity ., The T3Es of the group I P . syringae strains have a wider distribution in amino acid identity and a greater number of presence/absence polymorphisms than S . fredii or B . japonicum ., Even when the latter variation was excluded from analysis , S . fredii and B . japonicum exhibit significantly more amino acid conservation of T3Es than group I P . syringae , whereas there was only marginal difference between the rhizobial lineages ( Figure 5B ) ., To determine whether the levels of sequence conservation of T3E gene families differed relative to genes core to their respective genomes , we calculated and compared the within-family amino acid identity for the translated sequences of gene families core to each of the groups ( Figure 5C ) ., The T3E gene families were significantly more conserved in sequence in both groups of rhizobia ., In contrast , for the five group I P . syringae strains the translated sequences of the T3Es exhibited significantly lower amino acid identities as compared to the translated sequences of the core gene families ., Therefore , relative to their respective core genes , the T3E genes of rhizobial and P . syringae lineages differed , with the former displaying higher levels of sequence conservation and the latter having significantly lower conservation ., There is no clear relationship between T3Es and host range of pathogens 7 , 39 ., P . syringae strains that infect the same host possess substantially different collections of T3Es ., For example , P . syringae pairs , pvs ., tomato races DC3000 and T1 and lachrymans races 106 and 107 , share no more than 50% of their T3Es in common 7 , 40 ., The high variability in T3Es is in spite of the lower levels of genetic diversity detected relative to most pairs of rhizobial strains ( Figures 1 and S2; 9 ) ., It has been suggested that Xanthomonas pathovars with similar hosts share similar compositions of T3Es 41 , 42 ., However , the genome-wide diversity is unknown for these bacteria and furthermore , use of contemporary methods to study two Xanthomonas species has revealed a surprisingly high number of pseudogenized T3Es and divergence in T3E collections 43 , 44 ., For P . syringae , it is hypothesized that T3Es are capable of functioning in a range of plant species 39 ., The extensive host ranges for two of the rhizobial strains studied herein support this notion ., S . fredii NGR234 and USDA257 can infect 112 and 79 genera of host plants , respectively , many of which are not considered cultivated plants and are more apt to have high within-population genetic diversity 45 ., Support is further bolstered by the observation that P . syringae deleted of a T3E gene gains the ability to infect an otherwise non-host plant 46 ., Similarly , rhizobial mutants deficient in secretion of T3SS-associated proteins can gain new species of plants as hosts 19 , 24 , 25 ., To further test the potential for host range as a factor in the conservation of rhizobial T3Es , we compared their genetic patterns to those of T3Es from four P . syringae pathovars that like rhizobia , can infect legumes as hosts 7 ., First , we compared between the two P . syringae groups ., As expected , the within-group genetic diversity is similar ( Figures 1 and S2 ) ., The genetic patterns of T3Es of the legume pathovars do not deviate significantly from those of the group I strains ( Figures 1 , 4 , and S2 ) ., Finally , relative to core genes , the core T3Es of the legume pathovars have genetic patterns that are significantly different ( Figure 4C and 5C ) ., Thus , despite the fact that the legume pathovars are distributed between two P . syringae groups , they exhibit similar levels of genome-wide and T3E diversity as the group I strains ., Relative to the T3Es of S . fredii and B . japonicum , the T3Es of the legume pathovars are significantly more variable ., As was the case for comparisons to the group I strains , the representation of core , singletons , pseudogenes , and other T3E categories is significantly different between legume mutualist and legume pathovars ( Figure 4A and 4B ) ., Likewise , the T3Es of the legume mutualists have a significantly different distribution in percent amino acid identity within T3E families relative to those of the legume pathovars ( Figures 5A and 5B ) ., Therefore , we conclude that a difference in host range is not a likely explanation for the extreme contrasts in T3E conservation between mutualist and pathogen ., The type III secretion system is a key mechanism used by a diversity of bacterial mutualists to establish infections with their hosts ., We identified and validated type III effectors to test the two diametric frameworks of mutualist-host co-evolution ( Figures 2 and 3; Table S2 ) ., Rhizobial T3E genes show genetic patterns indicative of surprising conservation , pointedly contrasting the patterns consistent with the dynamic arms race model of co-evolution dogmatic for T3Es ( Figures 3–5 ) ., This finding is particularly striking in light of the observations that T3Es of mutualistic rhizobia are similar in regards to those of pathogens in having to maintain sufficiency in engaging and dampening PTI while avoiding ETI 15 , 17 , 24 , 47 ., Moreover , we demonstrated that the high conservation of T3Es in rhizobia relative to phytopathogens is not likely driven by differences in host range or phylogenetic diversity among genomes ( Figures 1 , S2 , 4 , and 5 ) ., The high conservation in sequence and the fact that most of the T3E loci are co-localized are also consistent with acquisition events by both species of rhizobia ., In B . japonicum , for example , most of the T3E genes are found distributed throughout an ∼700 kb-long symbiosis island ., However , analysis of B . japonicum USDA110 and USDA6 suggested that the symbiosis islands were acquired independently , arguing against a common genome innovation event 48 ., We favor an alternative explanation that the relative conservation of rhizobial T3Es reflects the selective pressures in these beneficial plant-microbe interactions ., It has been suggested that legume hosts exhibit less polymorphisms in loci that restrict nodulation , in contrast to the higher levels of polymorphisms observed in loci that mediate resistance against phytopathogens 49 ., Our data support this idea that novelty in mutualism can result in instability , specifically that rhizobial mutualists may be under pressure by the host that limits diversification 50 , 51 ., In this context , hosts select for the most beneficial rhizobial genotype and these consequently common genotypes are more likely to find a suitable host ., The type III effectors of S . fredii and B . japonicum thus exhibit mutualistic co-evolution with host defenses ., Bacterial strains used in this study were: S . fredii strains USDA207 and USDA257; S . fredii ( aka Rhizobium sp . ) NGR234; B . japonicum strains USDA6 , USDA110 , USDA122 , USDA123 , and USDA124; PtoDC3000 , its T3SS-deficient mutant ( ΔhrcC ) , and Escherichia coli DH5α ., Rhizobia strains and P . syringae were grown in modified arabinose gluconate media ( MAG ) or Kings B ( KB ) media , respectively , at 28°C ., E . coli DH5α was grown in Luria-Bertani ( LB ) media at 37°C ., Antibiotics were used at the following concentrations: 50 µg/ml rifampicin ( PtoDC3000 ) , 30 µg/ml kanamycin ( all bacterial strains ) , 50 µg/ml chloramphenicol ( B . japonicum strains ) , and 25 µg/ml gentamycin ( E . coli ) ., Genomic DNA was extracted from S . fredii strains USDA207 and USDA257 and B . japonicum strains USDA6 , USDA122 , USDA123 , and USDA124 using osmotic shock , followed by alkaline lysis and phenol-chloroform extraction ., We prepared 5 µg of DNA from each strain according to the instructions provided by the manufacturer ( Illumina , San Diego , CA ) ., Paired-end sequencing was done by the Center for Genome Research and Biocomputing Core Labs ( CGRB; Oregon State University , Corvallis , OR; Table S1 ) ., Velvet 0 . 7 . 55 was used to de novo assemble paired-end short reads 52 ., Multiple assemblies , using different parameters , were produced for each genome and the highest quality assembly was identified using methods described previously 53 ., Genomes were annotated using Xbase and further refined using the NCBI conserved domain database ( CDD; 54–61 ) ., The Mauve Aligner 2 . 3 ( default settings ) program was used to compare the draft and finished genomes and , in other instances , reorder contigs to reference sequences 62 ., To identify SNPs , we used Bowtie ver . 0 . 12 . 5 to align short reads to the finished genome sequence , allowing up to two mismatches 63 ., Reliable sequence differences were identified based on having coverage of ≥10 reads and ≥8 reads supporting the same alternative base call ., For P . syringae , we treated the publicly available genome sequences as true and incremented along the genome in 1 base pair increments , shearing in silico the genome into 32mers , and aligned the sequences to the indicated reference genome sequence ., Homologous sequences were identified using reciprocal BLASTP analysis ( e-value≤1×10−15; >50% length of sequence ) of translated sequences ( those <50 amino acids in length were excluded ) ., The Circos plot was generated using the Circos Table Viewer 64 ., Genome sequences were retrieved from http://www . ncbi . nlm . nih . gov/genome: S . fredii NGR234 ( NC_012587 ) , S . fredii USDA257 ( NC_018000 ) , B . japonicum USDA6T ( NC_017249 ) , B . japonicum USDA110 ( NC_004463 ) , the P . syringae pathovars , actinidiae ( Pan; AEAL00000000 ) , glycine ( Pgy R4; ADWY00000000 ) , lachrymans ( Pla 106; AEAM00000000 ) , morsprunorum ( Pmp; AEAE00000000 ) , phaseolicola ( Pph 1448a; NC_005773 ) , pisi ( Ppi R6; AEAI00000000 ) , syringae ( B728a; NC_007005 ) , tomato ( PtoDC3000; NC_004578 ) , and tomato ( Pto T1; ABSM00000000 ) ., Finished genome sequences from S . fredii USDA257 and B . japonicum USDA6T were used for post hoc analysis of genome assemblies 48 , 65 ., We used HAL ( default settings ) to identify clusters of orthologous genes and generate a whole-genome phylogeny of the 17 strains plus two δ-proteobacterial reference strains , Geobacter sulfurreducens PCA ( NC_002939 ) and Desulfovibrio vulgaris RCH1 ( NC_017310 ) , used as outgroups 66 ., PD values were calculated using the Picante R package 67 ., To calculate PD values for randomly assigned groups of three , four , five , and five strains , an ad hoc Perl script was used to randomly assign the 17 rhizobial and P . syringae strains into four groups ., The process was iterated 1000 times and PD values were calculated for each group per iteration ., Statistical significance was determined by comparing the observed PD values to the proportion of 1000 iterations that had higher or lower PD values than the observed PD values ., To identify the proportion of core genes for each group of strains , we identified the clusters of orthologous genes , generated by HAL , that were represented by all strains within each group ., Fishers exact test was used to compare the representations of T3E genes in the four categories for all possible pairs of bacterial groups 68 ., The Kolmogorov-Smirnov test was used to compare the distributions of percent amino acid identity of T3E genes for all pairwise comparisons 69 ., We developed a linear regression model that evaluates the average percent amino acid identity for both core and T3E families , using the core genes in the group I strains of P . syringae as the baseline:where Y\u200a=\u200athe response variable , percent amino acid identity; P\u200a=\u200a1 for the legume pathovars of P . syringae and P\u200a=\u200a0 otherwise; B\u200a=\u200a1 for the B . japonicum species and B\u200a=\u200a0 otherwise; S\u200a=\u200a1 for the S . fredii species and S\u200a=\u200a0 otherwise; E\u200a=\u200a1 for T3E families and E\u200a=\u200a0 for core gene families; ε\u200a=\u200arandom error ., Specifically , ß0 measures the average percent amino acid identity of the core genes in the group I strains of P . syringae; ß1 measures the difference in percent amino acid identity between the T3E families and the core gene families for the group I strains of P . syringae; ß2 , ß3 and ß4 measure the differences in percent amino acid identity for the core gene families for the legume pathovars of P . syringae , B . japonicum , and S . fredii , respectively , against that of the group I strains; ß5 , ß6 and ß7 allow the variation of the differences in percent amino acid identity between T3Es and core gene families across the groups; in particular , ß1+ß5 , ß1+ß6 and ß1+ß7 measure the differences in percent amino acid identity between the T3E families and the core gene families for the legume pathovars , B . japonicum , and S . fredii , respectively ., An F test was used to test the null hypotheses that the percent amino acid identity for within-family comparisons between translated T3E and core gene sequences are equal within bacterial groups: ß1\u200a=\u200a0 , ß1+ß5\u200a=\u200a0 , ß1+ß6\u200a=\u200a0 and ß1+ß7\u200a=\u200a0 ( F test with degrees of freedom 1 and 83712 ) ., A Bonferroni correction was used when applicable 70 ., We used sequences of 30 confirmed functional tts-boxes from B . japonicum , S . fredii and M . loti MAFF303099 to train a Hidden Markov Model 13 , 30 , 71 ., To identify candidate T3E genes , we identified CDSs downstream of tts-boxes with bit scores ≥5 . 0 , calibrated based on the identification of 11 functionally validated tts-boxes located on the pNGR234a plasmid 13 ., To be considered , CDSs had to be encoded on the same strand as the tts-box , either up to 10 kb downstream or until another CDS on the opposite strand was encountered ., TtsI-regulated operons , such as the nopB-rhcU operon of S . fredii NGR234 , can be substantial in length 72 ., We used BLASTX ( e-value≤1×10−15 ) to filter out CDSs with translated sequences homologous to components of the T3SS , proteins encoded by organisms that lack a T3SS , or proteins with general housekeeping functions ., We used BLASTN and sequences of candidate T3E-encoding genes to identify homologs from each of the eight genome sequences ( e-value cutoff≤1×10−15 ) ., T3Es were grouped into families based on BLASTP scores ≤1×10−5 across ≥60% the length of the protein 38 ., When all members of a family had amino acid identity ≥90% as determined using ClustalW , a single representative family member was chosen for testing 73 ., In families of <90% amino acid identity , members representative of the diversity were tested ., PCR , Gateway cloning into pDONR207 and the destination vector pDD62-Δ79AvrRpt2 , transformation into E . coli DH5α cells , and triparental mating into PtoDC3000 or ΔhrcC were done as previously described or according to the instructions of the manufacturer ( Invitrogen , Carlsbad , CA; 31 ) ., Infiltration and HR assays were done as previously described 31 ., Plants were grown in a controlled growth chamber environment ( 15-hour day at 22°C followed by 9-hour night at 20°C ) ., Experiments were replicated a minimum of three times .

Introduction, Results/Discussion, Materials and Methods

Two diametric paradigms have been proposed to model the molecular co-evolution of microbial mutualists and their eukaryotic hosts ., In one , mutualist and host exhibit an antagonistic arms race and each partner evolves rapidly to maximize their own fitness from the interaction at potential expense of the other ., In the opposing model , conflicts between mutualist and host are largely resolved and the interaction is characterized by evolutionary stasis ., We tested these opposing frameworks in two lineages of mutualistic rhizobia , Sinorhizobium fredii and Bradyrhizobium japonicum ., To examine genes demonstrably important for host-interactions we coupled the mining of genome sequences to a comprehensive functional screen for type III effector genes , which are necessary for many Gram-negative pathogens to infect their hosts ., We demonstrate that the rhizobial type III effector genes exhibit a surprisingly high degree of conservation in content and sequence that is in contrast to those of a well characterized plant pathogenic species ., This type III effector gene conservation is particularly striking in the context of the relatively high genome-wide diversity of rhizobia ., The evolution of rhizobial type III effectors is inconsistent with the molecular arms race paradigm ., Instead , our results reveal that these loci are relatively static in rhizobial lineages and suggest that fitness conflicts between rhizobia mutualists and their host plants have been largely resolved .

Rhizobia are an important group of bacteria that can enter into mutually beneficial symbiotic interactions with legume plants to fix atmospheric nitrogen ., However , in order to do so , a complex dialog involving the exchange of chemical and molecular signals must occur between partners ., Some species of beneficial rhizobia employ a type III secretion system , a well-characterized virulence mechanism used by pathogens to inject bacterial-encoded type III effector proteins directly into host cells to coerce the host into accommodating the microbe ., In this study , we generated draft genome sequences and employed computational as well as experimental methods to identify type III effectors from eight strains representing Sinorhizobium fredii and Bradyrhizobium japonicum ., We demonstrate that the type III effector genes of these rhizobial species are highly conserved in content with little diversity between strains ., This work is an important step towards understanding the roles for type III secretion systems and their effectors in mutualistic interactions .

gram negative, plant microbiology, microbial evolution, biology, microbiology, host-pathogen interaction, bacterial pathogens

journal.pgen.1000859

Structure, Function, and Evolution of the Thiomonas spp. Genome

In environments such as those impacted by acid mine drainage ( AMD ) , high toxic element concentrations , low levels of organic matter and low pH make growth conditions extreme ., AMD is generally characterized by elevated sulfate , iron and other metal concentrations , in particular , inorganic forms of arsenic such as arsenite ( As ( III ) ) and arsenate ( As ( V ) ) 1 , 2 ., While these waters are toxic to the majority of prokaryotic and eukaryotic organisms , a few Bacteria and Archaea are not only resistant to but also able to metabolize some of the toxic compounds present 1 ., Members of the Thiomonas genus are frequently found in AMD and AMD-impacted environments , as Thiomonas sp ., 3As and “Thiomonas arsenivorans” 2–6 ., These Betaproteobacteria have been defined as facultative chemolithoautotrophs , which grow optimally in mixotrophic media containing reduced inorganic sulfur compounds ( RISCs ) and organic supplements ., Some strains are capable of autotrophic growth and others are capable of organotrophic growth in the absence of any inorganic energy source 5 , 7 , 8 ., Recently described species and isolates include “Tm . arsenivorans” 9 , Tm ., delicata 10 , Thiomonas sp ., 3As 5 and Ynys1 11 ., Thiomonas sp ., 3As as well as other recently isolated strains from AMD draining the Carnoulès mine site ( southeastern France ) containing a high arsenic concentration ( up to 350 mg L−1 ) 3 , 12 , present interesting physiological and metabolic capacities , in particular an ability to oxidize As ( III ) ., Over the past few years an increasing number of genomes has been sequenced , revealing that bacterial species harbor a core genome containing essential genes and a dispensable genome carrying accessory genes 13 ., Some of these accessory genes are found within genomic islands ( GEIs ) 14 and have been acquired by horizontal gene transfer ( HGT ) ., These GEIs are discrete DNA segments ( from 10 to 200 kbp ) characterized by nucleotide statistics ( G+C content or codon usage ) that differ from the rest of the genome , and are often inserted in tRNA or tRNA-like genes ., Their boundaries are frequently determined by 16–20 bp ( up to 130 bp ) perfect or almost perfect direct repeats ( DRs ) ., These regions often harbor functional or cryptic genes encoding integrases or factors involved in plasmid conjugation or related to phages ., GEIs encompass other categories of elements such as integrative and conjugative elements ( ICE ) , conjugative transposons and cryptic or defective prophages ., Such GEIs are self-mobile and play an important role in genome plasticity 14 ., In almost all cases , GEIs have been detected in silico , by the comparison of closely related strains ., Nevertheless , the role of GEIs in genome plasticity has also been experimentally demonstrated in several pathogenic bacteria such as Staphylococcus aureus or Yersinia pseudotuberculosis 15 , 16 or in Pseudomonas sp ., strain B13 isolated from a sewage treatment plant 17 ., Deciphering dispensable genomes has revealed that the loss or gain of genomic islands may be important for bacterial evolution 18 ., Indeed , these analyses allow the determination of the genome sequence , called pan-genome or supragenome , not just of individual bacteria , but also of entire species , genera or even bacterial kingdom 19 , 20 ., These data result in debates on taxonomic methods used to define the bacterial species 21 , 22 , e . g . pathogens such as Streptococcus agalactiae 21 , 23 or environmental bacteria such as Prochlorococcus 24 , 25 or Agrobacterium 26 ., However , beyond these well-known and cultivable microorganisms , the diversity of bacteria , in particular those found in extreme environments , has hitherto been comparatively poorly studied ., Genome analysis of such extremophiles may yet reveal interesting capacities since these bacteria may express unexpected and unusual enzymes 27 ., Since the role of GEIs in extremophiles has not been yet well explored , little is known about their evolution ., In the present study , the genome of Thiomonas sp ., 3As was sequenced and analyzed ., It was next compared to the genome of other Thiomonas strains , either of the same species or of other species of the same genus ., This genome exploration revealed that Thiomonas sp ., 3As evolved to survive and grow in its particular extreme environment , probably through the acquisition of GEIs ., The genome of Thiomonas sp ., 3As comprises a 3 . 7 Mbp circular chromosome and a 46 . 8 kbp plasmid ( Table 1 ) ., The single circular chromosome contains 3 , 632 coding sequences ( CDSs ) ( Table 1 , Figure S1A ) ., The mean G+C content of the Thiomonas sp ., 3As genome is 63 . 8% ., However , the distribution along the genome revealed several regions with a G+C content clearly divergent from this mean value ( Figure S1 ) ., This suggests that several genomic regions are of exogenous origin ., Indeed , 196 genes having mobile and extrachromosomal element functions were identified in the genome , among which a total of 91 ISs ( Figure S1A , Table S1 ) representing 2 . 5% of total CDS ., None of these ISs were found as part of composite transposons , while several were identified as neighbors of phage-like site-specific recombinases ., The plasmid , pTHI , contains 68 predicted CDS ., 21 genes were found in synteny with genes carried by the R . eutropha JMP134 plasmid pJP4 , and among them , par/trf/pem genes necessary for plasmid partitioning , stability and replication ( Figure S1B ) ., These observations suggest that pTHI , as JMP134 , belongs to the IncP-1β group 28 ., pTHI contains 13 of the 14 genes involved in conjugation ( vir and tra genes ) and genes that could fulfill the function of the missing components were found on the chromosome ., Therefore , Thiomonas sp ., 3As may be able to express a complete Type IV secretory system ( T4SS ) of the Vir/Tra type required for pTHI conjugal transfer ., IncP-1β members are known to carry multi-resistance determinants and degradative cassettes 28 , and plasmid pTHI indeed carries a Tn3-related transposon ., This transposon contains part of a mercury resistance operon found in many other Gram negative bacterial transposon such as Tn21 , Tn501 and Tn5053 29 ., Thiomonas sp ., 3As is able to use organic compounds as a carbon source or electron donor 5 , 8 ., However , under certain conditions this bacterium may also be able to grow autotrophically 5 ., A complete set of cbb/rbc/cso genes involved in carbon fixation via the Calvin cycle , and genes involved in glycogen , starch and polyhydroxybutyrate ( PHB ) biosynthesis pathways were identified ( Figure 1 and 5 ) ., Fructose , glucose , several amino acid , C4-dicarboxylates , propionate , acetate , lactate , formate , ethanol and glycerol are potential carbon sources or electron donors , since genes involved in their import or degradation via the glycolysis , the Entner-Doudoroff , the TriCarboxylic Acid ( TCA ) or the “rubisco shunt” pathways are present in the genome ., The presence of all genes involved in the oxidative phosphorylation pathway ( Figure 1 ) suggests that Thiomonas sp ., 3As has a respiratory metabolism ., Moreover , since several genes coding for terminal oxidases ( i . e . cbb3 , bd or aa3 ) were found , this respiratory metabolism may occur over a wide range of oxygen tensions ., Finally , the presence of a nitrate reductase and of several formate dehydrogenases suggests that Thiomonas sp ., 3As is able to use nitrate anaerobically as an electron acceptor and formate as electron donor ., In the absence of carbohydrates , Thiomonas sp ., 3As is a chemolithotroph and may use reduced inorganic sulfur compounds ( RISCs ) as an electron donor 5 ., The presence of soxRCDYZAXB , dsr , sorAB , sqr and fccAB genes revealed that Thiomonas can oxidize thiosulfate , sulfite , S0 or H2S to sulfate ( Figure 1 ) 30 , 31 ., Thiomonas sp ., is a moderate acidophile ., Its optimum pH is 5 but this bacterium can withstand to pH as low as 2 . 9 ( Slyemi , Johnson and Bonnefoy , personal communication ) ., Thiomonas sp ., 3As pH homeostasis mechanisms may therefore be strictly controlled as previously described 32 , 33 ., First , genes encoding a potassium-transporting P type ATPase ( kdpABC ) are present in the Thiomonas sp ., 3As genome ., This ATPase could be involved in the generation of a positive internal potential produced by a greater influx of potassium ions than the outward flux of protons ., Second , to strengthen the membrane , likely by lowering membrane proton conductance , Thiomonas sp ., has cyclopropane fatty acids 5 ., Accordingly , two putative cfa genes encoding cyclopropane fatty acid synthase have been detected ., Third , cytoplasmic buffering can be mediated either by amino acid decarboxylation and/or by polyphosphate granules ., Genes encoding decarboxylases for lysine ( 4 CDS ) , phosphatidyl serine and glycine are present on Thiomonas sp ., 3As genome ., Moreover , urea ( formed from arginine by arginase ) may be degraded by urease ( ure genes ) or urea carboxylase and allophanate hydrolase ., Urease encoding genes are known to be involved in acid tolerance in Helicobacter pylori 34 ., Protons may be captured during polyphosphate synthesis ., Polyphosphate granules have indeed been observed in electron micrographs of thin sections of Thiomonas sp ., 3As 5 ., Genes involved in such mechanisms ( ppk , pap , ppx ) were found in 3As genome ., Fourth , primary and secondary proton efflux transporters were predicted by genome sequence analysis , including four putative Na+/H+ exchangers and voltage gated channels for chloride involved in the extreme acid resistance response in E . coli ( clcAB ) 35 ., Finally , the elimination of organic acids can lead to pH homeostasis ., Some organic acid degradation pathways have been detected in Thiomonas sp ., such as an acetyl-CoA synthetase-like ., Moreover , formate oxidation was observed ( Slyemi , Johnson and Bonnefoy , personal communication ) and two formate dehydrogenases are encoded by the Thiomonas sp ., 3As genome , these enzymes could convert acetate to acetyl-CoA and formate to CO2 and hydrogen , respectively ., The Carnoulès AMD contains a high concentration of heavy metals such as zinc or lead ., To resist to heavy metals , bacteria usually develop several resistance mechanism including toxic compounds extrusion pumps 36 or biofilm synthesis 37 ., Flagella are important for the first steps of biofilm formation and all genes involved in motility , twitching and chemotaxis , were found in its genome ., Thiomonas sp ., 3As is motile but , unlike H . arsenicoxydans , this motility was not affected by arsenic concentration ( Table S2 , 38 ) ., Finally , Thiomonas sp ., 3As is able to synthesize exopolysaccharides ( Table S2 ) , and one eps operon involved in their synthesis was identified in the genome , as well as two mdoDG clusters involved in glucan synthesis ., Several genes conferring resistance to cadmium , zinc , silver , ( cad , czc , and sil genes ) , chromium ( chr genes ) , mercury ( 2 mer operons encoding both MerA reductase but no organomercury lyase MerB ) , copper ( cop and cus genes ) and tellurite ( transporters THI_0898-0899 ) are likely involved in Thiomonas sp ., 3As heavy metal resistance ( Figure 1 ) ., Arsenic resistance in bacteria is partly due to the expression of ars genes , among which , arsC encodes an arsenate reductase , arsA and arsB encode an arsenite efflux pump , arsR encodes a transcriptional regulator 39 ., Thiomonas sp ., 3As is resistant to up to 50 mM As ( V ) and up to 6 mM As ( III ) ( Table S2 , 8 ) ., The analysis of the Thiomonas sp ., 3As genome revealed the presence of two copies of the ars operon , an arsRBC operon ( ars1 ) and an arsRDABC operon ( ars2 ) ., Thiomonas sp ., 3As is able to oxidize As ( III ) to As ( V ) 5 and this metabolism involves the arsenite oxidase encoded by aoxAB genes 5 ( Figure 1 ) ., It has been shown that arsenite is imported via the aquaglyceroporin GlpF in E . coli 40 ., However , as in H . arsenicoxydans 38 , no homologue of GlpF was identified in the Thiomonas sp ., 3As genome , suggesting that As ( III ) is imported via an unknown component ., As ( III ) is known to induce DNA damage and oxidative stress 41 , 42 ., 24 genes involved in such stress responses were identified in the Thiomonas sp ., 3As genome ( Figure 1 ) ., Moreover , this genome carries 54 genes involved in DNA repair ., However , this strain lacks some genes present in H . arsenicoxydans , such as alkB , whereas two genes involved in mismatch repair were duplicated ., Orthologs of genes that have been shown to be induced in response to arsenic in H . arsenicoxydans 38 were found in Thiomonas sp ., 3As , i . e . radA , recQ , ruvA , recA , xseA , polA , holB-like , dinB-like and parC ., The expression of polA has been previously shown to be induced in the presence of arsenic 8 , suggesting that the Thiomonas sp ., 3As response to arsenic include the expression of genes involved in DNA repair ., Several Thiomonas strains called CB1 , CB2 , CB3 and CB6 were isolated from the same environmental site as Thiomonas sp ., 3As ., The 16S rRNA/rpoA-based phylogeny of these isolates ( >97% nucleotide identity ) , as well as DNA-DNA hybridization experiments ( Figure 2A ) , revealed that they represent different strains of the same species ., All these strains are able to oxidize As ( III ) and are resistant to As ( III ) ( Table S2 ) ., Nevertheless , subtle physiological differences were observed ( Table S2 ) ., The existence of both phylogenetical relationships and physiological differences between these strains prompted us to perform a comparative genome analysis in order to address the evolution of Thiomonas strains ., Therefore , genome variability was searched for by investigating genetic similarities and diversities among these closely related Thiomonas strains , using a Comparative Genomic Hybridization ( CGH ) approach ( Figure 3 ) ., These experiments revealed the presence of a flexible CDS ( duplicated , absent or highly divergent ) pool in CB1 , CB6 CB3 and CB2 ( Figure 2B , Figure 3 , ArrayExpress database , accession number E-MEXP-2260 ) representing 2 . 5% , 3 . 2% , 24 . 1% and 23 . 1% of the genome of strain 3As ( Figure 2B ) , respectively ., Altogether , these experiments led to the definition of 919 dispensable CDS , i . e . absent or highly divergent in at least one strain , accounting for 25 . 3% of strain 3As genes ( Figure 2D ) ., The remaining conserved CDS ( 2713 CDS , 74 . 7% of the genome of strain 3As ) represent a common backbone of the “core” genes of this species ., In order to enlarge our comparative analysis , genomic similarities were similarly searched for in other Thiomonas species: an arsenite-oxidizing strain , “Tm . arsenivorans” , and two closely related strains that are unable to oxidize arsenite , Tm ., perometabolis and Thiomonas sp ., Ynys1 ( Table S2 , Figure 2 , ArrayExpress database , accession number E-MEXP-2260 ) ., No significant hybridization was observed with oligomers corresponding to the plasmid , suggesting that pTHI is absent in all these strains ., 18 . 4 , 37 . 9 and 53 . 6% of the 3As CDS were flexible in “Tm . arsenivorans” , Ynys1 and Tm ., perometabolis , respectively ., Altogether , 1571 CDS accounting for at least 43 . 3% of the Thiomonas sp ., 3As genome were found in the Thiomonas genus dispensable genome ( Figure 3D ) ., Finally , these CGH experiments revealed that the Thiomonas core genome contains 2 , 061 CDS ( 56 . 7% of the Thiomonas genome ) ., Interestingly , almost all genes involved in acid resistance described above , were found in this core genome , as for example genes involved in polyphosphate granule synthesis , cfa and kdp genes , genes encoding ion transporter amino acid decarboxylase , formate dehydrogenase and other hydrogenases ., One ars operon involved in arsenic resistance , i . e . ars1 , and almost all genes involved in DNA repair were also conserved in all strains ., Among the flexible pool , 19 regions ( ThGEI-A - ThGEI-S ) had similarities with GEIs found in other bacterial genomes , suggesting that they were possibly acquired by horizontal gene transfer:, ( i ) an abnormal deviation of the codon adaptation index ( CAI ) and the GC content at the 3rd nucleotide position of each codon ( GC3 ) was observed in these regions as compared to the rest of the genome ( Figure 3 ) ,, ( ii ) many of their genes formed syntenic blocks that differed from the general synteny observed in the rest of the genome ( Table S3 ) ,, ( iii ) genes with mobile and extrachromosomal element functions such as those coding for integrases were localized within these regions ,, ( iv ) these regions were present at the 3′-end of tRNA or miscRNA genes , and/or, ( v ) the borders of five deletions were verified in CB strains , by PCR and direct repeats ( 10 to 112 bp-long ) bordering these GEIs were found ( Table S3 ) ., Genes found in the 19 Thiomonas sp ., 3As GEIs and the syntenies they share with genes in other bacteria are shown in Table S3 ., Interestingly , 70 ( 76 . 9% ) of the 91 complete and partial ISs identified in the genome were located in genomic islands which represent only 21 . 5% of the genome ( Figure S1A ) ., In addition to the high numbers of ISs found in these GEIs , many hypothetical proteins as well as modification/restriction enzymes were encoded by these regions ., In ten GEIs , accessory genes are involved in a particular metabolism such as acetoin , atrazin , benzoate , ethyl tetra-butyl ether ( ETBE ) hydroxyisobutyrate phenylacetic acid and urea degradation ( ThGEI-E , ThGEI-C , ThGEI-S or ThGEI-R ) , or heavy metal resistance ( ThGEI-J , ThGEI-L , ThGEI-O ) ., Interestingly , several genes found in distinct GEIs shared high amino acid identity ( >70% , Figure S2 ) ., In addition , 47 genes found in the two regions ThGEI-C and ThGEI-S shared 100% identity ., Because of this duplication , a 7 kbp region in ThGEI-S could not be sequenced and this gap may correspond to duplicated genes of ThGEI-C ., These observations suggest that genomic rearrangements occurred between several GEIs ., Moreover , several islands seem to be composite , since some fragments of such islands are deleted or duplicated in Thiomonas strains ., Such composite structure may originate from insertion or excision of DNA elements in these GEIs , which involve integrase or excisionase ., This hypothesis is strengthened by the observation that 32 integrases were found in almost all GEIs except for ThGEI-B and ThGEI-R ., Some of such integrases are similar to phage integrases ., In addition , 2 excisionases are present in ThGEI-H and ThGEI-P and such genes were localized in the vicinity of tRNA , an additional phage-like character ., One GEI , ThGEI-J , contains a prophage region ( 55 . 6 kbp ) and a cluster of 6 heavy metal resistance genes ( 39 . 4 kbp , i . e . , cad , cus , czc and sil genes involved in resistance to Cd , Cu , Zn , Co and Ag ) ( Figure 4 ) ., The prophage region comprises 27 phage-related genes coding for structural and capsid or tail assembly proteins , replication , lysis and virulence factors ., No conserved synteny with any previously described prophage could be observed ., However , filamentous phage-like particles with icosahedral symmetry ( capsid diameter of approximately 100 nm ) and a various length tail ( >600 nm ) , were observed by TEM from Thiomonas sp ., 3As liquid cultures exposed to the phage lytic phase inducer mitomycin C . Similar phage-like particles were observed in growth culture supernatants from CB1 , CB3 , CB6 and “Tm . arsenivorans” ( Figure 4 ) but not from CB2 , Ynys1 and Tm ., perometabolis ( data not shown ) , in agreement with CGH results showing that the ThGEI-J is absent in these strains ( Figure 3 , Table S3 , ArrayExpress database , accession number E-MEXP-2260 ) ., These observations suggest that this prophage-like region may be functional in 3As , CB1 , CB3 , CB6 and “Tm . arsenivorans” under stress conditions , resulting in the formation of phage-like particles ., GEIs contribute to the adaptation of microorganisms to their ecological niches and participate in genome plasticity and evolution 14 ., Therefore , the environmental conditions may influence the loss or conservation of GEIs ., Such hypothesis was checked by searching for genome similarities between strains originated from similar environments , i . e . AMD ., To this aim , a hierarchical clustering was established based on genomic comparisons ( Figure 2C ) ., Interestingly , the clustering obtained was different from that of the 16S rRNA/rpoA-based phylogenetic trees ( Figure 2A ) ., Indeed , all strains that originated from AMD heavily loaded with arsenic , i . e . “Tm . arsenivorans” and strains 3As , CB1 , CB2 , CB3 and CB6 , grouped together , whereas Ynys1 and Tm ., perometabolis formed a distinct group ., Genes possibly dispensable for AMD survival were therefore searched for and we identified 2541 CDS conserved in all strains originated from AMD , and these CDS may constitute the “AMD” core genome of Thiomonas ., Interestingly , several genes present in the ThGEI-L and ThGEI-O were conserved in AMD-originated strains but absent in the other strains ( Figure 5 ) ., The ThGEI-L carries genes involved in panthotenate and biotine synthesis , and may confer auxotrophy to the strains carrying this island ., Moreover , genes encoding Co/Zn/Cd efflux pump were present in this GEI ., In addition , this island is particularly rich in proteins with GGDEF and EAL domains ., The GGDEF or EAL domain proteins are involved in either synthesis or hydrolysis of bis- ( 3′-5′ ) cyclic dimeric GMP ( c-di-GMP ) , an ubiquitous second messenger in the bacterial world that regulates cell-surface-associated traits and motility 43 , 44 ., Because of the presence of such genes in the vicinity of 2 genes involved in chemotaxis , this island may be important for Thiomonas strains to form biofilm , a cellular process involved in resistance to toxic compounds 37 ., Indeed , some of these genes are duplicated in CB2 and CB3 and these two strains were shown to develop better biofilm synthesis capacities ( Table S2 ) ., This island also carries several genes encoding integrases and components of T4SS , such as virB1 , virB4 , trbBCD , traCEFGI , mob and a pilE-like gene ., The presence of such genes suggests that this island originated from an integrative and conjugative element ( ICE ) that disseminates via conjugation 45 ., These observations suggest that this island may be still mobile ., ThGEI-O contains the aox and ars2 genes ( Figure 5 ) ., In addition , several other genes were found such as mer , cop , cus and cad genes involved in mercury , copper and cadmium resistance , respectively , cys involved in sulfate assimilation , and moe/moa genes involved in molybdenum cofactor biosynthesis as well as genes involved in exopolysaccharide production ., The synteny of the genes found in Thiomonas sp ., 3As ThGEI-O is not conserved in other arsenic-oxidizing bacteria ( Figure 5 ) ., Several genes present in this region are duplicated in CB1 and CB6 ( i . e . the cop and aox genes ) , or in CB3 ( i . e . mer , cop , cus , dsb , cys , ars , moe/moa , aox , and ptxB genes ) ., Only a single copy of this region is present in CB2 and 3As ., PCR amplification and sequencing revealed that this ThGEI-O island is located in a different genomic region in CB2 as compared to 3As ., Moreover , the aox and ars genes found in the ThGEI-O are duplicated in “Tm . arsenivorans” but absent in Ynys1 and Tm ., perometabolis ., Indeed , these two strains were unable to oxidize As ( III ) , their As ( III ) resistance was lower than that of the other strains , and gene PCR amplification of aox and ars2 failed with DNA extracted from these strains ( Table S2 ) ., Altogether , the presence of at least one copy of these genes in all six strains isolated from arsenic-rich environments ( i . e . 3As , CB strains and “Tm . arsenivorans” ) suggests that this GEI is of particular importance for the growth of Thiomonas strains in their toxic natural environment , AMD ., The evolutionary origin of the ThGEI-L and –O was investigated using two different approaches ., First , we performed the phylogenetic analysis of the 196 genes contained in these two islands ( Table S4 , Table S5 ) ., The resulting trees revealed that these genes have very different evolutionary histories suggesting that the formation of ThGEI-L and –O islands occurred through the recruitment of genes from various origins by HGT ( Table S4 , columns 2–5 ) ., Interestingly , the closest homologue of 30/75 and 22/121 3As genes , in ThGEI-L and –O respectively , is found in other Thiomonas species ( mainly Tm . intermedia ) , suggesting that the formation of these islands occurred prior to the diversification of the Thiomonas genus and is thus relatively ancient ., This hypothesis is supported by the global correspondence analysis ( COA ) performed on the entire genome ., Our results did not reveal any particular codon usage bias , strengthening the hypothesis that these ThGEIs are ancient in Thiomonas genus ( Figure S3 ) ., This may explained why the major genes of these two islands are present in 3As , CB1 , CB2 , CB3 , CB6 and “Thiomonas arsenivorans” , as for example , the ars2 operon and aox genes of the ThGEI-O ., The phylogenetic analysis of aox genes revealed that all Thiomonas sequences grouped together with relationships that are very similar to organism relationships inferred with rpoA ( Figure S4A and S4B ) ., This indicates that these genes were already present in the Thiomonas ancestor and vertically transmitted in this genus , but lost in Ynys1 and Tm perometabolis ., The phylogenetic analysis of the arsB genes , revealed that all Thiomonas sequences found in the ThGEI-O ( i . e . arsB2 from 3As , CB1 , CB2 , CB3 , CB6 and “Tm . arsenivorans” ) , grouped together but not with arsB1 genes that are part of the core genome of Thiomonas ., Moreover , the evolutionary histories of these two proteins are different: ArsB1 proteins belong in a group containing mainly Alpha-Proteobacteria , whereas ArsB2 seems more closely related to Gamma-Proteobacteria ( Figure S4C ) ., These observations revealed that the ars1 and ars2 operons were not acquired from the same source or at the same time ., The exploration of the Thiomonas sp ., 3As genome suggests that this strain has a wide range of metabolic capacities at its disposal ., Many of them may make this bacterium particularly well suited to survive in its extreme environment , the acidic and arsenic-rich waters draining the Carnoulès mine tailings , as for example biofilm formation and heavy metal resistance ., Moreover , some metabolic capacities are unique as compared to another arsenic-resistant bacterium , whose genome has been recently sequenced and annotated , H . arsenicoxydans , a strict chemoorganotroph , isolated from activated sludge 38 ., The first metabolic idiosyncrasy of Thiomonas sp ., 3As is its particular carbon and energy metabolic capacities ., Indeed , several organic or inorganic electron donors , such as reduced inorganic sulfur compounds 31 , could be used ., Second , some Thiomonas strains , i . e . CB1 , CB3 , CB6 and Tm ., arsenivorans , carry two copies of the aox operon ., As far as could be ascertained , this is the first example of aox gene duplication ., Finally , Thiomonas sp ., 3As is able to grow at pH 3 ., Several genes potentially involved in acid resistance were found in Thiomonas genome ., In addition , the Carnoulès toxic environment may cause severe DNA damage in Thiomonas sp ., 3As , since arsenic is a co-mutagen that inhibits the DNA repair system 41 ., DNA repair genes that have been previously shown to be induced in the presence of arsenic in H . arsenicoxydans were all found in Thiomonas sp ., 3As genome , and the expression of polA has been shown to be induced in the presence of arsenic 8 ., These observations suggest that this bacterium may respond to DNA damage ., Nevertheless , we can hypothesize that these stressful conditions may lead to genomic rearrangements in Thiomonas genome ., This could explain the important genomic diversity observed among the members of both the 3As species and the Thiomonas genus ., At the intra-species level , the dispensable genome defined by comparison of the CB strains with the 3As genome corresponds to 25 . 3% of Thiomonas sp ., 3As genome ., By comparison , this value is higher than that observed , with the same approach , in other bacteria such as S . agalactiae ( 18% ) 46 , lower than values calculated in the case of a pathogenic E . coli ( 32 . 4% ) 47 , and similar to the value obtained in Bacillus subtilis ( 27% ) 48 ., The value calculated for Thiomonas 3As and CB strains is very high , considering that these strains were isolated from the same site , closely related , and appear to share a recent common ancestor , as illustrated by our phylogenetical analyses ., Consequently , we observed that despite strong sequence identities of housekeeping genes such as 16S rRNA or rpoA , the whole genome DNA-DNA hybridization value was relatively low , close to or less than 70% , for strains CB2 and CB3 ., Conventionally , this should indicate that these bacteria belong to separate evolutionary lineages and must be considered as different species 49 ., However , the 16S rRNA-rpoA based analysis and CGH experiments revealed that the low DNA-DNA hybridization value correlates with the duplication or absence of several GEIs in these strains ., Consequently , we proposed that despite low DNA-DNA hybridization values , these five strains do indeed belong to the same species ., Similarly , the DNA-DNA hybridization values obtained with Thiomonas sp ., 3As as compared to strains Ynys1 and Tm ., perometabolis were very low , as previously observed 5 , 10 ., Altogether , the great genetic diversity observed in the present study by CGH experiments revealed that DNA-DNA hybridization method may not be appropriate to evaluate evolutionary lineages in Thiomonas strains ., In this respect the CGH approach seems to be a reliable phylogenetic tool for typing these strains , as suggested in previous studies on other bacteria 47 , 50 ., 19 GEIs constitute a large flexible pool of accessory genes that encode adaptive traits ., Some of these genes are not required for survival in AMD , since they were not found in all AMD-originated strain genome and correspond therefore to the dispensable gene pool ., On the other hand , CGH-based clustering analysis revealed a significant relationship between 3As , CB1-6 and “Tm . arsenivorans” , which originate from geographically distinct but similarly arsenic-rich environments ., The Thiomonas sp ., 3As strain and “Tm . arsenivorans” form two distinct groups on the basis of phylogenetical , physiological and genetic analyses ., Nevertheless , the percent of flexible CDS of Thiomonas sp ., 3As with “Tm . arsenivorans” , is relatively low ( 18 . 4% ) , as compared to the value obtained with Ynys1 and Tm ., perometabolis ( 37 . 9% and 53 . 6% , respectively ) ., This value obtained with “Tm . arsenivorans” was in the same order of magnitude as the value obtained with CB2 and CB3 ( 23% and 23 . 6% , respectively ) ., Altogether , 70% of the Thiomonas sp ., 3As genome was conserved among all strains originated from AMD ., Interestingly , two GEIs were conserved or duplicated in all these strains originated from AMD , i . e . ThGEI-O that carries the arsenic-specific operons ars2 and aox , and genes involved in heavy metal resistance , and ThGEI-L that carries several genes involved in heavy metal resistance , biofilm formation and/or motility ., Therefore , these GEIs shared by these species are presumably part of the AMD-originated Thiomonas core genome ., This observation suggests that the acquisition

Introduction, Results, Discussion, Materials and Methods

Bacteria of the Thiomonas genus are ubiquitous in extreme environments , such as arsenic-rich acid mine drainage ( AMD ) ., The genome of one of these strains , Thiomonas sp ., 3As , was sequenced , annotated , and examined , revealing specific adaptations allowing this bacterium to survive and grow in its highly toxic environment ., In order to explore genomic diversity as well as genetic evolution in Thiomonas spp ., , a comparative genomic hybridization ( CGH ) approach was used on eight different strains of the Thiomonas genus , including five strains of the same species ., Our results suggest that the Thiomonas genome has evolved through the gain or loss of genomic islands and that this evolution is influenced by the specific environmental conditions in which the strains live .

Recent advances in the field of arsenic microbial metabolism have revealed that bacteria colonize a large panel of highly contaminated environments ., Belonging to the order of Burkholderiales , Thiomonas strains are ubiquitous in arsenic-contaminated environments ., The genome of one of them , i . e . Thiomonas sp ., 3As , was deciphered and compared to the genome of several other Thiomonas strains ., We found that their flexible gene pool evolved to allow both the surviving and growth in their peculiar environment ., In particular , the acquisition by strains of the same species of different genomic islands conferred heavy metal resistance and metabolic idiosyncrasies ., Our comparative genomic analyses suggest that the natural environment influences the genomic evolution of these bacteria ., Importantly , these results highlight the genomic variability that may exist inside a taxonomic group , enlarging the concept of bacterial species .

microbiology/environmental microbiology, genetics and genomics/microbial evolution and genomics

journal.pgen.1007254

Large scale variation in the rate of germ-line de novo mutation, base composition, divergence and diversity in humans

Until recently , the distribution of germ-line mutations across the genome was studied using patterns of nucleotide substitution between species in putatively neutral sequences ( see 1 for review of this literature ) , since under neutrality the rate of substitution should be equal to the mutation rate ., However , the sequencing of hundreds of individuals and their parents has led to the discovery of thousands of germ-line de novo mutations ( DNMs ) in humans 2–6; it is therefore possible to analyse the pattern of DNMs directly rather than inferring their patterns from substitutions ., Initial analyses have shown that the rate of germ-line DNM increases with paternal age 4 , a result that was never-the-less inferred by Haldane some 70 years ago 7 , maternal age 6 , varies across the genome 5 and is correlated to a number of factors , including the time of replication 3 , the rate of recombination 3 , GC content 5 and DNA hypersensitivity 5 ., Previous analyses have demonstrated that there is large scale ( e . g . 1MB ) variation in the rate of DNM in both the germ-line 3 , 5 and the somatic tissue 8–12 ., Here we focus exclusively on germ-line mutations ., We use a collection of over 130 , 000 germ-line DNMs to address a range of questions pertaining to the large-scale distribution of DNMs ., First , we quantify how much variation there is at different scales and investigate whether the variation in the mutation rate at a large-scale can be explained in terms of variation at smaller scales ., We also investigate to what extent the variation is correlated between different types of mutation , and to what extent it is correlated to a range of genomic variables ., We use the data to investigate a long-standing question–what forces are responsible for the large-scale variation in GC content across the human genome , the so called “isochore” structure 13 ., It has been suggested that the variation could be due to mutation bias 14–18 , natural selection 13 , 19 , 20 , biased gene conversion 21–24 , or a combination of all three forces 25 ., There is now convincing evidence that biased gene conversion plays a role in the generating at least some of the variation in GC-content 26–28 ., However , this does not preclude a role for mutation bias or selection ., With a dataset of DNMs we are able to directly test whether mutation bias causes variation in GC-content ., The rate of divergence between species is known to vary across the genome at a large scale 1 ., As expected this appears to be in part due to variation in the rate of mutation 3 ., However , the rate of mutation at the MB scale is not as strongly correlated to the rate of nucleotide substitution between species as it could be if all the variation in divergence between 1MB windows was due to variation in the mutation rate 3 ., Instead , the rate of divergence appears to correlate independently to the rate of recombination ., This might be due to one , or a combination , of several factors ., First , recombination might affect the probability that a mutation becomes fixed by the process of biased gene conversion ( BGC ) ( reviewed by 26 ) ., Second , recombination can affect the probability that a mutation will be fixed by natural selection; in regions of high recombination deleterious mutations are less likely to be fixed , whereas advantageous mutations are more likely ., Third , low levels of recombination can increase the effects of genetic hitch-hiking and background selection , both of which can reduce the diversity in the human-chimp ancestor , and the time to coalescence and the divergence between species ., There is evidence of this effect in the divergence of humans and chimpanzees , because the divergence between these two species is lower nearer exons and other functional elements 29 , 30 ., And fourth , the correlation of divergence to both recombination and DNM density might simply be due to limitations in multiple regression; spurious associations can arise if multiple regression is performed on two correlated variables that are subject to sampling error ., For example , it might be that divergence only depends on the mutation rate , but that the mutation rate is partially dependent on the rate of recombination ., In a multiple regression , divergence might come out as being correlated to both DNM density and the recombination rate , because we do not know the mutation rate without error , since we only have limited number of DNMs ., Here , we introduce a test that can resolve between these explanations ., As with divergence , we might expect variation in the level of diversity across a genome to correlate to the mutation rate ., The role of the mutation rate variation in determining the level of genetic diversity across the genome has long been a subject of debate ., It was noted many years ago that diversity varies across the human genome at a large scale and that this variation is correlated to the rate of recombination 31–33 ., Because the rate of substitution between species is also correlated to the rate of recombination , Hellmann et al . 31 , 32 inferred that the correlation between diversity and recombination was at least in part due to a mutagenic effect of recombination , an inference that has been confirmed by recent studies of recombination 3 , 34 , 35 ., However , no investigation has been made as to whether variation in the rate of mutation explains all the variation in diversity , or whether biased gene conversion , direct and linked selection have a major influence on diversity at a large scale ., To investigate large scale patterns of de novo mutation in humans we compiled data from three studies which between them had discovered more than 130 , 000 autosomal DNMs: 105 , 385 from Jonsson et al . 36 , 26 , 939 mutations from Wong et al . 6 , and 11016 mutations from Francioli et al . 3 The datasets are henceforth referred to by the name of the first author ., We divided the mutations up into 9 categories reflecting the fact that CpG dinucleotides have higher mutation rates than non-CpG sites , and the fact that we cannot differentiate which strand the mutation had occurred on: CpG C>T ( a C to T or G to A mutation at a CpG site ) , CpG C>A , CpG C>G and for non-CpG sites C>T , T>C , C>A , T>G , C<>G and T<>A mutations ., The proportion of mutations in each category in each of the datasets is shown in Fig 1 ., We find that the pattern of mutation differs significantly between the studies ( Chi-square test of independence on the number of mutations in each of the 9 categories , p < 0 . 0001 ) ., This appears to be largely due to the relative frequency of C>T transitions in both the CpG and non-CpG context; a discrepancy which has been noted before37 , 38 ., In the data from Wong et al . 6 the frequency of C>T transitions at CpG sites is ~13% whereas it is ~16–17% in the other two datasets ., For non-CpG sites the frequency of C>T transitions is ~24% in all studies except that of Wong et al . in which it is 26% ., It is not clear whether these patterns reflect differences in the mutation rate between different cohorts of individuals , possibly because of age 3 , 4 , 6 or geographical origin 39 or whether the differences are due to methodological problems associated with detecting DNMs ., To investigate whether there is large scale variation in the mutation rate we divided the genome into non-overlapping windows of 10KB , 100KB , 1MB and 10MB and fit a gamma distribution to the number of mutations per region , taking into account the sampling error associated with the low number of mutations per region ., We focussed our analysis at the 1MB scale since this has been extensively studied before ., However , we show that the variation at 1MB forms part of a continuum of variation ., We also repeated almost all our analyses at the 100KB scale with qualitatively similar results ( these results are reported in supplementary tables ) ., We find that the amount of variation differs significantly between the three studies ( likelihood ratio tests: p < 0 . 001 ) , although , the differences are quantitatively small at the 1MB ( Fig 2 ) and 100KB ( S1 Fig ) scales ., The variation between datasets might be due to differences in age or ethnicity between the individuals in each study , or methodological problems–for example , there might be differences between studies in the ability to identify DNMs ., We can test whether callability is an issue in the Wong dataset because Wong et al . 6 estimated the number of trios at which a DNM was callable at each site ., If we reanalyse the Wong data using the sum of the callable trios per MB , rather than the number of sites in the human genome assembly , we obtain very similar estimates of the distribution: the coefficient of variation ( CV ) for the distribution is 0 . 27 when we use the number of sites and 0 . 24 when we use the sum of callable trios ., As expected the number of DNMs per site is significantly correlated between the datasets ( 1MB Francioli v Wong r = 0 . 15 , p<0 . 001; Francioli v Jonsson r = 0 . 19 p<0 . 001; Wong v Jonsson r = 0 . 29 , p<0 . 001 ) ., The correlation is weak , but this is likely to be in part due to sampling error ., If we simulate data assuming a common distribution , estimating the shape parameter as the mean CV of the distributions fit to the individual datasets , the mean simulated correlations are: Francioli v Wong r = 0 . 20; Francioli v Jonsson r = 0 . 29; Wong v Jonsson r = 0 . 41 ., This suggests that a substantial proportion of the variation is common to the three datasets , however in each case less than 5% of the simulated correlations are less than the observed correlation suggesting that some portion of the variation in the three datasets is uncorrelated ., The CV of the gamma distribution fitted to the density of DNMs is 0 . 18 , 0 . 27 and 0 . 15 for the Francioli , Wong and Jonsson datasets respectively ( Fig 2 ) ., The level of variation is significant ( i . e . the lower 95% confidence interval of the CV is greater than zero ) , however the level of variation is modest ( Fig 2 ) ., A gamma distribution with a coefficient of variation of 0 . 18 is one in which 90% of regions have a mutation rate within 30% of the mean ( i . e . if the mean is one , between 0 . 7 and 1 . 3 ) ., The gamma distribution fits the distribution of rates qualitatively quite well ( S2 Fig; S3 Fig for 100KB ) , even though a goodness-of-fit test rejects the model at both the 100KB and 1MB scales in all three datasets ( p<0 . 001 in all cases ) ., At the 1MB the observed distribution is more peaked than the fitted gamma distributed; there are too many regions with very low , very high and intermediate numbers of DNMs ., If we include estimates of the distribution for 10KB , 100KB and 10MB we find , as expected , that the variance in the mutation rate declines as the scale gets larger ( Figs 3 and 4 ) ., This is more marked for the Francioli dataset than for the Wong and Jonsson datasets ( Figs 3 and 4 ) ., If we plot the CV of the fitted gamma distribution against the window size we find that the log of the CV of the gamma distribution is approximately linearly related to the log of the window size for the Francioli and Wong datasets ( Fig 4 ) ; the relationship appears curvi-linear for the Jonsson dataset ., The fact that the CV declines gradually across scales suggests that the variation at the 1MB scale is part of a continuum of variation at different scales ., The linearity of the relationship in two of the datasets suggests that a simple phenomenon may underlie the variation at different scales ., If all the variation at the larger scales is explainable by variation at a smaller scale , then the CV at scale x should be equal to the CV at some finer scale , y , divided by the square-root of x/y; on a log-log scale this should yield a slope of -0 . 5 ., The slope for each dataset is shallower than this ( Francioli b = -0 . 25; Wong b = -0 . 10; Jonsson b = -0 . 16 ) ., This therefore suggests that there is variation at a larger scale that cannot be explained by variation at a smaller scale ., To test whether this is the case , we ran a series of one-way ANOVAs; testing variation at the 100KB scale using 10KB windows , 1MB using 100KB windows and 10MB using 1MB windows ., The results were significant for all datasets ( p<0 . 001 in all cases ) ., If we estimate the distribution for individual mutational types we find that in many cases the lower CI on the CV is zero; this might be because we do not have enough data to reliably estimate the distribution for each individual mutational type ., We therefore combined mutations into a variety of non-mutually exclusive categories ., In each case we estimated the distribution for the relevant category of sites–e . g . in considering the distribution of CpG rates we consider the number of CpG DNMs at CpG sites , not at all sites ., We find that the estimated distributions are similar for different mutational types except that there is rather more variation at CpG sites in the Francioli dataset ( Fig 3; 100KB results S1 Table ) ., Although the distributions are fairly similar for different mutational types , likelihood ratio tests demonstrate that there are significant differences between mutational categories ( S2 Table for 1MB and 100KB results ) ; this is particularly apparent for the Jonsson dataset , probably as a consequence of the size of this dataset ., Never-the-less the differences between different mutational categories are relatively small ., Given that there is variation in the mutation rate at the 1MB scale and that this variation is quite similar in magnitude for different mutational types , it would seem likely that the rate of mutation for the different mutational types are correlated ., We find that this is indeed the case ., We observe significant correlations between all categories of mutations in the three datasets ( Table 1; S3 Table for 100KB ) ., The correlations are weak but this is to be expected given the large level of sampling error ., To compare the correlation to what we might expect if the two categories of mutation shared a common distribution and were perfectly correlated , we simulated data under a common distribution , estimating the CV of the common distribution as the mean of the distributions fitted to the two mutational categories ., We find that generally the observed correlations are similar , and not significantly different , to the expected correlations ., In some cases , we observe that the simulated correlation is actually consistently weaker than the observed correlation; this may reflect the inadequacy of the gamma distribution in describing the distribution of rates ., The fact that the rates of Strong to Weak base pairs ( S>W ) and W>S mutation covary ( Table 1 ) suggests that mutational biases are unlikely to generate much variation in GC-content across the genome ., To investigate this further , we used two approaches to test whether there was variation in the pattern of mutation that could generate variation in GC content ., First , we used the DNM data for each window to predict the equilibrium GC content to which the sequence would evolve , fitting a model by maximum likelihood ( ML ) in which this equilibrium GC-content could vary across the genome ., The ML estimate for the mean equilibrium GC-content is similar in all datasets at ~0 . 32 ., The ML estimate and its 95% CIs for the standard deviation for the equilibrium GC-content are 0 . 02 ( 0 , 0 . 060 ) , 0 . 001 ( 0 , 0 . 036 ) and 0 . 011 ( 0 , 0 . 024 ) for the Francioli , Wong and Jonsson respectively; in each case confidence intervals encompass 0 , suggesting that a model with no variation in equilibrium GC-content fits the data well ., Furthermore , the upper confidence interval is small , suggesting that at most variation in the pattern of mutation generates little variation in GC-content ., However , the ML method does not rule out the possibility that there is some variation in the pattern of mutation ., Furthermore , the method does not take into account the difference in the mutation rate between CpG and non-CpG sites ., We therefore used a second approach in which we grouped windows together based on their current GC-content ., We then estimated the mutation rates for the 9 categories of mutation using the DNM data and used these estimated mutation rates in a simulation of sequence evolution , in which we evolved the sequence to its equilibrium GC content ., We find no correlation between the equilibrium GC content to which the sequence evolves and the current GC content ( Fig 5; S4 Fig for 100KB ) ., It has been suggested that the mutation rate at a site is predictable based on genomic features , such as replication time , by Michaelson et al . 5 , or the 7-mer sequence in which a site is found , by Aggarwala et al . 40 ., To investigate whether these models can explain the variation at large scales we used the models to predict the average mutation rate for each 100KB or 1MB region and correlated these predictions against the observed number of DNMs per site ., We find that the density of DNMs is significantly correlated to the rates predicted under the 7-mer model of Aggarwala et al . 40 ., This correlation is significantly positive for the Wong and Jonsson datasets , as we might expect , but significantly negative for the Francioli dataset ( Table 2; S4 Table for 100KB results ) ., To compare these correlations to what we might expect if the Aggarwala model explained all the variation at large scales , we simulated the appropriate number of DNMs across the genome according to this model ., The observed correlation is significantly smaller than the expected correlation for all datasets , however , the observed and expected correlations are quite similar for the Wong dataset suggesting that much of the variation in DNM density in this dataset is explainable by the model of Aggarwala et al . 40 ., However , the model explains almost none of the variation in the Jonsson dataset ., In contrast , the density of DNMs is significantly positively correlated to the predictions of the Michaelson model in the Francioli and Jonsson datasets , but not for the Wong dataset ., However , in all cases the correlation is substantially and significantly smaller than it could be if the model explained all the variation ( Table 2; S4 Table for 100KB results ) suggesting that this model fails to capture much of the variation at the 1MB and 100KB scales ., To try and understand why there is large scale variation in the mutation rate , we compiled a number of genomic variables which have previously been shown to correlate to the rate of germline or somatic DNM , or divergence between species: male and female recombination rate , GC content , replication time , nucleosome occupancy , transcription level , DNA hypersensitivity and several histone methylation and acetylation marks 3 , 5 , 9 , 41 , 42 ., Surprisingly , the three datasets yield different patterns of correlation ., The overall density of DNMs is significantly positively correlated to male and female recombination rates across all datasets , but otherwise there is no consistency ( Table 3; 100KB results S5 Table ) ; for example , DNM density is negatively correlated to replication time ( later replicating regions have higher mutation rates ) in the Francioli and Jonsson datasets , but positively correlated in the Wong dataset , and despite containing 10-times as much data , the correlation is weaker in the Jonsson than the Francioli dataset ., Overall , the correlations are more similar in their direction in the Francioli and Jonsson datasets ., Many of the genomic variables are correlated to each other ., If we use principle components to reduce the dimensionality , the first principle component ( PC ) explains 58% of the variation in the genomic variables , the second 13% , the third and fourth 6 . 9 and 5 . 7% of the variation ., We find that the density of DNMs is significantly negatively correlated to the first PC in the Francioli data ( r = -0 . 14 , p<0 . 001 ) , significantly positively in the Wong data ( r = 0 . 14 , p<0 . 001 ) and uncorrelated in the Jonsson data ( r = -0 . 013 , p = 0 . 54 ) ., All are significantly positively correlated to the second PC ( Francioli , r = 0 . 14 , p<0 . 001; Wong , r = 0 . 27 , p<0 . 001; Jonsson , r = 0 . 15 , p < 0 . 001 ) , uncorrelated to the third component and Wong and Jonsson are significantly correlated to the fourth component but in opposite directions ( Wong , r = -0 . 059 , p = 0 . 005; Jonsson , r = 0 . 1 , p<0 . 001 ) ., It is possible that the differences between Wong and the other datasets are due to biases in the ability to call DNMs ., However , analysing the Wong data using the number of callable trios at each site does not qualitatively alter the pattern of correlation in the Wong dataset ( Table 3 ) or the correlations to the principle components of the genomic features ( PC1 , r = 0 . 11 p<0 . 001; PC2 , r = 0 . 25 , p<0 . 001; PC3 , r = -0 . 019 , p = 0 . 37; PC4 , r = -0 . 048 , p = 0 . 019 ) ., To investigate whether these patterns are consistent across mutational types , we calculated the correlation between the density of each mutational type ( e . g . CpG C>T mutations at CpG sites ) and the first two PCs of the genomic features ., For the Francioli and Jonsson datasets the patterns are perfectly consistent; all mutational types , if they show a significant correlation , are significantly negatively correlated to the first PC , and significantly positively correlated to the second ( S6 Table ) ., For the Wong data , the patterns are more heterogeneous; all mutational types are positively correlated to the second PC , but some mutational types are significantly positively correlated to the first PC and others significantly negatively correlated ., In order to try and disentangle which factors might be most important in determining the rate of mutation we used stepwise regression ., We find , as expected , that the models selected for the three datasets are different ( Table 4 ) ; only male recombination rate is common to and correlated in the same direction in all three models ., The differences are not due to variation in the ability to call DNMs in the Wong dataset since repeating the analyses using the sum of callable trios rather than sites , does not alter the patterns ( Table 4 ) ., At the 100KB scale , replication time joins male recombination factor as a common factor in all three datasets ( S7 Table ) ., The differences between the three datasets could be due to paternal age since Francioli et al . 3 showed that the correlation between DNM density and replication time was only evident amongst individuals born to young fathers ( <28 years ) , and paternal age differs between the three studies: the average paternal age was 27 . 7 years in the Francioli dataset ( Laurent Francioli pers comm ) , 33 . 4 years in the Wong data 6 and 32 . 0 in the Jonsson data ( calculated from their supplementary data ) ., To investigate whether this could explain the differences between the datasets we divided the DNMs into those discovered in individuals with young ( <28 years ) and old fathers ( ≥28 years ) , and regressed the normalised DNM density ( dividing by the mean DNM density for each dataset in each age cohort ) against replication time and PC1 ., We find no evidence that the relationship between DNM density and replication time ( or PC1 ) is stronger in individuals born to young fathers in the Wong and Jonsson datasets ( Table 5 ) ., The amount of variation explained by the multiple regression models is small– 0 . 044 , 0 . 10 and 0 . 042 for Francioli , Wong and Jonsson respectively—but this might be expected given the small number of DNMs per MB and hence the large sampling error ., To investigate how much of the explainable variance the model explains we sampled rates from the gamma distribution fitted to the distribution of DNMs across the genome and generated DNMs using these rates and then correlated these simulated rates to the true rates ( i . e . those sampled from the gamma distribution ) ., The average coefficient of determination for the simulated data is 0 . 11 , 0 . 39 and 0 . 42 for the Francioli , Wong and Jonsson datasets respectively suggesting that the regression model explains ~37% , ~26% and ~10% of the explainable variance for the three datasets ., In all cases , none of the simulated datasets have a coefficient of determination that is as low as the observed ., The rate of divergence between species is expected to depend , at least in part , on the rate of mutation ., To investigate whether variation in the rate of substitution is correlated to variation in the rate of mutation we calculated the divergence between humans and chimpanzees , initially by simply counting the numbers of differences between the two species ., There are at least three different sets of human-chimpanzee alignments: pairwise alignments between human and chimpanzee ( PW ) 43 found on the University of California Santa Cruz ( UCSC ) Genome Browser , the human-chimp alignment from the multiple alignment of 46 mammals ( MZ ) 44 from the same location , and the human-chimp alignment from the Ensembl Enredo , Pecan and Ortheus primate multiple alignment ( EPO ) 45 ., We find that the correlation depends upon the human-chimpanzee alignments used and the amount of each 1MB window covered by aligned bases ( Fig 6 ) ., The correlation is significantly negative if we include all windows for the UCSC PW and MZ alignments at the 1MB scale , but becomes more positive as we restrict the analysis to windows with more aligned bases ., In contrast , the correlations are always positive when using the EPO alignments , and the strength of this correlation does not change once we get above 200 , 000 aligned bases per 1MB ., Further analysis suggests there are some problems with the PW and MZ alignments because divergence per MB window is negatively correlated to mean alignment length ( r = -0 . 31 , p < 0 . 0001 ) for the PW alignments and positively correlated ( r = 0 . 57 , p < 0 . 0001 ) for the MZ alignments ( S5 Fig ) ., The EPO alignment method shows no such bias and we consider these alignments to be the best of those available ., Therefore , we use the EPO alignments for the rest of this analysis ., To gain a more precise estimation of the number of substitutions we used the method of Duret and Arndt 21 , which is a non-stationary model of nucleotide substitution that allows the rate of transition at CpG dinucleotides to differ to than that at other sites ., As expected the divergence along the human lineage ( since humans split from chimpanzees ) is significantly correlated to the rate of DNMs ( Francioli , r = 0 . 20 p<0 . 001; Wong , r = 0 . 16 , p<0 . 001; Jonsson , r = 0 . 31 , p<0 . 001 ) ., However , the correlation between the rate of DNMs and divergence is not expected to be perfect even if variation in the mutation rate is the only factor affecting the rate of substitution between species; this is because we have relatively few DNMs and hence our estimate of the density of DNMs is subject to a large amount of sampling error ., To investigate how strong the correlation could be , we follow the procedure suggested by Francioli et al . 3; we assume that variation in the mutation rate is the only factor affecting the variation in the substitution rate across the genome between species and that we know the substitution rate without error ( this is an approximation , but the sampling error associated with the substitution rate is small relative to the sampling error associated with DNM density because we have so many substitutions ) ., We generated the observed number DNMs according to the rates of substitution , and then considered the correlation between these simulated DNM densities and the observed substitution rates ., We repeated this procedure 1000 times to generate a distribution of expected correlations ., Performing this simulation , we find that we would expect the correlation between divergence and DNM density to be 0 . 30 , 0 . 44 and 0 . 68 for the Francioli , Wong and Jonsson datasets respectively , considerably greater than the observed values of 0 . 20 , 0 . 16 and 0 . 31 respectively ., In none of the simulations was the simulated correlation as low as the observed correlation ., There are several potential explanations for why the correlation is weaker than it could be; the pattern of mutation might have changed 39 , 46–48 , or there might be other factors that affect divergence ., Francioli et al . 3 showed that including recombination in a regression model between divergence and DNM density significantly improved the fit of the model; a result we confirm here; the coefficient of determination when the sex-average recombination rate is included in a regression of divergence versus DNM density increases from 0 . 039 to 0 . 14 , 0 . 026 to 0 . 12 and 0 . 095 to 0 . 18 for the Francioli , Wong and Jonsson datasets respectively; similar patterns are observed for male and female recombination rates separately ., As detailed in the introduction there are at least four explanations for why recombination might be correlated to the rate of divergence independent of its effect on the rate of DNM:, ( i ) biased gene conversion ,, ( ii ) recombination affecting the efficiency of selection ,, ( iii ) recombination affecting the depth of the genealogy in the human-chimpanzee ancestor and, ( iv ) problems with regressing against correlated variables that are subject to sampling error ., We can potentially differentiate between these four explanations by comparing the slope of the regression between the rate of substitution and the recombination rate ( RR ) , and the rate of DNM and the RR ., If recombination affects the substitution rate , independent of its effects on DNM mutations , because of GC-biased gene conversion ( gBGC ) , then we expect the slope between divergence and RR to be greater than the slope between DNM density and RR for Weak>Strong ( W>S ) , smaller for S>W , and unaffected for S<>S and W<>W changes ., The reason is as follows; gBGC increases the probability that a W>S mutation will get fixed but decreases the probability that a S>W mutation will get fixed ., This means that regions of the genome with high rates of recombination will tend to have higher substitution rates of W>S mutations than regions with low rates of recombination hence increasing the slope of the relationship between divergence and recombination rate ., The opposite is true for S>W mutations , and S<>S and W<>W mutations should be unaffected by gBGC ., If selection is the reason that divergence is correlated to recombination independently of its effects on the mutation rate , then we expect all the slopes associated with substitutions to be less than those associated with DNMs ., The reason is as follows; if a proportion of mutations are slightly deleterious then those will have a greater chance of being fixed in regions of low recombination than high recombination ., If the effect of recombination on the substitution rate is due to variation in the coalescence time in the human-chimp ancestor , then we expect all the slopes associated with substitution to be greater than those associated with DNMs; this is because the average time to coalescence is expected to be shorter in regions of low recombination than in regions of high recombination ., Finally , if the effect is due to problems with multiple regression then we might expect all the slopes to become shallower ., Since the DNM density and divergences are on different scales we divided each by their mean to normalise them and hence make the slopes comparable ., The results of our test are consistent with the gBGC hypothesis; the slope of divergence versus RR is greater than the slope for DNM density versus RR for W>S mutations and less for S>W mutations ( Fig 7 ) ; we present the analyses using sex-averaged RR , but the results are similar for either male or female recombination rates , and for 100KB windows ( S6 and S7 Figs and S8 and S9 Tables ) ., These differences are signi

Introduction, Results, Discussion, Materials and methods

It has long been suspected that the rate of mutation varies across the human genome at a large scale based on the divergence between humans and other species ., However , it is now possible to directly investigate this question using the large number of de novo mutations ( DNMs ) that have been discovered in humans through the sequencing of trios ., We investigate a number of questions pertaining to the distribution of mutations using more than 130 , 000 DNMs from three large datasets ., We demonstrate that the amount and pattern of variation differs between datasets at the 1MB and 100KB scales probably as a consequence of differences in sequencing technology and processing ., In particular , datasets show different patterns of correlation to genomic variables such as replication time ., Never-the-less there are many commonalities between datasets , which likely represent true patterns ., We show that there is variation in the mutation rate at the 100KB , 1MB and 10MB scale that cannot be explained by variation at smaller scales , however the level of this variation is modest at large scales–at the 1MB scale we infer that ~90% of regions have a mutation rate within 50% of the mean ., Different types of mutation show similar levels of variation and appear to vary in concert which suggests the pattern of mutation is relatively constant across the genome ., We demonstrate that variation in the mutation rate does not generate large-scale variation in GC-content , and hence that mutation bias does not maintain the isochore structure of the human genome ., We find that genomic features explain less than 40% of the explainable variance in the rate of DNM ., As expected the rate of divergence between species is correlated to the rate of DNM ., However , the correlations are weaker than expected if all the variation in divergence was due to variation in the mutation rate ., We provide evidence that this is due the effect of biased gene conversion on the probability that a mutation will become fixed ., In contrast to divergence , we find that most of the variation in diversity can be explained by variation in the mutation rate ., Finally , we show that the correlation between divergence and DNM density declines as increasingly divergent species are considered .

Using a dataset of more than 130 , 000 de novo mutations we show that there is large-scale variation in the mutation rate at the 100KB and 1MB scales ., We show that different types of mutation vary in concert and in a manner that is not expected to generate variation in base composition; hence mutation bias is not responsible for the large-scale variation in base composition that is observed across human chromosomes ., As expected , large-scale variation in the rate of divergence between species and the variation within species across the genome , are correlated to the rate of mutation , but the correlation between divergence and the mutation rate is not as strong as it could be ., We show that biased gene conversion is responsible for weakening the correlation ., In contrast , we find that most of the variation across the genome in diversity can be explained by variation in the mutation rate ., Finally , we show that the correlation between the rate of mutation in humans and the divergence between humans and other species , weakens as the species become more divergent .

split-decomposition method, vertebrates, human genomics, animals, mammals, primates, multiple alignment calculation, mutation, substitution mutation, genome analysis, dna recombination, mammalian genomics, dna, gene conversion, epigenetics, chromatin, dna methylation, old world monkeys, research and analysis methods, monkeys, chromosome biology, gene expression, chromatin modification, dna modification, animal genomics, macaque, biochemistry, eukaryota, cell biology, computational techniques, nucleic acids, genetics, biology and life sciences, genomics, amniotes, computational biology, organisms

journal.pcbi.1000381

Functional Brain Networks Develop from a “Local to Distributed” Organization

In previous work regarding task-level control in adults , we applied rs-fcMRI to a set of regions derived from an fMRI meta-analysis that included studies of control-demanding tasks ., This analysis revealed that brain regions exhibiting different combinations of control signals across many tasks are grouped into distinct “fronto-parietal” and “cingulo-opercular” functional networks 21 , 36 ( see Table 1 and Figure 1 ) ., Based on functional activation profiles of these regions characterized in the previous fMRI study , the fronto-parietal network appears to act on a shorter timescale , initiating and adjusting top-down control ., In contrast , the cingulo-opercular network operates on a longer timescale providing “set-initiation” and stable “set-maintenance” for the duration of task blocks 37 ., Along with these two task control networks 21 , 36 , a set of cerebellar regions showing error-related activity across tasks 36 formed a separate cerebellar network ( Figure 1 ) ., In adults , the cerebellar network is functionally connected with both the fronto-parietal and cingulo-opercular networks 21 , 22 ., These functional connections may represent the pathways involved in task level control that provide feedback information to both control networks 22 , 36 ., Another functional network , and one of the most prominent sets of regions to be examined with rs-fcMRI , is the “default mode network” ., The default mode network ( frequently described as being composed of the bilateral posterior cingulate/precuneus , inferior parietal cortex , and ventromedial prefrontal cortex ) was first characterized by a consistent decrease in activity during goal-directed tasks compared to baseline 38 , 39 ., Resting-state fcMRI analyses have repeatedly shown that these regions , along with associated medial temporal regions , are correlated at rest in adults 15 , 16 , 32 , 40 ., While the distinct function of the default mode network is often linked to internally directed mental activity 39 , this notion continues to be debated 25 , 32 , 41–44 ., In two prior developmental studies , we used rs-fcMRI to examine the development of the task control and cerebellar functional networks 22 and , separately , the default mode network 32 ., The first study , addressing functional connectivity changes within and between the two task control networks and the cerebellar network 22 , showed that the structure of these networks differed between children and adults in several ways ( see 22 ) ., In general , many of the specific changes showed trends of decreases in short-range functional connections ( i . e . , correlations between regions close in space ) and increases in long-range functional connections ( i . e . , correlations between regions more distant in space ) ., We suggested that these global developmental processes support the maturation of a dual-control system and its functional connections with the cerebellar network 22 ., These results have now been replicated in a developmental resting connectivity study targeting sub-regions of the anterior cingulate 34 ., The development of the default mode network was independently examined in a separate analysis 32 ., In children , the default mode network was only sparsely functionally connected ., Many regions were relatively isolated with few or no functional connections to other default mode regions ., Over age , correlations within the default mode network increased and by adulthood it had matured into a fully integrated system ., Interestingly , as opposed to the task-control and cerebellar networks , very few short-range functional connections involving the default mode network regions existed in children ., Hence the numerous strong short-range functional connections that decreased with age when investigating the dual control networks were not seen within the default network ., In fact , some connections such as the functional connection between the ventromedial prefrontal cortex ( vmPFC; −3 , 39 , −2 ) and anterior medial prefrontal cortex ( amPFC; 1 , 54 , 21 ) regions , which are fairly close in space ( i . e . , short-range at ∼2 . 7 cm ) , had a substantial increase in correlation strength over development 32 ., The observation that different analyses suggested different developmental features suggests a need for a more nuanced and integrated characterization of the development of functional networks ., The goal of this manuscript is to employ several different network analysis tools to provide such a characterization ., Visualization techniques such as spring embedding , and quantitative measures , including ‘small world’ metrics and community detection algorithms , will be applied to these networks in an attempt to identify principles for the changes observed across development ., Because of the overlapping and sometimes inconsistent use of terminology between neuroscience and the computational sciences , we will briefly define two terms for the purposes of this paper ., The term “networks” will be used in the typical cognitive neuroscience formulation: a group of functionally related brain regions ( as described above ) ., The overall collection of regions ( encompassing all four “networks” ) will be referred to as the “graph . ”, Graph theory analyses were applied to 210 subjects , aged 7–31 , to investigate the emergence of temporal correlations in spontaneous BOLD activity between regions of the default mode , cerebellar , and two task-control networks ., For this initial analysis , average age-group matrices were created using a sliding boxcar grouping of subjects in age-order ( i . e . , group1: subjects 1–60 , group2: subjects 2–61 , group3: subjects 3–62 , etc . ) ., This generated a series of groups with average ages ranging from 8 . 48 years to 25 . 58 years ., Each of the groups average correlation matrices was converted into a graph , with correlations between regions greater than or equal to 0 . 1 considered as functionally connected ., In a first analysis , we used a visualization algorithm commonly used in graph theoretic analyses known as spring embedding that aids in the qualitative interpretation of graphs ( Figure 2 and Video S1 ) 45 ., In spring embedding , the positions of the nodes ( i . e . , regions ) in a graph are based solely on the strength and pattern of functional connections instead of their anatomical locations ., In this procedure , each functional connection between a pair of nodes is treated as a spring with a spring constant related to the strength of the specific correlation ., The entire system of pair-wise regional functional connections is then iteratively allowed to relax to the lowest global energetic state , i . e . , groups of nodes that are strongly interconnected will be placed close together even if anatomically distant ., By creating spring embedded graphs for each of the sliding boxcar groups in age-order , a movie representation can be made that shows the development of the network relationships ( from average age 8 . 48 to 25 . 48 years ) ( Video S1 ) ., The panels in Figure 2 provide snapshots from child , adolescent , and adult average ages in this movie ., In both Figure 2 and Video S1 , each node is color-coded in two ways: the outer border represents the general anatomical location ( i . e . , cerebral lobe ) of the node; the inner core color represents the coding by “function” as defined by a large number of fMRI studies ., One of the primary observations from the movie relates to this anatomical-functional distinction ., In children , regions appear to be largely arranged by anatomical proximity ., This arrangement can be seen in Figure 2 and Video S1 where , in children , regions can be readily grouped by cerebral lobe ( outline colors of spheres in Figure 2 and Video S1 ) ., Over age , as functional connections mature , the node arrangements change such that anatomically close regions are now largely distributed across the graph layout , in a pattern more aligned with the mature networks functional properties ( core colors of spheres in Figure, 2 ) 21 , 36–39 ., Thus , across development , local clusters of regions “segregate” from one another and “integrate” into more distributed adult functional relationships with more distant regions ., A group of regions in the frontal cortex provides a particularly salient example of segregation ., Frontal cortex contains regions that , in adults , are members of each of the task-control networks ( e . g . , dlPFC , frontal , dACC/msFC ) and the default network ( e . g . , vmPFC , amPFC ) ., As can be seen in Figure 2A ( and Video S1 ) , extensive correlations exist between most of these frontal regions in childhood ( see blue cloud Figure 2A ) ., Over the developmental window afforded by the current dataset , some of these strong “frontal-frontal” correlations begin to weaken ., With increasing age , regions in the frontal cluster segregate into 3 separate functional networks ., Accompanying this segregation is strong integration within the functional networks ., The default mode network provides the clearest example ., As illustrated in Figure 2B ( and in Video S1 ) , correlations between regions of the default mode network are weak ( or absent ) in children ( red cloud , Figure 2B ) ., Just as functional connections between the set of frontal regions are related to their anatomical proximity in children , the regions of the default mode network are each functionally connected to anatomical neighbors , and not to other members of the anatomically dispersed default mode network ., Over age , however , the functional connections between default mode network regions mature and the network integrates into a highly correlated system in adults ( Figure 2B and Video S1 ) ( also see 32 ) ., We note that these results were not specific to the 60-subject boxcar , and persist with smaller subject boxcars as well ( see Video S2 ) ., The qualitative observations noted above can be quantified using community structure detection tools ., Using such an approach is particularly important because of the bias inherent in relying on qualitative methods for deciding whether groups of regions that appear to be clustered are indeed clustered , and because of the a priori definitions of each network ., As stated by Newman: Among the many methods used to detect communities in graphs , the modularity optimization algorithm of Newman is one of the most efficient and accurate to date 46 ., This method uses modularity , a quantitative measure of the observed versus expected intra-community connections , as a means to guide assignments of nodes into communities ., We applied the modularity optimization algorithm to the group connectivity matrices derived from the sliding boxcars described above ., Measures of modularity ( Q ) were high , and did not show large changes across the age range ( Figure 3A and Figure S1 and Figure S2 ) ., This result was not dependent on any particular threshold ( Figure S1 ) ., Although comparable community structure was detected at all ages examined , the components of the communities varied by age ., As per our qualitative approach described above , in children , region clusters were largely arranged by cerebral lobe; while in adults , regions were largely clustered by their adult functional properties ( Figure 4A ) ., Again , this result was not unique to any particular threshold ( Figure 4B and 4C ) or size of boxcar ( Figure S3 ) ., We do note , however , that limited data points ( i . e . , subjects ) are available between the ages of 16 and 19 years ( see Materials and Methods ) and that our estimate of the specific transitions within this period should be interpreted with care ., As previously reported 22 , 34 , the segregation of closely apposed regions and the integration of distributed functional networks is associated with a general decrease in correlation strength between regions close in space and an increase in correlation strength between many regions distant in space ., This trend is shown in Figure 5 and also Figure S4 ., Long-range functional connections tend to be weak , but increase over time ( warm colors above the diagonal in Figure 5C and 5D and Figure S4C and S4D ) , integrating distant regions into functional networks ., Short-range functional connections tend to be stronger ( i . e . , higher correlation strength ) in children , yet those regions that do change predominantly become weaker over age ( cool colors below the diagonal in Figure 5A and 5B and Figure S4A and S4B ) ., However , there are some interesting nuances to this trend that deserve mention ., For instance , not all short-range functional connections decrease in strength over age ( Figure 5A and 5B and Figure S4A and S4B ) ., While few , some of the short-range functional connections , typically those in the same network , increase in strength over age ( Figure 5A and Figure S4A ) ., Similarly , although many long-range functional connections increase in strength , many others do not statistically change across development ( Figure 5C and , 5D and Figure S4C and S4D , grey connections ) ., In a seminal 1998 paper , Watts and Strogatz noted that the topology of many complex systems can be described as “small world” , a type of graph architecture that efficiently permits both local and distributed processing ., Graphs with a regular , lattice-like structure have abundant short-range connections , but no long-range connections ., Local interactions are thus efficient , but distributed processes involving distant nodes require the traversal of many intermediate connections ., Conversely , completely randomly connected graphs are fairly efficient at transferring distant or long-range signals across a network , but they are poor at local , short-range information transfer ., Watts and Strogatz , and others , often describe “small world” properties with two metrics: the average clustering coefficient and average path length of a graph ., The clustering coefficient measures how well connected the neighbors of a node are to one another ., The average path length measures the average minimum number of steps needed to go between any two nodes ., Lattices , optimized for local processes , have high average clustering coefficients but long average path lengths ., Conversely , random graphs , which have no preference for short-range connections , have low average clustering coefficients and short average path lengths , making them well suited for communication between distant nodes ., One of Watts & Strogatzs key insights was that by randomly rewiring a relatively small number of connections in a lattice graph ( i . e . , introducing a few long-range connections ) , a graph could retain its high average clustering coefficient , but dramatically reduce its average path length , thereby enabling efficient short- and long-range processes ., It is this hybrid graph topology ( i . e . , high clustering coefficients and short path lengths ) that matches the observed “small world” networks in many complex systems 47 ., As previously reported 21 , 48 , 49 , relative to comparable lattice and completely random graphs , the adult graph architecture showed high clustering coefficients and short path lengths , consistent with the ‘small world’ architecture ( Figure 3B and 3C ) ., Interestingly for these networks , in children ( i . e . , as early as age 8 ) , these metrics were quite similar to adults ( Figure 3B and 3C ) , and over age there was very little change in path lengths and clustering coefficients relative to comparable random and lattice graphs ., It was originally anticipated that path lengths would decrease over age as long-range anatomical connections were added ., Yet even at the youngest ages examined , path length was already quite short , near those of random graphs ., Importantly , these results were not dependent on any particular threshold ( Figure S5 ) ., We note that while the results shown here are largely descriptive , the error bars provided in Figure 3B and 3C constructed from random graphs underscores the difference between random configurations and the observed trends ., As early as 1875 spontaneous synchronized neural activity has been used to study various aspects of adult brain organization 50–53 ., However , despite the passing of over 130 years since its initial use , there remains uncertainty as to the role of intrinsic spontaneous brain activity in brain function ., In adults , spontaneous correlated activity has been suggested to be important for gating information flow 54 , building internal representations 43 , 44 , 54 , and maintaining mature network relationships 43 , 44 , 54 ., Much less work has been done in regards to development , but there are suggestions that spontaneous activity is important for the establishment of early cortical patterns ( e . g . , ocular dominance columns ) 55–58 and may over time represent ( in a Hebbian sense ) a history of repeated co-activation between regions 21 , 22 , 27 , 32 , 34 , 59 , 60 ., Within this framework , the changes in the correlation structure of spontaneous activity over development seen in this report may provide insight regarding the arrangement by which brain regions are communicating in children compared to adults ., If we consider the previously mentioned postulates , our results suggest that , typically , the most efficient way for children to respond to processing demands is to utilize more “local” level interactions as compared to adulthood ., That is , in childhood there is , relatively greater co-activation of anatomically proximal regions than for adults with similar processing demands ., A clear example of this is seen in Brown et al . 3 , where identical task performance on lexical processing tests strongly activates a large set of visual regions in children , but strong visual activation is much more restricted in adults ., These relationships may be reflected in correlated spontaneous activity measured via rs-fcMRI ., The correlations in our youngest children would then represent the anatomical and spontaneous activity-defined initial regional relationships plus 7 years of experience-dependent Hebbian processes tuning these developing connections ., The “local to distributed” organizing principle resonates with recent suggestions that perceptual and cognitive development involve the simultaneous segregation and integration of information processing streams 1 , 22 , 76 , 79 , 80 ., For instance , the “interactive specialization” hypothesis advanced by Johnson and colleagues , is consistent with these findings 1 , 81–83 ., Johnson points out that cortical regions and pathways have biased information processing properties at birth due to anatomic connectivity , yet they are much less selective than in adults ( i . e . , they are “broadly tuned” ) ., Interactive specialization predicts that shortly after birth , large sets of regions and pathways will be partially active during specific task conditions , However , as these pathways interact and compete with each other throughout development , selected regions will come online , be maintained , or become selectively activated or “tuned” as particular pathways dominate for specific task demands ., Thus , regional specialization relies on the evolving and continuous interactions with other brain regions over development ., If one extends this framework to the network level , the increases , decreases , and maintenance of correlation strengths seen between regions may reflect “specialization” of specific neural pathways to form the functional networks seen in adults ., The “local to distributed” developmental trajectory , discussed above , seems to be driven by an abundance of local , short range connections that generally decrease in strength over age as well as distant , long range connections that generally increase in strength over age ., Given the more prevalent short-range connections in children , we expected a more lattice-like structure , with high clustering coefficients and relatively high path lengths ., The results , however , clearly indicated that path lengths were near those of equivalent random graphs , and that the child functional networks are already organized as small world networks ., This result can be explained in the context of the re-wiring procedure discussed by Watts and Strogatz 47 ., Randomly rewiring a small percentage of local connections in a lattice has a mild linear effect on clustering coefficients , but a highly non-linear effect on path lengths ., This is to say , that by rewiring a small fraction of a lattices connections , substantial drops in path lengths can be seen , with almost no change in the clustering coefficient ., In late childhood , as shown in Figure 5 and Figure S2 , there are already a significant number of long-range short cuts present ., These long-range functional connections are likely responsible for the relatively short path lengths in the child group ., We anticipate that if the developmental trajectory of short and long-range functional connections were extended to younger ages , fewer long-range ‘short-cut’ functional connections would be present , and more short-range functional connections would exist ., Hence , the path lengths at these younger ages ( <7 years old ) would likely be longer ., Nevertheless , by 8 years old , the networks already display ‘small world’ properties similar to those of adult networks , indicating that efficient graph structures are already in place for both local and distant processing , though they are organized differently than in later development ., While we identified small world properties in both child and adult graphs , the size of the graph is relatively small with only 34 nodes ., Therefore , it is possible that with an increased number of nodes the specific results identified here will change , a possibility that will be addressed in further studies ., The regions used in the present analyses were all derived from adult imaging studies ., It seems likely that additional regions may be included in one or more of these networks in childhood ., In addition , individual differences with regards to the regions and networks chosen likely exist ., Future work that includes regions derived from studies using a child population and obtaining the functional connections within subjects from individually defined functional areas may refine the networks and developmental timecourses presented here 84 ., Of note , resting-state functional connectivity has been reported to be constrained by anatomical distance ( i . e . , correlations between regions decrease as a function of distance following an inverse square law ) 85 ., Thus , if a shift in this general bias occurred with development , then it is feasible that some of the changes seen here could be related to such a shift ., With this said , the specificity of the connection changes observed over age , the number of connections that run opposite to the general trends , and the similarity of the distance relationship in connectivity between children and adults when plotting all possible connections ( see Figure S6 ) , all suggest that the majority of changes observed here are not related to changes in this bias ., In addition , while there are now reports suggesting that changes observed over development with blood oxygen level dependent ( BOLD ) fMRI are not the product of changes in hemodynamic response mechanisms over age 86 , 87 , differences in the hemodynamic response function between children and adults could conceivably affect our results 88 ., A limitation of rs-fcMRI in general is the restricted frequency distribution that can be examined ., rs-fcMRI is used to measure correlations in a very low frequency range , typically below 0 . 1 Hz ., Dynamic changes in correlations in other frequency distributions could exist ( for example see 89 ) ., It is also possible that there are undetected developmental changes in power across frequency bands orthogonal to the changes visualized here ., The combination of other imaging and psychometric techniques with rs-fcMRI will likely help address these considerations ., Characterizing additional networks and how these changes map onto behavior will also help further characterize functional brain development ., Specifically , future work that demonstrates a direct relationship between behavior and the developmental trajectory seen here with rs-fcMRI , is presently needed to confirm ( or reject ) many of the theories presented here and elsewhere ., Importantly , consideration of these issues need not be limited to developmental studies , but should be considered whenever investigators compare groups with rs-fcMRI ., Nonetheless , the general results presented here represent a strong set of hypotheses to be tested in broader domains and larger-scale brain graphs ., First , that by age 8 years , regional relationships , as defined by rs-fcMRI , are organized as small-world-like networks , which , relative to adults , emphasize local connections ., Second , that for the same regions , adult networks show similar network metrics but with regional relationships that have a longer-range , more distributed structure reflecting adult functional histories ., In other words , the modular structure of large-scale brain networks will change with age , but even school age children will show relatively efficient processing architecture ., Subjects were recruited from Washington University and the local community ., Participants were screened with a questionnaire to ensure that they had no history of neurological/psychiatric diagnoses or drug abuse ., Informed consent was obtained from all subjects in accordance with the guidelines and approval of the Washington University Human Studies Committee ., fMRI data were acquired on a Siemens 1 . 5 Tesla MAGNETOM Vision system ( Erlangen , Germany ) ., Structural images were obtained using a sagittal magnetization-prepared rapid gradient echo ( MP-RAGE ) three-dimensional T1-weighted sequence ( TE\u200a=\u200a4 ms , TR\u200a=\u200a9 . 7 ms , TI\u200a=\u200a300 ms , flip angle\u200a=\u200a12 deg , 128 slices with 1 . 25×1×1 mm voxels ) ., Functional images were obtained using an asymmetric spin echo echo-planar sequence sensitive to blood oxygen level-dependent ( BOLD ) contrast ( volume TR\u200a=\u200a2 . 5 sec , T2* evolution time\u200a=\u200a50 ms , α\u200a=\u200a90° , in-plane resolution 3 . 75×3 . 75 mm ) ., Whole brain coverage was obtained with 16 contiguous interleaved 8 mm axial slices acquired parallel to the plane transecting the anterior and posterior commissure ( AC-PC plane ) ., Steady state magnetization was assumed after 4 frames ( ∼10, s ) ., Functional images were first processed to reduce artifacts 23 , 90 ., These steps included:, ( i ) removal of a central spike caused by MR signal offset ,, ( ii ) correction of odd vs . even slice intensity differences attributable to interleaved acquisition without gaps ,, ( iii ) correction for head movement within and across runs and, ( iv ) within run intensity normalization to a whole brain mode value of 1000 ., Atlas transformation of the functional data was computed for each individual via the MP-RAGE scan ., Each run then was resampled in atlas space ( Talairach and Tournoux , 1988 ) on an isotropic 3 mm grid combining movement correction and atlas transformation in one interpolation 91 , 92 ., All subsequent operations were performed on the atlas-transformed volumetric timeseries ., For rs-fcMRI analyses as previously described 16 , 23 , several additional preprocessing steps were used to reduce spurious variance unlikely to reflect neuronal activity ( e . g . , heart rate and respiration ) ., These steps included: ( 1 ) a temporal band-pass filter ( 0 . 009 Hz<f<0 . 08 Hz ) and spatial smoothing ( 6 mm full width at half maximum ) , ( 2 ) regression of six parameters obtained by rigid body head motion correction , ( 3 ) regression of the whole brain signal averaged over the whole brain , ( 4 ) regression of ventricular signal averaged from ventricular regions of interest ( ROIs ) , and ( 5 ) regression of white matter signal averaged from white matter ROIs ., Regression of first order derivative terms for the whole brain , ventricular , and white matter signals were also included in the correlation preprocessing ., These pre-processing steps likely decrease or remove developmental changes in correlations driven by changes in respiration and heart rate over age ., Resting state ( fixation ) data from 210 subjects ( 66 aged 7–9; 53 aged 10–15; 91 aged 19–31 ) were included in the analyses ., For each subject at least 555 seconds ( 9 . 25 minutes ) of resting state BOLD data were collected ., 34 previously published regions comprising 4 functional networks ( i . e . , cingulo-opercular , fronto-parietal , cerebellar , and default networks; see Table 1 and Figure, 1 ) were used in this analysis 16 , 21 , 22 , 37 ., For each region , a resting state timeseries was extracted separately for each individual ., For 10 adult subjects , resting data was continuous ., For the remaining 200 subjects , resting periods were extracted from between task periods in blocked or mixed blocked/event-related design studies 22 ., These concatenated-extracted rest periods were shown to be equivalent to continuous resting data in a recent study describing this method 23 ., In addition , several previous findings using this technique 21 , 22 , 32 have now been replicated using continuous resting blocks 27 , 33 , 34 and other continuous resting data 89 ., To examine the functional connections within and between the large set of regions used in this manuscript we chose to use graph theory ., Graph theory is particularly well suited to study large-scale systems organization across development , but requires the data be organized into specific correlation matrices ., To do this , for each of the 210 subjects , the resting state BOLD timeseries from each region was correlated with the timeseries from every other region , creating 210 square correlation matrices ( 34×34 ) ., Average group matrices were then created using a sliding boxcar grouping of subjects in age-order ( i . e . , group1: subjects 1–60 , group2: subjects 2–61 , group3: subjects 3–62 , … group151: subjects 151–210 ) , thus generating a series of groups with average ages ranging from 8 . 48 years old to 25 . 48 years old with each group composed of 60 subjects ., Average correlation coefficients ( r ) for each group were generated from the subjects individual matrices using the Schmidt-Hunter method for meta-analyses of r-values 21 , 85 , 93 ., In cases when the terms “child” or “adult” are used , the matrices or results referred to are the first and last of the sliding boxcar groups respectively , i . e . , the child group is the youngest 60 subjects , with an average age of 8 . 48 years old , and the adult group is the oldest 60 subjects , with an average age of 25 . 48 years old ., To generate a dynamic representation of the functional connections between regions across development , each of the groups correlation matrices was converted into a thresholded graph , such that correlations higher than r≥0 . 1 were considered connections , while correlations lower than the threshold were not connections ., For our initial analyses 21 , 22 , 32 graphs in child and adult groups were presented in either a pseudo-anatomical fashion or in their actual 3D positions ( in Talairach space ) ., Here we add another representation often used in graph theory - spring embedding ., In this procedure , a spring constant is added to all of the connections in the network allowing for the pairwise regional connections to relax to their lowest energetic state ., The algorithm applied in the present analysis is known as Kamada-Kawai 45 - one of the most commonly used strategies for displaying graph network data ., In brief , each functional connection between a pair of nodes is treated as a spring with a spring constant relat

Introduction, Results, Discussion, Materials and Methods

The mature human brain is organized into a collection of specialized functional networks that flexibly interact to support various cognitive functions ., Studies of development often attempt to identify the organizing principles that guide the maturation of these functional networks ., In this report , we combine resting state functional connectivity MRI ( rs-fcMRI ) , graph analysis , community detection , and spring-embedding visualization techniques to analyze four separate networks defined in earlier studies ., As we have previously reported , we find , across development , a trend toward ‘segregation’ ( a general decrease in correlation strength ) between regions close in anatomical space and ‘integration’ ( an increased correlation strength ) between selected regions distant in space ., The generalization of these earlier trends across multiple networks suggests that this is a general developmental principle for changes in functional connectivity that would extend to large-scale graph theoretic analyses of large-scale brain networks ., Communities in children are predominantly arranged by anatomical proximity , while communities in adults predominantly reflect functional relationships , as defined from adult fMRI studies ., In sum , over development , the organization of multiple functional networks shifts from a local anatomical emphasis in children to a more “distributed” architecture in young adults ., We argue that this “local to distributed” developmental characterization has important implications for understanding the development of neural systems underlying cognition ., Further , graph metrics ( e . g . , clustering coefficients and average path lengths ) are similar in child and adult graphs , with both showing “small-world”-like properties , while community detection by modularity optimization reveals stable communities within the graphs that are clearly different between young children and young adults ., These observations suggest that early school age children and adults both have relatively efficient systems that may solve similar information processing problems in divergent ways .

The first two decades of life represent a period of extraordinary developmental change in sensory , motor , and cognitive abilities ., One of the ultimate goals of developmental cognitive neuroscience is to link the complex behavioral milestones that occur throughout this time period with the equally intricate functional and structural changes of the underlying neural substrate ., Achieving this goal would not only give us a deeper understanding of normal development but also a richer insight into the nature of developmental disorders ., In this report , we use computational analyses , in combination with a recently developed MRI technique that measures spontaneous brain activity , to help us to understand the principles that guide the maturation of the human brain ., We find that brain regions in children communicate with other regions more locally but that over age communication becomes more distributed ., Interestingly , the efficiency of communication in children ( measured as a ‘small world’ network ) is comparable to that of the adult ., We argue that these findings have important implications for understanding both the maturation and the function of neural systems in typical and atypical development .

neuroscience/neurodevelopment, neuroscience/cognitive neuroscience

journal.pcbi.1002505

Insights into the Fold Organization of TIM Barrel from Interaction Energy Based Structure Networks

Proteins are amino–acid polymers capable of folding into unique three–dimensional functional states ., The information for the structure formation is contained within their amino–acid sequence 1 ., With an enormous amount of data available on genomic sequences in organisms and the structures of the proteins they encode , it has become evident that despite the large sequence space , the structure space is rather limited 2–4 ., It has been predicted that merely a few thousand protein folds are needed to generate the entire repertoire of the multimillion strong protein universe 5 , 6 ., The limited number of folds has been explained as a result of optimization of backbone packing 7 , 8 ., A recent analysis of the fold space showed that the atomic interaction network in the solvent–unexposed core of protein domains are fold–conserved , and that the network is significantly distinguishable across different folds , providing a “signature” of a native fold 9 ., As a common rule , homologous sequences generally take up similar folds and the sequence divergences are concomitantly accompanied by structural variations 10 ., However , increasing number of identified sequences and folds show a significant departure from this rule , i . e the same fold is able to house highly dissimilar protein sequences 11–14 ., Folds like the TIM ( Triosephosphate Isomerase ) barrel , Rossmann , αβ–plait , and all β–immunoglobins are taken up by divergent sequences thereby underscoring the availability of limited fold space ., These folds with their simple and symmetric architectures seem to be favorable folds for a large number of non–homologous sequences ., Such folds are of special interest since their investigation would provide profound insights into the principles governing protein folding and stability ., Although functional variations are related to structural variations , it has been established that proteins with disparate structures may retain their function during the course of their evolution as long as the local active site geometry is maintained 10 , 15 ., Triosephosphate Isomerase ( TIM ) Barrel is one the ancient folds with considerable sequence diversity 2 ., It is also one of the ubiquitously occurring enzymatic folds and hosts the most diverse enzymatic reactions catalyzing five of the six classes of biochemical reactions 16 , 17 ., Thus TIM barrel , possessing both structural and functional diversity , has appealed both structural biologists and biochemists equally over the years ., Factors responsible for its structural maintenance and functional diversity have been investigated in detail since its first structural discovery in 1975 16 , 18–24 ., The fold consists of an alternating helix–loop–strand secondary structure motif , where the strands assemble into the core β–barrel ., This β–barrel is therefore formed by parallel strands , which is a rarity in fold space 24 ., The outer rim of the barrel is maintained by helix–sheet and helix–helix interactions ., Evolutionary studies suggest that there are evidences for both divergent 23 and convergent 20 evolution of the TIM barrel proteins , and hence , its evolution is being highly debated ., A large number of computational studies have been carried on this fold , focusing mainly on their prevalence in the enzymes of various organisms catalyzing different functions , their structural and evolutionary properties 16 , 21–26 ., In this study we have explored the factors responsible for the stability of TIM fold taken up by dissimilar sequences ., Unlike earlier studies that focus on residue conservation , we have focused on interaction conservation as the basis of understanding the underlying structural determinants of the TIM fold ., Although this is a novel method , several concepts related to protein sequence-structure-function relationship have been explored and quantitative results have been presented in the literature ., For instance , evolutionary concepts were implemented in identifying pair-wise 27 and sets of residues , called as a “sectors” , that have undergone correlated mutations 28 in the protein sequences ., At the structure-dynamics level , coarse-grained network models have shown that proteins with similar architecture exhibit similar large-scale dynamic behavior 29 and the differences usually occur in regions where specific functions are localized ., Energetic coupling between residues has been investigated both experimentally by mutation followed by biochemical measurements 30 and from computational methods 31 ., The classical problem of studying the structure-function relationship in allostery has been addressed from protein structure network point of view 32–36 ., In essence the protein sequence-structure relationship and the structural changes accommodating their biological function have been investigated by a variety of methods ., Here , we have made the preliminary attempt to study the role of conserved interactions in stabilizing a fold by, ( a ) analyzing residue–residue interactions obtained from atomistic force fields;, ( b ) investigating the interactions and their threshold energy values at a global level by constructing Protein Energy Networks ( PEN ) ;, ( c ) obtaining a common PEN for a family of proteins ( f–PEN ) by structure based alignment followed by the construction of a common energy–weighted interaction matrix;, ( d ) using the f–PENs to study the conserved interactions responsible maintaining the fold and, ( e ) exploiting the conservation of interactions ( obtained from f–PENs ) to deduce phylogenetic relationship ( trees ) as opposed to the commonly practiced sequence based methods ., PENs are structure networks where the constituent amino–acids are the nodes and the edges represent the non–covalent interactions among them ., By representing the interactions as interaction energies ( obtained from molecular mechanics force fields ) , both the chemistry and the geometry of the amino–acids are better represented than other contact–based structure networks 37 , 38 ., We have used structural similarities between the remote homologues of TIM barrel fold to align their PENs to obtain information on the extent of interaction conservation among them ., The analysis of f–PENs has provided us a wealth of information in terms of the strength of interactions and their conservation ( at pair–wise as well as at the level of a collection of multiple interactions ) ., We have been able to identify the factors responsible for the stability of the different secondary structural interfaces in the TIM fold ., In general we have observed that the residues involved in high–energy interactions to have more conservation than the residues forming low–energy vdW dominated interactions ., We have seen that high–energy conserved interactions are present in the central β–barrel stabilizing it and in the catalytic loop regions helping in the functioning of the protein ., The interface between helices and sheets are dominated exclusively by low–energy interactions between non–conserved residues , thus contributing much to the sequence diversity ., We also observed that interaction conservation based phylogeny represents the structural and functional evolution better than those derived from sequence conservation ., The new outlook from “interaction conservation” has shed more light on the factors behind the fold organization of TIM fold by sequentially diverse homologues ., Such observations are unique and we believe that this method will pave an alternate way for understanding the basis of organization of other folds as well ., Furthermore , the information on interaction conservation can enable more controlled engineering of new proteins with enhanced structural/functional properties ., The TIM fold comprises three major secondary structural interfaces: the central β–barrel , α/β and α/α ( Figure 1a ) ., The central β–barrel is formed by staggered parallel β sheets forming the β/β interface and makes up the core of the fold ( Figure 1b ) ., The α/β interface flanks the barrel and is formed by the most common α–X–β motif ( where X can be any secondary structure like loops and β turns or even separate motifs ) ., The helices interact with each other to form the α/α interface facing the exterior ., It has been shown that the face of the fold with the C–terminal ends of the barrel and the adjoining loops contain the active–site residues , thus forming the catalytic face of the fold ( Figure 1b ) 18 ., As mentioned earlier TIM fold is rich in both sequential and functional diversity marking it a viable system for studying sequence–structure–function relationship ., The analysis of the Protein Energy Networks ( PENs ) provides a rationale to investigate the non–covalent interactions in proteins at various levels such as the interacting pairs ( edges ) , network of connected residues ( clusters ) , nodes connected by a large number of interactions ( hubs ) as a function of interaction energy ., The domains of the TIM barrel fold in the dataset ( Table S1 ) are represented as energy weighted structure networks ( PENs ) , in which the constituent amino–acids are considered as nodes and the edges are weighted based on the non–covalent interaction energies among the amino–acids ( Eq 3 , Methods Section ) ., Such a representation of PEN , capturing the non–covalent interaction energies at the atomic level , is capable of providing a consolidated view of the forces stabilizing the fold of the protein , yet retaining the details of individual interactions ., It is to be noted that highly favorable interactions ( for example , −25 kJ/mol ) will be referred to as “high–energy” interactions , whereas less favorable interactions ( for example , −10 kJ/mol ) will be referred to as “low–energy” interactions ., A range of unweighted PENes can be generated from the PEN using specific maximum energy cutoffs ( e ) to define the edges ( Eq 4 , Materials and Methods ) ., It was earlier noted from the PENs of a set of globular proteins that at low energies ( e>−10 kJ/mol ) the network is dominated by hydrophobic vdW interactions and above this value ( e<−10 kJ/mol ) , the electrostatic interactions starts dominating the edges in the PENs 38 ., The ljPENs are generated to focus exclusively on the vdW interactions by excluding the dominant terms of electrostatic interactions ., The largest cluster ( LC , see Materials and Methods ) profiles as a function of ‘e’ for both PENs and ljPENs are provided for the present dataset of 81 TIM barrel domains ( Figure S1 ) ., It is clear that the domains show three distinct network behaviors as a function of ‘e’ ( Figure S1a ) ., In the high–energy region ( e<−20 kJ/mol , henceforth denoted as pre–transition region ) , the LC size is small with the network connected by electrostatic interactions ., The size of the LC increases in the intermediate energy region ( −20<e<−10 kJ/mol , transition region ) following a sigmoidal profile by accruing low–energy vdW interactions and to encampass the whole protein in the low–energy region ( e>−10 kJ/mol , post–transition region ) , where the vdW interactions are dominant , tethering together local pockets of high–energy interactions ., The LC profile of ljPENs is similar to PENs except that the mid–transition point is around −7 kJ/mol ( Figure S1b ) , due to the absence of high–energy electrostatic interactions ., The TIM barrel domain is a common fold adopted by a large number of diverse sequences ., Here we ask the question whether these domains are stabilized by similar patterns of interactions ., Despite high sequence diversity we find common patterns of interactions of equivalent energies emerged when investigated at the family level ., The family level classification of the TIM fold was obtained from the SCOP database 39 ., We constructed family specific PENs for a chosen ‘e’ value ( f–PENes ) ( Figure, 2 ) and obtained the equivalent node/edge/network information from the multiple structural alignments of the constituent members ( Materials and Methods ) ., Each edge in the family specific network is given a commonality coefficient ( ccij ) value indicating the frequency of occurrence of that edge/interaction in the f–PENe ( Eq 5 and Figure 2f ) ., A ‘cc’ value of one corresponds to the presence and a ‘cc’ value zero represents the absence of interaction within a spatially similar position of the fold in all the members of a TIM family ., Thus various f–PENe ( cc ) can be generated for a specific family where f–PENe ( 1 . 0 ) represents interactions that are present in all the members of the fold and f–PENe ( 0 . 5 ) represents interactions that are present in at least half the members of the family ., In order to determine the role of an amino–acid ( node ) type in maintaining an interaction ( edge ) , we have used an Entropy based Conservation score ( EC ) for each node in the f–PEN ( see Methods Section 3 . 6 ) ., Generally if EC is greater than zero then there is a degree of conservation of that residue in the family , while a negative EC score shows that the residue is not conserved in that position ., Therefore , cc is a measure of “interaction conservation” between two nodes and EC is a measure of “residue conservation” of the nodes ., We have analyzed f–PENes in the dataset for edge distribution in different secondary structural interfaces namely the central β–barrel , α/β and α/α interfaces ., We further explore the network parameters like clusters and hubs in PENs and f–PENs to determine the maintenance of the fold architecture in the TIM fold despite low sequence homology ., In our analysis we principally focus on f–PENs at the pre–transition region ( ∼e<−18 kJ/mol , Figure S1a ) for studying the electrostatic contribution to the fold and the post–transition region of f–ljPENs ( ∼e<−8 kJ/mol , Figure S1b ) for obtaining the vdW contribution ., By analyzing the distribution of the conserved edges across different interfaces it is possible to determine how the fold is maintained irrespective of the residue conservation ., While the interaction–based studies discussed so far is a step above the residue level investigation , the network parameters like clusters and hubs go beyond pair–wise , by providing a collective view of multiple interacting residues ., For instance , even if common interacting pairs in a family of structures are not obvious , a collection of residues interacting at a threshold energy level at similar structural locations can be detected as clusters ., Therefore , we have utilized the PENs and f–PENs to study certain network properties like hubs and clusters to further understand the formation and stabilization of the fold ., One of the major implications in understanding protein sequence–structure–function relationship is that we can obtain a variety of evolutionary information ., Classically , existing phylogenetic methods exploit sequence conservation information to infer relationships and recent increase in structural data has resulted in the inclusion of structural features to deduce relationships between proteins 43 ., The most commonly used sequence conservation based methods fail to obtain correct relationships between remote homologues due to the misgivings of sequence alignment techniques in the “twilight region” of the sequence–structure space ., Here we deduce improved similarity relationships between remote homologues of the TIM fold through quantification of the similarity of interactions ( edges ) from their PENs ( details described in Materials and Methods ) ., Figure 6 shows the comparison of the cladograms ( a map of the hierarchical clusters ) obtained from the interaction based and sequence based techniques ., It can be readily seen that the interaction conservation based method clusters proteins of the same family under the same clade better than the sequence conservation based method ., It should be noted that the SCOP classification of families is based on sequence or structure or functional similarities ., The interaction based phylogeny matches very well with the SCOP classification than the sequence based method for the same dataset ., Despite low sequence identity ( ≤30% ) we were able to find domains that exhibited as high as ∼85% interaction conservation ( between d1r0ma1 and d1muca1 from DGDL family ) ., These observations show that the interaction based phylogenetic tree may be able to cluster the members of the family better than a residue based classification scheme ., Lockless and Rangathan 27 introduced a sequence-based method to investigate statistical interactions between residues ( Statistical Coupling Analysis ( SCA ) ) ., Later Halabi et al . , grouped these statistically correlated amino-acids into quasi-independent groups called sectors and studied their characteristics in Serine proteases 28 ., Here we have made the preliminary attempt to compare the interaction-energy based approach with the sequence based SCA approach ., We selected β-glycanase family of TIM fold for this comparison ., The interactions ( ≤−10 kJ/mol ) common to this family were identified and cross verified with correlated mutations obtained from SCA ., Although the correlation appeared to be weak at the pair-wise level , significant correlations are identified when the collective behavior of these correlated pairs are examined ., In other words , there is a significant match between the residues of the sector from SCA and the clusters obtained from the present energy based analysis ., The results have been pictorially depicted in Figure 7 ( details of the underlying calculations and comparison are provided in Table S2 and Table S3 ) ., Interestingly , the agreement is more in the regions stabilizing the structure ., The residues located more towards the function are identified by SCA and the PEN clusters encompass more of the residues required for the structural integrity ., Based on this reasonable correlation of the SCA sectors and PEN clusters , we emphasize the fact that the protein structures should be viewed as a collective entity and an examination of individual residues and pair interactions in isolation may not always provide a holistic view of the structure and function of proteins ., This feature was also reiterated by the coarse-grained network model studies on Rossmann-like domain proteins 29 ., A weak agreement of pair-wise correlations from SCA predictions with the biochemical experiments on double mutants of PDZ domain perhaps may be attributed to this reason ., Furthermore , fundamental issues like divergent 23 or convergent 20 evolution of proteins like TIM barrel , whose sequences are so diverse , has always been debated 16 ., Extensive investigation by complimentary approaches such as PEN , SCA and essential mode dynamics should be able to provide more clarity into such systems ., The sequence–structure relationship is a well–researched area , however , the factors that drive highly diverse sequences to fold into the same structure has not been well understood because of the apparent absence of consensus information from sequence similarity analyses ., Here we have taken an alternative approach in which we consider “interaction conservation” and analyze whether the preservation of interactions is an essential driving force in the formation of the fold rather than sequence conservation ., TIM barrel fold is one of the most popular folds that have a high sequence variability and functional diversity ., In this study we have analyzed non–homologous members of different families of the TIM fold and investigated various factors that contribute to the formation of the fold ., We have adapted the concept of interaction networks in order to study these protein structures from a global perspective ., Also , by using interaction energies we have realistically represented the residue–residue relationships in the network ., The subsequent methodology that exploits structural alignment to align the Protein Energy Networks ( PENs ) in a family of TIM fold has provided us with valuable information on the conservation of interactions in the family ., It was evident from our analyses of conserved interactions that the central β barrel is being stabilized by, ( a ) sequentially long–range conserved high–energy interactions and, ( b ) low–energy vdW interactions from residues of the neighboring strands interacting in tandem , in addition to the hydrogen–bonding network in the sheet ., Also , the analysis of the other interfaces like the α/β and the α/α show an absence of any high–energy conserved interactions , and being maintained exclusively by low–energy interactions ., In general we found that the residues involved in high–energy interactions are better conserved than low–energy interactions ., From our cluster analysis it was seen that the conserved interactions are not segregated into isolated interacting pairs but rather coalesce together to form a sub–network of interactions ., Our hub analysis has shown that the charged and the conserved residues are favorable to be hubs at higher energies , while hydrophobic residues with less conservation act as hubs at lower energies ., All these results suggest that, ( a ) the β barrel formation driven by high–energy interactions ( with the participating residues being conserved ) seem to be an important step in the organization of the TIM barrel;, ( b ) the formation of the other interfaces mainly by low–energy interactions ( with residue conservation being immaterial ) is a more canonical step in the fold formation common to all the folds of the α/β class , and can be taken up by a variety of sequences , thus contributing the high sequence diversity ., These conclusions concur with several experimental observations that suggest that while the α/β interfaces in TIM are resilient to mutations the β barrel is sensitive 18 , 40 , 41 , 44 ., We have analyzed the structural and functional relevance of conserved interactions in the regions involving loops in various TIM barrel families ., We found that loop based high–energy conserved interactions ( e<−20 kJ/mol ) are present near the active sites of a number of TIM barrel families ., This suggests that the loop based interactions are conserved during evolution to maintain the active site geometry for successful enzymatic functioning of the TIM proteins ., Therefore this method can be used in functional annotation of hypothetical proteins in cases where there are structural homologues but no sequence homologues ., Finally we exploited the concept of “interaction conservation” to construct a cladogram and compare it with the sequence based cladogram ., The outcome of analysis reinforces our assumption that it may be interaction conservation and not necessarily sequence conservation that determines the fold organization ., Our attempt to correlate our method with that of SCA suggests that there may be significant correlation between the sector residues and cluster residues ., However , extensive investigation by complimentary approaches such as PEN , SCA and Elastic Network Models ( ENM ) should be carried out and such an analysis will be able to provide more clarity to studying such protein systems ., The methodology of representing the protein structures as interaction energy based networks and using structural alignments to align these networks has provided us a very convenient handle to study structure homology among sequentially diverse proteins , from a network point of view ., We were able to study the salient features that stabilize the TIM fold using this method , and also analyze how interaction conservation can play an important role in the formation of this fold ., We believe that this methodology can shed valuable knowledge on the fold maintenance by remote homologues and pave way for useful de novo design and analysis of protein folds ., The dataset used in this analysis is composed of domains from the TIM fold given by Structural Classification Of Proteins ( SCOP ) 39 ., The coordinates for the domains are obtained from ASTRAL 45 ., The domains are sorted into their respective families as given in SCOP ., The sequence identity within the members of each family is less than 30% ., The culling of domains with higher sequence identity was done using cd–hit 46 ., All the families constitute at least three members ( except HMGL like domains ( HMGL ) and Adenosine/AMP deaminase ( ADA ) families , ( see Table S1 ) ) ., The dataset consisting of 19 families with 81 domains is presented in Table S1 ., The secondary structural elements ( SSE ) for each domain were assigned using DSSP 47 ., Structure network construction requires the coordinates of the interacting amino acids ( nodes ) and a criterion to define the interactions ( edges ) ., A purely geometry based all-atom interaction can be deduced from the crystal structure , which we had used to describe the Protein Structure Networks ( PSNs ) 48 ., Recently , we have considered the chemistry in greater detail by explicitly considering the interaction energy between residues 38 ., Although qualitative results are expected to be similar from both formalisms , PEN has the advantage of capturing subtle details of importance , whereas the PSN approach has the advantage of being simple to adopt ( Figure S7 ) ., The interaction energies can be obtained on a single structure or on an ensemble of structures of a given protein ., The set of structures can be obtained from experiments ( X-ray crystallography , Nuclear Magnetic Resonance ) under different environment or by simulations from a single starting conformation ., In the case where the conformational changes are small , a set of conformations will provide a statistically relevant average structure and in the case of large conformational change , it is advantageous to study them independently to characterize the structural variations in different states of the same protein , for example to understand the effect of ligand binding ., In this study we have used Molecular Dynamics ( MD ) simulations to obtain the structure ensemble for each of the TIM domains ., We have considered the crystal structures for all the proteins in the dataset ( Table S1 ) and subjected them to minimization and Molecular Dynamics simulations for a brief time interval ( 20 ps ) to obtain interaction energies in equilibrium ., In our earlier studies we have shown that the correlation between interaction energies calculated using the equilibrated structures from 2 ns simulations and 20 ps simulations was around 90% 38 ., The MD simulations were performed using GROMACS ( GROningen MAChine for Simulations ) 49 for just 20 ps and structure ensemble for each domain is obtained by sampling its trajectory every 1 ps ., The average interaction energies among the amino–acids are computed using the structure ensemble thus obtained ., Selenomethionines ( MSE ) present in certain domains like d1pbga_ and d1uwsa_ from Glycosyl hydrolase family ( F1GH ) were converted to Methionine and missing atoms in the residues were generated using Swiss PDB viewer 50 ., The best conformations for both the modified and the built residues recommended by the Swiss PDB viewer from its rotamer library were used ., The details of the construction of PEN are given in Vijayabaskar and Vishveshwara 38 ., Briefly , the non–bonded interaction energies ( Eij , Eq, 1 ) between all pairs of residues were obtained as a summation of the electrostatic ( given by columbic potential , Eq, 2 ) and van der Waals ( given by the Lennard Jones ( LJ ) potential , Eq, 3 ) interaction energies averaged over the structure ensemble ., PEN is constructed with amino–acids as nodes , and with edges drawn between all pairs of residues except the sequential neighbors ., The edges are weighted with the calculated Eij ., ljPENs take into account only the van der Waals ( vdW ) interactions ( i . e Eij\u200a=\u200aVLJ ) ., Unweighted networks ( PENe and ljPENe ) can be obtained for a specific maximum energy cutoff ‘e’ as given in Eq 4 ., ( 1 ) ( 2 ) ( 3 ) ( 4 ) Steps involved in the construction of the family specific PEN ( f–PEN ) by alignment of the PENes of its members is given in detail in Figure 2 ., Domains in a family are structurally aligned using MUSTANG ( MUltiple STructural AligNment AlGorithm ) 51 ( Figure 2b ) ., A family specific Multiple Structure based Sequence Alignment ( MSSA ) was obtained for all the members of a given family and the residues that are aligned in the MSSA are referred to as Equivalent residues ., Residues that were not structurally super–imposable were compensated within the alignment using gaps ( Figure 2c ) ., The PENes are remapped using the equivalent node information obtained from the MSSA ( Figure 2d ) ., The gaps in the MSSA are introduced as virtual nodes in the corresponding PENes , such that the edge weights of a virtual node to all other nodes in the PEN were highly unfavorable ( Eij\u200a=\u200a100 kJ/mol where either i or j is a virtual node ) ( Figure 2d ) ., The remapped PENes are then aligned to form the family specific PEN ( f–PENe ) ( Figure 2e ) such that the nodes are equivalent and edges exists only if they were present in any of the realigned PENes ( Figure 2f ) ., In a f–PENe , the values ( X , Eq, 5 ) of the edges can vary from 0 to M , where 0 represents the absence of an edge in all the members of the f–PENe and M represents the edge being present in all members ., Therefore each edge is given a commonality coefficient ( ccij , Eq 5 ) , and it represents the measure of the frequency of occurrence of an edge between equivalent nodes within the members of a family ., ( 5 ) where X is the total number of members having the edge between nodes ‘i’ and ‘j’ with interaction energy better than ‘e’ , Aeij is the element of the adjacency matrix of the remapped PENe and M is the total number of members in the family ( Figure 2e ) ., Thus , a family specific PEN can be denoted as f–PENe ( cc ) where ‘e’ is the interaction energy cutoff used to generate PENes for all the members of the family and edges are constructed only if their ccij is better than ‘cc’ ., The f–PENe ( cc ) consists of both equivalent and virtual nodes and represents spatially conserved interactions across the members of that family ., In fact both the ‘e’ and ‘cc’ values can be used as weights in order to construct a weighted matrix ., However , in this study , we have considered un-weighted matrix at given values of ‘e’ and ‘cc’ ., Entropy based Conservation scores ( EC ) for each alignment position in the MSSA were obtained using AL2CO 52 ., In this method the entropy is normalized with the mean and standard deviation ., Thus better the entropy score , the more conserved the amino–acids are at that position ., A network similarity matrix ( S ) for any two members ‘a’ and ‘b’ in the dataset is constructed as given in Eq, 6 . S is an adjacency matrix which takes a value of 1 if the interaction energies between equivalent residues in the MSSA are similar ., The Similarity Score ( SSab ) between the PENs of any two members in the dataset is derived as given in Eq, 7 . This value is the fraction of edges that is conserved between the two members ., The distance matrix ( D , Eq 8 ) with each row and column representing a domain in the dataset , is used to construct the phylogenetic tree ., ( 6 ) ( 7 ) ( 8 ) where Ea and Eb are PENs of any two members in the dataset that are remapped based on their pairwise MSSA , and N is the total number of nodes in the remapped PENs ., The concept of structure conservation is often used in structural alignment methods 53 , 54 ., For instance , an alignment based on dynamic characteristics of structurally similar but functionally distinct proteins have been reported earlier 29 ., The identification of energetically similar edges in two proteins done in the present study , can also serve as a basis for alternate method of structural alignment , although it is not pursued in this study ., Clusters were generated using Depth First Search ( DFS ) algorithm 55 ., Family specific clusters in a family of TIM fold are

Introduction, Results/Discussion, Materials and Methods

There are many well-known examples of proteins with low sequence similarity , adopting the same structural fold ., This aspect of sequence-structure relationship has been extensively studied both experimentally and theoretically , however with limited success ., Most of the studies consider remote homology or “sequence conservation” as the basis for their understanding ., Recently “interaction energy” based network formalism ( Protein Energy Networks ( PENs ) ) was developed to understand the determinants of protein structures ., In this paper we have used these PENs to investigate the common non-covalent interactions and their collective features which stabilize the TIM barrel fold ., We have also developed a method of aligning PENs in order to understand the spatial conservation of interactions in the fold ., We have identified key common interactions responsible for the conservation of the TIM fold , despite high sequence dissimilarity ., For instance , the central beta barrel of the TIM fold is stabilized by long-range high energy electrostatic interactions and low-energy contiguous vdW interactions in certain families ., The other interfaces like the helix-sheet or the helix-helix seem to be devoid of any high energy conserved interactions ., Conserved interactions in the loop regions around the catalytic site of the TIM fold have also been identified , pointing out their significance in both structural and functional evolution ., Based on these investigations , we have developed a novel network based phylogenetic analysis for remote homologues , which can perform better than sequence based phylogeny ., Such an analysis is more meaningful from both structural and functional evolutionary perspective ., We believe that the information obtained through the “interaction conservation” viewpoint and the subsequently developed method of structure network alignment , can shed new light in the fields of fold organization and de novo computational protein design .

Proteins are polymers of amino-acids that fold into unique three-dimensional structures to perform cellular functions ., This structure formation has been shown to depend on the amino-acid sequences ., But examples of proteins with diverse sequences retaining a similar structural fold are quite substantial that we can no longer consider such phenomenon as exceptions ., Therefore , this non-canonical relationship has been studied extensively mostly by studying the remote sequence similarities between proteins ., Here we have attempted to address the above-mentioned problem by analyzing the similarities in the spatial interactions among amino-acids ., Since the protein structure is a resultant of different interactions , we have considered the proteins as networks of interacting amino-acids to derive the common interactions within a popular structural fold called the TIM barrel fold ., We were able to find common interactions among different families of the TIM fold and generalize the patterns of interactions by which the fold is being maintained despite sequence diversity ., The results substantiate our hypothesis that interaction conservation might by a driving factor in fold formation and this new outlook can be used extensively in engineering proteins with better biophysical characteristics .

protein structure, biology, computational biology, macromolecular structure analysis

journal.pntd.0005797

Temperature modulates dengue virus epidemic growth rates through its effects on reproduction numbers and generation intervals

Dengue virus ( DENV ) is a mosquito-borne pathogen that infects hundreds of millions of people each year across as many as 128 countries 1 ., Along with numerous other arthropod-borne viruses ( arboviruses ) , including chikungunya and Zika viruses 2 , 3 , DENV causes epidemics with considerable public health impact ., Rapidly growing , intense epidemics can overwhelm healthcare systems 4 , leaving those infected without adequate medical treatment and with a significantly elevated risk of mortality to a disease that is seldom fatal when proper treatment is available 5 ., A number of factors can lead to variability in the frequency and severity of arbovirus epidemics , including importation probability 6 , host susceptibility 7 , and climatic conditions 8 ., In particular , temperature is known to be a major driver of spatial and temporal variability in arbovirus transmission , as indicated by empirical studies of relationships between temperature and several epidemiologically important vector and pathogen traits , including mosquito lifespan 9–11 , incubation time of the pathogen in the mosquito 9 , 10 , 12 , the rate at which mosquitoes engage in blood feeding 9 , 13 , and mosquito density 14 ., Analyses of the effects of temperature on vector-borne pathogen transmission have focused primarily on the basic reproduction number R0 through the effects of temperature on the aforementioned vector and pathogen traits 11 , 15 , 16 ., Defined as the average number of secondary infections arising from a primary infection in a fully susceptible population , R0 is a fundamentally important epidemiological quantity , because it is informative about the conditions under which a pathogen can invade , or be eliminated from , a host population ., The generation interval , which is the period of time separating sequential infections , is the temporal analogue of R0 ., Through a fundamental mathematical relationship 17 , R0 and the generation interval are related to the epidemic growth rate r , which is defined as the per capita change in incidence per unit time and characterizes the dynamics of early-stage epidemic growth in a susceptible population ., Because the relationship between r and temperature has never been characterized for arboviruses , there is little scientific basis for understanding how epidemic growth rates may be related to temperature ., Our goal was to quantify the effects of temperature on DENV epidemic growth rates by first establishing a probabilistic description of DENV generation intervals as a function of temperature ., We then combined our generation interval calculations with a temperature-dependent formulation of the basic reproduction number , R0 , and solved for the epidemic growth rate r as a function of temperature ., This new capability to calculate r as a function of temperature allowed us to identify temperature ranges that maximize r and to classify regions by their potential for increasing or decreasing epidemic growth rates based on their current and future temperatures ., Our results and the accompanying code are made freely available online at https://github . com/asiraj-nd/arbotemp to facilitate the incorporation of temperature-dependent descriptions of these quantities into future studies ., We define the generation interval as the elapsed time between a primary human infection and a secondary human infection deriving from that primary human infection via two bites from the same individual mosquito 18 ., To derive a quantitative , probabilistic description of the generation interval for dengue , we adapted an existing framework that defines the generation interval as a sum of random variables for each of four sequential , constituent phases of the transmission cycle 19 ., Similar to a recent analysis for Plasmodium falciparum malaria 20 , we furthermore quantified each of these phases of the transmission cycle as dependent on temperature ( Fig 1 ) ., Following Huber et al . 20 , we defined these phases as: ( 1 ) the intrinsic incubation period ( IIP ) ; ( 2 ) the period between onset of symptoms in humans and subsequent transmission to mosquitoes ( human-to-mosquito transmission period , HMTP ) ; ( 3 ) the extrinsic incubation period ( EIP ) ; and ( 4 ) the period between a mosquito becoming infectious and subsequent transmission to humans ( mosquito-to-human transmission period , MHTP ) ( Fig 1 ) ., Below , we describe the derivation and parameterization of each of these phases of the transmission cycle as four independent random variables based on available data 13 , 21 , 22 ., To obtain a single random variable describing the generation interval as a whole , we took the sum of the four constituent random variables in Fig 1 by applying the convolution theorem , which involves taking the inverse Fourier transform of the product of the Fourier transforms of each random variable 23 ., The basic reproduction number ( R0 ) is defined as the average number of secondary infections in humans originating from a single primary human infection introduced into a fully susceptible population ., We used the formal definition of R0 for mosquito-borne pathogens based on a set of classic “Ross-Macdonald” assumptions 29 , which takes the temperature-dependent form, R0 ( T ) =m ( T ) bca ( T ) 2e−μ ( T ) n ( T ) μ ( T ) γ ,, ( 1 ), where m ( T ) is the mosquito-to-human ratio as a function of temperature T , μ ( T ) is the mean daily mortality rate of adult mosquitoes at temperature T , b and c are human-to-mosquito and mosquito-to-human infection probabilities , a ( T ) is the mosquito biting rate as a function of temperature , 1/γ is the average duration of infectiousness in humans , and n ( T ) is the mean extrinsic incubation period at temperature T . We note that the mean daily mortality rate of adult mosquitoes , μ ( T ) , is the inverse of the mean for the MHTP distribution used in obtaining the generation interval distribution , while the mean extrinsic incubation period , n ( T ) , is the mean for the EIP distribution , also used in obtaining the generation interval distribution ., Our parameterization of the ratio c/γ equaled the integral of the non-normalized HMTP curve describing the infectiousness of humans to mosquitoes over time 22 , as noted in the section describing the generation interval ., The parameter b did not appear in our description of the generation interval , because it affects only the magnitude of transmission ( i . e . , R0 ) rather than its timing ( i . e . , generation interval ) ., This parameter is poorly understood empirically , so we chose a value of b = 0 . 4 consistent with a previous model 30 ., We described biting rate a as a function of temperature T ( i . e . , a ( T ) ) using two temperature-dependent estimates based on the average duration of the Ae ., aegypti gonotrophic cycle 9 , 31 , similar to how gonotrophic period was incorporated into the generation interval ., This process involved weighting the temperature-dependent length of the first cycle and the temperature-dependent length of each subsequent cycle based on the probability of the mosquito surviving to a given number of cycles ( see S1 Appendix for mathematical derivation ) ., To capture one potential effect of temperature on the ratio of mosquitoes to humans m , we assumed that m ( T ) = λ / μ ( T ) consistent with equilibrium assumptions of a mosquito population with adult mortality rate μ ( T ) and constant parameter λ , which is the ratio of the daily rate of adult female mosquito emergence and the number of humans subject to feeding by the mosquitoes represented by m ( T ) 32 ., Because values of λ are highly variable in space and time for reasons other than temperature variation , we examined the sensitivity of the value of λ across a range of values 0 . 0–0 . 5 ., We arrived at 0 . 5 as an upper limit for λ by dividing an upper limit for R0 based on independent estimates ( maximum of 7 . 8 33 ) by all other terms on the right-hand side of Eq 1 ( 19 . 73 at 32 . 5°C ) ., This is equivalent to assuming that one new adult female mosquito emerges from larval habitats every other day for each human at risk of biting within a given population ., To account for uncertainty associated with values of R0 that we calculated , we generated 1 , 000 Monte Carlo samples from the uncertainty distributions of each model parameter as described in each of the references 9 , 12 , 13 in which those parameters were originally described ., For μ ( T ) and n ( T ) , we took random draws of their parameters consistent with published descriptions of uncertainty in the parameters of these functions from their original sources 13 , 14 ., For a ( T ) , we used nonlinear least-squares estimates of the first gonotrophic period’s ρ parameter in the model by Focks et al . 9 by refitting it to their data , resulting in mean 8 . 83x10-3 and standard deviation 3 . 8x10-4 ., We assumed similar uncertainties ( standard deviation ) around the ρ parameter for the second gonotrophic period proposed by Otero et al . 31 ., We then took random draws from normal distributions describing uncertainties in these two parameters and weighted the resulting two temperature-dependent biting rates ( inverses of the gonotrophic periods ) according to the probability of the mosquito surviving to a given number of gonotrophic cycles , as described in S1 Appendix ., A summary of parameters and their default values is available in S4 Table ., Given temperature-dependent formulations of R0 ( T ) and the DENV generation interval g ( t ) described above , we solved for the corresponding epidemic growth rate r ( T ) as a function of temperature by applying the result, 1R0 ( T ) =∫0∞e−r ( T ) tg ( t ) dt, ( 2 ), from Wallinga and Lipsitch 17 ., Although this does not yield an explicit relationship between r and T that can be probed analytically , it does provide a way of numerically characterizing the impacts of temperature on r ., We further note that this approximation of r ( T ) assumes a fully susceptible , well-mixed population of mosquitoes and hosts ., We first derived a formulation of the generation interval for dengue , stochastic variability therein , and its dependence on temperature based on the assumptions described above ., We then performed analyses of the relationship between temperature and r , including identification of the temperature that maximizes r and how incremental changes in r driven by changes in temperature can be attributed to distinct contributions from changes in R0 versus changes in the generation interval ., For comparison with our detailed formulation of the epidemic growth rate r , we examined two approximations of the generation interval commonly used in transmission models: a fixed-length generation interval and an exponentially distributed generation interval ., For each , we considered two formulations: one with a mean generation interval of 16 days 34 and one with temperature-dependent mean generation interval as calculated using our method ., We next considered how average monthly temperature data at 5 km x 5 km resolution for each month of the year based on historical records ( average for 1950–2000 ) 35 may change epidemic growth rates under climate change scenarios ., For this analysis , we used three different scenarios for mean temperature in 2050 ( average for 2040–2060 ) corresponding to Representative Concentration Pathways ( RCPs ) that describe a set of alternative trajectories for the atmospheric concentration of key greenhouse gases: RCP 8 . 5 , high greenhouse gas concentration scenario; RCP 6 . 0 , medium baseline ( or high mitigation ) scenario; and RCP 4 . 5 , intermediate mitigation scenario 36 ., We obtained gridded population estimates for the year 2000 from the Global Rural/Urban Mapping Project 37 and for 2050 by projecting values from 2000 onward according to medium-fertility population projections for each country 38 ., We excluded regions from this analysis where Ae ., aegypti occurrence probabilities fall below 0 . 8 , a threshold value that separates two distinct modes of local occurrence probabilities globally 39 , 40 ., Potential for diurnal temperature fluctuations to influence DENV transmission has been suggested by temperature effects on extrinsic incubation period ( EIP ) and mosquito survival 10 ., We examined potential effects of diurnal temperature fluctuations on the generation interval , basic reproduction number R0 , and epidemic growth rate r by introducing an 8°C diurnal temperature range ( DTR ) around all mean temperatures ., We assumed a sinusoidal progression within the day with a decreasing exponential curve at night 9 , 41 ., We also assumed an absolute maximum temperature for Ae ., aegypti survival of 37 . 73°C over three consecutive hours and 40 . 73°C in any single hour , as well as a maximum temperature of 45 . 9°C in any hour of the day for DENV incubation to take place , similar to assumptions of another recent model of temperature-dependent viral transmission by Ae ., aegypti mosquitoes 42 ., We developed a probabilistic description of the DENV generation interval by sequentially summing random variables associated with each phase of the transmission cycle ( Fig 1 ) ., Allowing each of these component random variables to depend on temperature ( Fig 2A ) resulted in a description of the generation interval that was itself strongly dependent on temperature and captured variability and uncertainty in the underlying components ( Fig 2B ) ., For example , mean generation interval halved from 30 to 15 days with a change in temperature from 25 to 35°C ., Sensitivity of the mean generation interval to changes in temperature was nonlinear , with steeper changes at more extreme temperature values ( Fig 2B ) due to increasing steepness of the relationships between temperature and the component random variables ( Fig 2A ) ., The basic reproduction number , R0 , was also sensitive to temperature , as it includes the same temperature-dependent random variables as the generation interval ., At low temperatures , increases in temperature caused a steady increase in R0 due to a shortening extrinsic incubation period and increasing biting frequency ( Fig 2A and 2C ) ., Beyond a peak temperature of 32 . 5°C , R0 decreased rapidly with increasing temperatures due to rapidly increasing mosquito mortality ( Fig 2A and 2C ) ., This result contrasted with a lower peak temperature ( ~29°C ) that was obtained in our analysis ( not shown in figures ) under an assumption that biting rate did not depend on temperature ., Effects of temperature on the DENV generation interval and R0 contributed to similar effects on epidemic growth rate , r ., Under mean estimates of model parameters , r increased with temperature until it peaked at 33°C ( Fig 2D ) ., Under 1 , 000 Monte Carlo samples of model parameters , peak temperature for r varied within a relatively narrow band with 95% of values falling between 32 . 6 and 33 . 2°C ( Fig 3 ) ., As both R0 and the generation interval are temperature-dependent , changes in r due to temperature occur through both components ., At a constant mosquito emergence rate λ , changes in R0 accounted for the majority of changes in r , although changes in the generation interval accounted for a greater degree of change near extreme and peak temperature regions ( Fig 4; S1 Fig ) ., Allowing for diurnal temperature fluctuations ( 8°C daily temperature range for all mean temperature values ) shortened the mean generation interval and increased its variance relative to a scenario with no diurnal temperature fluctuation ( S2 Fig ) ., Similarly , R0 decreased when DTR was considered , as the temperature at which R0 peaks decreased from 32 . 5 to 30 . 9°C due to the effect of daily temperature extrema ( under DTR ) on mosquito survival ( S3 Fig ) ., The combined effect of these changes on epidemic growth rate was a slight decrease , while the temperature at which the epidemic growth rate peaks remained close to its value under a scenario with no diurnal temperature fluctuation ( Fig 3; S2–S4 Figs ) ., Because a fully detailed generation interval distribution is beyond the capabilities of many commonly used modeling frameworks 43 , we examined the correspondence between epidemic growth rates r calculated under our detailed approach and under four less detailed approximations of the generation interval that are commonly used in transmission models ., A fixed-length generation interval yielded a consistently better approximation of our detailed calculations of r as a function of temperature than did an exponentially distributed generation interval ( Fig 5A vs . 5B ) ., Calculations of r under the fixed-length approximation tended to match calculations of temperature-dependent r under the detailed generation interval distribution particularly well in temperature ranges of significance to epidemics ( i . e . , where r > 0 ) ( Fig 5B ) ., For both fixed-length and exponential generation interval distributions , allowing their mean values to follow the temperature-dependent model improved their correspondence with our detailed formulation of temperature-dependent r ( Fig 5A & 5B vs . 5C & 5D ) ., These differences in r resulting from different assumptions about the distribution and temperature dependence of r could be of significance to epidemic projections , given that differences in r as small as 0 . 01 can lead to differences in incidence projections of an order of magnitude only a few months into an epidemic ( Fig 6 ) ., Our result that the temperature threshold for maximum r was relatively constant around 33°C ( 95% CI: 32 . 6–33 . 2°C ) offers a useful reference point ., In a given area and with other factors held constant , an increase in temperatures beyond this threshold would imply first a rise and then a fall in r between present and future ., An increase in temperatures that never exceeds this threshold would imply an increase in r between present and future ., At most times of year in most regions of the world that are suitable for DENV transmission , temperature increases by 2050 are expected to fall into the latter category ( i . e . , remaining below 33°C ) , suggesting that temperature changes could increase epidemic growth rates in those areas ( S1–S3 Tables , S5–S16 Figs ) ., On the other hand , temperature increases by 2050 in regions such as India and the African Sahel are expected to exceed 33°C during April-June , potentially resulting in lower epidemic growth rates in those areas during a portion of the year ( S1–S3 Tables , S5–S16 Figs ) ., The central advance that we have made is the development of a probabilistic description of the generation interval for dengue virus ( DENV ) that is based on first principles of transmission , synthesizes pertinent data for DENV and Ae ., aegypti , and characterizes the generation interval as a function of temperature ., Although there is little data with which to independently validate our calculations , the mode of our generation time distribution at optimal temperatures for transmission ( approximately 16 days at 28–32°C ) accords with independent estimates of this quantity based on statistical analyses of spatiotemporal dengue case data from Thailand ( 15–17 days ) 34 ., Combining this result with a temperature-dependent description of the basic reproduction number , R0 , we obtained a temperature-dependent description of the epidemic growth rate , r ., All of these quantities were estimated explicitly for DENV but are also relevant for other arboviruses such as chikungunya and Zika , given their similar ecology and given that many of the parameters we used are not specific to any one virus but instead to their common vector ., The generation interval has a wide range of applications in epidemiology , including the identification of sources of infection 44 , the establishment of causal linkages between cases 45 , and the characterization of temporal variation in transmission 8 , 46 ., These and other studies have typically assumed a static generation interval of either fixed length 47 or with some standard statistical distribution 48 ., Our result that the generation interval for DENV is not static but is instead highly dynamic with respect to temperature highlights that transmission models for DENV and other arboviruses could be systematically inaccurate by excluding temperature-dependent effects ., Future work will be needed to address the existence and significance of any such inaccuracies , but our results about the sensitivity of r to the form of the generation interval and temperature dependency therein suggest that these effects could be substantial ., Our calculations of R0 are consistent with the notion that temperature plays an important role in determining optimal conditions for transmission ( i . e . , peak R0 at 32 . 5°C ) and for delimiting conditions where transmission is sustainable ( i . e . , R0 > 1 , Fig 2B and 2C ) ., However , these results are only valid for a given value of the ratio of new adult mosquitoes to humans λ , which we allowed to vary within a plausible range due to the fact that it depends on a wide range of factors other than temperature ., In particular , λ depends on the availability and quality of aquatic habitats for mosquitoes 49 and sociocultural factors that affect contact between people and mosquitoes 50 ., Some studies have used temperature-based R0 calculations to delimit geographic ranges of other vector-borne diseases such as malaria 16 , but we used R0 solely as part of an intermediate step to link the generation interval with epidemic growth rates ., Although R0 is important for quantifying threshold conditions for pathogen persistence , it is not well suited for characterizing temporal dynamics of transmission 51 ., By combining temperature-dependent descriptions of R0 and the generation interval , our results offer a new way to characterize the intensity of dengue epidemics as a function of temperature ., One common concern about analyses based on R0 , and estimates of r based on R0 , is whether they are relevant beyond the context of a novel pathogen in a fully susceptible population ., Estimates of r based on the effective reproduction number , R 17 , offer a more generalizable alternative to estimates of r based on R0 , which is what we have considered in this study ., To consider how the distinction between R0 and R might impact our results , we note the relationship R = R0S , where S is the proportion of a population that is susceptible ., This linear relationship between R0 and R implies that extrapolating our results below S = 1 should result in behavior similar to how our temperature-dependent estimates of r vary with changes in λ , given that λ also affects R0 linearly ., Perhaps most importantly , this reasoning implies that the temperature at which epidemic growth rates peak should be applicable across contexts in which either the susceptible fraction S or the mosquito-human ratio λ vary ., Still other factors affecting R0 and r could vary across contexts—e . g . , species or strain differences 52—that could be important for some future applications ., One limitation of our approach is that the precise value of the temperature threshold for maximum r could be subject to revision as understanding of the relationships between temperature and transmission parameters improves ., In previous work 15 , revised assumptions about the effects of temperature on transmission parameters were shown to affect prior understanding of the relationship between temperature and R0 for malaria ., Independently validating our calculations with epidemic data could be one way to address these uncertainties , but epidemic growth rates based on case reports can be difficult to compare across sites ., Even if factors such as temperature are consistent across sites , still others may vary and have major impacts on epidemic growth rates , including mosquito abundance 39 , population immunity 53 , and reporting rates 54 ., Due to these and other variations across locations , Johansson et al . 55 found no detectable association between temperature and large-scale epidemic dynamics ., Our results make important progress towards being able to resolve the roles of and complex interactions among these factors in future studies ., Based on current understanding of relationships between temperature and transmission parameters , our result that r consistently peaks around 33°C ( 95% CI: 32 . 6–33 . 2°C ) led us to examine which populations globally could remain below , newly exceed , or further surpass this temperature under future climate change scenarios ., We found that most people currently living in areas at risk for DENV transmission could be subject to increased epidemic growth rates by 2050 under a range of scenarios about future temperature increases ., For most DENV-endemic areas , this would have little effect on the overall burden of disease , which is already high , but it could affect transmission dynamics , making epidemics more intense ., At the same time , there are a number of important caveats to bear in mind about these projections ., First , transmission depends not only on temperature but also other abiotic variables , such as rainfall , in complex ways 56 ., Second , the effects of abiotic variables may be outweighed by changes in human factors , such as economic development , urbanization , demography , and population immunity 57 , 58 ., Third , long-term projections of dengue are highly variable and conflicting 59 , making the long-term effects of any single change such as temperature nearly impossible to anticipate ., Although r will vary across different regions for different reasons , our finding that temperature changes under future climate change could elevate epidemic intensity of dengue in some areas suggests a categorically new way in which climate change might impact infectious disease transmission 60 ., Our quantification of these effects focused on DENV , but these results also offer tentative , but plausible , estimates of how epidemics of other viruses transmitted by Ae ., aegypti mosquitoes , such as chikungunya and Zika , might be impacted under future climate change ., Our qualitative results apply even more broadly , implying that temperature has the potential to shape multiple aspects of vector-borne parasite life history and to influence multiple aspects of the temporal dynamics of associated diseases .

Introduction, Materials and methods, Results, Discussion, Conclusion

Epidemic growth rate , r , provides a more complete description of the potential for epidemics than the more commonly studied basic reproduction number , R0 , yet the former has never been described as a function of temperature for dengue virus or other pathogens with temperature-sensitive transmission ., The need to understand the drivers of epidemics of these pathogens is acute , with arthropod-borne virus epidemics becoming increasingly problematic ., We addressed this need by developing temperature-dependent descriptions of the two components of r—R0 and the generation interval—to obtain a temperature-dependent description of r ., Our results show that the generation interval is highly sensitive to temperature , decreasing twofold between 25 and 35°C and suggesting that dengue virus epidemics may accelerate as temperatures increase , not only because of more infections per generation but also because of faster generations ., Under the empirical temperature relationships that we considered , we found that r peaked at a temperature threshold that was robust to uncertainty in model parameters that do not depend on temperature ., Although the precise value of this temperature threshold could be refined following future studies of empirical temperature relationships , the framework we present for identifying such temperature thresholds offers a new way to classify regions in which dengue virus epidemic intensity could either increase or decrease under future climate change .

Recurrent , rapidly intensifying epidemics of dengue–the world’s most prevalent mosquito-borne viral disease–pose a challenge to healthcare systems throughout the tropical and subtropical world ., An acute disease that tends to respond well to proper treatment , the sometimes intense nature of dengue epidemics has been known to overwhelm healthcare systems and elevate the morbidity and mortality of patients left without adequate medical treatment under peak epidemic conditions ., Here , we quantify the temperature dependence of dengue epidemic intensity by quantifying two distinct determinants of epidemic growth rate: the average number of secondary infections arising from each primary infection and the average time between successive infections in humans ., Our results show that the time between successive infections in humans decreases steadily with increasing temperatures , whereas the average number of secondary infections peaks at intermediate temperatures ., Altogether , this suggests a peak temperature for dengue epidemic intensity ., Applying this result to global temperature projections under future climate change scenarios suggests that dengue epidemics in many regions of the world could become more intense under future temperature increases .

death rates, invertebrates, dengue virus, medicine and health sciences, pathology and laboratory medicine, chikungunya infection, infectious disease epidemiology, demography, atmospheric science, pathogens, tropical diseases, microbiology, random variables, animals, viruses, mathematics, rna viruses, neglected tropical diseases, insect vectors, climate change, infectious diseases, medical microbiology, epidemiology, microbial pathogens, disease vectors, insects, probability theory, arthropoda, people and places, mosquitoes, climatology, flaviviruses, earth sciences, viral pathogens, biology and life sciences, species interactions, physical sciences, viral diseases, organisms

journal.pgen.1002885

Transgene Induced Co-Suppression during Vegetative Growth in Cryptococcus neoformans

In genetically modified plants and fungi , introduced transgenes are in some cases silenced and can also cause silencing of endogenous genes if they share sufficient homology 1 ., This phenomenon has been termed co-suppression or homology-dependent gene silencing ( HDGS ) , and usually occurs when multiple copies of particular sequences are present in the genome 2 , 3 ., Depending on the mechanistic level at which silencing occurs , HDGS has been distinguished into two types: transcriptional gene silencing and post-transcriptional gene silencing ( TGS and PTGS , respectively ) 1 ., During TGS , repression of transcriptional initiation is normally achieved by establishing an epigenetic inactivation state characterized by altered methylation patterns or chromatin structure ., In contrast to TGS , genes silenced by PTGS are normally transcribed but their transcripts do not accumulate , as a consequence of rapid degradation ., An RNAi degradation process is central to PTGS , in which a dsRNA intermediate homologous to the target gene or transgene is generated and processed into siRNAs of 21–25 nucleotides ., siRNAs subsequently guide gene silencing in a sequence-specific manner 4 , 5 ., In fungi , a co-suppression like phenomenon referred to as quelling 6 , 7 occurs in Neurospora crassa vegetative tissue , while other related silencing phenomena , repeat-induced point mutation ( RIP ) 8 and methylation induced pre-meiotically ( MIP ) 9 , 10 are active in the premeiotic phase during the sexual cycles of N . crassa , Ascobolus immersus , and Coprinopsis cinerea , respectively ., Quelling , RIP , and MIP are well-documented cases of HDGS showing variability in developmental timing , and are also mechanistically heterogeneous ., MIP in A . immersus and C . cinerea and RIP in N . crassa involve a similar molecular mechanism to inactivate repetitive sequences , except that repeat sequences are methylated in MIP 11 , whereas RIP involves inactivating the methylated sequences by introduction of C-to-T ( G-to-A ) transition mutations 12 ., Quelling is an RNAi-related PTGS process that is induced by siRNAs and requires the core RNAi components , including Argonaute , Dicer-like proteins , and RNA-dependent RNA polymerase ( RdRP ) 5 , 13 , 14 , 15 ., In quelling , the silenced loci can act in trans , leading to silencing of all homologous genes ., Co-suppression phenomena similar to quelling have been widely described in Arabidopsis , Drosophila , and C . elegans 16 , 17 ., In many cases , these phenomena are induced by either highly expressed single copy transgenes or highly repetitive transgene arrays present in the genome , and gene silencing is often linked to an RNAi pathway 13 , 18 , 19 ., In addition to regulating transgene expression , co-suppression is also considered a host genome defense mechanism acting against invading parasitic selfish DNA , such as transposons and viruses ., For example , Piwi-interacting RNAs ( piRNAs ) that bind to the Piwi proteins of the Argonaute superfamily are required for silencing transposons in the animal germline via an RNAi-PTGS process 20 , 21 ., Additionally , viruses carrying homology to nuclear sequences can trigger PTGS in plants , resulting in silencing of both viral genes and endogenous genes 22 , 23 ., Transgene-induced gene silencing has also been observed in Cryptococcus neoformans ., In a previous study , we reported the discovery of sex induced silencing ( SIS ) in C . neoformans , a novel silencing mechanism triggered by a tandem multi-copy insertion of a SXI2a-URA5 transgene 24 ., SIS is highly active during the sexual cycle and requires the central components of the RNAi pathway to silence target genes post-transcriptionally 24 ., The frequency of SIS is increased with higher transgene copy number , but becomes saturated at ∼50% when more than three copies of a transgene are integrated into the genome ., In addition to inactivating the transgene arrays , SIS also functions to squelch transposon activity during the sexual cycle , which is reflected in the observation that an increased transposition/mutation rate was detected in RNAi mutant progeny 24 ., In summary , the SIS RNAi pathway has been proposed as a meiotic mechanism to guard genome integrity ., Here we report a robust transgene-induced silencing phenomenon homologous to SIS , but occurring during C . neoformans vegetative ( asexual ) growth ., This silencing is more related to co-suppression in plants and quelling in N . crassa because no sexual cycle is involved ., To distinguish this process from SIS , we named this silencing phenomenon asexual co-suppression ., Cryptococcus co-suppression is also initiated by the integration of a transgene array into the genome , and inactivates genes that are homologous to DNA sequences introduced by the transgene ., We observed that the silencing efficiency reached ∼90% in the case of more than 25 copies of a transgene ., The fact that the suite of RNAi machinery components and the RPA complex are necessary for asexual co-suppression indicates that asexual co-suppression may share a similar molecular mechanism with quelling in N . crassa ., Notably , when the introduced transgene array contains regions homologous to different target genes , the silencing rates may vary and are gene-specific ., These findings provide evidence that a quelling-like asexual co-suppression pathway operates in C . neoformans , and occurs at different efficiencies with regard to different target genes ., Similar to SIS , asexual co-suppression could be a major mechanism involved in silencing transposons and repetitive sequences in C . neoformans during vegetative growth ., CPA1 and CPA2 are two homologous genes encoding two conserved cyclophilin A proteins , Cpa1 and Cpa2 , in C . neoformans ., They are both located on chromosome 2 , linked 21 . 01 kb apart , and share 85% nucleotide identity in the coding regions ( Figure 1A ) ., Previously , we disrupted the CPA1 and CPA2 genes , individually and in combination , and determined the functions of cyclophilin A in C . neoformans 25 ., Both the Cpa1 and Cpa2 cyclophilin A proteins are the targets of the immunosuppressive and antifungal natural product cyclosporine A ( CsA ) ; either cpa1 or cpa2 single mutant strains remain sensitive to CsA at 37°C but cpa1 cpa2 double mutants are completely resistant to CsA ., Thus , both Cpa1 and Cpa2 mediate CsA inhibition of calcineurin and inhibit growth at 37°C ., In addition , the cpa1 cpa2 mutant is sterile in genetic crosses and formed almost no heterokaryotic filaments , basidia , or basidiospores 25 ., Paradoxically , during screens to isolate cpa1 single mutant strains following introduction of a cpa1::ADE2 disruption allele ( Figure 1A ) into the ade2 strain M049 , we observed that ∼25% of Ade+ transformants were resistant to CsA ., The cpa1::ADE2 disruption allele is composed of the C . neoformans ADE2 gene with 5′ and 3′ UTRs inserted into a full length cpa1 gene at an internal BalI restriction site ( Figure 1A ) ., The Ade+ transformant strain PPW23 remained sensitive to CsA whereas strains PPW22 and PPW25 exhibited various degrees of CsA resistance with PPW22 being the most CsA resistant and viable on YPD medium containing 100 µg/ml CsA at 37°C ( Figure 1B ) ., CsA resistance exhibited by some of the transformants was unstable and these colonies were variegated , forming white and red colonies/sectors indicative of repression or loss of the ADE2 marker gene ( Figure 1C ) ., PCR analyses indicated that all of these isolates carried the wild-type length CPA1 and CPA2 genes ( Figure 2A ) ., Furthermore , we sequenced the CPA1 and CPA2 genes from the transformed strains , and found no mutations ., Thus , in no case was the CsA resistance phenotype attributable to the disruption or mutation of CPA1 or CPA2 ., Based on PCR analyses ( Figure 2A ) , the PPW22 and PPW23 isolates contain an intact cpa1::ADE2 transgene allele , as well as the wild type CPA1 gene ., Isolate PPW25 also carries the transgene , but PCR amplification failed with a primer matched to the 5′ region of the CPA1 gene , suggesting that mutations or rearrangements affecting this region occurred during integration ( Figure 2A ) ., Taken together , we propose that silencing of the endogenous CPA1 and CPA2 genes is triggered by introduction of ectopic transgene ( s ) that share significant homology with both genes ., The Ade+ transformant strains PPW22 , PPW23 , and PPW25 varied in their ability to grow on YPD medium with CsA at 37°C , even though all contain the cpa1::ADE2 transgene allele ., We explored the hypothesis that the silencing frequency of CPA1 and CPA2 is affected by the location/copy number/arrangement of the transgenic alleles present in the genome ., In addition to PPW22 , PPW23 , and PPW25 , we analyzed four other transformed strains containing the cpa1::ADE2 transgene to test our hypothesis ., Southern analyses with probes to CPA1 and ADE2 were conducted to compare the genomic structures at the cpa1::ADE2 locus in these transformed strains ., As shown in Figure 2B , the endogenous CPA1 gene hybridized to a CPA1 specific probe in PPW22 , PPW23 , PPW25 , PPW26 , and PPW27 but not in the cpa1 mutant strains PPW51 , PPW52 , and PPW75 ., Furthermore , similar Southern patterns corresponding to the transgene were observed among PPW22 , PPW23 , PPW26 , PPW27 and the cpa1 mutant ., PPW25 exhibited a different transgene hybridization pattern , consistent with the PCR result noted above ( Figure 2A ) , indicating that mutations or gene rearrangement occurred during integration of the cpa1::ADE2 transgene allele in this isolate ., The signals derived from the ectopically integrated cpa1::ADE2 alleles are 1 . 25 kb ( ADE2 probe ) and 1 . 9 kb ( both ADE2 and CPA1 probes ) , which were more intense in PPW22 , PPW26 , and PPW27 compared with the signals from the endogenous CPA1 or ADE2 genes , indicating that multiple copies of the transgenes have integrated into these genomes ( Figure 2B ) ., We further determined the copy number of the cpa1::ADE2 transgene by quantitative PCR analyses ., As shown in Figure 2C , PPW22 , the isolate exhibiting a high degree of CsA resistance , carries ∼25 copies of the transgene , while only three copies were found in PPW25 and one copy in PPW23 ( Figure 2C ) ., The transgenes in PPW22 are stably inherited during mitotic growth as shown in Figure S1 ., In addition , PPW26 , PPW27 , and the cpa1 mutant strains PPW51 and PPW52 all contain various copy numbers of the transgene: 35 copies in PPW26 , 65 copies in PPW27 , 7 copies in PPW51 , and 10 copies in PPW52 , respectively ., We also performed Southern analysis with an ADE2 probe to explore the possible distribution of the transgenes in the genome ., As shown in Figure 3A , an ∼4 . 2 kb band with high intensity was observed in PPW22 , PPW26 , PPW27 , and PPW52 , suggesting that the transgenes are likely arranged as tandem repeats in these strains ( Figure 3A and 3B ) ., For PPW51 , although it carries a 7-copy transgene , it did not yield an intense hybridization band at 4 . 2 kb , indicating a different transgene arrangement ., To define the location of the transgene , we conducted pulsed-field gel electrophoresis and chromosomal Southern blots with probes directed against the ADE2 and CPA1 genes ., The chromosomes in the ade2 , PPW22 , PPW23 , PPW25 , and PPW26 lanes that hybridized to the CPA1 probe were the same ones that hybridized to the rDNA probe on a duplicate filter , indicating that the transgenes in PPW22 , PPW23 , PPW25 , and PPW26 are located on chromosome 2 , on which the rDNA gene cluster and the CPA1/CPA2 genes reside 26 ( Figure 3C ) ., This interpretation is further supported by the finding that hybridization with an ADE2-specific probe revealed that the location of the cpa1::ADE2 transgene in the four transformants is on chromosome 2 ., PPW27 is distinct from other transformed strains in that the transgenes are located on a chromosome smaller than chromosome 2 ., We then explored the correlation between the silencing efficiency and the transgene copy number by measuring spontaneous CsA resistance after strains bearing different copies of the transgene were grown on rich ( YPD ) medium ., With ∼65 copies of the transgene , PPW27 exhibited the highest silencing rate ( ∼95% ) ., PPW22 ( ∼25 copies ) and PPW26 ( ∼35 copies ) showed a 90% silencing frequency , whereas the frequency was reduced to ∼0 . 1% in PPW25 , in which three transgenes were present ., No CsA resistant colonies were observed from PPW23 , which contains a single transgene ( Table 1 ) ., Interestingly , in the cpa1 mutant strains PPW51 and PPW52 that contain similar copy numbers of the transgene ( 7 vs . 10 ) , they exhibited very different silencing efficiency: 0 . 002% in PPW51 and 1% in PPW52 ., Taken together , our results showed a correlation between the copy number of the transgene and the intensity of co-suppression; however , this correlation may not be strict , as indicated by the low silencing rate in PPW51 ., In this strain , other factors such as the arrangement of the transgene may also affect silencing frequency ., We also found that the high copy number of the transgene did not always lead to a high silencing frequency of the transgenic ADE2 genes ., For instance , we observed ∼10% red ade− colonies derived from PPW22 and PPW26 ( Figure 1C and Figure S2 ) ., Only PPW27 exhibited a high silencing rate for the ADE2 gene ., In addition , quantitative real-time RT-PCR showed that the ADE2 expression levels were much higher in PPW22 compared with those in the wild-type strain , while the levels of CPA1 were virtually undetectable in PPW22 , indicating that silencing of the endogenous CPA1 gene and the transgenic ADE2 gene might be two distinct events ( Figure S3 ) ., In strains PPW25 , cpa1 cpa2 , and AAC1 ( gpa1::ADE2 ) 27 , which contain different copy numbers of the ADE2-based transgene , expression levels of the ADE2 gene were also high and in accord with the numbers of the ADE2 transgenes , suggesting lower silencing rates of the transgenic ADE2 genes ( Figure S3 ) ., Our results suggest that ADE2 is less sensitive to silencing , and higher copy numbers or a specific transgene location may be required for its efficient silencing when compared with silencing of CPA1/2 ., The observation that strain PPW22 was completely CsA resistant albeit encoding wild-type copies of the CPA1 and CPA2 genes indicates that introduction of the cpa1::ADE2 transgene allele might result in silencing of both the CPA1 and CPA2 genes ., To test this hypothesis , a northern blot was performed with a probe that hybridizes to both the CPA1 and CPA2 genes , which are highly conserved and share ∼93% overall nucleotide identity in the coding regions , including regions of 100% identity spanning two regions of 26 and 339 nucleotides ( Figure 4A ) ., Similar to the cpa1 cpa2 double mutant strain , little or no CPA1 or CPA2 mRNA was detected in the PPW22 silenced strain ., By contrast , much more CPA1/2 mRNA accumulated in strains PPW23 and PPW25 with lower copy number transgene arrays ( Figure 4B ) ., Furthermore , western blotting using antiserum raised against Cpa1 and that cross-reacts with Cpa2 25 documents that the Cpa1 and Cpa2 proteins are both considerably reduced in abundance in PPW22 ( Figure 4C ) ., In the fungus N . crassa , multiple repeated transgenes often induce silencing through methylation , and the methylation inhibitor 5-azacytidine can increase the reversion frequency of some silenced strains 28 ., Our observations suggest that a similar process does not operate in C . neoformans ., First , growth of PPW22 or PPW25 cells on medium containing 5-azacytidine did not restore CsA sensitivity ., Second , analysis of Southern blots on DNA digested with the methylation-dependent DNA endonuclease McrBC did not provide any evidence that methylated DNA was present ( data not shown ) ., We then tested whether asexual co-suppression is dependent on the RNAi silencing pathway , similar to quelling in N . crassa that occurs during vegetative growth 29 ., Because generation of siRNA is a hallmark of RNAi pathways , we examined the presence of siRNA by northern blotting ., When probed with a 32P-labeled sense CPA1 transcript , abundant siRNA of ∼22 nt were observed in PPW22 ( 25 transgenes ) , whereas only a very modest level was detected in PPW25 ( 3 transgenes ) , and none occurred in PPW23 ( 1 transgene ) or the ade2 strain M049 ( Figure 4D ) ., The observation of antisense siRNAs of CPA1 suggests that siRNAs are derived from a dsRNA precursor and may function as a trans-acting factor to silence both CPA1 and CPA2 , by virtue of the high sequence identity CPA2 shares with CPA1 ., In a duplicate blot probed with a 32P-labeled sense ADE2 probe , no siRNA was detected specific for ADE2 ., Hybridization signals to ADE2 siRNAs were only observed in samples of RNA extracted from red ( ade− ) colonies produced by variegation of strains PPW22 and PPW25 from white ( Ade+ ) to red ( ade− ) ( Figure 4E ) ., Overall , these findings indicate that silencing of the endogenous CPA1/CPA2 and transgenic ADE2 genes both involve an RNAi pathway and the two silencing events appear to occur at different frequencies ., Previous studies have shown that the C . neoformans serotype A strain H99 genome encodes one Argonaute ( Ago1 ) , two Dicers ( Dcr1 and Dcr2 ) , and one RNA dependent RNA polymerase ( Rdp1 ) 24 , 30 ., To further verify that the RNAi machinery is required for silencing , we deleted these components in the PPW22 strain ., For each gene disruption , at least two independent deletion mutants were obtained and analyzed ( Table S1 and Figure S4 ) ., As shown in Figure 5A , independently isolated ago1Δ , rdp1Δ , and dcr2Δ mutations completely reversed the CsA resistance phenotype to CsA sensitivity ., Accordingly , the expression levels of the CPA1/CPA2 genes and the Cpa1 and Cpa2 protein levels were restored to the WT levels in these RNAi mutant strains ( Figure 5B and 5C ) ., Additionally , ago1Δ , rdp1Δ , and dcr2Δ mutations blocked the generation of small RNAs corresponding to CPA1/2 in PPW22 ( Figure 5D ) ., Deletion of DCR1 had little or no influence on the CsA phenotype and the dcr1Δ mutant still contained a similar amount of siRNA as in the DCR1 wild-type PPW22 , suggesting that Dcr1 plays a minor role in the asexual co-suppression RNAi pathway compared with Dcr2 ., This is in agreement with our previous findings that Dcr2 plays the major role and Dcr1 a minor role in transgene-mediated sex-induced silencing 24 ., Our finding that multiple copies of transgenes induce silencing of endogenous genes during vegetative growth of C . neoformans is similar to quelling in N . crassa 29 ., The molecular mechanism of quelling has been extensively studied ., It has been proposed that the production of transgene-specific aberrant RNA transcripts is the critical initial step in generating dsRNA precursors during an RNAi mediated pathway 29 ., Recent studies have shown that an RNA-dependent RNA polymerase ( RdRP ) and a single-stranded DNA binding protein complex ( RPA ) play important roles in the production of aberrant RNA and the subsequent generation of dsRNA 31 , 32 ., RPA is a conserved eukaryotic heterotrimeric complex critical for DNA replication and repair 33 , 34 ., In mammals , it is composed of three subunits , Rpa70 , Rpa32 , and Rpa14 , named according to their respective molecular masses 34 ., The observation that, 1 ) Rpa70 interacts with RdRP 32 and, 2 ) the RPA complex is required for quelling in N . crassa 31 prompted us to test if RPA is involved in transgene-induced asexual co-suppression in C . neoformans ., By analyses of the Cryptococcus genome database , we identified orthologs of RPA70 ( CNAG_01144 . 2 ) and RPA32 ( CNAG_01316 . 2 ) in C . neoformans as reciprocal best BLAST hits with the N . crassa RPA genes ., We have been unable to identify an RPA14 ortholog thus far ., We first sought to generate rpa deletion mutations in the PPW22 transgenic strain to examine if the RPA complex functions in silencing ., No progeny bearing an rpa70Δ mutant allele were obtained after sporulation of two independent RPA70/rpa70Δ diploid strains ( Figure S5 ) , and thus RPA70 is essential ( see Figure S6 ) ., We therefore constructed a “Decreased Abundance by mRNA Perturbation” ( DAmP ) allele of rpa70 in strain PPW22 to reduce the mRNA expression levels of RPA70 ., An rpa32Δ haploid mutant was found to be viable ., As shown in Figure 6A and 6B , independently isolated mutants bearing rpa70-DAmP or rpa32Δ mutations significantly reduced or abolished the CsA resistance silencing phenotype and restored the mRNA levels of CPA1 and CPA2 ., In addition , siRNAs corresponding to the CPA1/CPA2 genes were undetectable in the rpa70-DAmP and rpa32Δ mutants ( Figure 6C ) ., These results confirm that RPA is an important factor in the vegetative transgene silencing pathway ., In addition , sexual development was defective during rpa32Δ×rpa32Δ bilateral mating: less hyphae formed , fewer basidia were produced , and no spore chains were observed ( Figure S7 ) ., This sporulation defect makes it more difficult to assess whether RPA also plays a role during SIS ., The RNAi machinery is evolutionary conserved in a wide variety of fungal species 35 ., Quelling , which was identified in N . crassa as the first RNAi silencing phenomenon reported in the fungal kingdom , is induced by transformation with transgenes that are homologous to an endogenous target 6 , 7 ., Even transgenes containing only fragments of genes led to inactivation of the corresponding native genes ., Although RNAi silencing has been discovered in C . neoformans 30 , 36 , no similar robust transgene-induced silencing analogous to the canonical process of quelling had been reported ., Hitherto , efficient and stable RNAi silencing during C . neoformans vegetative growth has always been achieved by introducing inducible constructs for dsRNA expression , including a hairpin RNA-expressing plasmid and an opposing-dual promoter system 36 , 37 ., We have previously reported an RNAi-related repeat silencing mechanism triggered by tandem multicopy transgenes in C . neoformans , whereas the silencing is most effective ( ∼50% ) during sexual development and occurs at a very low frequency ( ∼0 . 1% ) during vegetative growth 24 ., SIS-RNAi silencing is important in defending the genome from transposons , as is quelling in N . crassa ., Several lines of evidence support that RNAi silencing in C . neoformans also plays a role in transposon suppression during vegetative growth ., First , our deep sequence siRNA library obtained from a mitotically growing strain reveals that abundant RNA sequences share homology with repetitive transposable elements 24; second , a number of repetitive transposons are highly expressed or active in an rdp1 mutant strain based on a comparative transcriptome analysis 24 , 30 ., However , the silencing rate during vegetative growth ( 0 . 1% ) we observed previously is much lower than SIS ( 50% ) when silencing URA5 transgene arrays ., Thus , the question here is whether there is an effective transgene-induced silencing during vegetative growth in Cryptococcus , like quelling in N . crassa ., This study provides evidence that asexual co-suppression can be highly effective: transformed strains containing high copy numbers of the cpa1::ADE2 transgene exhibited ∼90% silencing of the endogenous CPA1 and CPA2 genes ., In this scenario , we hypothesize that RNAi may effectively suppress some highly repetitive transposons during vegetative growth , and SIS , a higher efficiency silencing mechanism deployed during the sexual cycle , prevents progeny from suffering assault by an even wider range of transposons and would otherwise occur when the two genomes of the mating parents are brought into contact with each other ., We found that individual transformed strains bearing different copy numbers of cpa1::ADE2 displayed variable efficiency of RNA interference: the silencing rate of the CPA1 and CPA2 genes reached ∼90% in the presence of 25 copies of the transgene while much lower or no silencing occurred in the transformants containing three- or a single-copy transgenes ., The correlation between transgene copy number and silencing efficiency has also been observed in SIS 24 ., However , in both cases , this correlation is not strict , as we have observed in strain PPW51 with multiple copies of transgenes a low silencing rate and SIS is saturated at a 50% with three or more copies of a transgene ., We hypothesize that in addition to the copy number , the transgene arrangement and location may also contribute to the silencing frequency ., The presence of a high copy number of transgenes has been frequently found to be a prerequisite to trigger effective silencing ., Two explanations have been proposed in plants ., First , a high copy number of transgenes is necessary because transgenic mRNA accumulation needs to reach a threshold to induce specific degradation of both endogenous and transgenic mRNA 38 ., However , evidence has been found in contradiction with this argument ., For example , weakly or negligibly transcribed transgenes can efficiently induce PTGS 3 and high expression levels of a transgene are not sufficient to trigger gene silencing in N . crassa 7 ., Thus , we favor an alternative hypothesis proposed by English et al . and Cogoni et al . 6 , 39 that a qualitatively aberrant feature of transgenic DNA or RNA ( aDNA or aRNA ) , rather than a high level of transgenic RNA accumulation , can trigger gene silencing ., One plausible model is that RNA expressed from one repeat strand invades the DNA template of a neighboring repeat to form an aberrant RNA-DNA hybrid ., This process could be facilitated by DNA/RNA helicases , RPA single strand DNA binding proteins , and both RdRP and DdRP activities ., However , after silencing is initiated , the silencing efficiency is affected by additional factors , including growth conditions ( vegetative vs . meiotic growth ) , and target gene expression levels ( discussed more below ) ., We also noted that silencing of transgenic ADE2 genes occurred at a much lower frequency compared with silencing of the endogenous CPA1/2 genes in some transformed strains ( PPW22 and PPW26 ) , although both were introduced by the same transgene array ., Robust silencing of both CPA1/2 and ADE2 was observed only in PPW27 , with the presence of ∼65 copies of the cpa1::ADE2 transgene at a different location ., We hypothesize that ADE2 is less sensitive to silencing , and inactivation of ADE2 may require either higher copy numbers or a stricter location of the transgene ., A similar case has also been observed when an interference plasmid incorporating portions of CAP59 and ADE2 between the opposing promoters was introduced to simultaneously silence both genes; however , colonies exhibiting only one mutant phenotype were found 36 ., Based on these observations , we hypothesize that several factors may determine such variation in frequency ., The first is that various mRNA expression levels between the target genes may dictate the silencing frequency ., To observe a given mutant phenotype , for example CsA resistance compared to adenine auxotrophy , the mRNA must drop below a threshold that may differ for each target depending on its mRNA transcript levels and mechanism of action and regulation ., This hypothesis can also explain our previous finding that silencing of transgenic URA5 is extremely low during mitosis , but centromeric repetitive transposons that are barely expressed can be silenced by RNAi 24 ., The other reasons for varying silencing efficiency could be the intrinsic characteristics of the genes determined by their base compositions , genome locations , and sequence context ., The genes that are prone to form secondary structures could be more resistant to RNase activity , or the position of a gene locus within the nucleus may affect the efficiency of homology pairing ., A central question related to transgene silencing is: how do cells produce repetitive transgene-specific dsRNA to initiate RNAi ?, As discussed above , our observations support that it is not just the accumulation of transgenic RNA but rather the large tandem repetitive transgenic DNA itself that can be recognized and initiate aberrant RNA ( aRNA ) and dsRNA synthesis ., With regard to quelling in N . crassa , dsRNA is generated by QDE-1 using the single strand repetitive DNA as the template because QDE-1 can act as both a DNA-dependent RNA polymerase ( DdRP ) and as an RNA-dependent RNA polymerase 31 , 40 ., More importantly , it is the RPA complex that recruits QDE-1 to the transgenic DNA repeats and promotes dsRNA formation 31 ., Interestingly , we found that two components in the RPA complex , Rpa70 and Rpa32 , are required for asexual co-suppression , suggesting that RPA may play a conserved role during transgene silencing in fungi ., It will be of further interest to investigate how similar asexual co-suppression and quelling are at the molecular level , including examining whether Rdp1 in C . neoformans interacts with the RPA complex and whether it can function both as an RdRP and as a DdRP ., Alternatively , a novel DdRP may remain to be identified that is responsible for the recognition of aberrant DNA repeats and to generate the initial dsRNA in C . neoformans ., The C . neoformans MATα strain H99 has been previously described 41 , 42 ., M049 is an ade2 strain derived from H99 following gamma irradiation , involving a chromosomal translocation event within the ADE2 locus 41 , 42 ., All other strains used in this study are listed in Table S1 ., Yeast cells were grown and maintained on yeast extract-peptone-dextrose ( YPD ) media ., Synthetic dextrose ( SD ) medium lacking adenine and YPD medium containing CsA ( 100 µg/ml ) were used to test whether isolates of interest are auxotrophic for adenine or resistant to cyclosporine A . Mating of C . neoformans was conducted on 5% V8 juice agar medium ( pH\u200a=\u200a5 ) or Murashige and Skoog ( MS ) medium minus sucrose ( Sigma-Aldrich ) , as previously reported 43 ., A standard overlap PCR approach was used to disrupt genes of interest 44 ., All mutations were generated in the serotype A background with the dominant NAT or NEO selectable markers ., The overlap PCR products were introduced into the genome of recipient strains by biolistic transformation 45 ., When constructing the DAmP mutant alleles , the NEO selectable marker was inserted immediately after the stop codon of the RPA70 open reading frame by transformation with an overlap PCR product encoding the NEO marker flanked at each end with homology to the targeted locus 46 , 47 ., Primers that were used for amplification of the 5′ and 3′ flanking regions of each gene disruption cassette are listed in Table S2 ., Transformants were initially screened by PCR and Southern blot analyses were then conducted to identify a single integration at the desired locus ., Quantitative real-time PCR assays were performed with primers specific to the actin gene ACT1 and the ADE2 and CPA1 genes ( Table S2 ) to determine the copy numbers of the cpa1::ADE2 transgene ., DNA of the wild-type strain H99 and transformed strains PPW22 , PPW23 , PPW25 , PPW26 , PPW27 , PPW51 , PPW52 , and PPW75 that contain the cpa1::ADE2 transgene were used as templates ., PCR was conducted with Brilliant SYBR Green QPCR Master Mix ( STRATAGENE ) and the relative quantity of the ADE2 or CPA1 gene determined by the ΔΔCt method according to the following equations: ( Ct value means the threshold cycle ) ΔCt\u200a=\u200aCt ( target ) - Ct ( normalize ) and ΔΔCt\u200a=\u200aΔCt ( experimental ) - ΔCt ( control ) Comparative expression level\u200a=\u200a2−ΔΔCt ., The PFGE was carried out using a BioRad CHEF-DR II System ., The plugs for the CHEF gel electrophoresis were prepared as previously described 42 ., To separate chromosomal DNA , plugs were embedded

Introduction, Results, Discussion, Materials and Methods

Introduction of DNA sequences into the genome often results in homology-dependent gene silencing in organisms as diverse as plants , fungi , flies , nematodes , and mammals ., We previously showed in Cryptococcus neoformans that a repeat transgene array can induce gene silencing at a high frequency during mating ( ∼50% ) , but at a much lower frequency during vegetative growth ( ∼0 . 2% ) ., Here we report a robust asexual co-suppression phenomenon triggered by the introduction of a cpa1::ADE2 transgene ., Multiple copies of the cpa1::ADE2 transgene were ectopically integrated into the genome , leading to silencing of the endogenous CPA1 and CPA2 genes encoding the cyclosporine A target protein cyclophilin A . Given that CPA1-derived antisense siRNAs were detected in the silenced isolates , and that RNAi components ( Rdp1 , Ago1 , and Dcr2 ) are required for silencing , we hypothesize that an RNAi pathway is involved , in which siRNAs function as trans factors to silence both the CPA1 and the CPA2 genes ., The silencing efficiency of the CPA1 and CPA2 genes is correlated with the transgene copy number and reached ∼90% in the presence of >25 copies of the transgene ., We term this transgene silencing phenomenon asexual co-suppression to distinguish it from the related sex-induced silencing ( SIS ) process ., We further show that replication protein A ( RPA ) , a single-stranded DNA binding complex , is required for transgene silencing , suggesting that RPA might play a similar role in aberrant RNA production as observed for quelling in Neurospora crassa ., Interestingly , we also observed that silencing of the ADE2 gene occurred at a much lower frequency than the CPA1/2 genes even though it is present in the same transgene array , suggesting that factors in addition to copy number influence silencing ., Taken together , our results illustrate that a transgene induced co-suppression process operates during C . neoformans vegetative growth that shares mechanistic features with quelling .

The development of gene transfer methods allows the production of transgenic lines in myriad eukaryotes ., Frequently , transgenic DNA is integrated into the genome and transmitted as a heritable Mendelian trait ., However , the introduced transgenes are in some cases not expressed ( silenced ) ., In addition , transgenes can also provoke silencing of endogenous genes with which they share sequence homology ., This phenomenon was first observed in plants and named co-suppression ., In fungi the best-documented co-suppression phenomenon occurs in vegetative tissue of the filamentous fungus Neurospora crassa and is termed quelling ., Here we report a robust asexual co-suppression pathway that operates in the pathogenic fungus Cryptococcus neoformans and shares molecular components with quelling ., Compared with the sex induced silencing ( SIS ) phenomenon previously discovered in C . neoformans , which efficiently silences genes during mating ( ∼50% ) but not during vegetative growth ( ∼0 . 2% ) , asexual co-suppression operates efficiently during vegetative growth to suppress transgene expression and may also silence transposons and other repetitive sequences .

mycology, fungi, functional genomics, rna interference, gene expression, genetics, microbial pathogens, epigenetics, biology, genomics, microbiology, genetics and genomics, gene function

journal.pgen.1004794

RNA Processing Factors Swd2.2 and Sen1 Antagonize RNA Pol III-Dependent Transcription and the Localization of Condensin at Pol III Genes

Mitotic chromosome condensation is essential for genome integrity ., When defective , chromosomes often remain entangled and fail to segregate properly in anaphase ., A key driver of chromosome condensation is the highly conserved condensin complex ., Condensin is made of five sub-units ( SMC2Cut14 , SMC4Cut3 , CAP-D2Cnd1 , CAP-GCnd3 and CAP-HCnd2 , name of the human protein followed by its name in fission yeast ) and it is one of the main components of mitotic chromosomes 1 ., In vitro , purified condensin can introduce positive supercoils into a relaxed plasmid in the presence of topoisomerase I 2 , 3 ., These observations support the idea that condensin shapes mitotic chromosomes by changing the topology of chromatin around its binding sites ., However , the mechanisms underlying the association of condensin with chromatin remain poorly understood ( reviewed in 4 ) ., Several studies have illustrated the paradoxical relationships linking gene transcription and the localization of condensin ., From pro- to eukaryotes , condensin is preferentially enriched at highly transcribed genes 5 , 6 , 7 , 8 , suggesting that some highly conserved transcription-associated feature ( s ) that predate ( s ) the appearance of nucleosomes help to recruit condensin ., However , experiments in yeast indicated that RNA polymerases must be silenced before condensin can bind , at least at repetitive sequences such as the rDNA or the sub-telomeres 9 , 10 ., These somewhat contradictory observations could potentially be reconciled if one hypothesizes that a by-product of the transcription process facilitates the recruitment of condensin ., In this study , we have considered that such a by-product could be R-Loops or transcription-associated topological stress ., R-Loops result from the formation of stable DNA:RNA hybrids in the genome ., As a consequence of the hybridization of the RNA to the template , the non-transcribed strand of the DNA remains single-stranded ( reviewed in 11 ) ., Interestingly , the hinge domain of the Smc2/Smc4 heterodimer in condensin shows high affinity in vitro for single-stranded DNA 12 , 13 ., Moreover , a recent study proposed that chromatin is less accessible to restriction enzymes in mutants where R-Loops accumulate , consistent with the idea that R-Loop formation favours chromatin compaction 14 ., Interestingly , fission yeast condensin can disassemble DNA:RNA hybrids in vitro 15 and its chicken counterpart localizes to CpG islands 6 , which constitute major R-Loop forming regions in the genome 16 ., Taken together , these observations support the idea that R-Loops and condensin could interact functionally in vivo 14 ., According to the twin supercoiled domain model , high rates of transcription induce positive supercoiling of the chromatin in front of the elongating polymerase , whilst negative supercoiling accumulate upstream of the polymerase 17 ., As such , highly expressed genes represent regions of the genomes that accumulate topological stress ., As confirmed in vivo recently , this stress is monitored by topoisomerase I and topoisomerase II 18 , 19 , 20 ., Interestingly , in vitro assays have indicated that condensin binds preferentially to positively supercoiled plasmids in the presence of ATP 21 ., Whether or not this transcription-associated topological stress contributes to the binding of condensin in vivo has not been addressed ., In order to clarify the functional relationships between transcription and chromosome condensation , we recently carried out a genetic screen in fission yeast to identify deletions of transcription-associated factors that would rescue a condensin deficiency 22 ., For this , we isolated loss-of-function mutations that could rescue the thermo-sensitivity of the condensin mutant cut3-477 23 ., Two of the mutations we isolated were the deletions of swd2 . 2 ( swd2 . 2Δ ) and sen1 ( sen1Δ ) 22 ., Swd2 . 2 is a non-essential component of the Cleavage and Polyadenylation Factor ( CPF ) , the complex responsible for 3′end maturation of RNA Pol II transcripts in yeast ( reviewed in 24 ) , where it acts to maintain the proper levels of CPF-associated phosphatases 22 ., Fission yeast Sen1 is the homologue of human Senataxin and has been shown to unwind DNA:RNA hybrids in vitro 25 ., Budding yeast Sen1 is involved in transcription termination 26 but its role in fission yeast has not been characterized ., Here we show that both factors act directly at Pol III-transcribed genes to limit the association of condensin and the accumulation of topological stress ., Furthermore , topological stress at Pol III-transcribed genes facilitates the association of condensin when Swd2 . 2 and Sen1 are missing ., On their own , the deletions of swd2 . 2 ( swd2 . 2Δ ) and sen1 ( sen1Δ ) partly restored growth of cut3-477 cells at the restrictive temperature ( Figure 1A ) and reduced the proportion of anaphase cells displaying chromosome segregation defects ( Figure 1B ) ., Combining both deletions ( sen1Δswd2 . 2Δ ) resulted in a stronger suppressor effect ( Figure 1AB ) ., The double mutant sen1Δswd2 . 2Δ also suppressed the other condensin mutant cut14-208 ( Figure S1 ) ., Strikingly , Chromatin Immunoprecipitation ( ChIP ) analysis in cycling cell populations showed that the localization of condensin was altered at specific loci when Swd2 . 2 and Sen1 were both missing: its recruitment increased significantly at genes transcribed by RNA Pol III ( Gln . 04 , Met . 07 , Ser . 13 , Pro . 09 , Tyr . 04 , Gly . 05 , 5S rRNA , Arg . 04 on Figure 1C ) , whereas it was significantly reduced at the rDNA arrays ( 18S&Rfb2 ) ., The binding of condensin remained unaffected at kinetochores ( cnt1 ) or at highly transcribed Pol II genes ( Act1 , Adh1 , Fba1 and SPAC27E2 . 11c ) ., The sequences of all the primers used in this study are available on Table S1 ., The mitotic indexes of both cell populations ( swd2 . 2+sen1+ and swd2 . 2Δsen1Δ ) were comparable ( Figure 1D ) , ruling out that the changes in the association of condensin are due to indirect , cell-cycle defects ., These data established that Sen1 and Swd2 . 2 act to limit the localization of condensin at Pol III-transcribed genes ., The reasons why the association of condensin at the rDNA arrays is reduced in the absence of Swd2 . 2 and Sen1 will be explained elsewhere ., We found previously that Swd2 . 2 associates with Pol III-transcribed genes and that lack of Swd2 . 2 restored the localization of condensin at Pol III-transcribed genes in the condensin-deficient mutant cut3-477 22 ., Here , we show that Sen1 is also significantly enriched at Pol III-transcribed genes and that its binding is independent of Swd2 . 2 ( Figure 2A ) ., Furthermore , affinity purification of Sen1 followed by mass-spectrometry analysis of its associated proteins identified most sub-units of the RNA Pol III complex as its most stable binding partners ( Table S2 ) ., We confirmed this interaction by showing that the RNA Pol III sub-unit Rpc25 co-precipitates with Sen1 ( Figure 2B ) ., Note however that Sen1 did not co-precipitate with Sfc6 , a sub-unit of TFIIIC ( Figure S2 ) , a complex required for the association of RNA Pol III with chromatin 27 ., ChIP analysis showed that the association of Rpc25 with chromatin was significantly increased in the absence of Sen1 ( Figure S3 ) or in swd2 . 2Δsen1Δ cells ( Figure 2C&D ) ., In swd2 . 2Δsen1Δ cells , the stabilization of RNA Pol III on chromatin was associated with an increase in the steady-state level of tRNAs , as detected by RT-qPCR analysis ( Figure 2E ) ., Taken together , these experiments concur to show that Swd2 . 2 and Sen1 play a direct role at Pol III-transcribed genes , where they limit the association of RNA Pol III and the accumulation of transcripts ., These results show that the accumulation of condensin at Pol III-transcribed genes in swd2 . 2Δsen1Δ cells is concomitant with an enhanced transcriptional activity ., It was recently argued that budding yeast Sen1 limits the accumulation of DNA:RNA hybrids , including at Pol III-transcribed genes 28 ., Fission yeast Sen1 similarly was shown to display a DNA:RNA helicase activity in vitro 25 ., These observations and the additional arguments detailed in the introduction prompted us to test the possibility that R-Loops could represent a transcription by-product facilitating the association of condensin with chromatin ., We speculated that lack of Sen1 and Swd2 . 2 could result in the accumulation of R-Loops at Pol III-transcribed genes where they might contribute to increase the association of condensin ., To establish whether or not R-Loops form at Pol III-transcribed genes in fission yeast , we first monitored by ChIP the chromatin association of RNase H1 , one of the endogenous enzymes known to disassemble R-Loops ., More specifically , we introduced at the endogenous locus a point mutation ( D129N ) in the fission yeast RNase H1 ( Rnh1 ) , because the same mutation was shown to weaken the catalytic activity of human RNase H1 29 ., Consistent with this , the D129N mutation did stabilize the interaction of Rnh1 with Pol III-transcribed genes ( Figure 3A ) ., Furthermore , the interaction of Rnh1D129N with Pol III-transcribed genes was lost upon over-expression in vivo of RnhA , the RNase H1 enzyme from E . coli ( Figure 3B ) ., Upon over-expression , RnhA itself did not stably associate with Pol III-transcribed genes ( Figure S4 ) , showing that the loss of Rnh1D129N from Pol III-transcribed genes upon over-expression of RnhA cannot be explained by its mere replacement by bacterial RnhA ., Finally , Figure S5 shows that the association of Rnh1D129N with the rDNA repeats increased significantly in the absence of topoisomerase I ( top1Δ ) , consistent with the observations reported previously that lack of Top1 triggers the accumulation of R-Loops at rDNA in budding yeast 30 ., This confirmed that Rnh1D129N was able to detect significant changes in R-Loop accumulation ., Taken together , these data show that ChIP with Rnh1D129N is a reliable way to identify R-Loop forming regions in fission yeast ., We sought to confirm the formation of R-Loops at genes transcribed by RNA Pol III using another approach ., A method that is commonly used to map R-Loop forming regions in yeast is to perform ChIP using the S9 . 6 antibody because of its high affinity for DNA:RNA hybrids 31 ., ChIP requires formaldehyde cross-linking followed by sonication of the chromatin ., We found that the ability of S9 . 6 to detect R-Loops generated after transcription in vitro was greatly diminished both by formaldehyde cross-linking and by sonication ( Figure S6 ) ., We do not know at this stage whether this is because R-Loops are partly destroyed by these treatments or because these treatments reduce the affinity of the antibody for R-Loops ., To circumvent these issues , we extracted genomic DNA from unfixed cells , digested soluble RNA using RNase A and sheared the DNA using a cocktail of restriction enzymes ( see Methods ) ., Dot blot analysis using the S9 . 6 antibody confirmed that our procedure largely preserved R-Loops ( Figure 3C ) ., We then performed DNA:RNA immuno-precipitation ( DRIP ) using the S9 . 6 antibody in stringent conditions , in the presence of 500 mM NaCl ., As expected , the DRIP signal at 18S , the canonical R-Loop forming region within the rDNA repeats 30 , increased significantly in the absence of RNase H1 and RNase H2 ( rnh1Δrnh201Δ cells ) and disappeared almost entirely upon treatment of the genomic DNA with commercial RNase H ( Figure 3D ) ., On the contrary , the DRIP signal detected at a non-transcribed region NT ( chr I , 3009300-3009500 , 32 ) remained low both in rnh1Δrnh201Δ cells and upon treatment with RNase H . Those controls demonstrated that the signals we detected using DRIP were specific ., In agreement with the results obtained using ChIP of Rnh1D129N as a reporter for the presence of R-Loops , we detected strong DRIP signals at Pol III-transcribed genes in the absence of RNase H1 and RNase H2 ( Figure 3D ) ., In conclusion , the two methods we have set up to map R-Loop forming regions establish that R-Loops are a prominent feature of Pol III-transcribed genes in fission yeast ., Using ChIP of Rnh1D129N , we established that R-Loops accumulate to similar levels at Pol III-transcribed genes in cycling cells ( >90% of interphase cells ) and in cells synchronized in early mitotis ( Figure S7A ) ., Consistent with this , ChIP established that the association of RNA Pol III with chromatin is largely maintained in mitosis ( Figure S7B ) ., Taken together these experiments support the idea that transcription at Pol III-transcribed genes is maintained in mitosis , at a time when condensin is loaded on chromosomes in fission yeast ., Finally , lack of Swd2 . 2 and Sen1 resulted in a small but significant increase in the formation of R-Loops at some but not all Pol III-transcribed genes ( Figure S8 ) ., Note however that this increase could be due to the fact that Pol III transcription is stimulated in the absence of Swd2 . 2 and Sen1 ( Figure 2C&D ) ., As such , these observations therefore do not prove that Swd2 . 2 and Sen1 antagonize R-Loop formation at Pol III-transcribed genes directly ., To establish whether R-Loops at Pol III-transcribed genes could contribute to the accumulation of condensin , we prevented the formation of stable R-Loops by over-expressing RnhA ., ChIP analysis showed that over-expression of RnhA did not reduce the amount of condensin recruited at Pol III-transcribed genes in swd2 . 2Δsen1Δ cells ( Figure 3E ) or in wild-type mitotic cells ( Figure S9 ) ., These data concur to demonstrate that stable , long-lived R-Loops play little or no part in recruiting condensin ., Note that over-expression of RnhA did not interfere either with the association of RNA Pol III ( Figure S10A ) or Sen1 ( Figure S10B ) ., Because Xenopus condensin shows greater affinity in vitro for positively supercoiled DNA 21 , we speculated that the cue facilitating the accumulation of condensin at Pol III-transcribed genes in the absence of Swd2 . 2 and Sen1 could be local topological constraints ., Consistent with an increase in topological stress in swd2 . 2Δsen1Δ cells , ChIP analysis detected strong accumulation of topoisomerase I ( Top1 ) at most loci ( Figure 4A ) , although the protein levels of Top1 remained unaffected ( Figure 4B ) ., We also detected enhanced accumulation of topoisomerase II ( Top2 ) , mostly at Pol III-transcribed genes ( Figure 4C ) , when the protein levels of Top2 remained unaffected ( Figure 4D ) ., Transcription-associated topological stress was recently shown to destabilize nucleosomes 19 ., At some but not all Pol III-transcribed genes that we tested , we detected a significant reduction in the recruitment of histone H3 ( Figure 4E ) in swd2 . 2Δsen1Δ cells , which is consistent with the local depletion of nucleosomes ., The concomitant accumulation of Top1 and Top2 and the depletion of nucleosomes suggest that topological stress is greater at Pol III-transcribed genes in swd2 . 2Δsen1Δ cells ., We speculate that the increased transcription of Pol III-transcribed genes in swd2 . 2Δsen1Δ cells could contribute at least in part to this enhanced topological stress ., As R-Loops unwind the DNA , it was possible that the abundance of R-Loops formed at Pol III-transcribed genes ( Figure 3 ) could contribute to this topological stress ., To test this possibility , we monitored by ChIP the localization of Top2 upon over-expression of RnhA ., Surprisingly , the localization of Top2 was not altered at Pol III-transcribed genes upon over-expression of RnhA , whilst it was reduced at the Pol I-transcribed 18S ( Figure S11 ) ., This suggested that the impact of R-Loop formation on the surrounding chromatin depends on where in the genome R-Loops form ., Based on these results , we envisaged two possible models to explain the increased localization of condensin at Pol III-transcribed genes in the absence of Swd2 . 2 and Sen1: either the accumulation of Top1 and/or Top2 helps to recruit and/or stabilize condensin , or topological stress facilitates the association of condensin at Pol III-transcribed genes ., We previously identified the deletion of Top1 ( top1Δ ) as a suppressor of cut3-477 22 , suggesting that the accumulation of Top1 that results from lack of Swd2 . 2 and Sen1 is unlikely to facilitate the association of condensin with chromatin ., Figures 5A&B show that the triple deletion swd2 . 2Δsen1Δtop1Δ was a better suppressor of cut3-477 than the double deletion swd2 . 2Δsen1Δ ., This genetic evidence suggested that failure to monitor topological stress in top1Δ cells might facilitate the association/function of condensin ., In support of this , ChIP analysis showed that there was a small but significant increase in the association of condensin at most Pol III-transcribed genes in cells deleted for Swd2 . 2 , Sen1 and Top1 ( swd2 . 2Δsen1Δtop1Δ cells ) ( Figure 5C ) ., Taken together , these data support the following model: the absence of Swd2 . 2 and Sen1 increases the transcriptional activity at Pol III-transcribed genes and this might contribute to enhance local topological constraints ., These constraints , either directly or indirectly , contribute to recruit or maintain condensin at Pol III-transcribed genes ( Figure 5D ) ., To establish whether topological stress was sufficient to stimulate the association of condensin with chromatin , we monitored the association of condensin in the temperature-sensitive Top2 mutant top2-191 ( 33 ) at the semi-restrictive temperature of 28°C ., This analysis showed that the association of condensin was not significantly disrupted in these conditions ( Figure S12A ) ., Similarly , lack of Top1 on its own did not significantly impact the association of condensin ( Figure S12B ) ., Taken together , these observations suggest that topological stress on its own is not sufficient to stimulate the association of condensin with chromatin ., In order to explain that condensin localizes to highly expressed genes from pro- to eukaryotes , whatever the RNA polymerase involved , we first hypothesized that a transcription by-product could facilitate the association of condensin with chromatin ( see Introduction ) ., We speculated that this mechanism could represent the ancestral way of recruiting condensin to chromatin ., Complementary cis-acting factors would then have evolved to stabilize the interaction of condensin with specific loci , as shown previously ( reviewed in 4 ) ., In this study we specifically considered two transcription by-products as potential condensin-attracting features: R-Loop formation and transcription-associated topological stress ., Both features have been described both in pro- and eukaryotes and they generate structures ( single-stranded DNA and positive supercoiling ) for which condensin has been shown to display high affinity in vitro ., Our data are not consistent with the idea that stable R-Loops could be involved in recruiting condensin ., Similarly , topological stress on its own was not sufficient to disrupt the localization pattern of condensin ., However , our data show that topological stress facilitated the association of condensin at Pol III-transcribed genes when Swd2 . 2 and Sen1 were missing ., These observations are consistent with the recent demonstration that supercoiling at highly expressed genes contributes to the establishment of topological domains and small-range chromosome compaction in Caulaboacter crescentus 34 ., How could topological stress create a better binding site for condensin at Pol III-transcribed genes in the absence of Swd2 . 2 and Sen1 ?, First , condensin might simply have a higher affinity for supercoiled chromatin , as suggested by the observation that condensin associates preferentially in vitro with positively supercoiled plasmids 21 ., Alternatively , or in addition , topological stress might work by facilitating nucleosome eviction 19 ., Consistent with the latter , budding yeast condensin associates preferentially with nucleosome-free regions , especially at Pol III-transcribed genes 12 ., To explain that lack of Top1 only facilitates the association of condensin at Pol III-transcribed genes when Swd2 . 2 and Sen1 are missing , we speculate that the level of topological stress has to go over a certain threshold in order to attract/stabilize condensin ., This threshold would be reached in the chromatin around Pol III-transcribed genes when Swd2 . 2 and Sen1 are missing but not when Top1 only is missing ., The biology of R-Loops is a rapidly expanding field of investigation , and many observations now demonstrate that R-Loops control genome stability and gene expression in multiple ways ( reviewed in 35 ) ., It is therefore essential to establish reliable methods to map R-Loop forming regions in genetically tractable organisms such as yeast to address the many functions of R-Loops in vivo ., We presented evidence that the commonly used S9 . 6 ChIP method to map R-Loop forming regions in yeast is challenged by the fact that R-Loops , or at least their recognition by the S9 . 6 antibody , are partly sensitive to formaldehyde cross-linking and sonication ., To circumvent this problem , we have developed two reliable alternatives to map R-Loop forming regions in fission yeast ., Both of our methods concur to demonstrate that RNA-Pol III transcribed genes are major R-Loop forming regions in fission yeast ., R-Loops have also been detected at Pol III-transcribed genes in budding yeast ( 28 ) , suggesting that R-Loop formation is a conserved feature of Pol III transcription , at least in yeast ., We would like to argue that the two methods we have set up are complementary: not only do they map R-Loop forming regions but their use in parallel can also give information regarding the stability of R-Loops formed at different loci ., Our data show that RNase H1 is most abundant at Pol III-transcribed genes throughout the cell-cycle , suggesting that R-Loops are constantly formed and detected by RNase H1 there ., Our data also show that over-expression of RnhA in vivo counter-acts R-Loop formation more efficiently at Pol III-transcribed genes than within the rDNA for example ( 18S , Figure 3B ) ., On the contrary , DRIP only yields significant signals at Pol III-transcribed genes when RNase H1 and RNase H2 are missing ( rnh1Δrnh201Δ cells ) , whilst the DRIP signals at the rDNA ( 18S ) are significant in wild-type cells , when RNase H1 and RNase H2 are fully active ., At Pol III-transcribed genes , DRIP signals increase 10-20 fold in rnh1Δrnh201Δ cells , whilst they only increase ∼3-fold at the rDNA ( 18S ) ., Our interpretation of these data is that R-Loops formed at 18S are stable and a relatively poor substrate for RNase H1 , whilst R-Loops formed at Pol III-transcribed genes are unstable and a good substrate for RNase H1 ., A corollary to these observations is that DRIP is probably better suited to detect long-lived , stable R-Loops ., This might explain why DRIP did not detect significant R-Loop formation at Pol III-transcribed genes in human cells ( 16 , 36 ) ., We conclude that using both R-Loop mapping methods in parallel could provide indications of the relative stability of R-Loops at different loci ., The reasons why R-Loops formed at Pol III-transcribed genes are labile are still unclear but we speculate that R-Loops formed at Pol III-transcribed genes might be smaller than those formed at the 18S because the Pol III transcription units are much smaller ., Further studies will be required to understand the consequences of R-Loop formation at Pol III-transcribed genes and how the half-life of an R-Loop might influence its function ., R-Loop formation has been shown to be associated with increased phosphorylation of histone H3 on Serine 10 and reduced chromatin accessibility 14 ., In turn , the phosphorylation of histone H3 on Serine 10 facilitates the interaction between adjacent nucleosomes , thereby promoting chromatin compaction 37 ., We showed previously that to constitutively increase the levels of histone H3 phosphorylated on Serine 10 by deleting PP1 phosphatase ( dis2Δ ) was not sufficient to significantly improve chromosome segregation when condensin was deficient 22 , suggesting that H3-S10-mediated chromatin compaction cannot compensate for the deficiency of condensin ., Here we presented evidence that stable R-Loops do not significantly contribute to the recruitment of condensin ., Taken together , these observations concur to establish that R-Loop-mediated chromatin compaction is distinct from condensin-mediated chromosome condensation ., Our data also suggest that the action of condensin is more fundamental to building a mitotic chromosome than R-Loop-mediated chromatin compaction ., Our data have highlighted unexpected ways by which proteins involved in the metabolism of RNA can affect chromosome segregation and genome integrity ., Published data demonstrated conclusively that mutations in such factors in general and in Sen1 in particular resulted in chromosome instability ( CIN ) in yeast , in a mechanism involving R-Loop formation antagonizing replication fork progression ( 38 , 39 and reviewed in 35 ) ., Here on the contrary , our data show that deletions of two such factors , Swd2 . 2 and Sen1 , facilitate the segregation and stability of chromosomes when condensin is deficient , in a mechanism that does not require stable R-Loop formation ., In addition , our data show that Swd2 . 2 and Sen1 keep topological stress under control at Pol III-transcribed genes ., We speculate that the enhanced transcription at Pol III-transcription associated with lack of Swd2 . 2 and Sen1 could contribute to such stress ., However , we cannot exclude the possibility that RNA Pol III-dependent transcription is also defective in other ways that could explain the accumulation of topological stress when Swd2 . 2 and Sen1 are missing ., The answer to this question will require further studies ., Beautiful in vitro approaches demonstrated unequivocally that budding yeast Sen1 contributes to transcription termination of some RNA Pol II transcripts ( 26 ) ., It is not yet known whether fission yeast Sen1 has the same function ., As fission yeast Sen1 is not essential for viability whilst its budding yeast counterpart is , it is possible that the function of Sen1 has diverged in fission yeast ., This idea is supported by our data showing that RNA Pol III is likely to be the most stable binding partner of Sen1 in fission yeast and that Sen1 antagonizes Pol III-dependent transcription ., On the contrary , a recent study aimed at identifying the binding partners of RNA Pol III in budding yeast did not identify Sen1 , suggesting that the interaction between Sen1 and RNA Pol III is not as stable and/or abundant in budding yeast 40 ., Further work is required to understand the function of fission yeast Sen1 at Pol III-transcribed genes ., Previous studies had concluded that the inhibition of RNA Pol I or RNA Pol II in mitosis was a pre-requisite for the binding of condensin at repetitive sequences 9 , 10 , suggesting that a processive RNA polymerase is a hindrance to the binding of condensin on chromatin ., Here we challenge this idea by showing that an enhanced recruitment of condensin at Pol III-transcribed genes is associated with an increase in the expression of the same genes ., These data show that , at least at Pol III-transcribed genes , an active polymerase is not an obstacle for the binding of condensin ., A complete list of all of the strains used in this study is given in Table S3 ., Standard genetic crosses were employed to construct all strains ., Rnh1-GFP , Sen1-GFP , and Top1-3flag were generated using a standard PCR procedure ., To obtain Rnh1D129N , Rnh1 was PCR amplified and cloned into pCRII ( Life technologies ) ., Site-directed mutagenesis was then used to mutate the residue D129 into N ( GAC to AAC ) using Quickchange protocols ( Stratagene ) ., Overlapping PCR was used to add a C-terminus GFP tag and a cassette of resistance to kanamycin ( KanR ) to the mutagenized Rnh1 in order to integrate the mutagenized Rnh1 at the endogenous Rnh1 locus ., After yeast transformation , proper integrants were selected by PCR and western blot and were sequenced to verify the presence of the mutation ., The plasmid over-expressing RnhA tagged with 1xFLAG at its N-terminus was obtained from Eun Shik Choi and Robin Allshire ( WTCCB , Edinburgh , UK ) ., In order to stably integrate the plasmid in the genome , it was linearized by digestion with MluI and then transformed in to yeast according to standard procedures ., 1 , 5 . 108 cells were treated with 1% formaldehyde ( Sigma ) at 17°C for 30′ ., After extensive washes with cold PBS , cells were frozen in liquid Nitrogen ., Frozen cells were then broken open using a RETSCH MM400 Mill and then resuspended in cold lysis buffer ( Hepes-KOH 50 mM pH 7 , 5 , NaCl 140 mM , EDTA 1 mM , Triton 1% , Na-deoxycholate 0 , 1% , PMSF 1 mM ) ., The lysats were then sonicated at 4°C using a Diagenode sonicator ., Immuno-precipitation was done overnight at 4°C using Protein A-coupled Dynabeads previously incubated with the anti-GFP A11122 antibody ( Invitrogen ) or using Protein G-coupled Dynabeads previously incubated with the anti-myc 9E10 antibody ( Sigma ) according to the manufacturers instructions ., Beads were washed successively with ( 5′ incubation on rotating wheel ) : Wash I buffer ( 20 mM Tris pH 8 , 150 mM NaCl , 2 mM EDTA , 1% Triton-X100 , 0 , 1% SDS ) , Wash II buffer ( 20 mM Tris pH 8 , 500 mM NaCl , 2 mM EDTA , 1% Triton-X100 , 0 , 1% SDS ) and Wash III buffer ( 20 mM Tris pH 8 , 1 mM EDTA , 0 , 5% Na-deoxycholate , 1% Igepal , 250 mM LiCl ) ., After two additional washes in TE pH 8 , the beads were resuspended in 10% Chelex resin ( Biorad ) and incubated at 98°C for 10′ ., After addition of 2 µL of 10 mg/mL of proteinase K , the mixture was incubated at 43°C for 1 hour , then for another 10 mn at 98°C ., After centrifugation , the supernatant was collected and analyzed by qPCR ., 8 . 108 cells were frozen in liquid nitrogen , broken open using a RETSCH MM400 Mill and then resuspended in cold lysis buffer ( Hepes-KOH 50 mM pH 7 , 5 , NaCl 140 mM , EDTA 1 mM , Triton 1% , Na-deoxycholate 0 , 1% ) ., After phenol/chloroform purification and ethanol precipitation , the DNA was resuspended in TE pH 8 and split into two samples ., Both samples were digested with BsrGI , EcoRI , HindIII , SspI and XbaI according to the manufacturers instructions and RNase H was added to one of the two samples ., After digestion , each sample was divided into two and incubated overnight at 4°C in IP buffer ( 100 mM MES pH 6 , 6 , NaCl 500 mM , 0 , 05% Triton , 2 mg/mL BSA ) in the presence of either Protein A-coupled Dynabeads or Protein A-coupled Dynabeads previously incubated with the S9 . 6 antibody according to the manufacturers instructions ., The beads were then washed three times in IP buffer ., After two additional washes in TE pH 8 , the beads were resuspended in 10% Chelex resin ( Biorad ) and incubated at 98°C for 5′ ., After addition of 2 µL of 10 mg/mL of proteinase K , the mixture was incubated at 43°C for 30′ , then for another 5′ at 98°C ., After centrifugation , the supernatant was collected and analyzed by qPCR ., Immunoprecipitation was carried out as described previously 22 , except that cells were broken open using a RETSCH MM400 Mill ., To purify Sen1-associated proteins ( Table S2 ) , a protein extract was prepared from 109 cells expressing GFP-tagged Sen1 from the endogenous locus ., After immuno-precipitation with 15 µL of magnetic beads , the beads were washed three times with 1 mL of lysis buffer and twice with 1 mL of PBS containing 0 , 02% Tween ., The beads samples were then subjected to in-solution reduction , carbamidomethylation and tryptic digestion ., After acidification with 10%Trifluoroacetic Acid the samples were centrifuged 3 times to eliminate the beads ., Peptide sequences w

Introduction, Results, Discussion, Materials and Methods

Condensin-mediated chromosome condensation is essential for genome stability upon cell division ., Genetic studies have indicated that the association of condensin with chromatin is intimately linked to gene transcription , but what transcription-associated feature ( s ) direct ( s ) the accumulation of condensin remains unclear ., Here we show in fission yeast that condensin becomes strikingly enriched at RNA Pol III-transcribed genes when Swd2 . 2 and Sen1 , two factors involved in the transcription process , are simultaneously deleted ., Sen1 is an ATP-dependent helicase whose orthologue in Saccharomyces cerevisiae contributes both to terminate transcription of some RNA Pol II transcripts and to antagonize the formation of DNA:RNA hybrids in the genome ., Using two independent mapping techniques , we show that DNA:RNA hybrids form in abundance at Pol III-transcribed genes in fission yeast but we demonstrate that they are unlikely to faciliate the recruitment of condensin ., Instead , we show that Sen1 forms a stable and abundant complex with RNA Pol III and that Swd2 . 2 and Sen1 antagonize both the interaction of RNA Pol III with chromatin and RNA Pol III-dependent transcription ., When Swd2 . 2 and Sen1 are lacking , the increased concentration of RNA Pol III and condensin at Pol III-transcribed genes is accompanied by the accumulation of topoisomerase I and II and by local nucleosome depletion , suggesting that Pol III-transcribed genes suffer topological stress ., We provide evidence that this topological stress contributes to recruit and/or stabilize condensin at Pol III-transcribed genes in the absence of Swd2 . 2 and Sen1 ., Our data challenge the idea that a processive RNA polymerase hinders the binding of condensin and suggest that transcription-associated topological stress could in some circumstances facilitate the association of condensin .

Failure to condense chromosomes prior to anaphase onset can lead to genome instability ., The evolutionary-conserved condensin complex drives chromosome condensation , probably by changing the topology of chromatin around its binding sites ., Condensin localizes to regions of high transcription , suggesting that some transcription-associated feature ( s ) direct its association with chromatin ., Here we considered that transcription-dependent DNA:RNA hybrids or topological stress could be involved in recruiting condensin ., Our data show that condensin is indeed enriched at regions accumulating DNA:RNA hybrids but that they are not involved in its recruitment ., Rather , we identify a mutant combination where increased transcription by RNA Pol III is associated locally with stronger topological stress ., Strikingly the localization of condensin is dramatically enhanced at the same loci and we show that topological stress contributes to this enhanced association ., Our data strengthen the idea that transcription creates the environment necessary to recruit condensin in mitosis .

chromosome structure and function, gene regulation, cell cycle and cell division, cell processes, dna transcription, mitosis, fungi, chromatin, schizosaccharomyces, chromosome biology, gene expression, schizosaccharomyces pombe, molecular biology, yeast, biochemistry, rna, rna processing, cell biology, transcriptional termination, genetics, biology and life sciences, organisms, chromosomes

journal.pgen.1000369

Expression of the Multiple Sclerosis-Associated MHC Class II Allele HLA-DRB1*1501 Is Regulated by Vitamin D

Multiple sclerosis ( MS ) is a common inflammatory disease of the central nervous system characterized by myelin loss , axonal pathology , and progressive neurological dysfunction 1 ., The aetiology of MS is unknown , however it is clear that genetic and environmental components are important 1 , 2 ., The only genetic association with MS in Northern Europeans had been with extended MHC haplotypes , especially those containing HLA-DRB1*1501 3 ., The interleukin 7 receptor ( IL7RA ) , interleukin 2 receptor ( IL2RA ) , ecotropic viral integration site 5 ( EVI5 ) and kinesin family member 1B ( KIF1B ) genes have recently been shown to be additional MS susceptibility loci 4 , 5 , 6 , 7 ., The largest of these , KIF1B , has a relatively small effect size ( odds ratio ( OR ) =\u200a1 . 3 ) ., The MHC ( OR\u200a=\u200a5 . 4 ) is the key susceptibility locus in MS and other susceptibility genes identified to date appear to contribute little to overall risk 3 ., The principal MHC class II haplotype that increases MS risk in individuals of Northern European descent is HLA- DQB1*0602-DQA1*0102 -DRB1*1501-DRB5*0101 8 , although other HLA-DRB1 haplotypes have important influences on risk by epistatic interactions 9 , 10 , 11 , 12 ., Intense linkage disequilibrium within the MHC has frustrated attempts at fine mapping and no precise susceptibility locus has been identified 9 , 13 ., Twin studies have established that monozygotic ( MZ ) twin concordance is significantly greater than for dizygotics ( DZ ) ., In the study by Willer and colleagues concordance was 25 . 3% and 5 . 4% respectively 14 ., The observation that most MZ twin pairs are discordant for MS suggests environmental , stochastic factors or both but the most striking illustration of the importance of the environment in MS susceptibility is the 5-fold difference in MS risk between Tasmania and Queensland 15 ., In the Northern Hemisphere , MS prevalence shows a north-south gradient , mirrored by a south-north gradient in the southern hemisphere ( reviewed by 16 ) ., In accordance with the disease geography , sunlight , specifically through its role in generating active vitamin D , has been proposed as a key environmental factor for the disease 17 ., Circumstantial evidence to support this comes from studies showing that MS patients are deficient in vitamin D 18 and that dietary vitamin intake reduces disease risk 19 ., Additionally , a pooled analysis of over 40 , 000 patients from Canada , Great Britain , Denmark , and Sweden showed that fewer people with MS were born in November and more in May 20 , highlighting a risk factor that varies seasonally ., Vitamin D is primarily known for its critical role in calcium homeostasis , however recent evidence has highlighted many actions on immune and central nervous system development and function 21 ., These have contributed to the notion that this is how vitamin D affects MS risk , although direct links have not yet been identified ., Vitamin D is a secosteroid hormone synthesized in the skin or ingested in the diet ., Intake from dietary sources accounts for a much smaller proportion of total vitamin D , mainly owing to its rarity in foods 22 , 23 ., During exposure to sunlight , ultraviolet B ( UVB ) radiation ( 290–315 nm ) is responsible for photolyzing 7-dehydrocholesterol , the precursor of vitamin D3 , to previtamin D3 which , in turn , rapidly spontaneously isomerizes to vitamin D3 22 , 23 ., Vitamin D3 is biologically inert and requires hydroxylation in the liver to 25-hydroxyvitamin D3 ( 25 ( OH ) D ) ., Once formed , this major circulating form of vitamin D3 is further hydroxylated in the kidney to its active form , 1 , 25-dihydroxyvitamin D3 ( 1 , 25 ( OH ) 2D ) , by 25-hydroxyvitamin D-1α-hydroxylase ( 1-OHase ) ., Recently it has been recognized that most tissues in the body ( including the brain , thymus and cells of the immune system ) also possess the 1-OHase enzyme ., Thus numerous tissues in the body have the capacity to locally produce 1 , 25 ( OH ) 2D 22 , 23 ., Most biological effects of 1 , 25-dihydroxyvitamin D3 or calcitriol , are mediated by the vitamin D receptor ( VDR ) ., This receptor is a member of the steroid receptor super-family and influences the rate of transcription of vitamin D responsive genes by acting as a ligand activated transcription factor that binds to vitamin D response elements ( VDREs ) in gene promoters 21 ., Early studies had provided evidence for an effect of vitamin D on HLA gene expression 24 , 25 , although no specific mechanism has been characterised ., Here we examined the hypothesis of a direct interaction between vitamin D and MS associated MHC class II genes ., Genetic variation characteristic of the most significant risk haplotypes for MS , those bearing HLA-DRB1*15 , includes a functional vitamin D response element ( VDRE ) in the proximal promoter region of HLA-DRB1 ., This provides a mechanism linking the major environmental and genetic risk factors for MS ., Using the sequence for the HLA-DRB1*15 haplotype carried by the homozygous lymphoblastoid cell line PGF we scanned in silico for VDREs using Jaspar 26 with a profile score threshold of 80% ., We analysed the entire genomic sequence of the HLA-DRB1 , HLA-DQA1 and HLA-DQB1 genes as well as 5 kb upstream of the transcriptional start sites of these genes to include promoter regions ., VDREs exhibit a multitude of sequence variations , providing a spectrum of binding affinities for VDR , thus enabling these elements to respond to differing concentrations of VDR/1 , 25 ( OH ) 2D 22 ., The analysis revealed only one potential VDRE located in the proximal promoter region immediately 5′ to the transcriptional start site of HLA-DRB1 ( Figure 1 ) ., IL2RA and IL7RA were also searched in silico for potential VDR binding sequences; no putative VDREs were found ., The occurrence and conservation of the putative VDRE element identified in the PGF sequence was examined in individuals with the HLA-DRB1*15 MS risk allele ., The HLA-DRB1 promoter was resequenced in 322 HLA-DRB1*15 homozygous individuals , both MS affected and unaffected ., An additional 168 individuals homozygous for other HLA-DRB1 alleles were also sequenced ., The putative VDRE was present on all HLA-DRB1*15 bearing haplotypes with no variants found which disrupted the VDRE consensus sequence ., In contrast , a number of nucleotide changes were found within the 15 base pairs of the VDRE on all non-HLA-DRB1*15 haplotypes ., For example , nearly all ( 98% of 57 sequenced individuals ) of HLA-DRB1*04 , HLA-DRB1*07 and HLA-DRB1*09 haplotypes , all of which are non-MS associated alleles in the Canadian population 10 , carried the sequence GGGTGGAGAGGGGTCA ., This sequence was predicted to function less effectively as a VDRE than the one on HLA-DRB1*15 bearing haplotypes according to Jaspar 26 ., The modestly MS associated haplotype , HLA-DRB1*17 , differed from HLA-DRB1*15 at the VDRE in 50% of the individuals sequenced ., The putative VDRE in the HLA-DRB1 promoter was investigated for ability to bind the vitamin D receptor in vitro using an electrophoretic mobility shift assay ( EMSA ) ., Upon addition of recombinant VDR and retinoic acid receptor beta ( RXR , a co-regulator of VDR binding and transactivation 22 ) to a radiolabelled probe spanning the putative VDRE in the HLA-DRB1 promoter , two protein-DNA complexes on EMSA were observed ( Figure 2 , lane 2 ) ., Both complexes were specifically competed with 10 to 100-fold molar excess of unlabelled VDRE probe ( Figure 2 , lanes 3–5 ) , while 10 to 100 fold molar excess of an unrelated probe containing an early growth response ( EGR ) factor binding site had no effect ( Figure 2 , lanes 6–7 ) ., Finally , addition of a polyclonal antibody directed against VDR specifically retarded complex I , resulting in a supershift of the upper complex ( Figure 2 , lane 8 ) ., This data showed the putative VDRE in the HLA-DRB1 promoter corresponding to the HLA-DRB1*15 haplotype could bind recombinant VDR/RXR with high specificity in vitro ., When probes corresponding to the HLA-DRB1*04/07/09 variant VDRE were used , significantly lower affinity binding was found ( data not shown ) ., Whether or not the VDR is recruited to the VDRE in the HLA-DRB1 gene promoter was examined ex vivo ., Chromatin immunoprecipitation ( ChIP ) experiments were performed using lymphoblastoid cells bearing the HLA-DRB1*15 haplotype ( the PGF cell line ) which were either unstimulated or stimulated for 24 hours with 1 , 25-dihydroxyvitamin D3 and then cross-linked in the presence of formaldehyde ., Immunoprecipitation was performed using antibodies against VDR ., The VDR bound DNA fragments were then recovered after reversal of protein-DNA crosslinking and analysed by PCR using primers specific for the HLA-DRB1 promoter ., A representative agarose gel is shown in Figure 3 ., This revealed clear evidence of binding by VDR to the HLA-DRB1 promoter when compared to input chromatin and mock antibody controls for cells with the HLA-DRB1*15 haplotype , complementing the in vitro data from the EMSA experiments ., The VDRE was then investigated to see if it modulated levels of gene expression in vitro ., Reporter gene constructs were engineered in which −181 to +53 of the HLA-DRB1 gene sequence was placed upstream of a pGL3 luciferase reporter ., pGL3_DRB1prom had the complete −181 to +53 sequence , pGL3_DRB1prom_hap1 had the same sequence as pGL3_DRB1prom but the VDRE replaced with the HLA-DRB1*04/07/09 VDRE and pGL3_DRB1prom_del had the 15 base pair VDRE sequence specifically deleted ., These constructs were then transiently transfected into Raji B cells ., A renilla luciferase reporter construct driven by the thymidine kinase promoter ( pRL_TK ) was co-transfected to normalise luciferase activity ., pGL3_DRB1prom had significantly higher basal reporter gene activity than pGL3_DRB1prom_del ( P\u200a=\u200a0 . 03 on paired t-test , two tailed ) ., After stimulation with 1 , 25-dihydroxyvitamin D3 , there was a significant 1 . 6 fold increase in luciferase activity with pGL3_DRB1prom ( P\u200a=\u200a0 . 002 ) , but no significant change with pGL3_DRB1prom_del ( P\u200a=\u200a0 . 12 ) , nor pGL3_DRB1prom_hap1 ( P\u200a=\u200a0 . 58 ) ( Figure 4 ) ., To investigate any effect of vitamin D on the cell surface expression of HLA-DRB1 , the HLA-DRB1*15 homozygous lymphoblastoid cell line PGF and the HLA-DRB1*07 homozygous lymphoblastoid DBB cell line were stained with anti-HLA-DRB1 antibody ., PGF cells constitutively expressed HLA-DRB1 at higher levels then DBB ( average geometric mean fluorescence intensity ( MFI ) PGF\u200a=\u200a97 . 1 , DBB\u200a=\u200a42 . 8 , P\u200a=\u200a0 . 0002 ) ., Upon addition of 1 , 25-dihydroxyvitamin D3 , there was a 1 . 3 fold increase in the expression of HLA-DRB1 in PGF cells ( P\u200a=\u200a0 . 031 on paired t-test , two tailed ) but no significant difference in the expression of HLA-DRB1 in DBB cells ( P\u200a=\u200a0 . 10 ) ., While the role of the environment is clearly important in determining MS risk , the relevant underlying mechanism ( s ) have remained elusive and there has been no experimental support for a direct environment-gene interaction ., Although differences in Epstein-Barr virus infection are seen when MS patients are compared to controls , extensive searches for specific viral infections have failed to confirm direct involvement ., 2 ., Where appropriate data is available , the amount of winter sunlight parallels the range of MS prevalence , and high sunlight exposure is associated with low disease prevalence 2 ., The effects of migration between high and low risk geographic regions have been examined in several populations ( e . g . UK immigrants to South Africa , or Asian and Caribbean immigrants to the UK ) ., These studies show that MS risk is influenced by the migrants country of origin 27 ., Despite the limits of small sample sizes , a ‘critical age’ has been hypothesized: immigrants who migrate before adolescence acquire the risk of their new country , while those who migrate after retain the risk of their home country ., Dietary difference for vitamin D intake ( oily fish consumption ) plausibly explains the striking exception to MS latitudinal risk in Norway 2 ., As familial aggregation is genetically determined 28 , environmental factors thus appear to be operative at a broad population level , perhaps acting at a young age 27 and/or during gestation 20 ., A good candidate for an environmental factor that influences MS disease risk is vitamin D . We approached the candidacy of vitamin D by searching first for vitamin D response elements within the MHC class II region ., Specifically we investigated the major candidate genes in the disease associated locus , HLA-DRB1 , HLA-DQA1 and HLA-DQB1 and identified a consensus binding site for VDR next to the HLA-DRB1 gene ., This was the only VDRE we found and strikingly it shows haplotype-specific differences , being highly conserved in the major MS associated haplotype HLA-DRB1*15 dominant in Northern European populations , but not conserved among non-MS associated haplotypes ., This was itself circumstantial evidence supporting a vitamin D role in the functional characteristics of this haplotype ., The identified VDRE lies close to the highly conserved MHC class II specific regulatory SXY module ., This module comprises S , X and Y regulatory elements important for constitutive , and indirectly for IFN-γ-induced , expression of HLA class II genes co-ordinated by the MHC class II transactivator MHC2TA 29 ., The VDRE was highly conserved on HLA-DRB1*15 haplotypes ( no mutations on over 600 chromosomes ) suggesting a selective pressure to maintain this response element for the HLA-DRB1*15 allele ., Variants were found to some extent on all other non HLA-DRB1*15 haplotypes ., The results may additionally/alternatively reflect the ancestral origin of the HLA-DRB1*15 ( DR51 ) haplotype 30 which displays the strongest linkage disequilibrium among the MHC class II haplotypes 31 ., We note the association between this haplotype and MS risk is characteristic of Northern European populations , the ones most vulnerable to vitamin D deficiency 2 ., EMSA experiments using recombinant proteins demonstrated that in vitro VDR can bind specifically to the putative VDRE in the proximal HLA-DRB1 promoter found on the HLA-DRB1*15 haplotype ., ChIP data showed specific enrichment of the region spanning the VDRE in VDR immunoprecipitated samples relative to input and mock antibody controls , demonstrating that the vitamin D receptor was recruited to this haplotype in this ex vivo model system ., Finally , transient transfection and flow cytometric assays established that the VDRE present in the HLA-DRB1 promoter can influence gene expression and imparts 1 , 25-dihydroxyvitamin D3 sensitivity to HLA-DRB1*15 ., The variant VDRE present on other , non-MS associated HLA-DRB1 haplotypes was not responsive to 1 , 25-dihydroxyvitamin D3 ., A T cell repertoire with millions of specificities provides surveillance against a multitude of foreign pathogens 32 ., An inherent danger in recognizing so many foreign proteins is the potential to respond to self-proteins ., To circumvent this problem T cells are scrutinised for self-reactivity as they mature in the thymus with deletion of those posing the greatest threat ( central deletion ) 32 ., One constraint on central deletion is the requirement for the relevant autoantigen to be present in the thymus ., Whether or not these are expressed as proteins at levels sufficient to induce T cell deletion is not clear ., Given the results of this study , variable expression of HLA-DRB1 could affect central deletion of autoreactive T cells ., It is plausible that a lack of vitamin D in utero or early childhood can affect the expression of HLA-DRB1 in the thymus , and impacting on central deletion ., For MS , in HLA-DRB1*15 bearing individuals , a lack of vitamin D during early life could allow auto reactive T cells to escape thymic deletion and thus increase autoimmune disease risk ., Indeed it has been shown that antigen presentation in the thymus of VDR knock-out mice is impaired 33 ., However the mechanism for a HLA- vitamin D interaction remains unclear as is the timing and tissue in which such interactions might occur ., A major selective pressure on skin pigmentation is thought to have been vitamin D deficiency with progressively lighter skin pigmentation at increasing distance from the equator related to variation in intensity of ultraviolet radiation with latitude 34 ., The presence of a VDRE specific to HLA-DRB1*15- bearing haplotypes , present at high allele frequencies among Northern Europeans , suggests a possible role for vitamin D in selection at this locus ., The intriguing possibility that vitamin D responsiveness rather than any antigen-specificity determines the increased MS risk of the HLA-DRB1*15 haplotype warrants consideration and can be tested in the infrequent haplotypes bearing the VDRE on other non-HLA-DRB1*15 haplotypes ., In summary , we have identified and functionally characterised a vitamin D response element ( VDRE ) in the HLA-DRB1 promoter region ., These studies imply direct interactions between HLA-DRB1 , the main susceptibility locus for MS , and vitamin D , a strong candidate for mediating the environmental effect ., This study provides more direct support for the already strong epidemiological evidence implicating sunlight and vitamin D in the determination of MS risk ., Given that a high frequency of vitamin D insufficiency in the general population has been observed 35 , our data support the case for supplementation during critical time periods to reduce the prevalence of this devastating disease ., All participants in the study were ascertained through the ongoing Canadian Collaborative Project on the Genetic Susceptibility to MS ( CCPGSMS ) 36 ., Subject ascertainment , genotyping and sequencing has been previously described 9 , 10 , 37 ., Each participating clinic in the CCPGSMS obtained ethical approval from the relevant institutional review board , and the entire project was reviewed and approved by the University of British Columbia and the University of Western Ontario ., EMSAs were performed as previously described 38 ., The VDRE probe comprised of the annealed sense and antisense strands of the nucleotide sequence agctGTGGGTGGAGGGGTTCATAG , the EGR probe agctAAATCCCCGCCCCCGCGATGGA and the VDRE variant probe agctGTGGGTGGAGAGGGGTCATAG ., Full length recombinant purified VDR and recombinant purified RXR beta were purchased from Invitrogen , and polyclonal VDR antibody from Affinity Bioreagents ., Radioactivity was quantitated with the Packard Cyclone phosphorimager , and analyzed with Optiquant ( Perkin Elmer Life Sciences ) ., Values were compared using the Chi square test ., The lymphoblastoid cell line PGF was cultured in RPMI-1640 medium supplemented with 10% fetal bovine serum , 0 . 2 mM L-glutamine at 37°C in 5% humidified CO2 ., 60×106 cells were harvested unstimulated or after stimulation with 0 . 1 uM calcitriol ( Sigma ) ., Cells were crosslinked using a 1% formaldehyde buffer for 15 minutes at room temperature , quenched with glycine and chromatin prepared as previously described 39 ., Chromatin was sheared by sonication in the presence of 212–300 microns glass beads ( Sigma ) at 4°C using a double step microtip attached to a Branson 450 Sonifier with coupler ( Branson ) in 30 second bursts ( six pulses at 40% ) with the samples cooled on ice for 1 minute between pulses ., Sonicated chromatin was then processed and subject to immunoprecipitation as previously described 39 using magnetic ‘Dynabeads M-280’ ( Dynal ) precoated with anti rabbit IgG to which the primary antibody VDR was bound ( Affinity Bioreagents ) ., We followed the buffer used for immunoprecipitation and subsequent washes as described 40 ., Following reversal of crosslinks , RNase A and Proteinase K digestion , DNA was extracted using phenol-chloroform and amplified by PCR with separation on a 2 . 0% agarose gel ., The primers used for PCR were: forward- GCAACTGGTTCAAACCTTCC and reverse- GTCCCCAGACAAAGCCAGT ., Cycling conditions were: 95°C for 10 minutes; a touchdown of 14 cycles ( 95°C for 30 seconds; 61°C with −0 . 5°C per cycle , for 30 seconds; 72°C for 30 seconds ) ; 35 cycles of 95°C for 30 seconds , 53 . 5°C for 30 seconds , 72°C for 30 seconds; 72°C for 7 minutes ., The plasmids were constructed by inserting the promoter region ( −181 to +53 ) of the human HLA-DRB1 gene ( pGL3_DRB1prom with the VDRE sequence ( chr6:32 , 665 , 500–32 , 665 , 760 ) , pGL3_DRB1prom_del with the VDRE sequence deleted ( chr6:32 , 665 , 500–32 , 665 , 559 combined with chr6:32 , 665 , 575–32 , 665 , 760 ) ) into the pGL3 reporter plasmid ., Two independent plasmid preparations were used in transient transfection experiments for each construct ., Raji B cells were cultured in RPMI-1640 medium supplemented with 10% fetal bovine serum , 0 . 2 mM L-glutamine at 37°C in 5% humidified CO2 ., Lipofectamine-LTX and PLUS reagent ( Invitrogen ) were used for transient transfection of expression constructs , following the manufacturers protocol ., pRL_TK was co-transfected to normalize for transfection efficiency ., When indicated , cells were stimulated with 0 . 1 uM calcitriol ( Sigma ) for 24 hours ., Cells were harvested after 24 hours and lysed in 500 ul of 1× lysis buffer ( Promega ) and analyzed using the Dual-Luciferase reporter assay kit ( Promega ) and a Turner luminometer model 20 ( Promega ) following the manufacturers protocol ., Paired t-tests were used to compare expression values ., Each transfection was carried out 12 times in total ., The lymphoblastoid cell lines PGF ( International Histocompatibility Workshop number IHW09318 ) and DBB ( IHW09052 ) were cultured in RPMI-1640 medium supplemented with 10% fetal bovine serum , 0 . 2 mM L-glutamine at 37°C in 5% humidified CO2 ., 1×106 cells were harvested unstimulated or 24 hours after stimulation with 0 . 1 uM calcitriol ( Sigma ) in three biological replicates ., Cells were stained with either a FITC conjugated monoclonal anti-human HLA-DR antibody ( Sigma , F1902 ) or a FITC conjugated isotype control antibody ( Sigma , F6522 ) for 30 minutes at room temp , then washed with 2% BSA in PBS and re-suspended in 1 mL of 2% paraformaldehyde ., Cells were analysed using CyAn flow cytometer ( Dako ) .

Introduction, Results, Discussion, Materials and Methods

Multiple sclerosis ( MS ) is a complex trait in which allelic variation in the MHC class II region exerts the single strongest effect on genetic risk ., Epidemiological data in MS provide strong evidence that environmental factors act at a population level to influence the unusual geographical distribution of this disease ., Growing evidence implicates sunlight or vitamin D as a key environmental factor in aetiology ., We hypothesised that this environmental candidate might interact with inherited factors and sought responsive regulatory elements in the MHC class II region ., Sequence analysis localised a single MHC vitamin D response element ( VDRE ) to the promoter region of HLA-DRB1 ., Sequencing of this promoter in greater than 1 , 000 chromosomes from HLA-DRB1 homozygotes showed absolute conservation of this putative VDRE on HLA-DRB1*15 haplotypes ., In contrast , there was striking variation among non–MS-associated haplotypes ., Electrophoretic mobility shift assays showed specific recruitment of vitamin D receptor to the VDRE in the HLA-DRB1*15 promoter , confirmed by chromatin immunoprecipitation experiments using lymphoblastoid cells homozygous for HLA-DRB1*15 ., Transient transfection using a luciferase reporter assay showed a functional role for this VDRE ., B cells transiently transfected with the HLA-DRB1*15 gene promoter showed increased expression on stimulation with 1 , 25-dihydroxyvitamin D3 ( P\u200a=\u200a0 . 002 ) that was lost both on deletion of the VDRE or with the homologous “VDRE” sequence found in non–MS-associated HLA-DRB1 haplotypes ., Flow cytometric analysis showed a specific increase in the cell surface expression of HLA-DRB1 upon addition of vitamin D only in HLA-DRB1*15 bearing lymphoblastoid cells ., This study further implicates vitamin D as a strong environmental candidate in MS by demonstrating direct functional interaction with the major locus determining genetic susceptibility ., These findings support a connection between the main epidemiological and genetic features of this disease with major practical implications for studies of disease mechanism and prevention .

Multiple Sclerosis ( MS ) is a complex neurological disease with a strong genetic component ., The Major Histocompatibility Complex ( MHC ) on chromosome 6 exerts the strongest genetic effect on disease risk ., A region at or near the HLA-DRB1 locus in the MHC influences the risk of MS . HLA-DRB1 has over 400 different alleles ., The dominant haplotype of Northern Europe , marked by the presence of DRB1*1501 , increases risk of MS by 3-fold ., The environment also plays a key role in MS . The most striking illustration of this is the geographical distribution of the disease in populations matched for ethnicity ., This has led to the proposal that sunshine , and in particular , vitamin D , is an environmental factor influencing the risk of MS . Circumstantial evidence supporting this comes from studies showing the involvement of vitamin D in immune and nervous system function ., The current investigation sought to uncover any relationship between vitamin D and HLA-DRB1 ., It was found that vitamin D specifically interacts with HLA-DRB1*1501 to influence its expression ., This study therefore provides more direct support for the already strong epidemiological evidence implicating sunlight and vitamin D in the determination of MS risk , and implies that vitamin D supplementation at critical time periods may be key to disease prevention .

public health and epidemiology/epidemiology, immunology/genetics of the immune system, neurological disorders/multiple sclerosis and related disorders, immunology/autoimmunity

journal.ppat.1001094

A Novel Family of Toxoplasma IMC Proteins Displays a Hierarchical Organization and Functions in Coordinating Parasite Division

The phylum Apicomplexa contains numerous obligate intracellular pathogens that are the cause of serious disease in humans and animals , greatly influencing global health and causing significant economic loss worldwide ., The phylum includes Plasmodium falciparum , the causative agent of malaria which claims 1–2 million human lives annually , and Toxoplasma gondii , a pathogen that infects more than thirty percent of the worlds population and causes severe neurological disorders and death in immunocompromised individuals 1 ., Most of the drugs used to treat apicomplexans target metabolic pathways or the chloroplast-derived apicoplast 2 , 3 , 4 , but these parasites also possess elaborate and unique structures that are required for replication and invasion and thus represent attractive new targets for therapeutic intervention ., Apicomplexans are grouped with dinoflagellates and ciliates in the alveolata infrakingdom 5 ., The unifying morphological characteristic of this group is the presence of alveoli: membrane sacs located beneath the plasma membrane ., Molecular phylogenetic data supports this grouping , as does the identification of a conserved family of articulin-like membrane skeleton proteins , the alveolins , which associate with alveoli in all three phyla 6 , 7 ., While the presence of alveoli is conserved , each of these groups has adapted this peripheral membrane structure for different cellular functions to fit their distinct niches ., In dinoflagellates , the alveoli sometimes contain cellulose-based plates that function as protective armor 8 ., In contrast , ciliate alveoli are calcium storage devices thought to play roles in regulation of cilia , exocytosis from cortical organelles known as extrusomes , and control of cytoskeletal elements 9 , 10 , 11 ., In apicomplexans , the alveoli in conjunction with an underlying filamentous network are termed the inner membrane complex ( IMC ) 12 , 13 ., Flattened alveoli underlie the entirety of the plasma membrane except for a small gap at the apex and base of the cell 14 ., These cisternae are organized into a patchwork of rectangular plates capped by a single cone-shaped plate at the apex of the cell ., Freeze-fracture studies of the IMC plates expose a lattice of intramembranous particles ( IMPs ) , an arrangement that suggests an association with proteins of the underlying filamentous network and subtending cortical microtubules 15 , 16 , 17 ., Together , these features of the IMC are the foundation for a unique form of gliding motility used for host cell invasion and also serve as the scaffold for daughter cell formation during division 18 , 19 ., Toxoplasma tachyzoites replicate by endodyogeny , a process of internal cell budding that produces two daughters within an intact mother parasite ., Following centriole duplication , daughter cell formation begins with the concurrent assembly of an apical and basal complex 20 ., Although these two structures consist of cytoskeletal components that will eventually cap opposite ends of the mature parasite , they are initiated in close spatial and temporal proximity ., IMC construction then proceeds by the extension of the basal complex away from the daughter apical complex , generating a bud into which replicated organelles are packaged ., Parasite division is completed by a number of maturation steps terminating with the adoption of the maternal plasma membrane 21 ., The apical , cone-shaped cisterna is unique in form and presumably the earliest membrane component deposited into the nascent IMC 19 ., A number of cytoskeletal IMC markers localize to a region at the parasite apex thought to correspond to this apical-most IMC plate ., A GFP fusion of the dynein light chain , TgDLC , can be detected in an apical cap region but predominantly localizes to the conoid and is also found in the basal complex , spindle poles and centrioles ., TgCentrin2 , the most divergent of the three Toxoplasma centrin homologues , labels the preconoidal rings and a peripheral ring of ∼6 annuli located at the lower boundary of the TgDLC cap ., It has been suggested that these annuli lie at the juncture between the apical cap plate and the flanking set of IMC plates 20 ., Additionally , PhIL1 , a cytoskeletal IMC protein of unknown function , is detected throughout the IMC but strongly enriched in the apical cap and basal complex 22 ., Only a few proteins are known to directly associate with the IMC membranes ., These include a number of proteins associated with gliding motility 23 , 24 , 25 , as well as the heat shock protein Hsp20 26 and one isoform of the purine salvage enzyme hypoxanthine-xanthine-guanine phosphoribosyltransferase 27 ., Thus , despite the central role of this conserved membrane system in apicomplexan biology , little is known of its composition , organization , and construction ., We present here a family of proteins unique to the Apicomplexa that localize to three distinct sub-compartments of the Toxoplasma IMC ., ISP1 localizes to a region corresponding to the apical cap , ISP2 occupies a central IMC region , and ISP3 resides in both the central IMC region and a basal IMC compartment ., ISP1 and 3 are early markers for bud formation and label previously unobserved daughter IMC structures in the absence of parasite cortical microtubules , indicating that microtubules are not required for initial assembly of IMC membranes ., We show that the ISPs are initially targeted to the IMC by conserved residues predicted for coordinated myristoylation and palmitoylation in the extreme N-terminus of each of these proteins ., Interestingly , deletion of ISP1 results in the relocalization of ISP2 and 3 to the apical cap , demonstrating an interactive , hierarchical targeting among this family of proteins to these distinct sub-compartments of the IMC ., Finally , disruption of ISP2 results in a severe loss of parasite fitness and dramatic defects in daughter cell formation ., Although the ISP2 knockout parasites ultimately compensate for these defects , this data shows an important role for these proteins in the coordination of daughter cell assembly ., We previously generated a panel of monoclonal antibodies against a mixed fraction of T . gondii organelles 28 ., One of the antibodies , 7E8 , stains a cone-shaped structure at the periphery of the apical end of the parasite ( Figure 1A ) ., This staining pattern extends from a gap at the extreme apex ( Figure 1A , arrow ) ∼1 . 5 µm along the length of the parasite , a localization suggestive of the apical IMC plate observed by electron microscopy 14 ., Colocalization with TgCentrin2 shows that 7E8 staining is delimited at its apex and base by this apical cap marker , indicating that 7E8 does indeed detect a protein associated with the anterior-most IMC plate ( Figure 1D ) ., During early endodyogeny , 7E8 staining is visible in daughter parasites as a pair of small rings within each mother parasite ( Figure 1B , arrows ) ., As daughter formation proceeds , this structure enlarges and extends to form the apical cap seen in mature tachyzoites ( Figure 1C ) ., The association with forming daughter scaffolds together with the extreme apical gap further suggests that 7E8 labels the apical sub-compartment of the IMC ., We also frequently observe 7E8 staining a single dot near the basal border of the cone ( Figure 1C , arrow ) which is distinct from TgCentrin2 annuli ( Figure 1D , inset ) ., Western blot analysis of Toxoplasma lysates with mAb 7E8 revealed a single band at ∼18 kDa ( Figure 1E ) ., We used the 7E8 antibody to isolate its target protein by immunoaffinity chromatography ., The isolated protein was separated by SDS-PAGE ( Figure 1F ) , digested with trypsin , and seven peptides were identified by mass spectrometry corresponding to the hypothetical T . gondii protein TGGT1_009340 ( Figure 1G ) ., EST and cDNA sequencing confirmed that the gene model is correct ., Due to its unique localization , we named this protein IMC Sub-compartment Protein 1 ( ISP1 ) ., Examination of the 176 amino acid sequence of ISP1 reveals that it contains a high number of charged residues ( ∼30% ) ., While there are a relatively large number of ESTs encoding ISP1 , the protein lacks conserved domains that could suggest its function ., The protein contains a glycine at position two , which is predicted to be myristoylated 29 as well as a pair of cysteines at positions seven and eight strongly predicted to be palmitoylated 30 ., Since ISP1 lacks a predicted signal peptide or transmembrane domain , these residues suggested a mechanism for IMC membrane association ., BLAST analysis of the ISP1 sequence revealed orthologues across the apicomplexan phylum , including Neospora , Theileria , Cryptosporidia , Babesia , and Plasmodium ( Figure S1 ) ., Orthologues were also found in Eimeria by BLAST against EST libraries ( data not shown ) ., ISP1 also showed significant homology in its C-terminal region to CP15/60 , a poorly characterized putative surface glycoprotein in Cryptosporidia 31 , 32 ., No ISP1 orthologues were identified outside of the phylum indicating that this protein is restricted to the Apicomplexa ., BLAST analysis of the T . gondii genome using the ISP1 sequence identified two additional hypothetical proteins with considerable sequence similarity to ISP1 , which we named ISP2 ( TGGT1_058450 ) and ISP3 ( TGGT1_094350 ) ( Figure 2A ) ., The greatest degree of sequence similarity between these three proteins exists within the C-terminal two-thirds of their sequences ., The N-terminal regions of the proteins are more divergent , but each contain a conserved glycine at position two as well as a pair of conserved cysteines predicted to be myristoylated and palmitoylated , respectively ( Figure 2A , boxed residues ) ., ISP2 additionally contains a third cysteine at position five predicted to be palmitoylated ., Similar to ISP1 , these proteins are highly charged and have a relatively large number of corresponding ESTs ., OrthoMCL analysis of the ISPs indicates two ortholog groups within Apicomplexa ., ISP1 and ISP2 segregate with one group while ISP3 segregates with another ( Figure S1 ) ., The Toxoplasma genome may encode a fourth ISP family member ( TGGT1_063420 ) , although it does not segregate with any OrthoMCL group ., This predicted protein lacks the conserved glycine and cysteine residues present in the N-termini of other ISP proteins ., Only a single EST is present for TGGT1_063420 , indicating that it is poorly expressed relative to the other ISPs , and thus it was not investigated further ., To localize ISP2 and ISP3 in T . gondii , we expressed each gene under the control of its endogenous promoter with a C-terminal HA epitope tag ., Intriguingly , ISP2 localizes to a previously unrecognized central sub-compartment of the IMC , which begins at the base of the ISP1 apical cap and extends approximately two-thirds the length of the cell ., The apical boundary of this compartment is delineated by the TgCentrin2 annuli ( Figure 2B ) ., The posterior boundary has a jagged edge suggesting it corresponds to discrete IMC plates ( Figure 2D , arrows ) ., While the ISP2 signal terminates near the end of the subpellicular microtubules , the termini for these two structures are not identical ( Figure S3 , WT ) ., Antisera raised against recombinant ISP2 confirmed this central IMC sub-compartment localization , ensuring that exclusion of ISP2 from the apical cap and basal IMC is not an artifact of epitope tagging ( Figure S2A ) ., Similar to ISP2 , ISP3 stains the central section of the IMC ., However , ISP3 staining extends to the posterior end of the complex , identifying a third sub-compartment of the IMC ( Figure 2C ) ., A small gap in ISP3 staining is observed in the posterior region similar to that seen for other IMC proteins 23 ., Antisera raised against recombinant ISP3 gave a poor signal by IFA , but was sufficient to confirm localization to both the IMC central and basal sub-compartments ( Figure S2B ) ., As with ISP1 , ISP2 and ISP3 are visible in forming daughter parasites ., Whereas the maternal signals of ISP1 and ISP2 appear to remain stable throughout endodyogeny , the maternal ISP3 signal rapidly attenuates with the onset of endodyogeny while it concentrates in daughters ( Figure 2E ) ., Attenuation of ISP3 in mothers and enrichment in daughters was also observed with our polyclonal antibody , indicating this is not the result of a C-terminal processing event that removes the HA epitope tag ( Figure S2C ) ., Thus , ISP3 provides an excellent marker for bud initiation , growth , and maturation during endodyogeny ( Figure 2E and Video S1 ) ., The observations that the ISPs are visible at the periphery of forming daughters prior to adoption of the maternal plasma membrane and that gaps are present at the extreme apex and base suggests an association with the IMC ., To confirm IMC association , we treated extracellular parasites with Clostridium septicum alpha-toxin ., This vacuolating toxin causes a dramatic separation of the plasma membrane and the underlying IMC , enabling differential localization of these closely apposed membrane systems 33 ., In toxin-treated parasites , the ISP proteins segregate with the IMC and not with the plasma membrane , confirming that the ISPs are indeed IMC proteins ( Figure 3A–B ) ., To ascertain if the ISPs are embedded in the IMC protein meshwork that includes the articulin-like protein IMC1 , we performed detergent extractions of extracellular parasites in 0 . 5% NP-40 ., In these conditions , each ISP was solubilized similar to the control protein ROP1 , while IMC1 remained in the insoluble pellet fraction ( Figure 3C ) ., This extraction profile demonstrates that the ISPs are not embedded in the detergent resistant protein meshwork that underlies the IMC membranes ., We disrupted microtubules in intracellular parasites to assess whether the underlying microtubules influence ISP localization ., Apicomplexan microtubules are selectively susceptible to disruption by dinitroanilines , such as oryzalin 34 ., After 40 hours of 2 . 5 µM oryzalin treatment , all tubulin is unpolymerized and dispersed ., Without spindle microtubules ( mitosis ) and subpellicular microtubules ( budding ) , productive daughter formation repeatedly fails resulting in an undivided , amorphous mother cell with a polyploid DNA content 35 ( Figure 4A ) ., Intriguingly , we observe ISP1 labeling numerous small rings that are centrally located within oryzalin-treated parasites ( Figure 4A , inset ) of approximately the same dimensions as ISP1 early daughter buds in untreated , replicating parasites ( compare with Figure 1B , arrows ) ., Since polymerization of subpellicular microtubules is essential to drive bud extension , these rings likely represent failed attempts to build new daughter buds 36 ., A larger peripheral patch of ISP1 with a central hole is also observed , likely representing the original parent apical cap ( Figure 4A , arrows ) ., While ISP2 was not observable in these early bud rings ( Figure 4B ) , we did detect ISP3 in these structures within oryzalin-treated parasites ( Figure 4C , inset arrows ) , suggesting that both the apical cap and remaining IMC sub-domains are formed independently of microtubules at a very early stage of bud development ., While membrane skeleton proteins are likely candidates for providing the foundation for these structures , we were unable to detect the articulin-like protein IMC1 in these early bud rings , even at lower oryzalin concentrations ( 0 . 5 µM ) that only disrupt cortical microtubules ( Figure 4D and Video S2 ) ., The greatest sequence similarity within the ISP family is present in the C-terminal two-thirds of the proteins while the N-terminal region is more divergent ( Figure 2A ) , thus we reasoned that the unique targeting of each ISP family member might be controlled by its N-terminal region ., To test if the N-terminal region of ISP1 is necessary for targeting , we eliminated the first 63 residues to create a truncated protein fused to YFP ., ISP164–176-YFP does not target to the IMC but is instead distributed throughout the cytoplasm and nucleus , showing that this N-terminal region is necessary for apical cap targeting ( Figure 5A ) ., To determine if the ISP1 N-terminal region is sufficient for targeting , we fused the first 65 residues of ISP1 ( containing the putative acylation sequence and divergent N-terminal region ) to YFP and expressed this construct in Toxoplasma ., The ISP11–65-YFP fusion traffics to the apical cap in an identical fashion to endogenous ISP1 ( Figure 5B ) ., To further narrow the N-terminal region required for apical cap targeting , we generated an additional fusion of the first 29 residues of ISP1 ( containing the putative acylation sequence ) to YFP ., This fusion also traffics in a manner identical to full length ISP1 ( Figure 5C ) , demonstrating that this N-terminal domain is both necessary and sufficient for apical cap targeting ., To assess targeting of ISP2 and ISP3 , we also created fusions of their N-terminal regions ( residues 1–41 and 1–36 respectively ) to YFP ., The ISP31–36-YFP fusion targets to the central and basal sub-compartments of the IMC but is restricted from the apical cap ( Figure 5D ) , showing that this region is sufficient for proper sub-compartment targeting ., In contrast , ISP21–41-YFP localized to the entire IMC , overlapping with endogenous ISP1 in the apical cap and extending into the basal IMC sub-compartment ( data not shown ) ., To ensure this change in targeting for ISP21–41 was not an artifact of the YFP fusion , we replaced YFP with an HA tag ( shown to have no effect on the targeting of full length ISP2 , Figure 5E ) ., The ISP21–41-HA protein also localized throughout the IMC ( Figure 5F ) , demonstrating that the N-terminal domain of ISP2 is sufficient for targeting to the IMC , but not for correct sub-compartment localization ., Protein myristoylation occurs co-translationally through the action of an N-myristoyl transferase 37 ., This modification is sufficient to promote transient association with membranes for otherwise cytosolic proteins ., This weak membrane affinity can then be stabilized by addition of one or more palmitoylations through the action of a palmitoyl acyltransferase ( PAT ) , effectively locking a protein into a target membrane system in a mechanism known as “kinetic trapping” ., The ISPs each contain a second position glycine followed by cysteines within the first 10 residues that are predicted to be myristoylated and palmitoylated , respectively ( Figure 2 , boxed residues ) ., We mutated the glycine and cysteine residues in HA epitope tagged ISP constructs to examine their effect on targeting ., As predicted by the kinetic trapping model , mutation of the second position glycine to an alanine abolished IMC targeting in each family member ( Figure 6 and Figures S3 and S4 , G2A ) , resulting in proteins distributed throughout the cytoplasm ., Mutation of the cysteine residues to serine was performed individually and together ., While only minor defects in targeting were observed with individual cysteine mutations , mutation of both cysteines abolished ISP1 and ISP3 targeting ( Figure 6 and Figure S4 ) ., In the case of ISP2 , targeting was only abolished when all three cysteines were coordinately mutated ( Figure S3 ) ., While coordinated cysteine mutants of the ISPs are distributed in the cytoplasm similar to G2A mutants , we also often observed perinuclear staining that is especially concentrated just apical of the nucleus ( arrows , Figure 6 and Figure S3 and S4 ) ., Presumably , myristoylation of these proteins still occurs , but without palmitoylation , these mutants are left to transiently sample the different membrane systems within the cell and therefore may appear concentrated as they associate with the ER and Golgi membranes present in this region ., These results demonstrate that these residues are essential to ISP sorting and indicate that coordinated acylation of the ISPs is responsible for IMC membrane targeting ., To assess the function of ISP1 , we disrupted the ISP1 gene by homologous recombination ( Figure 7A ) ., We identified clones which lacked ISP1 expression by IFA and Western blot ( Figure 7B–C ) , indicating successful disruption of the ISP1 locus and demonstrating that ISP1 is not necessary for in vitro propagation of T . gondii ., Disruption of ISP1 did not result in any gross defect in parasite growth ., However , we were surprised to find that both ISP2 and ISP3 were relocalized in the Δisp1 strain ., In the parental strain , ISP2 staining terminates sharply at the ring of TgCentrin2 annuli bordering the base of the apical cap ( Figure 7D , arrowheads ) ., However , in Δisp1 parasites , ISP2 staining extends past this border , relocalizing to the apical cap sub-compartment of the IMC ( Figure 7D ) ., Apical cap relocalization is also observed for ISP3 in the Δisp1 strain ( Figure 7E ) ., To ensure the ISP2 and ISP3 relocalization to the apical cap is truly a result of the absence of ISP1 , we reintroduced the ISP1 gene with a C-terminal YFP fusion into the Δisp1 strain ., This fusion protein targets correctly to the apical cap and , importantly , reestablishes the wild-type localization of ISP2 ( Figure 8A , insets ) and ISP3 ( data not shown ) , excluding them from the apical cap ., Thus , ISP1 exhibits a gate-keeping effect on ISP2 and 3 , preventing access to the apical cap and establishing a hierarchy of protein targeting among these IMC sub-compartments ., To determine if ISP1 performs a broader scaffolding function within the apical cap , we evaluated the localization of TgDLC1 using a GFP fusion; however , we observed no change in the localization of this protein in the absence of ISP1 ( data not shown ) ., Given the ability of ISP1 to exclude other family members from the apical cap , we exploited our ISP11–65-YFP construct to determine whether or not the N-terminal region that is sufficient for apical cap targeting also plays a role in exclusion from this compartment ., Expression of this construct in Δisp1 parasites does not result in exclusion of ISP2 ( Figure 8B ) or ISP3 ( data not shown ) from the apical cap , demonstrating that distal sequences present in the more conserved regions of ISP1 ( residues 66–176 ) are necessary for exclusion ., To further assess whether the C-terminal region from another ISP family member could substitute for the ISP1 C-terminal domain and function in exclusion , we constructed a hybrid protein containing the N-terminal 65 amino acids of ISP1 and the C-terminal region of ISP2 ( residues 43–160 ) fused to YFP ., Similar to the ISP11–65-YFP construct , the ISP1N/2C-YFP chimera targets to the apical cap but does not exclude ISP2 ( Figure 8C ) or ISP3 ( data not shown ) ., These results demonstrate that the exclusion activity of the C-terminal region of ISP1 is specific to this family member and cannot be replaced by the complementary region from ISP2 ., We created an additional chimera consisting of the N-terminal region of ISP2 ( residues 1–41 ) fused to the C-terminal region of ISP1 ( residues 67–176 ) ., While the N-terminal region of ISP2 alone targets YFP or HA throughout the IMC ( Figure 5F ) , inclusion of the C-terminal region of ISP1 restricts the localization to the apical cap and central regions of the IMC ( Figure 8D , see discussion ) ., In parasites expressing this chimera , ISP2 and 3 are mostly relocalized into the base portion of the IMC ( Figure 8E–F , brackets ) ., The fact that the ISP1 C-terminal region is able to exhibit exclusion activity against the other ISPs when artificially targeted to other domains of the IMC strengthens the conclusion that the ISP1 C-terminal region constitutes an ISP exclusion domain ., To further investigate the function of the ISP proteins , we disrupted the genes encoding ISP2 and ISP3 by homologous recombination ., To accomplish this , we employed a recently developed Δku80 parasite strain that is highly efficient at homologous recombination 38 ., We first removed HPT from the Ku80 locus by homologous recombination and negative selection using 6-thioxanthine , creating Δku80Δhpt strain parasites ., We then used this strain to disrupt ISP2 or ISP3 and confirmed these deletions by IFA ( not shown ) and Western blot ( Figure 9A and Figure S5 ) ., In contrast to our findings for Δisp1 parasites , localization of other ISP family members was unchanged in both Δisp2 and Δisp3 strains ( data not shown ) ., While no gross phenotype was seen in Δisp3 parasites , the Δisp2 strain parasites were obviously defective in growth as the knockout was rapidly lost from transfected populations and its isolation required cloning early following transfection ., To assess this loss in fitness , we performed competition growth assays between parent and Δisp2 parasites by mixing these strains in culture and monitoring the culture composition at each passage ., The parental strain rapidly out competed the Δisp2 parasites , confirming a severe fitness loss in these parasites ( Figure 9B ) ., Further analysis by IFA revealed that Δisp2 parasites display a number of defects in parasite division ., Most frequently , we observed the construction of >2 daughters per mother cell in each round of endodyogeny with some parasites assembling as many as 8 daughters ( Figure 9C ) ., To quantify this defect , we stained for ISP1 , an early marker for bud formation during endodyogeny , and counted vacuoles containing parasites undergoing endodyogeny and assembling >2 buds ., As expected , we saw a dramatic increase in the number of parasites producing more than two daughters in the Δisp2 strain ( Figure 9D ) ., Neither Δisp1 or Δisp3 parasites showed any aberration in daughter cell assembly compared to wild-type parasites ( data not shown ) ., Assembly of >2 daughters in Δisp2 parasites sometimes occurred around a single polyploid nucleus with karyokinesis accompanying budding ( bottom left parasite , Figure 9C ) while other parasites assembled the spindle apparatus and underwent karyokinesis without budding , resulting in a mother parasite with two nuclei ( Figure 9E ) ., We also observe parasites containing two discrete nuclei in the process of budding >2 daughters ( outlined parasites , Figure 9F ) ., Less frequently , we observed a catastrophic failure of Δisp2 parasites to appropriately segregate nuclei , resulting in anucleate zoids and nuclei extruded in the vacuole ( Figure 9G ) ., These vacuoles also show major defects in apicoplast segregation with a few cells receiving both a nucleus and an apicoplast while some received only an apicoplast and others received neither ., Finally , some vacuoles with nuclear segregation defects contained many immature buds within the vacuole ( Figure 9H ) ., These buds appear to be outside of any intact parasite and it is unclear if they were initiated within a mother cell and then somehow liberated into the vacuolar space or if they were the result of a budding event that was initiated within the vacuolar space itself ., In these vacuoles , several elongated apicoplasts are strung throughout the vacuolar space , associated with the extracellular buds and nuclei ., Surprisingly , the Δisp2 parasites recovered from both the fitness and replication defects after approximately two months of culture ( data not shown ) , preventing complementation by genetic rescue ., To ensure these phenotypes are specific to the disruption of ISP2 and not the consequence of any off target effects , we generated a second independent Δisp2 line ., This line displayed the same loss of fitness and cell division defects , indicating these phenotypes are specifically linked to disruption of the ISP2 locus ( data not shown ) ., Alveoli are the unifying morphological feature among ciliates , dinoflagellates and apicomplexans where these unique membrane stacks have been adapted to suit these divergent organisms in vastly different niches ., In apicomplexans , the membrane stacks ( the IMC ) have been exploited to provide unique and critical roles in parasite replication , motility and invasion ., Freeze-fracture studies reveal a highly sophisticated arrangement of IMC plates with dissimilar organization of IMPs in the apical versus lower plates indicating compositional differences between these regions 14 ., Identification of the ISPs clearly demonstrates that the protein constitution of the membrane cisternae is not uniform ., The ISP compartments have sharp boundaries ( Figure 2B–D ) , suggesting that they correspond to discrete cisterna or groups thereof ( Figure 10A ) ., ISP1 localizes to the apical cap compartment that is delimited by TgCentrin2 and thus represents the first membrane associated protein of this apical-most IMC plate ., Previously , the cytoskeleton-associated proteins PhIL1 and TgDLC1 were shown to localize in part to the apical cap region 20 , 22 ., The C-terminal half of PhIL1 is sufficient for apical cap localization and also for retaining cytoskeletal association ., This portion of the protein lacks predicted transmembrane domains or acylation signals , indicating that it links directly to a sub-domain of the cytoskeleton independent of the membrane stacks ., Electron micrographs of detergent-extracted parasites show substantial differences in the cytoskeletal filaments in this region ( e . g . thicker filaments and a parallel instead of interwoven arrangement ) , indicating that distinct sub-domains exist in both the IMC membranes and underlying network 13 ., Localization of ISP2 and 3 revealed two additional sub-compartments of the IMC that have not been previously observed: a central compartment labeled by ISP2 and a basal compartment labeled by ISP3 ., The abutment of ISP2 and ISP3 staining against the posterior end of the apical cap likely corresponds to the junction between the apical cap and the rectangular plates constituting the remainder of the IMC ., The presence of TgCentrin2 annuli at this border is striking as centrins are calcium-binding contractile proteins known to play a role in the duplication of microtubule organizing centers 39 ., While the ISP3 sub-compartment clearly terminates at the posterior end of the IMC , it is unclear what accounts for the basal boundary of the ISP2 sub-compartment which lies approximately two-thirds down the length of the parasite ., One possibility is an association with the cortical microtubules that also terminate in this region 40 ., However , the microtubules and ISP2 signal do not consistently terminate at the same point ., Alternatively , the signal termination may correspond to another junction of IMC plates and the exclusion of ISP2 from the basal region of the IMC may reflect another point of hierarchical targeting , as we discovered for ISP1 in the apical cap ., While ISP1 and 2 are both retained in mother parasites during endodyogeny , ISP3 maternal staining dissipates as daughter parasites form ., The strong ISP3 signal in early buds along with the rapid attenuation of ISP3 signal in the mother during endodyogeny provides an unhampered view of the membranes of the daughter buds ( Figure 2E and Video S1 ) ., Expression of IMC proteins is tightly regulated during the cell cycle including the ISPs , which show an expression profile similar to that of IMC1 ( Michael White , personal communication ) ., Thus , the bright ISP3 staining in daughters and concomitant loss of signal in mother cells could be due to synthesis in daughters and degradation in mothers ., Alternatively , since palmitoylation is a reversible lipid modification , recycling by de-palmitoylation at the parent IMC and re-palmitoylation at daughter IMCs could account for the ISP3 dynamics observed ., ISP1 and 3 are localized to numerous ring structures in oryzalin-treated parasites , indicating that initiation of bud IMC assembly repetitively occurs under these conditions and is not dependent on microtubules ., Microtubule polymerization is essential for cell division and cortical microtubule extension is thought to drive bud growth , ex

Introduction, Results, Discussion, Materials and Methods

Apicomplexans employ a peripheral membrane system called the inner membrane complex ( IMC ) for critical processes such as host cell invasion and daughter cell formation ., We have identified a family of proteins that define novel sub-compartments of the Toxoplasma gondii IMC ., These IMC Sub-compartment Proteins , ISP1 , 2 and 3 , are conserved throughout the Apicomplexa , but do not appear to be present outside the phylum ., ISP1 localizes to the apical cap portion of the IMC , while ISP2 localizes to a central IMC region and ISP3 localizes to a central plus basal region of the complex ., Targeting of all three ISPs is dependent upon N-terminal residues predicted for coordinated myristoylation and palmitoylation ., Surprisingly , we show that disruption of ISP1 results in a dramatic relocalization of ISP2 and ISP3 to the apical cap ., Although the N-terminal region of ISP1 is necessary and sufficient for apical cap targeting , exclusion of other family members requires the remaining C-terminal region of the protein ., This gate-keeping function of ISP1 reveals an unprecedented mechanism of interactive and hierarchical targeting of proteins to establish these unique sub-compartments in the Toxoplasma IMC ., Finally , we show that loss of ISP2 results in severe defects in daughter cell formation during endodyogeny , indicating a role for the ISP proteins in coordinating this unique process of Toxoplasma replication .

Apicomplexans are the cause of important diseases in humans and animals including malaria ( Plasmodium falciparum ) , which claims over a million human lives each year , and toxoplasmosis ( Toxoplasma gondii ) , which causes birth defects and neurological disorders ., These parasites possess a unique cortical system of membrane sacs arranged on a cytoskeletal meshwork , together referred to as the inner membrane complex ( IMC ) ., The IMC is the anchor point for the gliding motility machinery necessary for host invasion and also a scaffold around which new parasites are constructed during replication ., Here we have uncovered new insights into the organization and function of this structure by identifying and characterizing ISP1-3 , a family of proteins that define novel sub-compartments within the Toxoplasma IMC ., Residues predicted for myristoylation and palmitoylation are critical in the membrane targeting of these proteins , suggesting that multiple palmitoyl acyltransferase activities reside within the IMC and dictate its organization ., Surprisingly , ISP1 is required for proper sub-compartment sorting of ISP2 and 3 , revealing a novel hierarchical targeting mechanism for the organization of this membrane system ., Disruption of ISP2 results in defects during endodyogeny and a dramatic loss in parasite fitness , revealing that the ISP proteins play an important role in coordinating parasite replication .

cell biology/membranes and sorting, cell biology/microbial growth and development, infectious diseases/protozoal infections, microbiology/cellular microbiology and pathogenesis, cell biology/cytoskeleton

journal.pgen.1004102

Comprehensive Functional Annotation of 77 Prostate Cancer Risk Loci

The basic goal of research into human genetics is to connect variation at the genetic level with variation in organismal and cellular phenotype ., Until recently , inferences about such connections have been limited to the kind associated with heritable disorders and developmental syndromes ., Such variations often turn out to be the result of disruptions to protein coding sequences of critical enzymes for an affected pathway ., Recent advances in genomics and medicine have begun to illuminate a sea of variation of a more subtle variety , not always the result of mutation of protein coding sequences ., In particular , genome-wide association studies ( GWAS ) have identified thousands of variants associated with hundreds of disease traits 1 ., These variants , typically encoded by single nucleotide polymorphisms ( SNPs ) , are given landmark status and called ‘index-SNPs’ ( they are also frequently referred to in the literature as ‘tag-SNPs’ ) as the reference for disease or phenotype association in that region ., The vast majority of these variants reside within intergenic or intronic regions 2 , prompting at least two new avenues of inquiry:, 1 ) What is the nature and scope of risk encoded at these ‘non-coding’ loci ?, , and, 2 ) What are the target genes , and how do these alterations account for increased risk in a disease ?, At present , little is known regarding the functional mechanisms of the common variant susceptibility loci in non-coding regions ., For one , there are many genetically correlated variants that—to varying degrees—may account for the risk associated with each index-SNP ., It is unclear whether more than one variant carries functional consequences relevant to the risk that was reported ., In addition , we are only beginning to understand the nature of non-coding regions as revealed by histone modifications and other chemical signatures on chromatin ., Efforts to fill this void are underway , notably by the ENCODE consortium 3 , whose goal it is to catalog all the major chromatin biofeatures , including histone modifications , accessible chromatin and transcription factor bound regions in the form of digital footprinting and ChIP-seq for transcription factors , among others ., Currently , a mosaic of annotations for all the known histone modifications and 119 different transcription factors has been released for 147 cell types , including an androgen-sensitive prostate adenocarcinoma cell line isolated from lymph-node metastasis , called Lymph Node Cancer of the Prostate ( LNCaP ) 4–6 ., Insights into cancer biology of the prostate have already begun to emerge from this work ., For example , risk polymorphisms for the 8q24 locus have been extensively characterized in our lab and others 7 , 8 ., We propose that by identifying all the variants that are in linkage disequilibrium with GWAS SNPs and subsequently filtering down to those present within genome-wide functional annotations we will identify the most likely causal susceptibility variants within regulatory elements that can be tested for their functional significance ., We previously developed the R-Bioconductor package Funci–SNP} 2 which performs these operations , including the linkage disequilibrium calculations , based on data from the 1 , 000 genomes project ( www . 1000genomes . org 9 ) automatically ., With the advent of Funci–SNP} and similar tools such as RegulomeDB 10 , performing annotations of this type becomes possible , and indeed essential to understanding the candidate variations that may underlie risk for disease ., Post-GWAS analyses of breast cancer 11 for example identified putative functional variants using Funci{SNP} and genome-wide chromatin biofeature data for breast epithelia-derived cell lines as described above , but this level of detail is lacking for prostate cancer ., In that study , we catalogued and assessed the correlated functional variants at 72 breast cancer risk loci and performed preliminary enrichment analysis of motifs ., We identified over 1 , 000 putative functional SNPs , most of which were in putative enhancers ., We provide here a similar analysis for prostate cancer , extending the previous work and introducing some improvements to the downstream analyses ., We also present some new ChIP-seq datasets to add to ENCODE ., In order to identify variants that are in linkage disequilibrium with 77 prostate cancer risk loci ( defined as all significant GWAS , replication study and post-GWAS identified variants , see Table 1 for references ) , that are also relevant to the biology of prostate epithelia , we employed our bioinformatics tool , Funci{SNP} 2 to integrate biofeatures with 1000 genomes data 9 ( see Methods for a detailed list of biofeatures ) ., For the LNCaP cell line , genome-wide data are generally available both with and without androgen treatment ., Since the androgen receptor is a driver of prostate cancer 12 , we included both conditions where possible ., We also considered protein coding exons , and untranslated regions with miRcode target sequences ., Importantly , we also included the index-SNPs in our analysis ., We note that some critical datasets were not available when we initiated our studies ., For example , ChIP-seq data for the histone modification H3K27Ac was not available for LNCaP cells ., This is a mark of active enhancers , which are extremely cell-type specific ., Although other marks , such as DNase I hypersensitivity or H3K4me1 , can reveal regions of open chromatin , they do not identify active enhancers ., Therefore , we performed ChIP-seq for H3K27Ac in LNCaP cells , after a period of incubation in charcoal-stripped serum ( i . e . androgen depleted ) followed by exposure to vehicle control or physiological levels of the androgen dihydrotestosterone ( 10 nM DHT ) ., For LNCaP treated with vehicle ( minus DHT ) we observed 57 , 623 peaks , with an average peak height of 32 tags and median height of 22 tags , and a range of 9 to 212 tags ., The average peak width was 2 , 233 bp ., For LNCaP post-androgen stimulation , we observed 60 , 752 peaks , with an average peak width of 2 , 267 bp ., Overall the relative tag density and peak width distribution was extremely similar between the two conditions ( see Figure 1 , top and middle panels ) ., A plot of peak height vs . peak width reveals a linear relationship in log space ( Figure 1 , bottom panel ) ., Because we wanted to limit our studies to robust enhancers , we chose the top 25 , 000 peaks , which have a tag density of for use in Funci{SNP} ., This cutoff marks an inflection point where the number of tags increases geometrically over background ( Figure S1 ) ., A comparison of the top 25 , 000 H3K27Ac peaks detected before and after induction with DHT revealed an 84% overlap ( see Figure S2 ) , suggesting that only a small percentage of all H3K27Ac peaks are responsive to hormone treatment ., We also wished to include transcription factor binding data in our analyses ., Although there were data available for ChIP-seq of androgen receptor ( AR ) , FOXA1 and NKX3-1 , data for TCF7L2— another transcription factor with a proposed role in prostate- and other cancers 13— was not available ., Therefore we performed ChIP-seq for TCF7L2 in LNCaP ., We chose the top 15 , 000 peaks , with an average peak height of 57 tags and a range of 23 to 229 tags and an average peak width of 432 bp ., These properties are also displayed graphically in Figure, 1 . TCF7L2 binding sites were also highly enriched in the center of TCF7L2 ChIP-seq peaks ( Figure S3 ) ., Using Funci{SNP} , we identified 49 , 305 SNPs that were correlated in the population in which the original index SNP was reported within prostate epithelial chromatin biofeatures , of which only 727 had an value greater than or equal to 0 . 5 ( Figure 2A ) ., The most common SNP annotations are associated with H3K27-acetylation ( 385 SNPs ) and the other enhancer marks H3K4-monomethylation ( 231 SNPs ) and LNCaP DNaseI hypersensitivity ( 268 SNPs , see Figure 2B ) ., A complete visualisation of correlated SNPs with and all associated biofeatures are available on the UCSC genome browser; furthermore all custom tracks may be downloaded in bed format via the table browser therein: http://genome . ucsc . edu/cgi-bin/hgTracks ?, hgS_doOtherUser=submit&hgS_otherUserName=hazelett&hgS_otherUserSessionName=pca ., After identifying SNPs in primary biofeatures , we grouped them according to putative functional classes for further analysis ., We identified 30 SNPs in putative promoter regions −1000 bp to +100 bp relative to transcription start sites , 663 SNPs in putative enhancer regions , 4 SNPs in microRNA target sequences within or UTRs , and 27 SNPs in coding exons ( Figure 2C ) ., To directly observe the relationships of the annotations to each SNP across the entire set , we performed unsupervised clustering on the matrix of biofeatures and SNPs ( Figure 2D ) ., The resulting cluster diagram neatly captures the functional categories , but also reveals a cluster of SNPs in regions bound by multiple transcription factors ., Perhaps most importantly , Figures 2C and 2D clearly show that the majority of variation associated with risk for prostate cancer resides within what we have defined as putative risk enhancers ., We identified 27 exon SNPs in linkage disequilibrium with index SNPs for prostate cancer ( Figure 2B & 2C ) ., Of these SNPs , 13 encoded missense substitutions in coding exons , 14 encoded synonymous substitutions , and 0 corresponded to nonsense condons or other types of lesions ( Table 2 ) ., We conducted a preliminary exploration of the potential effects of the 11 missense variants using publically available software packages PROVEAN 14 , SIFT 15 , Polyphen2 16 , and SNAP 17 ., The results of this analysis are summarized in Table, 2 . All four algorithms predicted that a single index-SNP , the rare variant rs138213197 , encoding a Glycine to Glutamine substitution at position 84 of the homeobox transcription factor HOXB13 , has a deleterious effect ., Two other missense variants , rs2452600 ( ) and rs7690296 ( ) , correlated to index SNP rs17021918 , encoded potentially damaging changes in the PDLIM5 gene ., Three of four algorithms predicted rs2452600 to be damaging or non-neutral , and rs17021918 was only predicted to be non-neutral by SNAP ., Three missense variants in the MLPH gene were not predicted to be deleterious , but were highly correlated to each other and only weakly correlated to index SNP rs2292884 , raising the possibility that together they form a haplotype that weakens or damages protein function ., We next identified 29 and UTR SNPs , of which 4 occur within microRNA target element regions ., We cross referenced against highly conserved , high-scoring elements defined by miRcode 18 ., Index SNP rs4245739 was located within a miR target sequence in the UTR of the MDM4 gene ., This SNP was previously reported in functional annotation of iCOGS 19 for prostate cancer , esophogeal squamous cell carcinoma 20 and is a functional variant in breast cancer 21 ., The other three variants affect putative target sequences in the HAPLN1 , SLC22A3 , and FOXP4 genes , and are also of potential interest ( see Table 3 for details ) ., In order to identify putative functional variants within proposed enhancer and promoter regions , 663 SNPs from enhancers and 30 SNPs from promoters were queried against 87 positional weight matrices ( PWM ) compiled from Factorbook 22 ( see Methods ) ., Factorbook includes response element definition for the FOXA family of transcription factors , TCF7L2 , MYC , and GATA1 and -3 among others ., In addition we used PWMs from Homer 23 for FOXA1 , the androgen receptor ( AR ) and NKX3-1 ., We identified a subset of 509 variants in putative enhancers and 20 variants in promoter regions that disrupt response elements ( see UCSC genome-browser http://genome . ucsc . edu/cgi-bin/hgTracks ? hgS_doOtherUser=submit&hgS_otherUserName=hazelett&hgS_otherUserSessionName=pca ) ., For both promoters and enhancers we also identified a subset of disruptive variants that target response elements for factors of special interest to prostate cancer , namely AR , FOXA1 , NKX3-1 , TCF7L2 , MYC , GATA1 and GATA3 ., There were 6 SNPs in promoters and 177 in enhancers for this short list of PCa-specific factors ., These findings for PCa response elements are summarized in Figure, 3 . There are many densely situated independent risk loci in the 8q24 . 21 region centromeric of the MYC oncogene 19 , 24–34 , which therefore warranted additional consideration ., Figure 4 displays the region zoomed in to Mb ., Because 5C chromatin conformation capture data are available for the 8q24 region in LNCaP through ENCODE 3 , we examined the relationship of these data to our risk enhancers ., A circos plot showing interacting regions with the highest tag densities ( see histogram inset with dotted cutoff in Figure 4 ) reveals extensive overlap between putative risk enhancers and sites of intrachromasomal interaction ., Several SNPs effecting FOXA1 and ETS1 transcription factor binding sites in the vicinity of the POU5F1B locus are located within putative enhancer regions that interact in a complex manner with each other , with the POU5F1B coding region , and with both the MYC and FAM84B genes ., Another locus , the PCAT1 non-coding gene , has several SNPs affecting MYC , ETS1 and TCF7L2 candidate binding sites that potentially interact with the MYC gene locus ( Figure 4 ) ., Another putative enhancer situated between PCAT1 and CCAT1 non-coding RNA genes interacts with the enhancer telomeric of POU5F1B pseudogene and also with MYC ., It is striking from this view that 7 of the 16 index SNPs ( rs7837688 , rs1447295 , rs445114 , rs16902094 , rs188140481 , rs10086908 , rs12543663 ) do not overlap any biofeatures or chromatin 5C capture data , whereas the correlated enhancer SNPs with response element disruptions do ., These variants cluster within 5C-interacting regions despite having been filtered with LNCaP biofeatures , which are distributed evenly throughout the region ( see for example DNase I and FOXA1 tracks in Figure 4 ) ., These data are consistent with the hypothesis that some GWAS hits have no direct effect , but instead are correlated to nearby functional variants ., After the Funci{SNP} analysis , many index SNPs had redundant associations with correlated SNPs ., We examined each locus carefully to determine the number of unique and independent risk loci ., Starting from a list of 91 SNPs as input to Funci{SNP} , we determined that there were 77 loci that were independent ., We tabulated the independent risk loci in sequential order ( Table 1 ) in the genome ., In 25 of the 77 risk loci , we also were able to examine the LD structure for index SNPs that have been reported in two ethnic groups ., For these SNPs , we asked whether some SNPs had higher correlation with the index SNP in both GWAS-tested populations ( see Table 1 for population ) ., For example rs1512268 near the NKX3-1 gene , which reached genome-wide significance for both African and European populations ( see Table 1 for references ) , was correlated to 15 other SNPs at , but a single SNP , rs1606303 was highly correlated at in populations with both African and European ancestry ( Figure 5 ) ., Thus , we have also identified subsets of SNPs in the supplementary materials for rs12621278 ( Figure S4 ) , rs7584330 ( Figure S5 ) , rs17021918 ( Figure S6 ) , rs7679673 ( Figure S7 ) , rs12653946 ( Figure S8 ) , rs1983891 ( Figure S9 ) , rs339331 ( Figure S10 ) , rs9364554 ( Figure S11 ) , rs10486567 ( Figure S12 ) , rs6983267 ( Figure S13 ) , rs7127900 ( Figure S14 ) , rs10896449 ( Figure S15 ) , rs11228565 ( Figure S16 ) and rs8102476 ( Figure S17 ) present in different ethnic groups ., Nine other loci , at rs2710647 , rs6465657 , rs13252298 , rs7000448 , rs817826 , rs1571801 , rs10993994 , rs5759167 and rs5919432 did not have any SNPs at in both populations ., It is possible that the likeliest functional SNP in these cases is the index SNP ., One remaining SNP , rs5945572 in the NUDT11 region , was identified in African and European populations ( see Table 1 for refs . ) , and also correlated to the same three SNPs as two other index SNPs , rs1327301 and rs5945619 ., However , rs1327301 and rs5945619 , which were identified in Europeans ( see Table 1 for refs . ) surprisingly were not correlated to rs5945572 in Africans ., Two of the three correlated SNPs encode disruptions of MYC ( rs28641581 ) and AR ( rs4907792 , marked for functional followup , see below ) binding sites in putative enhancers ., Therefore , we hypothesize that all three index SNPs in this region are correlated to these other functional SNPs as the primary source of risk , and that together they constitute a single independent risk locus ( #76 in Table 1 ) ., We next asked whether the 663 enhancer SNPs were enriched for disruption in any of the 87 PWMs chosen from Factorbook and Homer ., In other words , we wanted to know whether disruption of any specific transcription factor response elements was associated with GWAS SNPs at greater than expected frequency ., We approached this question in two ways ., First , we asked whether response element disruptions were enriched against a background of randomly selected SNPs ., In order to ensure that we were drawing inference from the background distribution we drew samples ( ) of random SNPs ( ) , counted the number of motif disruptions for each of the 87 factors , and bootstrapped a 95% confidence interval on each PWM ., After applying the Bonferroni correction for multiple hypotheses , no factors remained significant ( Figure 6 , ) ., Second , we hypothesized that LNCaP cell-specific enhancer regions might differ from random SNPs in the relative abundance of some motifs , and therefore might be a more appropriate background ., To test this , we repeated the procedure of random selection of SNPs , this time filtering by the same genomic regions used in our Funci{SNP} analysis to define putative enhancers ., Figure 6 shows the relationship of the estimates to random background vs . random draws from LNCaP biofeatures ., To make the results comparable between different motifs , we expressed the observed motif disruptions as a statistic ., This statistic is a ratio of the difference in counts of disrupted motifs from the mean to the standard deviation ( see Methods , eq . 2 ) ., None of the factors of special interest in prostate cancer , i . e . MYC , FOXA , AR , GATA1 or 3 , ETS1 , TCF7L2 , and NKX3-1 , were enriched compared to LNCaP background ., The regression line ( in blue ) clearly indicated significant deviation from the line of unity , suggesting greater similarity of the GWAS correlated SNPs to random LNCaP biofeature SNPs compared to background , consistent with our hypothesis ., A Shapiro-Wilk test for normality revealed that the scores from LNCaP and random background are normally distributed ( and respectively ) ., Hence , the observed deviations were largely within the range of what we expected given a random sample of SNPs in LNCaP-specific biofeatures ., Prostate cancer is driven by androgen receptor signaling 12 , and is likely also influenced by basic cellular processes that contribute to other cancers 35 , 36 ., Therefore there are two classes of potential targets ., The first is the nearest gene ( s ) to the risk lesion , the exact location of which is somewhat uncertain but lies in a region of probability with a local maximum at the index-SNP ., In this category there are known oncogenes and tumor suppressors ., The second class , which does not exclude the first , comprises genes that are known targets of regulation by the androgen receptor ., We first took an inventory of nearby genes to the 77 risk loci ( see Table 1 ) and analyzed gene ontology enrichment using the annotation clustering tool at the DAVID bioinformatics site 37 ., The highest enrichment was for transcription factors ( enrichment score 4 . 08 , Figure 7A ) ., Overall , 20 DNA-binding transcription factors are directly associated with 35 out of 77 independent prostate cancer GWAS loci: HNF1B , AR , CTBP2 , RFX6 , OTX1 , HOXB13 , PAWR , FOXP4 , ZNF652 , ZBTB38 , VDR , NCOA4 , JAZF1 , NKX3-1 , VGLL3 , MDM4 , MYC , KLF4 , KLF5 and HDAC7 ., By inspection , we also identified at least 10 additional transcription factors within 500 kb of 9 other GWAS loci , that are also reasonable candidates for contributing to prostate cancer risk: SOX13 , ZFP36L2 , ATOH8 , DLX1 & DLX2 ( same locus ) , GATA2 , SKIL , SP8 , ASCL2 , and DPF1 ., Enrichment of broader categories of genes including transcriptional regulation ( enrichment score 3 . 44 ) , negative regulation of transcription ( enrichment score 2 . 52 ) , transcription and RNA metabolism ( enrichment score 2 . 06 ) , nuclear compartment annotations ( enrichment score 2 . 00 ) , and zinc-finger proteins ( enrichment score 1 . 46 ) was observed ., We also detected enrichment for genes involved in male gonad and sex differentiation ( enrichment score 1 . 53 , Figure 7B ) and gland development and branching morphogenesis clusters ( enrichment score 1 . 40 ) ., The DAVID website suggests 1 . 3 as an approximation for an equivalent of the group non-log 0 . 05 value cutoff 38 ., These findings suggest that genes involved in the regulation of transcription and the differentiation of male gonad structures may be overrepresented in genomic regions with heightened risk for prostate cancer ., In our second analysis we selected all nearby androgen regulated genes within 500 kb of putative functional variants ., There were 36 androgen regulated genes near 18 independent risk loci , including several from the list of transcription factors discussed in the previous section: MYC , GATA2 , NCOA4 , ZBTB38 , ZNF652 , NKX3-1 ., Other non-transcription factor genes were notable for being both androgen regulated and among the nearest in proximity to the GWAS hit , including KLK3 ( otherwise known as prostate serum antigen PSA ) , IGF2R , CHMP2B , BMPR1B , and the cell cycle reglator Cyclin D1 ( CCND1 ) ., Table 4 lists the genes and their relative expression in androgen-stimulated LNCaP cells ., To test the hypothesis that one or more of our putative functional polymorphisms disrupts a true transcription factor response element , we evaluated a sample of the enhancers in an in vitro heterologous enhancer-reporter luciferase assay in LNCaP cells ., In the absence of good prior information , we could not predict the magnitude of the effect of a variant at a single nucleotide in a strong consensus binding site on enhancer activity ., In order to obtain reliable inference on basal enhancer activity and response to androgen for possibly very slight changes , we eliminated other sources of variation such as plasmid preparation , batch and transfection effects ., Thus , we sampled evenly over this parameter space ( ) and used a hierarchical bayesian model to estimate the true enhancer activity and androgen ( DHT ) response , as well as the effect of SNP alleles on both ( see Methods , equation 3 ) ., The first enhancer containing rs113057513 , which encodes a consensus androgen response element ( Figure 8A ) near the androgen receptor gene , showed slightly elevated luciferase activity of 17 . 9% ( ) for the G allele after DHT treatment ( Figure 8D ) ., However , the difference is not biologically relevant and there was no basal activity for this enhancer relative to the negative controls ., In contrast to the enhancer at the AR gene locus , the enhancers near NUDT11 ( Figure 8B ) and in an intron of the JAZF1 transcriptional repressor gene ( Figure 8C ) showed a strong induction of - and -fold , respectively ., Even more strikingly , both SNPs had highly significant allele specific differences in DHT-induction ., Of the three enhancers that we tested , which all contain SNPs affecting a putative ARE , the enhancer containing rs10486567 in JAZF1 showed 10-fold elevated basal activity relative to controls ( Figure 8C ) ., All three enhancers showed significantly increased activity in the presence of DHT ( Figure 8D ) ., The NUDT11-enhancer at rs4907792 has either a T or a C allele ., The C allele creates a reasonably good androgen response element by the middle C of the ACA motif , whereas the T disrupts it ( see sequence logos , Figure 8B ) ., In our luciferase assay , we did not detect a difference between alleles in basal activity , however the T allele is weaker by an estimated 1 . 8-fold relative to the C allele after induction with DHT ., This 80% difference in the activity of the two alleles suggests that rs4907792 is critically important to the androgen sensitivity of this enhancer , and that the C allele of rs4907702 has more activity than the T allele ., For the JAZF1 enhancer , we detected a very significant difference of 1 . 39-fold ( 95% credible range of differences 1 . 21–1 . 61 ) in basal activity between the G and the A allele ( Figure 8C , salmon bars ) ., This particular locus is bound by the tumor suppressor NKX3-1 and the oncogene FOXA1 in LNCaP cells ( Figure 8C , gbrowse view ) and the SNP itself affects a critical residue in the response elements of both factors ( see logos in Figure 8C ) ., Thus , one version of rs10486567 , encoding a G , creates a strong consensus NKX3-1 response element at this position ., The alternate version of the SNP , encoding an A , destroys the NKX3-1 site in favor of an equally strong FOXA1 site ., Androgen Receptor also binds to the locus ( Figure 8C ) in LNCaP cells , and it is flanked by H3K4-monomethyl and H3K27-acetylation signals , providing additional evidence for this locus as a true enhancer ., Consistent with a role for androgen signaling at this enhancer , we observed a 6 . 7-fold induction for the A allele after DHT treatment ., We also detected significant allele-specific differences in DHT induction of 1 . 28-fold between A and G ( 95% credible range of differences 1 . 09–1 . 47 ) , with the A allele being the strongest ., Thus , there is an estimated mean difference of 28% in the magnitude of the androgen effect between the A and G alleles of rs10486567 ., Therefore , the risk associated with the C allele of rs4907792 creates a stronger androgen response element and increased NUDT11 expression by eQTL analysis 39 ., Interestingly , the risk associated with the G allele of rs10486567 in the JAZF1 intron creates an NKX3-1 binding site while destroying a FOXA1 binding site in line with the DHT-dependent decrease in enhancer activity; we would hypothesize that JAZF1 is likely a tumor suppressor influenced by this enhancer ., We have presented here the most comprehensive account and annotation of GWAS risk loci for prostate cancer that have been reported to date ., We believe that this has value not only as a framework upon which to test new hypotheses , but to stimulate other bioinformatics efforts going forward ., In the following sections we will discuss the implications of our findings with respect to the mechanisms of disease risk and the biology of human enhancers in such regions ., Finally , we will explore some possible approaches for discovery of true functional SNPs by experimental means , including this work ., One of our primary motivations for using Funci{SNP} is that it restricts the number of correlated SNPs to those with biofeatures in the relevant cell type ., We have chosen biofeatures associated with coding exons , microRNA regulatory targets , and most importantly , enhancers ., Some loci may confer risk by alternative mechanisms , such as ncRNA , but as these are not well understood at this time , we think it best to postpone that analysis until it becomes practical ., Furthermore , the vast majority of GWAS variants and their correlates lie well outside the regions where primary sequence features of that type ( i . e . exon annotations ) are present , hence we believe that many important risk variants will be identified within enhancer regions ., There are at least two other types of potential regulatory variation that are difficult to capture with this type of analysis ., One is alterations to the primary sequence that , by mechanisms which have yet to be elucidated , alter the pattern of nucleosome spacing or histone modification ., It is known that some sequences contribute to nucleosome positioning in chromatin 40–42 ., A second mechanism that we have not explored in our annotation is the effect of such polymorphisms on DNA methylation at CpG sites ., Such polymorphisms may contribute to variation in gene expression levels 43 ., Another issue is that many identified GWAS associations consist of common variants with only slightly elevated risk ( odds ratios in the range of 1 . 02 to 1 . 8 ( see Figure S18 ) ., We anticipate that such small magnitude of risk is associated with very small changes in the regulation of certain key genes ., Since many of the genes associated with risk loci are key regulators of development and cellular biology ( e . g . MYC ) , such disruptions are necessarily tissue specific and mild so as to confer slightly elevated risk over a lifetime , and perhaps with cumulative effects or environmental interaction ., So far the vast majority of GWAS risk that has been reported does not affect protein coding regions ., Indeed , as much as 77% of GWAS variation is associated with DNAse I hypersensitivity sites 44 ., Our findings are consistent with this: 663 of 727 SNPs are located in enhancers ., Moreover , 509 of these SNPs potentially disrupt known transcription factor response elements , vs . only 13 SNPs encoding putative missense mutations in proteins ., Our analysis of the missense variations in our correlated and index SNPs suggests that it is possible that a few of them encode damaging mutations , but this was by no means the unanimous conclusion from the various algorithms we tried ., The only clearly damaging variant was rs138213197 , which encodes a change from Glycine to Glutamate in the HOXB13 gene , and was previously reported to be associated with a high risk of prostate cancer 45 ., This result was also recently confirmed in a GWAS 46 ., Expression of HOXB13 is critical for mammalian prostate development 47 , and likely involved in carcinogenesis of the prostate as a tumor suppressor 48 , 49 ., The allele frequency of this variant is very low ( ) , possibly suggesting lower fitness in utero ., Furthermore the risk allele has an odds ratio of 4 . 42 46 and individual carriers are likely to contract prostate cancer at an earlier age 45 ., Nonetheless , it remains possible that even milder variants in one of the other proteins that we have catalogued in Table 2 also contribute to risk ., It will be necessary to do follow-up allele replacement experiments either in cell lines or in other model systems , e . g . mouse to determine the contribution to cellular or disease phenotype , if any ., In order to zero in on which SNPs are likely to be functional and causal , we need to know which of the putative enhancer regions are most likely to be true enhancers ., This information will come from a variety of sources including computational models using ENCODE data ., In addition , chromatin conformation capture experiments that elucidate the intrachromosomal looping , which brings transcription factors into association with the PolII complex at promoters and thereby promotes gene transcription will be vital to this effort ., ENCODE has provided some limited 5C chromatin interaction data for the MYC region , which we have superimposed on our Funci{SNP} results in Figure 4 ., These data show a clear relationship between the Funci{SNP} results and regions of chromatin that interact with both MYC and other genes in the region ., Despite the fact that chromatin biofeatures are scattered evenly throughout the region , the correlated SNPs appear to fall only within these special regions where intramolecular chromatin interactions are apparent ., It is also notable that the specialist transcription factors AR and NKX3-1 are restrict

Introduction, Results, Discussion, Conclusion, Materials and Methods

Genome-wide association studies ( GWAS ) have revolutionized the field of cancer genetics , but the causal links between increased genetic risk and onset/progression of disease processes remain to be identified ., Here we report the first step in such an endeavor for prostate cancer ., We provide a comprehensive annotation of the 77 known risk loci , based upon highly correlated variants in biologically relevant chromatin annotations— we identified 727 such potentially functional SNPs ., We also provide a detailed account of possible protein disruption , microRNA target sequence disruption and regulatory response element disruption of all correlated SNPs at ., 88% of the 727 SNPs fall within putative enhancers , and many alter critical residues in the response elements of transcription factors known to be involved in prostate biology ., We define as risk enhancers those regions with enhancer chromatin biofeatures in prostate-derived cell lines with prostate-cancer correlated SNPs ., To aid the identification of these enhancers , we performed genomewide ChIP-seq for H3K27-acetylation , a mark of actively engaged enhancers , as well as the transcription factor TCF7L2 ., We analyzed in depth three variants in risk enhancers , two of which show significantly altered androgen sensitivity in LNCaP cells ., This includes rs4907792 , that is in linkage disequilibrium ( ) with an eQTL for NUDT11 ( on the X chromosome ) in prostate tissue , and rs10486567 , the index SNP in intron 3 of the JAZF1 gene on chromosome 7 ., Rs4907792 is within a critical residue of a strong consensus androgen response element that is interrupted in the protective allele , resulting in a 56% decrease in its androgen sensitivity , whereas rs10486567 affects both NKX3-1 and FOXA-AR motifs where the risk allele results in a 39% increase in basal activity and a 28% fold-increase in androgen stimulated enhancer activity ., Identification of such enhancer variants and their potential target genes represents a preliminary step in connecting risk to disease process .

In the following work we provide a complete summary annotation of functional hypotheses relating to risk identified by genome wide association studies of prostate cancer ., In addition , we present new genome-wide profiles for H3K27-acetylation and TCF7L2 binding in LNCaP cells ., We also introduce the concept of a risk enhancer , and characterize two novel androgen-sensitive enhancers whose activity is specifically affected by prostate-cancer risk SNPs ., Our findings represent a preliminary approach to systematic identification of causal variation underlying cancer risk in the prostate .

genome-wide association studies, cancer genetics, genetic mutation, genomics, genetic association studies, human genetics, genetics, gene regulation, epigenetics, molecular genetics, biology, computational biology, molecular cell biology, population biology, genetics of disease, mutational hypotheses, histone modification

journal.pcbi.1004275

Automated High-Throughput Characterization of Single Neurons by Means of Simplified Spiking Models

In vitro patch-clamping is the gold standard used to investigate the intrinsic electrophysiological properties of neurons , but remains labour intensive and requires a trained experimentalist with high technical skills ., In the last years , several platforms have been developed that automatize electrophysiological recordings for ion-channel screening and drug discovery 1 ., Most of the existing platforms are , however , designed to record from mammalian cell lines or oocytes in which ion-channels of interest are artificially expressed 2 , 3 ., In the near future , this technology is likely to be transferred to more complex setups , such as in vitro brain slices ., High-throughput electrophysiology can be pushed forward with in vivo whole-cell patch-clamp recordings that are , at least partially , automatized 4 ., With this technique , three to seven minutes are sufficient for a trained technician or a robot to automatically identify a cell and form a gigaohm seal of the same quality as achieved by an electrophysiologist 4 ., This technological advance represents an important step towards high-throughput electrophysiology in vivo or on in vitro brain slices ., To make sense of the large amount of data that automated patch-clamp can produce , adequate computational tools and experimental protocols have to be developed ., Traditional protocols for single-neuron characterization rely on current-clamp injections of stimuli ( e . g . , square current pulses , ramps of current ) that are specifically designed to extract a small number of parameters ( e . g . , membrane time constant , firing threshold ) ., While this is a valid approach , the input currents adopted in these experiments are artificial and strongly differ from the signals that single neurons process in vivo ., Moreover , the choice of the parameters used for single-neuron characterization is arbitrary and different parameters are generally estimated in separate sets of experiments ., In this study , an alternative method is proposed in which the electrophysiological properties of neurons are characterized by means of simplified neuron models ., Ideally , a single-neuron model should be sufficiently complex and flexible to capture , by a single change of parameters , the spiking activity of different neurons , but also simple-enough to allow robust parameter estimation 5 , 6 ., Detailed biophysical models with stochastic ion channel dynamics can in principle account for every aspect of single-neuron activity; however , due to their complexity , they require high computational power 5 , 7–9 ., While systematic fitting of detailed biophysical models is possible 10–15 , most of the existing methods assume the knowledge of all the parameters that determine the dynamics of the ion channels included in the model ., Overall , a reliable and efficient fitting procedure for detailed biophysical models is not known 6 ., In a second class of spiking neuron models , which we call simplified threshold models , the biophysical mechanisms relevant for neural computation are not explicitly modeled , but are accounted for by phenomenological ( i . e . , effective ) descriptions 16 , 17 ., Despite their simplicity , threshold models are surprisingly good at predicting single-neuron activity 6 , 18–25 , at least for the case of single-electrode somatic stimulation ( but see 26 , 27 ) ., Nowadays , simplified threshold models are mainly used in large-scale simulations to study the emergent properties of neural circuits 28 , 29 ., By taking a different perspective , we will demonstrate that the same models can also serve an equally important purpose , namely to characterize the electrical properties of single neurons ., In this view , simplified threshold models are interpreted as computational tools to automatically compress the information contained in a voltage recording into a set of unique and meaningful parameters ., Summarizing the information of complex voltage recordings can in turn enable systematic comparisons , clustering and identification of cell types ., Finally , in patch-clamp experiments aimed at studying detailed aspects of the neuronal dynamics , automated online identification of neurons could allow for on the fly implementation of specific stimulus sets , which are best suited for the neuron under study ., After demonstrating that a limited amount of data , and little computing time , are sufficient to fit and validate our previous Generalized Integrate-and-Fire model ( GIF , see 30 , 31 ) , we introduce an experimental protocol that , combined with automated patch-clamp technology , could make automated high-throughput single-neuron characterization possible ., On the experimental side , the protocol relies on in vitro somatic injections of rapidly fluctuating currents that mimic natural inputs received in vivo at the soma of neurons ., On the computational side , the protocol is based on Active Electrode Compensation 32 , 33 , GIF model parameter extraction 30 , 31 and the spike-train similarity measure M d * 34 ., These computational methods are combined and implemented in a Python toolbox ( freely available at wiki . epfl . ch/giftoolbox ) ., The validity of our approach is finally demonstrated with two applications:, i ) in silico recordings obtained by simulating the activity of a multi-compartmental conductance-based model; and, ii ) in vitro recordings from L5 pyramidal neurons obtained using manual patch clamping ., We found that fitting and validating a GIF model takes approximatively five minutes ., Considering the time required to automatically establish a patch-clamp seal , the complete characterization of a single neuron can therefore be achieved in around ten minutes ., We conclude that GIF models are useful not only for network simulations , but also for rapid and systematic single-neuron characterization ., The GIF model discussed in this study 31 , 37 is a leaky integrate-and-fire model augmented with a spike-triggered current η, ( t ) , a moving threshold γ, ( t ) and the escape rate mechanism 38 , 39 for stochastic spike emission ( Fig 1A ) ., This model is able to predict both the spiking activity and the subthreshold dynamics of individual neurons ( Fig 1B ) , and it is flexible enough to capture the behavior of different neuronal cell types 37 ., In the model , the subthreshold membrane potential V, ( t ) evolves according to the following differential equation:, C V ˙ ( t ) = - g L ( V ( t ) - E L ) - ∑ t ^ j < t η ( t - t ^ j ) + I ( t ) , ( 1 ), where the parameters C , gL and EL define the passive properties of the neuron , I, ( t ) is the input current and { t ^ j } are the spike times ., According to Eq 1 , the passive properties of the membrane are described by an exponential filter κ ( t ) = R τ m exp ( − t τ m ) , with R = g L − 1 being the cell resistance and τm = RC being the membrane timescale ( Fig 1A ) ., Each time an action potential is fired , an intrinsic current with stereotypical shape η, ( t ) is triggered ., By convention , the spike-triggered current η, ( t ) is hyperpolarizing when its amplitude is positive and depolarizing otherwise ., Currents triggered by different spikes accumulate and produce spike-frequency adaptation , if η, ( t ) > 0 ( or facilitation , if η, ( t ) < 0 ) ., The functional shape of η, ( t ) varies among neuron types 37 ., Consequently the time course of η, ( t ) is not assumed a priori but is extracted from intracellular recordings ., Each time a spike is emitted , the numerical integration is stopped during a short absolute refractory period Tref and the membrane potential is reset to V ( t ^ j + T ref ) = V reset ., Spikes are produced stochastically according to a point process with conditional firing intensity λ ( t∣V , VT ) , which exponentially depends on the momentary difference between the membrane potential V, ( t ) and the firing threshold VT, ( t ) 22 , 39 , 40:, λ ( t | V , V T ) = λ 0 · exp ( V ( t ) - V T ( t ) Δ V ) , ( 2 ), where λ0 has units of s−1 , so that λ, ( t ) is in Hz and ΔV defines the level of stochasticity ., According to Eq 2 , if ΔV ≠ 0 , the probability of a spike to occur at a time t ^ ∈ t ; t + Δ t is given by:, P ( t ^ ∈ t ; t + Δ t ) = 1 - exp ( - ∫ t t + Δ t λ ( s ) d s ) ≈ λ ( t ) Δ t ., ( 3 ), In the limit ΔV → 0 , the model becomes deterministic and action potentials are emitted at the precise moment when the membrane potential crosses the firing threshold ., Importantly , the value of ΔV is extracted from experimental data ., Finally , the dynamics of the firing threshold VT, ( t ) is given by:, V T ( t ) = V T * + ∑ t ^ j < t γ ( t - t ^ j ) , ( 4 ), where V T * is a constant and γ, ( t ) describes the stereotypical time course of the firing threshold after the emission of an action potential ., Since the contribution of different spikes accumulates , the moving threshold defined in Eq 4 constitutes an additional source of adaptation ( or facilitation ) ., Similar to η, ( t ) , the functional shape of γ, ( t ) is not assumed a priori but is extracted from intracellular recordings ., All model parameters are summarized in Table 1 ., Given the intracellular voltage response Vdata, ( t ) evoked in vitro by a controlled input current Itr, ( t ) , all of the GIF model parameters are extracted from experimental data ( training set ) using a three-step procedure ( Fig 2 ) that we previously introduced 30 , 31 ., A detailed description of the fitting procedure can be found in the Materials and Methods section ., In Step 1 ( Fig 2 , Step 1 ) , the experimental spike train S data = { t ^ j } is first defined as the collection of instants t ^ j at which Vdata, ( t ) crossed a certain threshold from below ., The average spike shape VSTA, ( t ) is then obtained by computing the spike-triggered average ( STA ) of Vdata, ( t ) ., Depending on the cell type ( i . e . , depending on the average spike shape ) , the absolute refractory period Tref is fixed to a certain value and the reset potential is computed as Vreset = VSTA ( Tref ) ., In the GIF model , a period of absolute refractoriness can alternatively be implemented by setting the first milliseconds of the spike-triggered threshold movement γ, ( t ) to very large values ., For this reason , as long as Tref remains smaller than the shortest interspike interval ( ISI ) observed in the data , its precise value is not critical ., A sensible choice is to set Tref about twice the spike width at half maximum ., In Step 2 ( Fig 2 , Step 2 ) , the first-order temporal derivative of the experimental voltage V ., data ( t ) is estimated by finite differences and the parameters θsub = {C , gL , EL , η, ( t ) } determining the membrane potential dynamics are extracted by fitting Eq 1 on V ., data ( t ) ., This is done by exploiting the knowledge of the experimental voltage Vdata, ( t ) and the external input Itr, ( t ) ., To avoid a priori assumptions on the functional shape of the spike-triggered current , η, ( t ) is expanded in a linear combination of rectangular basis functions ., Consequently , optimal parameters minimizing the sum of squared errors between V ., ( t ) and V ., data ( t ) can be efficiently obtained by solving a multilinear regression problem 22 ( cf . Eqs 17–18 ) ., In Step 3 ( Fig 2 , Step 3 ) , the parameters estimated so far are first used to compute the subthreshold membrane potential of the model V ^ model ( t ) ., For that , Eq 1 is numerically solved by enforcing adaptation currents η, ( t ) at all the observed spike times { t ^ j } ., Given V ^ model ( t ) , the parameters θ th = { V T * , Δ V , γ ( t ) } defining the firing threshold dynamics ( cf . Eqs 2–4 ) are then extracted by maximizing the probability ( i . e . , the log-likelihood ) of the experimental spike train Sdata, ( t ) being produced by the GIF model ( cf . Eqs 20–21 ) ., Similar to η, ( t ) , the spike-triggered threshold movement is extracted by expanding γ, ( t ) in a linear combination of rectangular basis functions ., Since the parameters λ0 and V T * are redundant , λ0 is fixed to 1 Hz ., With the exponential function in Eq 2 , the log-likelihood to maximize is guaranteed to be a concave function of θth41 and the optimization problem can be solved using standard gradient ascent techniques ., The method used in this last step closely resembles the standard GLM fitting procedure 35 , 36 ., However , here , by exploiting the information contained in the subthreshold dynamics of the membrane potential , the maximum likelihood approach is specifically used to infer the dynamics of the firing threshold ., In contrast to GLMs , the GIF model can consequently disentangle adaptation processes mediated by intrinsic currents and threshold movements ., To obtain a high-throughput pipeline for GIF model parameter extraction , the method described in the previous section has to be complemented with a validation protocol designed to automatically detect and discard trials in which the fitting procedure fails ., Good spiking neuron models should be able to accurately predict the occurrence of individual action potentials with millisecond precision 6 ., To take into account the stochastic nature of single neurons 42 , we designed a validation protocol based on the measurement of the model performance in predicting spike emission probability ., After the acquisition of the training dataset used for parameter extraction , a new set of recordings ( test dataset ) is performed in which single neurons are stimulated repetitively with a test current Itest, ( t ) ., The resulting set of experimental spike trains is then compared against a set of spike trains predicted by repetitive simulations of the GIF model ., To obtain a quantitative measure of the model’s predictive power , the similarity M d * 34 between the two sets of spike trains is computed ( Materials and Methods ) ., M d * takes values between 0 and 1 , where M d * = 0 indicates that the model is unable to predict any of the experimental spikes and M d * = 1 indicates a perfect match ., Importantly , M d * avoids the small-sample bias known to occur when measuring the similarity between small groups of spike trains as well as the deterministic bias known to favor noise-free models 34 ., To estimate the amount of data required to perform GIF model parameter extraction , we first tested our fitting procedure on an artificial training set generated by simulating the response of a GIF model to a fluctuating current I, ( t ) ., The choice of reference parameters ( Fig 3A–3D , black ) was based on previous results 31 ., In particular , both the spike-triggered current η, ( t ) and the threshold movement γ, ( t ) were defined as a linear combination of K = 26 log-spaced rectangular basis functions approximating a power-law decay over 5 seconds 31 , 43 ., Overall , the reference model had 59 parameters: 31 were related to the subthreshold dynamics and 28 to the firing threshold ., The input current I, ( t ) used to build the artificial training set was generated at ΔT −1 = 20 kHz by numerically solving the stochastic differential equation τ I ., = − I + I 0 + 2 τ σ ( t ) ξ ( t ) in discrete time, I ( t + Δ T ) = I ( t ) + I 0 - I ( t ) τ · Δ T + 2 σ 2 Δ T τ · 𝓝 ( 0 , 1 ) , ( 5 ), where ξ, ( t ) is a Gaussian white-noise process generated by independently sampling from a Normal distribution 𝓝 ( 0 , 1 ) , τ = 3 ms is the characteristic timescale on which the input fluctuates , I0 defines the mean input and σ, ( t ) is the time-dependent standard deviation of I, ( t ) ., Ornstein-Uhlenbeck processes ( i . e . stationary filtered Gaussian processes ) have been extensively used to model the input current received in vivo at the soma of neocortical neurons 44 ., Here , we relaxed the assumption of stationarity by modulating the variance of the input with a periodic oscillation 43 given by:, σ ( t ) = σ 0 ( 1 + Δ σ sin ( 2 π f t ) ) , ( 6 ), where σ0 and Δσ are constants and f = 0 . 2 Hz is the modulation frequency ., An input current with non-stationary statistics drives the neurons through different regimes producing broad ISI distributions that better constrain the fit of adaptation processes ., The input parameters I0 , σ0 and Δσ were adjusted to generate an artificial training set in which the GIF model emitted spikes at an average firing rate of 10 Hz oscillating over 5 seconds between around 7 and 13 Hz ., The fitting procedure illustrated in Fig 2 was then applied to recover the reference parameters of the GIF model used to generate the artificial dataset ( Fig 3A–3D , black ) ., To estimate the amount of data required to guarantee a high degree of accuracy , this operation was repeated several times by varying the size of the training set Ttr ( i . e . , the duration of the input current I, ( t ) ) ., Fig 3A–3D shows a comparison between the reference parameters and the results obtained by fitting a training set of Ttr = 10 seconds ( gray ) and Ttr = 100 seconds ( red ) ., Overall , we found that 100 seconds were sufficient to accurately recover the reference parameters ., To quantify the accuracy of the fit , we computed the mean error ϵparam on model parameters as a function of Ttr ( see Materials and Methods ) ., The results indicate that the minimum amount of data required for accurate parameter extraction is 30–40 seconds ., In particular , we found that 100 seconds were sufficient to limit the error to ϵparam < 2 . 0% ( Fig 3E , top ) ., The great accuracy with which the fitted model was able to predict the spiking activity of the reference model ( M d * = 0 . 998 ) confirmed the goodness of this fit ( Fig 3E , middle ) ., To achieve high-throughput and perform parameter extraction on the fly , it is crucial to minimize the computing time ( CPU time ) required for the fit ., We measured the CPU time as a function of the training set duration Ttr ( Fig 3E , bottom ) and we found that accurate parameter extraction from a training set of Ttr = 100 seconds requires around 60 seconds of computing ., We concluded that GIF model parameter extraction is suitable for high-throughput ., A second time-consuming procedure that has to be analyzed is the validation protocol ., To quantify the predictive power of the fitted model , the reference model was stimulated with repetitive injections of a test current Itest, ( t ) generated according to Eqs 5–6 ., To estimate the number of repetitions ntest and the duration Ttest of the test current required to obtain a reliable estimate of the model predictive power , the similarity measure M d * was computed multiple times using different values of ntest and Ttest ( Fig 3F ) ., On average , the value of M d * was independent of both the input current duration and the number of repetitions , confirming that the spike-train metrics M d * successfully eliminates the small sample bias 34 ., We measured the variability of M d * across validation procedures performed with different realizations of Itest, ( t ) and found that the reliability of M d * increased with both the number of repetitions ntest and the duration of the test current Ttest ( Fig 3F ) ., Spike-triggered processes can last for several seconds 31 , 43 ., This sets a constraint on the minimal duration of both the test current Itest, ( t ) and the interstimulus interval ., By taking into account these constraints , we concluded that , while respecting high-throughput constraints , a validation protocol based on nine injections of a 10-second current guarantees a reliable estimation of the model’s predictive power ( Fig 3F ) ., Based on the results reported in the previous section , we designed a protocol for the fit and the validation of GIF models on in vitro intracellular recordings ( Fig 4 ) ., The protocol is conceptually divided in two phases ., In the first part , a training set is acquired by recording the single-neuron response to a fluctuating input Itr, ( t ) lasting for Ttr = 100 seconds and generated according to Eqs 5–6 ., These data are then used for parameter extraction ., In the second part of the protocol , nine repetitive injections of a new 10-second current Itest, ( t ) are performed with an interstimulus interval of 10 seconds , so as to allow the cell to recover ., These data ( test set ) are then used to quantify the predictive power of the GIF model with the spike-train similarity measure M d * ., Since all the computations required for parameter extraction and model validation can be performed on the fly , the whole protocol requires 5 minutes and is suitable for high-throughput ., Current-clamp experiments in which the same electrode is used both for stimulating and recording from single neurons are biased due to the voltage drop across the electrode 32 ., To remove this bias , intracellular recordings are preprocessed using a technique called Active Electrode Compensation ( AEC , refs . 32 , 33 , see Materials and Methods ) ., To perform AEC , the filtering properties of the electrode have to be estimated ., For that , an additional 10-second subthreshold current injection is performed before the acquisition of the training set ( Fig 4 ) ., A different class of models used to describe the electrical activity of individual neurons includes the so called multi-compartment conductance-based models ( or detailed biophysical models ) ., In contrast to point-neuron models , detailed biophysical models account for the intricate morphology of both dendritic and axonal arborizations and explicitly describe the dynamics of a large variety of ion channels mediating active currents ., Both aspects are likely to play a role in single-neuron information processing 45 , 46 ., A detailed biophysical model ( DBM ) has recently been proposed that captures several features of L5b thick-tufted pyramidal neurons 14 ., In particular , this model includes active dendrites and describes the interactions between Na+-spiking at the soma , back-propagating action potentials and Ca2+-spikes generated at the distal apical dendrites ., To validate our procedure for high-throughput single-neuron characterization , the protocol described in Fig 4 was tested in silico by simulating the DBM response to a set of current injections ( Fig 5A , see Materials and Methods ) ., The input parameters were calibrated to obtain an average firing rate of 10 Hz with slow rate fluctuations between 7 and 13 Hz ., Moreover , to model stochastic spike emission , a source of noise was introduced by corrupting the input current with some additive white-noise ( see Materials and Methods ) ., Capturing the DBM spiking response to dendritic injections goes beyond the scope of this study ., Since we are ultimately interested in automatic somatic patching , all in silico experiments were preformed by delivering the current at the somatic compartment ( Fig 5A ) ., DBM somatic recordings were then used to perform GIF model parameter extraction ( Fig 5B–5D ) ., Compared with previous results from in vitro recordings in L5 pyramidal neurons 30 , 31 , the membrane filter κ, ( t ) was characterized by a relatively short timescale ( τm = 6 . 7 ms , s . d . 0 . 1 ms , Fig 5B ) ., GIF model parameter extraction also revealed the presence of a long-lasting adaptation current ( Fig 5C ) as well as a long-lasting spike-triggered movement of the firing threshold ( Fig 5D ) ., Consistent with the tendency of L5b pyramidal neurons to produce bursts of action potentials ( ref . 14 and Fig 5G ) , the activation of the spike-triggered current was not instantaneous ., According to cable theory 47 , the large number of dendritic branches explicitly modeled in the DBM , is expected to manifest itself in a membrane filter κ, ( t ) decaying over multiple timescales ., To verify the accuracy of the single-exponential assumption and to compare the GIF model performance against a reference model , we also used the in silico recordings to fit a GLM 35 , 36 , ( Fig 5E and 5F ) ., In the GLM , the linear filter κGLM, ( t ) acting on the input current is not assumed a priori to be an exponential function and its shape is extracted from experimental data using a non-parametric method ( see Materials and Methods ) ., We found that the GLM filter κGLM, ( s ) and the membrane filer κ, ( t ) of the GIF model were in good agreement ( Fig 5E ) , suggesting that complex dendritic morphologies weakly affect temporal integration at the somatic compartment ., Further quantitative evidence was provided by fitting κGLM, ( t ) with a single exponential function and comparing the resulting timescale against τm ( Fig 5B , inset ) ., The GLM spike-history filter hGLM, ( t ) extracted from in silico recordings ( Fig 5F ) was also in good agreement with the effective adaptation filter h, ( t ) of the GIF model 31 , 37:, h ( t ) = ∫ 0 ∞ κ ( s ) η ( t - s ) d s + γ ( t ) ., ( 7 ), This result confirmed that hGLM, ( t ) combines , but cannot disentangle , the effects of the adaptation current η, ( t ) and the movement of the firing threshold γ, ( t ) ., In contrast to GIF models , GLMs do not model absolute refractoriness with a dead time followed by a voltage reset ., This explains why , during the first milliseconds , hGLM, ( t ) is much larger than h, ( t ) ( Fig 5F ) ., Finally , consistent with previous results that in L5 pyramidal neurons spike-frequency adaptation occurs on multiple timescales 31 , 43 , we noticed that both h, ( t ) and hGLM, ( t ) were approximatively linear on double logarithmic scales ( Fig 5F , inset ) ., The predictive power of both the GIF model and the GLM was then assessed on a test set obtained by simulating the DBM response to nine repetitive injections of a new 10-second current ( Fig 6A ) ., Both models achieved a similar performance and were able to predict around 80% of the spikes emitted by the DBM ( temporal precision Δ = 4 ms; M d * = 0 . 80 , s . d . 0 . 01 , GIF; M d * = 0 . 79 , s . d . 0 . 01 , GLM; Fig 6B ) ., Compared to the GLM , the GIF model presented two advantages ., First , the GIF model , but not the GLM , explicitly modeled the dynamics of the membrane potential and could therefore predict the DBM subthreshold voltage with an average root mean squared error ( RMSE ) of 3 . 4 mV , s . d . 0 . 03 mV ( variance explained ϵV = 74 . 3 % , s . d . 1 . 1%; Fig 6C ) ., Second , the time required to perform parameter extraction was faster for the GIF model than for the GLM ( TCPU = 86 s , GIF; TCPU = 143 s , GLM ) ., Repeating the entire protocol by varying the duration of Itr, ( t ) confirmed that a training set of Ttr = 100 s was sufficient to ensure convergence of the fitting procedure ( Fig 6D ) ., Overall , these results suggest that , despite their simplicity , modern point-neuron models are capable of predicting most of the spikes emitted by a detailed biophysical model in response to complex somatic current injections ., To confirm the results reported in the previous section , the protocol for high-throughput single-neuron characterization was further tested using standard current-clamp in vitro recordings from L5 pyramidal neurons ( see Materials and Methods ) ., At the beginning of the experiment , the input current was calibrated to obtain an average firing rate of 10 Hz with amplitude fluctuations between 7 and 13 Hz ., For that , we set Δσ = 0 . 5 , I0 = σ0 and adjusted I0 in order to obtain an average firing rate of around 10 Hz ., While this simple approach works well for L5 pyramidal neurons , different cell types might require a more involved calibration protocol in which I0 ≠ σ0 ., In these cases , an alternative solution consist of:, i ) temporarily setting I0 = 0 pA and looking for two values of σ0 , denoted σ 0 min and σ 0 max , giving rise to subthreshold voltage fluctuations σV of desired magnitudes ( e . g . , σ V min ≈ 3 mV and σ V max ≈ 7 mV ) ;, ii ) set σ 0 = ( σ 0 min + σ 0 max ) / 2 and Δ σ = ( σ 0 max − σ 0 min ) / 2 σ 0;, iii ) adjust I0 to obtain an average firing rate of around 10 Hz ., Since the same patch-clamp electrode was used to simultaneously stimulate and record from single neurons , the acquired signal Vrec, ( t ) is a biased version of the real membrane potential Vdata, ( t ) 32 , 33 ., This bias is due to the voltage drop Ve, ( t ) across the patch-clamp electrode and was removed using a technique called Active Electrode Compensation ( AEC , see Materials and Methods and Fig 7A ) ., In AEC 32 , 33 , the electrode is modeled as an arbitrarily complex linear filter κe, ( t ) estimated at the beginning of the experiment from the optimal linear filter κopt, ( t ) between a 10-second subthreshold current Isub, ( t ) and the recorded response Vsub, ( t ) ( Fig 7B ) ., For all subsequent injections , we estimated the voltage drop across the electrode Ve, ( t ) by convolving the input current with the electrode filter κe, ( t ) ( Fig 7C ) ., We finally recovered the membrane potential Vdata, ( t ) by subtracting Ve, ( t ) from the recorded signal Vrec, ( t ) ( Fig 7A and 7D ) :, V data ( t ) = V rec ( t ) - V e ( t ) ., ( 8 ) According to our high-throughput protocol , the training set was compensated only after its complete acquisition ., With this strategy , the time-consuming procedure required to estimate the electrode filter can be performed during the acquisition of the training set ( see Fig 4 ) , limiting the total duration of the protocol ., Consistent with previous results 31 , we found that the electrode filter κe, ( t ) decayed on a very rapid timescale τe = 0 . 54 ms , s . d . 0 . 11 ms ( Fig 7C ) ., Consequently , AEC acted on the raw data as a low-pass filter with a cutoff frequency of around 2 kHz ., After AEC , the in vitro recordings acquired from ten different L5 pyramidal neurons ( Fig 8A ) were used to perform GIF model parameter extraction ( Fig 8B–8E ) ., All of the extracted parameters were consistent with the ones previously obtained by fitting the GIF model on in vitro recordings from L5 pyramidal neurons responding to a mean-modulated input 31 ., The parameters describing the passive properties of the membrane ( i . e . , the resting membrane potential EL , the membrane timescale τm and the input resistance R ) revealed the presence of cell-to-cell variability ( Fig 8B and 8E ) and were on average consistent with previous results obtained using standard characterization protocols based on current-step injections 48 ., Our characterization approach further showed that , in L5 Pyr neurons , spike-frequency adaptation is mediated by a long-lasting adaptation current featuring a power-law decay ( Fig 8C; see also ref . 31 ) , which possibly results from the combined action of multiple channels operating on different timescales 49 ., In standard protocols for single-neuron characterization , spike-triggered currents are generally assessed indirectly by measuring the spike after-hyperpolarization ( AHP ) induced by an artificial current pulse designed to evoke one or more action potentials ( see , e . g . , refs . 49 , 50 ) ., Importantly , with our characterization method the magnitude and the time-course of adaptation currents mediating AHPs can be measured simultaneously along with the other GIF model parameters , while neurons are processing in vivo-like inputs ., Finally , our characterization protocol showed the presence of large firing threshold movements triggered by the emission of action potentials and lasting for several hundreds of milliseconds ( Fig 8D ) ., The dynamical properties of the firing threshold have been previously shown to be cell-type specific 30 , 51 and functionally relevant 52 , but are generally not considered in standard characterization protocols ., To allow for a comparison , we also used the in vitro recordings to perform GLM parameter extraction ( Fig 8F–8H , see Materials and Methods ) ., Confirming the results reported in the previous section , the effective spike-history filter h, ( t ) of the GIF model obtained by combining the spike-triggered current η, ( t ) and threshold movement γ, ( t )

Introduction, Results, Discussion, Materials and Methods

Single-neuron models are useful not only for studying the emergent properties of neural circuits in large-scale simulations , but also for extracting and summarizing in a principled way the information contained in electrophysiological recordings ., Here we demonstrate that , using a convex optimization procedure we previously introduced , a Generalized Integrate-and-Fire model can be accurately fitted with a limited amount of data ., The model is capable of predicting both the spiking activity and the subthreshold dynamics of different cell types , and can be used for online characterization of neuronal properties ., A protocol is proposed that , combined with emergent technologies for automatic patch-clamp recordings , permits automated , in vitro high-throughput characterization of single neurons .

Large-scale , high-throughput data acquisition is revolutionizing the field of neuroscience ., Single-neuron electrophysiology is moving from the situation where a highly skilled experimentalist can patch a few cells per day , to a situation where robots will collect large amounts of data ., To take advantage of this quantity of data , this technological advance requires a paradigm shift in the experimental design and analysis ., Presently , most single-neuron experimental studies rely on old protocols—such as injections of steps and ramps of current—that rarely inform theoreticians and modelers interested in emergent properties of the brain ., Here , we describe an efficient protocol for high-throughput in vitro electrophysiology as well as a set of mathematical tools that neuroscientists can use to directly translate experimental data into realistic spiking neuron models ., The efficiency of the proposed method makes it suitable for high-throughput data analysis , allowing for the generation of a standardized database of realistic single-neuron models .

journal.pcbi.1004145

A Dynamical Phyllotaxis Model to Determine Floral Organ Number

How to determine the numbers of body parts is a fundamental problem for the development of complete body structures in multicellular organisms ., Digit numbers in vertebrates are evolutionarily optimized for the specific demands of the organism 1; the body-segment number in insects is constant despite the evolutionarily diversified gene regulation in each segment 2–4; and five petals are indispensable to forming the butterfly-like shape that is unique to legume flowers 5 ., Studies of animal structures , such as vertebrate limbs and insect segments , strongly suggest that crosstalk between pre-patterns ( e . g . , morphogen gradients ) and self-organizing patterns underlies the developmental process of organ-number determination 6–13 ., In plant development , a self-organization based on the polar transport of the phytohormone auxin 14–16 is conserved among seed plants 17 and seems to be the main regulator of the development of a hierarchal body plan , called a shoot , consisting of a stem and lateral organs such as leaves ., The number of concentration peaks in most self-organizing patterns , such as Turing pattern and the mechanisms proposed for plant-pattern formation , is proportional to the field size 15 , 18 , 19 ., Despite having a diversified field size for floral-organ patterning , the eudicots , the most diverged clade among plants , commonly have pentamerous or tetramerous flowers containing five or four sepals and petals ( the outer floral organs ) , respectively , and rarely have other numbers of organs 20 , 21 ., Here , we focus on the developmental properties that so precisely and universally determine the floral organ numbers through self-organizing processes ., Phyllotaxis , the arrangement of leaves around the stem , provides insight into floral development , because studies of floral organ-identity determination 22 have verified Goethe’s foliar theory , which insists that a flower is a short shoot with specialized leaves 23 ., Phyllotaxis is mainly classified into two types: spiral phyllotaxis , which has a constant divergence angle and internode length , and whorled phyllotaxis , which has several leaves at the same level of a stem 24 ., For spiral phyllotaxis , Hofmeister described a hypothesis of pattern formation in 1868 24 , which we summarize in three basic rules: the time periodicity of primordia initiation , the initiation of a primordium at the largest available space at the edge of the meristem ( the undifferentiated stem-cell region ) , and the relative movement of primordia in a centrifugal direction from the apex due to the growth of the stem tip ., Following that hypothesis , numerous mathematical models incorporating contact pressure 25 , 26 , inhibitor diffusion 27 , reaction-diffusion 18 , 28 , and mechanical buckling of the epidermis 29 , 30 were proposed to explain the observed phyllotactic patterns ., Over the past ten years , these mathematical models were tested and interpreted in light of modern molecular biology ., Several studies have suggested that the competitive polar transport of the auxin accounts for two of Hofmeister’s rules , the periodicity of initiation and the initiation at the largest space , which together are capable of reproducing both spiral phyllotaxis and whorled phyllotaxis 15 , 16 , 31 ., Despite their simple rules and uncertain molecular basis , the phyllotaxis models can account for several of the quantitative properties observed in organ patterning ., For example , one model showed that the divergence angle between successive leaves is 180 degrees for the first and second leaves , 90 degrees for the second and third leaves , and oscillating thereafter , converging to the golden angle , 137 . 5 degrees , which agrees with the phyllotaxis of true leaves in Arabidopsis thaliana after the two cotyledons 32 , 33 ., Similar oscillatory convergence to a particular divergence angle occurs in the sepal primordia of the pentamerous flower of Silene coeli-rosa , Caryophyllaceae ., In S . coeli-rosa , the divergence angle is 156 degrees at first , and then it oscillates , converging on 144 degrees 34 ., The golden angle also appears in the floral organs of several Ranunculaceae species 35 , 36 ., The agreements between the phyllotaxis models and actual floral development suggest that mathematical models can give useful clues to the underlying mechanisms of not only phyllotaxis but also floral organ patterning ., There are at least three fundamental differences , however , between real floral development and the phyllotaxis models ., The first difference is the assumption of constant primordium displacement during tip growth , which comes from Hofmeister’s hypothesis and has been incorporated into most phyllotaxis models ., Although the helical initiation has been thought to always result in spiral phyllotaxis , many eudicots form the whorled-type sepal arrangements in their blooming flowers subsequent to helical initiation 37 ( Fig 1; e . g . , Caryophyllaceae 34 , Solanaceae 38 , Nitrariaceae 39 , and Rosaceae 40 ) ., The remnants of helical initiation are more obvious in the pseudo-whorls ( e . g . , Ranunculaceae 41 ) , where the distance between each organ primordium and the floral center varies slightly even in the whorls of mature flowers , which usually have more varied floral organ numbers 20 , 35 , suggesting that post-meristematic modifications of primordia positions 42 play an essential role in generating the whorled arrangement and determining the floral organ number during floral development ., In contrast , most phyllotaxis models have assumed constant growth of the primordia , so that the whorls appear only after the simultaneous initiation of several primordia 19 ., The second difference comes from the fact that floral development is a transient process , whereas most phyllotaxis models have focused on the steady state of the divergence angle ., Although the golden angle ( 137 . 5 degrees ) is quite close to the inner angle of regular pentagon ( 144 degrees ) , the developmental convergence from 180 degrees ( cotyledon ) to 137–144 degrees in phyllotaxis requires the initiation of more than five primordia , both in A . thaliana leaves and in the mathematical models 16 , 33 ., In contrast , the divergence angle between the second and third sepal primordia in pentamerous eudicot flower development is already close to 144 degrees 34 ., The third difference comes from the accuracy of the floral organ number in many eudicots ., Although the polar auxin-transport model reproduced both wild-type and mutant A . thaliana floral organ positioning 43 , the organ number in the model was more variable , even with an identical parameter set ( Fig 3 in 43 ) , than that in experimental observations ( Table 1 in 44 ) ., Moreover , among eudicot species , the appearance of pentamerous flowers is robust , despite the diversity of the meristem size and the outer structures , including the number and position of outside organs such as bracts 20 ., Together , the differences between real floral development and previous phyllotaxis models indicate that floral development requires additional mechanisms to determine the particular organ number ., To resolve the inconsistencies between the earlier models and actual floral development , we set out a simple modeling framework , integrating Hofmeister’s rules with two additional assumptions , namely , the repulsion between primordia that can repress primordium growth and the temporal decrease in initiation inhibition of new primordium , which were proposed independently in the contact pressure model 25 , 45 , 46 and the inhibitory field model 33 , 47 , 48 , respectively , for phyllotaxis ., First , when we incorporated mutual repulsion among primordia into the growth process , a whorled-type pattern emerged spontaneously following the sequential initiation of primordia ., The mutual repulsion obstructed the radial movement of a new primordium after a specific number of primordia arose , causing a new whorl to emerge ., The number of primordia in the first whorl tended to be four or eight ., Second , when we assumed that older primordia have less influence on the initiation of a new primordium , the pentamerous whorl arrangement , which is the most common arrangement in eudicot flowers , became dominant ., We analytically show the conditions for the development of tetramerous and pentamerous whorls , and we predict possible molecular and physiological underpinnings ., Following the earlier models 49 , we represented the meristem as a circular disc with radius R0 and the primordia as points ( Fig 2A ) ., A new primordium arises at the point along the edge of the meristem ( R0 , θ ) , in polar coordinate with the origin at the meristem center , where θ gives the minimum value of the inhibition potential Uini ., As one of the simplest setups for sequential initiation 37 , we followed the assumption of earlier models for spiral phyllotaxis 49 , which state that new primordia arise sequentially with time intervals τ , as opposed to the simultaneous initiation studied previously for whorled phyllotaxis 19 ( Fig 1 ) ., Although the structures outside of the flower , such as bracts and other flowers , as well as the position of the inflorescence axis , may affect the position of organ primordia , the pentamerous whorls appear despite their various arrangement 20 ., Therefore , as the first step of modelling of floral organ arrangement , we assumed that whorl formation is independent of any positional information from structures outside of the flower ., Thus , we calculated the inhibition potential only from floral organ primordia which are derived from a single floral meristem ., The potential functions for the initiation inhibition by preexisting primordia have been extensively analyzed in phyllotaxis models 16 , 47 , 49 ., The potential decreases with increasing distance between an initiating primordium and the preexisting primordia account for the diffusion of inhibitors secreted by the preexisting primordia 27 , 50 , and the polar auxin transport in the epidermal layer , as proposed in previous models of phyllotaxis 15 , 16 , 31 and the flowers 43 ., We employed an exponential function exp ( −dij/λini ) as a function of θ , where dij denotes the distance between a new primordium i and a preexisting primordium j at ( rj , θj ) as, d i j = R 0 2 + r j 2 - 2 R 0 r j cos ( θ - θ j ) ⋅ ( 1 ) The function decreases spatially through the decay length λini exponentially , induced by a mechanism proposed for the polar auxin transport , i . e . , the up-the-gradient model 15 , 16 ., Up-the-gradient positive feedback amplifies local auxin concentration maxima and depletes auxin from the surrounding epidermis , causing spatially periodic concentration peaks to self-organize 15 , 16 and thus determine the initiation position of the primordia 51 ., The amplification and depletion work as short-range activation and long-range inhibition , respectively 52 , which are common to Turing patterns of reaction-diffusion systems 18 ., Since the interaction of local maxima in the reaction-diffusion systems follows the exponential potential 53 , 54 , the up-the-gradient model likely explains the exponential potential between the auxin maxima , while the rigorous derivation requires further research ., The decay length λini depends not only on the ratio of the auxin diffusion constant and the polar auxin-transport rate 15 but also on other biochemical parameters for polar transport and the underlying intracellular PIN1 cycling 55 ., Another mechanism , referred to as the with-the-flux model 56 , 57 , has been proposed for the polar auxin transport ., Although with-the-flux positive feedback can also produce spatial periodicity , the primordia position corresponds to auxin minima 57 , which is inconsistent with observations 51 ., On the other hand , the with-the-flux mechanism can explain auxin drain from the epidermal layer of the primordia to internal tissue 58 ., Since the drain gets stronger as the primordia mature 58 , 59 , the auxin drain could cause decay of the potential depending on the primordia age ., The auxin decrease in maturing organs can also be caused by controlling auxin biosynthesis 60 , 61 ., Therefore , we integrated another assumption , namely that the inhibition potential decreases exponentially with the primordia age at the decay rate α ( Fig 2B ) ., Temporally decaying inhibition was proposed previously to represent the degradation of some inhibitors 47 , 48 and account for various types of phyllotaxis by simple extension of the inhibitory field model 33 ., Taken together , the potential at the initiation of the i-th primordium is given by, U i n i ( θ ) = ∑ j = 1 i - 1 exp ( - α ( i - j - 1 ) ) exp ( - d i j λ i n i ) ⋅ ( 2 ) Most phyllotaxis models have assumed , based on Hofmeister’s hypothesis , that the primordia move outward at a constant radial drift depending only on the distance from the floral center without angular displacement , which makes helical initiation result in spiral phyllotaxis 49 ., Here , we assumed instead that all primordia repel each other , even after the initiation , except for movement into the meristematic zone ( Fig 2C ) following observation of the absence of auxin ( DR5 expression ) maxima at the center of the floral bud ( e . g . , 62 ) ., Even at the peripheral zone away from the meristem , the growth is not limited ., Hence there is no upper limit for the distance between primordia and the center ., The repulsion exerted on the k-th primordium is represented by another exponentially decaying potential when there are i primordia ( 1 ≦ k ≦ i ) :, U g , k ( r , θ ) = ∑ j = 1 , j ≠ k i exp ( - d k j λ g ) , ( 3 ), where the decay length , introduced as λg , can differ from λini ., The primordia descend along the gradient of potential Ug to find a location with weaker repulsion ., The continuous repulsion can account for post-meristematic events such as the mechanical stress on epidermal cells caused by the enlargement of primordia 63 , 64 or the gene expression that regulates the primordial boundary 42 ., The present formulation ( Eq 3 ) is similar to the contact pressure model , which has been proposed for re-correcting the divergence angle after initiation 25 , 45 , 46 ., Another type of post-initiation angular rearrangement has been modeled as a function of the primordia age employed as i −j −1 in the present model ( Eq 2 ) and the distance between primordia with some stochasticity 65 ., Eq 3 accounts for not only the angular rearrangement but also the radial rearrangement with stochasticity in both directions as will be described in the next subsection ., We modeled the initiation process numerically by calculating the potential Uini ( Eq 2 ) for angular position θ incremented by 0 . 1 degree on the edge of the circular meristem ., We introduced a new primordium at the position where the value of Uini took the minimum , provided that the first primordium is initiated at θ = 0 ., We modeled the growth process by using a Monte Carlo method 66 to calculate the movement of primordia in the outside of the meristem depending on the potential Ug , k ( Eq 3 , Fig 2C ) ., After the introduction of a new primordium , we randomly chose one primordium indexed by k from among the existing primordia and virtually moved its position ( rk , θk ) to a new position ( r k ′ , θ k ′ ) in the outer meristem ( r k , r k ′ ≥ R 0 ) ., The new radius r k ′ and the angle θ k ′ were chosen randomly following a two-dimensional Gaussian distribution whose mean and standard deviation were given by the previous position ( rk , θk ) and by two independent parameters , ( σr , σθ/rk ) , respectively ., Whether or not the k-th primordium moved to the new position was determined by the Metropolis algorithm 66; the primordium moved if the growth potential ( Eq 3 ) of the new position was lower than that of the previous position ( i . e . , U g , k ( r k ′ , θ k ′ ) < U g , k ( r k , θ k ) ) ., Otherwise , it moved with the probability given by, P M P = exp ( - β Δ U g ) , ( 4 ), where Δ U g = U g , k ( r k ′ , θ k ′ ) − U g , k ( r k , θ k ) and β is a parameter for stochasticity ., This stochasticity represents a random walk biased by the repulsion potential ., A case PMP = 0 represents that primordia movement always follows the potential ( ΔUg < 0 ) ., The first primordium stays at the meristem edge r = R0 until the second one arises when PMP = 0 because the growth potential is absent , while it can move randomly outside of the meristem when PMP ≠ 0 ., To maintain the physical time interval of the initiation process at τ steps for each primordium , the number of iteration steps in the Monte Carlo simulation during each initiation interval was set to iτ , where i denotes the number of the primordia ., We also studied the movement following Ug by numerical integration ( fourth-order Runge-Kutta method ) of ordinary differential equations to confirm the independence of the numerical methods ( S1 Fig ) ., All our programs were written in the C programming language and used the Mersenne Twister pseudo-random number generator ( http://www . math . sci . hiroshima-u . ac . jp/m-mat/MT/emt . html ) 67 ., Because the initiation time interval is constant , one possible scenario for forming a whorled pattern should involve decreasing or arresting the radial displacement of primordia ( Fig 1 , forth row ) ., Therefore , we focused on the change in radial position and velocity to find the whorled arrangement , while angular positions were not taken into account in the present manuscript ., Numerical simulations showed that several whorls self-organized following the sequential initiation of primordia ., Although several previous phyllotaxis models showed the transition between a spiral arrangement following sequential initiation and a whorled arrangement following simultaneous initiation 15 , 16 , 19 , they were not able to reproduce the emergence of a whorled arrangement following sequential initiation , which is the situation observed in many eudicot flowers ( Fig 1 ) 34 , 37 , 38 , 40 , 41 ., In the present model , a tetramerous whorl appeared spontaneously that exhibited four primordia almost equidistant from the meristem center ( Fig 2D , left and middle ) , by arresting radial movement of the fifth primordium at the meristem edge until the seventh primordium arose ( arrowhead in Fig 2D , right ) ., Likewise , subsequent primordia produced the same gap in radial distance for every four primordia ( Fig 2D , middle and right ) , leading to several whorls comprising an identical number of primordia ( Fig 2D ) ., The radial positions of all primordia were highly reproducible despite stochasticity in the growth process ( error bars in Fig 2D–2F , middle and right ) ., Therefore , we identified the whorled arrangement by radial displacement arrest ( arrowhead in Fig 2D , right ) ., The initiation order and angle of the first tetramerous whorl in the model reproduced those observed in A . thaliana sepals 68 ( S2A Fig ) ., The first primordium scarcely moved from the initiation point until the second primordium arose because growth repulsion was absent ., The second primordium arose opposite the first , whereas the third and fourth primordia arose perpendicular to the preceding two ., The angular position of the primordia did not change once the whorl was established because the primordia within a whorl blocked the angular displacement by the growth potential Ug ( S3 Fig ) ., Introducing mutual repulsion among the primordia throughout the growth process caused the whorled arrangement to spontaneously emerge ( Fig 2D ) ., This was in contrast to the model of constant growth in which all primordia move away depending only on the distance from the floral apex 49 ., A study of post-meristematic regulation by the organ-boundary gene CUP-SHAPED COTYLEDON2 ( CUC2 ) showed that A . thaliana plants up-regulating CUC2 gene have an enlarged primordial margin and have whorled-like phyllotaxis following the normal helical initiation of primordia 42 , suggesting that repulsive interactions among primordia after initiation are responsible for the formation of the floral whorls ., In the present model , the meristem size R0 controls the transition from non-whorled ( Fig 2E ) to whorled arrangement ( Fig 2D ) ., Radial spacing of the primordia was regular when R0 was small ( Fig 2E , middle ) because the older primordia pushed any new primordium across the meristem ( Fig 2E , left ) , causing continuous movement at the same rate ( Fig 2E , right ) ., Above a threshold meristem size R0 , a tetramerous whorl appeared spontaneously ., The primordium number within each whorl increased up to eight with increasing R0 , but the number tended to be more variable ( S2B Fig ) ., In the A . thaliana mutant wuschel , which has a decreased meristem size , the pattern of four sepals does not have square positions at the stage when the wild-type plant forms a tetramerous sepal whorl 69 ., Conversely , the clavata mutant , which has an increased meristem size , has excessive floral organs with larger variation 69 ., Our model consistently reproduced not only the transition from the non-whorled arrangement ( Fig 2E ) to the tetramerous whorled arrangement ( Fig 2D ) but also the variable increase in the primordia number within a whorl as the meristem size R0 increased ., The pentamerous whorl stably appeared in the presence of temporal decay of initiation inhibition ( α > 0 in Eq 2 ) ., The whorls comprising five primordia appeared in the same manner as the tetramerous whorls , namely , via the locking of the sixth primordium at the initiation site ( Fig 2F , right; S2C Fig ) ., In order to study the organ number within each whorl extensively , known as the merosity 70 , we counted the number of primordia existing prior to the arrest of primordium displacement , which corresponds to the merosity of the first whorl ( arrowheads in Fig 2D and 2F , right ) ., We defined arrest of primordium displacement as occurring when the ratio of the initial radial velocity of a new primordium immediately after initiation to that of the previous primordium was lower than 0 . 2 ., The definition does not affect the following results as long as the ratio is between 0 . 1 and 0 . 6 ., We found that the key parameter for merosity is the relative value of R0 normalized by the average radial velocity V = σ r / 2 π ( see S1 Text ) and the initiation time interval τ ( Fig 3 ) ., The arrest of radial displacement did not occur below a threshold of R0/Vτ ( the left region colored red in Fig 3A ) , whereas the whorled arrangement appeared above the threshold value of R0/Vτ ., As R0/Vτ increased further , tetramery , pentamery , hexamery , heptamery , and octamery appeared , successively ( Fig 3A ) ., The present model showed dominance of special merosity , i . e . , tetramery and octamery in the absence of temporal decay of inhibition ( α = 0 in Eq 2; Fig 3A ) ; pentamery in the presence of temporal decay ( α > 0; Fig 3B and 3C ) , in contrast to previous phyllotaxis models for whorled arrangement in which the parameter region leading to each level of merosity decreased monotonically with increasing merosity 19 ., The major difference between α = 0 and α > 0 was that θ3 , the angular position of the third primordium , took an average value of 90 degrees when α = 0 ( arrowhead in Fig 3A bottom magenta panel ) and decreased significantly as α increased ( arrowhead in Fig 3A bottom cyan panel ) ., In a pentamerous flower Silene coeli-rosa , the third primordium is located closer to the first primordium than the second one 34 ., This is consistent with the third primordium position at α > 0 , indicating the necessity of α , as we will discuss in the next section ., The parameter region R0/Vτ for pentamery expanded with increasing α , whereas the border between the whorled and non-whorled arrangements was weakly dependent on α ( Fig 3C ) ., The tetramery , pentamery , and octamery arrangements were more robust to R0/Vτ and α than the hexamery and heptamery arrangements ., Dominance of the particular number also appears in the ray-florets within a head inflorescence of Asteraceae 71 , in which radial positions show the whorled-type arrangement 72 ., Meanwhile , the leaf number in a single vegetative pseudo-whorl transits between two to six by hormonal control without any preference 73 ., Moreover , the transition between the different merosities occurred directly , without the transient appearance of the non-whorled arrangement ., This is in contrast to an earlier model 19 in which the transition between different merosity always involved transient spiral phyllotaxis ., The fact that the merosity can change while keeping its whorled nature in flowers ( e . g . , the flowers of Trientalis europaea74 ) supports our results ., To our knowledge , ours is the first model showing direct transitions between whorled patterns with different merosities as well as preferences for tetramery and pentamery , the most common merosities in eudicot flowers ., To further validate our model of the pentamerous whorl arrangement , we quantitatively compared its results with the radial distances and divergence angles in eudicot flowers ., Here we focus on a Scanning Electron Microscope ( SEM ) image of the floral meristem of S . coeli-rosa , Caryophyllaceae ( Fig 4A–4C ) 34 , because S . coeli-rosa exhibits not only five sepals and five petals in alternate positions , which is the most common arrangement in eudicots , but also the helical initiation of these primordia , which we targeted in the present model ., In addition , to our knowledge , this report by Lyndon is the only publication showing a developmental sequence for both the divergence angle Δθk , k+1 = θk+1 −θk ( 0 ≤ Δθk , k+1 < 360 ) and the ratio of the radial position , rk/rk+1 , referred to as the plastochron ratio 75 , in eudicot floral organs ., Reconstructing such developmental sequences of both radial and angular positions is an unprecedented theoretical challenge , while those which describe the angular position alone for the ontogeny of spiral phyllotaxis ( 180 degree , 90 degree and finally convergence to 137 degree 16 , 33; the ‘M-shaped’ motif , i . e . , 137 , 275 , 225 , 275 and 137 degrees 76 , 77 ) have been reproduced numerically ., By substituting the initial divergence angle between the first and second sepals of S . coeli-rosa into Δθ1 , 2 = 156 but not any plastochron data into the simulation ( θ1 = 0 and θ2 = 156 degree ) , we numerically calculated the positions of the subsequent organs ( Fig 4D ) ., The observed divergence angle Δθ2 , 3 = 132 degree indicates α > 0 , because Δθ2 , 3 = Δθ1 , 3 = ( 360−156 ) /2 = 102 degree at α = 0 , in the present model setting r1 ≅ r2 ., Even when r1 > r2 , the divergence angle was calculated as Δθ2 , 3 = 113 degree ( r1 = R0+2Vτ , r2 = R0+Vτ , R0 = 1 , Vτ = 0 . 14 , and λini = 0 . 05 estimated from the S . coeli-rosa SEM image 34; see S4 Fig for detail ) , which is still less than the observed value ., As α became larger , the inhibition from the second primordium became stronger than that from the first one , making Δθ2 , 3 consistent with the observed value in S . coeli-rosa ( Fig 4E , top ) ., For the subsequent sepals and petals , the model faithfully reproduced the period-five oscillation of the divergence angle and the plastochron ratio until the ninth primordium ( Fig 4E ) , notably in the deviation of the divergence angle from regular pentagon ( 144 degree ) and the increase of plastochron ratio at the boundary between the sepal and petal whorls ., Moreover , a similar increase in the plastochron ratio occurred weakly between the second and third primordia in the first whorl ( closed arrowhead in Fig 4E ) , indicating a hierarchically whorled arrangement ( i . e . , whorls within a whorl ) ., Such weak separation of the two outer primordia from the three inner ones within a whorl is consistent with the quincuncial pattern of sepal aestivation that reflects spiral initiation in many of eudicots with pentamerous flowers ( e . g . , Fig 2D–E in 21 ) ., Even with an identical set of parameters , the order of initiation in the first pentamerous whorl can vary depending on the stochasticity in the growth process ., The variations of the initiation order in simulations may be caused by the absence of the outer structure , because the axillary bud seems to act as a positional information for the first primordia in S . coeli-rosa floral development ( Fig 4B ) ., The positioning of the five primordia in the first whorl was reproducible in 70% of the numerical replicates , within less than 20 degrees of that in S . coeli-rosa or that of the angles in a regular pentagon ., Mismatches in the inner structure ( from the tenth primordium , i . e . , the last primordium in petal whorl ) might be due to an increase in the rate of successive primordia initiation later in development 35 , which we did not assume in our model ., The agreements between our model and actual S . coeli-rosa development of sepals and petals in both the angular and the radial positions suggests that the S . coeli-rosa pentamerous whorls are caused by decreasing inhibition from older primordia ., A possible mechanism to arrest the radial displacement of a new primordium , a key process for whorl formation ( arrowheads in Fig 2D and 2F ) , involves an inward-directed gradient of the growth potential Ug , k ( Eq 3 ) of a new primordium so that its radial movement is prevented ., To confirm this for tetramerous whorl formation ( Fig 3A ) , we analytically derived the parameter region such that the radial gradient of the growth potential at the angle of the fifth primordium Ug , 5 ( Eq 3 ) , which is determined by the positions of the preceding four primordia , is inward-directed ., For ease in the analytical calculation , we set α = 0 and PMP = 0 ., The first four primordia positions were intuitively estimated ( see S2 Text ) as, r 1 = R 0 + 3 τ V , θ 1 = 0 r 2 = R 0 + 3 τ V , θ 2 = 180 r 3 = R 0 + 2 τ V , θ 3 = 90 r 4 = R 0 + τ V , θ 4 = 270 , ( 5 ), which agreed with the numerical results with an error of less than several percent regardless of the parameter spaces ., Hereafter we demonstrate a case Vτ = 6 . 0 ., The position of the fifth primordium derived from the positions of four existing primordia ( Eq 5 ) becomes θ5 = 90 when R0 ≤ 2 , whereas θ5 ∼ 135 when R0 > 2 ( S5 Fig ) ., Next , we calculated the potential for the fifth primordium in radial direction by substituting Eq 5 and the position of the fifth primordium θ5 into Eq 3 ., The function becomes, U g , 5 ( r , θ 5 ) = ∑ j = 1 4 exp ( - d 5 j λ g ) = ∑ j = 1 4 exp ( - r j 2 + r 2 - 2 r j r cos ( θ j - θ 5 ) λ g ) ⋅ ( 6 ), The potential exhibits a unimodal ( 2 < R0 < 10; Fig 5A ) or bi-modal ( R0 < 2 , R0 > 10; Fig 5B and 5C ) shape ., At R0 < 10 , the potential gradient at the initiation position of the fifth primordium ∂Ug , 5 ( r , θ5 ) /∂r∣r = R0 is outward-directed ( Fig 5A ) , providing almost constant growth resulting a non-whorled arrangement in the simulations ( Fig 3A , red region ) ., At R0 > 10 , we defined the radial position of the local maximum closest to the fifth primordium as rmax ( open arrowhead in Fig 5B and 5C; red squares in the upper half of Fig 5D ) and the local minimum as rmin ( blue circles in Fig 5D; 0 < rmin < rmax ) ., The potential gradient ∂Ug , 5 ( r , θ5 ) /∂r∣r = R0 has a negative value when R0 < rmin or rmax < R0 ( Fig 5C ) , causing the fifth primordium to constantly move outward ., On the other hand , the potential gradient is positive , i . e . , directed inward ( Fig 5B ) , when rmin < R0 < rmax ( between the two solid arrowheads in Fig 5D ) , causing the arrest of radial movement of the fifth primordium ., The values of rmin and rmax , analytically calculated as function of R0 and τ ( solid black line in Fig 5E ) , were faithfully consistent with the parameter boundaries between the non-whorled pattern and the tetramerous-whorled pattern and between the tetra

Introduction, Model, Results/Discussion

How organisms determine particular organ numbers is a fundamental key to the development of precise body structures; however , the developmental mechanisms underlying organ-number determination are unclear ., In many eudicot plants , the primordia of sepals and petals ( the floral organs ) first arise sequentially at the edge of a circular , undifferentiated region called the floral meristem , and later transition into a concentric arrangement called a whorl , which includes four or five organs ., The properties controlling the transition to whorls comprising particular numbers of organs is little explored ., We propose a development-based model of floral organ-number determination , improving upon earlier models of plant phyllotaxis that assumed two developmental processes: the sequential initiation of primordia in the least crowded space around the meristem and the constant growth of the tip of the stem ., By introducing mutual repulsion among primordia into the growth process , we numerically and analytically show that the whorled arrangement emerges spontaneously from the sequential initiation of primordia ., Moreover , by allowing the strength of the inhibition exerted by each primordium to decrease as the primordium ages , we show that pentamerous whorls , in which the angular and radial positions of the primordia are consistent with those observed in sepal and petal primordia in Silene coeli-rosa , Caryophyllaceae , become the dominant arrangement ., The organ number within the outmost whorl , corresponding to the sepals , takes a value of four or five in a much wider parameter space than that in which it takes a value of six or seven ., These results suggest that mutual repulsion among primordia during growth and a temporal decrease in the strength of the inhibition during initiation are required for the development of the tetramerous and pentamerous whorls common in eudicots .

Why do most eudicot flowers have either four or five petals ?, This fundamental and attractive problem in botany has been little investigated ., Here , we identify the properties responsible for organ-number determination in floral development using mathematical modeling ., Earlier experimental and theoretical studies showed that the arrangements of preexisting organs determine where a new organ will arise ., Expanding upon those studies , we integrated two interactions between floral organs: ( 1 ) spatially and temporally decreased inhibition of new organ initiation by preexisting organs , and ( 2 ) mutual repulsion among organs such that they are “pushed around” during floral development ., In computer simulations incorporating such initiation inhibition and mutual repulsion , the floral organs spontaneously formed several circles , consistent with the concentric circular arrangement of sepals and petals in eudicot flowers ., Each circle tended to contain four or five organs arranged in positions that agreed quantitatively with the organ positions in the pentamerous flower , Silene coeli-rosa , Caryophyllaceae ., These results suggest that the temporal decay of initiation inhibition and the mutual repulsion among growing organs determine the particular organ number during eudicot floral development .

journal.ppat.1003286

Post-Transcriptional Regulation of the Trypanosome Heat Shock Response by a Zinc Finger Protein

When living organisms are exposed to temperatures above their growth optima , they respond by increased synthesis of heat-shock proteins ., In eukaryotes as diverse as animals , ciliates and plants , heat-shock protein expression is controlled by heat-shock transcription factors , whose activation enables them to bind conserved heat-shock elements in the promoters of heat-shock protein genes and activate their transcription 1 , 2 , 3 , 4 ., Trypanosoma brucei and related Kinetoplastid protists must also adapt to different temperatures: they multiply both in mammals , with temperatures varying from 32°C to 38°C depending on species and body location ( see e . g . 5 , 6 , 7 ) , and in arthropod vectors in which the temperature variations are much greater ( e . g . 8 ) ., In Kinetoplastids , however , the regulation relies exclusively on post-transcriptional mechanisms ., Transcription is polycistronic 9 , 10 , and individual mRNAs are produced by trans splicing and polyadenylation 11 , 12 ., The final cytoplasmic RNA level is determined by the rates of processing , transport from the nucleus , and degradation 13 ., For most trypanosome mRNAs , the rate of degradation is a critical determinant of expression 14 ., Two forms of T . brucei are routinely studied in the laboratory: the bloodstream form ( found in the mammalian host , cultivated axenically in vitro at 37°C ) and the procyclic form ( found in the midgut of the Tsetse fly vector , cultivated axenically at 27°C ) ., Upon transfer of procyclic forms to 41°C , transcription by RNA polymerase II is gradually shut down 15 and trans splicing is inhibited 16 ., The overall level of translation also decreases , as shown by reduced in vivo 35S-methionine labelling and the collapse of polysome profiles 17 ., This is partly due to rapid mRNA degradation , as judged both by profiling of total mRNA 17 and examination of specific transcripts 18; and it is partly due to effects on translation 17 ., After heat-shock , poly ( A ) binding protein and several translation factors accumulate in granules 17 ., Meanwhile , the mRNAs encoding HSP83 and the major cytosolic HSP70 remain stable and continue to be translated 17 , 18 ., Although the trypanosomes are able to recover from a 41°C heat shock lasting up to 2 hours , it is not known whether the heat-shock response is required for the recovery ., Indeed , it is not known whether the trypanosome heat-shock response has any selective advantage ., T . brucei has five virtually identical genes encoding the major cytosolic HSP70 that are arranged in a tandem array 19 ( unfortunately collapsed to one locus , Tb927 . 11 . 11330 , in the genome assembly ) and are constitutively co-transcribed 20 , 21 ., Using reporters , it was shown that sequence elements in the HSP70 3′-untranslated region ( 3′-UTR ) are responsible for the stability of the mRNA after heat-shock 18 , 22 ., Similar observations were also made for HSP70s of the Kinetoplastids Trypanosoma cruzi 23 and Leishmania infantum 24 , 25 ., The multiple copies of the HSP83 genes ( encoding the major Kinetoplastid HSP90 homologue ) are also in a tandem array ., The 3′-UTR of Leishmania HSP83 mRNA is important for both mRNA stability and increased translation during heat-shock 26 , and it was proposed that temperature-induced changes in RNA secondary structure might play a role in regulation 27 ., Post-transcriptional mechanisms are also responsible for heat-induced increases in Leishmania HSP100 mRNA 28 , 29 ., The stability , localization and translation states of eukaryotic mRNAs are influenced by proteins that bind to them ., For example , in mammalian cells , tristetraprolin ( also called TTP , Tis11a , and Zfp36 ) , and BRF1 and BRF2 ( Butyrate response factors 1 and 2 ) bind to AU-rich elements with a consensus of UAUUUAUU; they recruit components of the mRNA degradation machinery , promoting mRNA decay 30 ., These three proteins , together with related proteins from other Opisthokonts ( together called the “Tis11 family” ) , possess two C8C5C3H zinc finger domains separated by a linker of about 10 amino acids ., Immediately preceding the zinc finger domain is a six-residue conserved sequence , R/K-Y-K/R-T-E/K-L , which strongly influences the sequence specificity of RNA binding 31 ., The activities of TTP and BRF proteins are regulated by phosphorylation , and are critical for control of inflammation and cell proliferation in mammals 30 ., Other proteins compete for binding to the AU-rich element and promote mRNA stability 32 ., T . brucei has forty-nine CCCH zinc finger proteins , some of which have been implicated in control of gene expression 33 , 34 , 35 , 36 , 37 , 38 , 39 ., So far , however , none has been shown to have a destabilising function ., In search of possible destabilising proteins , we looked for predicted trypanosome proteins with the Tis11 consensus ., We here show that the protein with the best match , ZC3H11 , indeed binds to mRNAs containing an AUU sequence element but that - in contrast to the situation in mammalian cells - the consequence is an increase in mRNA abundance ., Most interestingly , ZC3H11 appears to be a master regulator of stress response mRNAs ., To find CCCH proteins that might be involved in post-transcriptional gene regulation in T . brucei , we scanned all of them for the Tis11 consensus ., The best matches were in ZC3H11 , ZC3H12 and ZC3H13; ZFPs 1–3 also showed some similarity ( Figure 1A ) ., ZC3H11 ( locus Tb927 . 5 . 810 ) consists of 364 amino acids , and has a predicted molecular weight of 39 . 6 kDa ., The zinc finger starts at residue 70 , and is preceded by the Tis11 consensus RYKTKL ., The ZC3H11 gene is found in all available Kinetoplastid genomes ., Most sequence identity is concentrated around the zinc finger: comparing all available proteomes , the 6mer has consensus RYKTK ( L/Y/F ) ., Some additional conserved patches are the sequence H ( N/D ) PY around residue 200 of T . brucei ZC3H11 and a serine-rich region near the C-terminus ( Supplementary Figure S1 ) ., Alignment of the ZC3H11 zinc finger with those of other Tis11 family proteins showed that some of the residues required for interaction with AU-rich elements were conserved ., From a crystal structure of BRF2 ( Tis11-d ) with UUAUUUAUU 31 , it was found that each zinc finger of BRF2 specifically binds the sequence UAUU ., Using the residue numbering in Figure 1A , and numbering the 4-nt bound RNA as U ( 1 ) -A ( 2 ) -U ( 3 ) -U ( 4 ) , Tyr18 intercalates between the Us in the 3rd and 4th positions ( U3 and U4 ) , and Phe26 intercalates between U1 and A2 ., These residues are conserved in ZC3H11 ., Specificity for U1 was conferred by backbone hydrogen bonds with ( Asn/His ) 25 and Glu5; these residues are not conserved in ZC3H11 , which has a basic residue at position 5 , like C . elegans MEX5 ( Figure 1A ) ., In contrast , A2 is hydrogen-bonded by Leu6 and Arg8 , which are conserved ., A notable difference between ZC3H11 and other Tis11-family zinc fingers is the presence of a novel Asp residue at position 21 in place of the conserved glycine ., Attempts to generate a polyclonal antibody that could detect ZC3H11 in parasite lysates failed ., In order to detect ZC3H11 in trypanosomes , we therefore integrated a sequence encoding an N-terminal V5-epitope tag 40 in frame with one of the ZC3H11 open reading frames ( ORFs ) ., Since the 3′-UTR is conserved by this procedure , expression levels are expected to be approximately normal unless the tag affects protein stability ., In both procyclic forms , which we normally grow at 27°C , and the bloodstream stage , grown at 37°C , the V5-ZC3H11 fusion protein was detected as an extremely faint band that migrated at about 60 kDa instead of the expected 40 kDa ( Figure 1B & C , lane 2 ) ., The abundance of V5-ZC3H11 was , however , dramatically increased upon heat shock ., In procyclic forms , induction of ZC3H11 was transient at 37°C ( Figure 1B lanes 3–6 ) , but stronger and more extended at 41°C; at later time points , some smaller products appeared ( Figure 1B lanes 7–10 ) ., Since the longer incubations resulted in a decline in cell viability , the faster-migrating bands could indicate either proteolytic degradation or the removal of posttranslational modifications ., For bloodstream forms , incubation at 43°C led to a rapid induction of V5-ZC3H11 although the cells started to die within an hour ( Figure 1C lanes 3 , 4 ) ., Further experiments showed that in addition to elevated temperatures , mild translational stress from low concentrations of puromycin also increased V5-ZC3H11 expression in both developmental stages ( Figure 1B lane 11–13 and 1C lanes 5 , 6 ) ., In a preliminary attempt to determine the mechanism of this expression regulation , we incubated cells with lactacystin ( not shown ) or MG132 ( Figure 1D ) to inhibit the proteasome ., Indeed , the amount of ZC3H11 increased ( Figure 1D ) ., This might mean that the protein is normally rapidly degraded by the proteasome , but is stabilised upon heat shock or puromycin stress ., Alternatively , proteasome inhibition could be acting as another sort of stress , with ZC3H11 protein increasing by another mechanism ., We also expressed ZC3H11-myc in cells with V5-ZC3H11 ., Reciprocal pull-downs revealed no evidence for dimerization ( not shown ) ., To determine the nature of the possible post-translational modifications , we incubated cell lysates with λ-phosphatase before electrophoresis ., The 60 kDa bands ( Figure 1E , upper panel , lane 1 ) collapsed to one band at approximately 50 kDa ( Figure 1E , upper panel , lane 3 ) ; this was prevented by addition of phosphatase inhibitors ( Figure 1E , upper panel , lane 4 ) , showing that the 10 kDa migration difference was indeed due to phosphorylation ., Interestingly , the extra band seen after heat shock at 41°C had a similar migration ( Figure 1B lanes 8–10 , Figure 1C , lane 3 ) , suggesting that it too might have been dephosphorylated ., For an N-terminal fragment containing the first 128 residues of ZC3H11 , extending 39 residues beyond the zinc finger , a similar pattern was observed ( Figure 1E , lower panel ) , except that a portion appeared unmodified ., The N-terminal fragment ran as several bands between 23 kDa and 17 kDa , which all collapsed to the lowest band upon phosphatase treatment ., This indicates that residues in the N-terminal region can be phosphorylated ., The low abundance of the V5-in situ tagged protein precluded localisation by microscopy or cell fractionation ., We therefore instead looked at the location of ZC3H11 bearing a tandem affinity purification ( TAP ) tag , inducibly expressed from a strong RNA polymerase I promoter in procyclic forms ., ZC3H11-TAP was clearly excluded from the nucleus and found in the cytoplasm in somewhat granular structures ( Figure 1F ) ., The presence of an IgG-binding domain in the tag prevented us from looking for colocalisation with stress granule markers ., Inducibly expressed ZC3H11 with a C -terminal myc tag gave similar results but with a much fainter signal: a rather granular cytoplasmic immunofluorescence which became marginally brighter after a one-hour heat shock ( Supplementary Figure S1B ) ., To find out which mRNAs were bound by ZC3H11 , we inducibly expressed myc-tagged ZC3H11 in procyclic trypanosomes , precipitated the protein using anti-myc antibody , and compared bound and unbound RNAs by RNASeq ., The twenty-four most strongly enriched transcripts are listed in Table 1 and the full list is in Supplementary Table S1 , sheet, 1 . Strikingly , more than half of the strongly bound mRNAs were implicated in the stress response ., Thirteen of them encoded a full set of chaperones required for protein refolding ., All classes of cytosolic HSPs were represented - HSP70 , HSP83 ( HSP90 family ) , HSP100 , HSP110 and HSP20 ., Also , mRNAs encoding putative homologues of co-chaperones were present: DnaJ ( HSP40 ) proteins , a FKBP/TPR domain protein , stress induced protein 1 ( STI1 ) and cyclophilin-40 ., Three additional bound mRNAs encoded the mitochondrial chaperone HSP60 , a copper chaperone for cytochrome c , and the glutaredoxin GRX2 , which protects against oxidative stress ., Notably , mRNAs encoding chaperones for co-translational folding ( TRiC complex ) or organellar import ( mitochondrial HSP70 and ER-resident BiP were not enriched ., Among bound transcripts that do not encode annotated chaperones , the most notable encoded GPEET procyclin ., Five bound mRNAs encoded proteins of unknown function ., We next analysed the 3′-UTRs of all bound transcripts for enriched motifs ., We found a striking enrichment of an ( AUU ) n repeat motif ( Figure 2A , supplementary Table S1 , sheet 2 ) ., Of the 22 most strongly bound mRNAs , 14 contained perfect ( AUU ) 4 repeats and four more had a repeat of 11 nt ( Table 1 ) : they included all but one ( CYP40 ) of the chaperone mRNAs ., A scan of the whole genome revealed 325 genes with a good match to the 12mer ( AUU ) 4 sequence in their predicted 3′-UTR , of which only 44 were at least two-fold enriched in the ZC3H11-bound fraction ., Although the 3′-UTRs used in the analysis may , in some cases , be incorrect , this result shows that the presence of an ( AUU ) repeat alone is not sufficient to give ZC3H11 binding ., The putative AUU repeat binding motif is interesting because the sequence bound by a single Tis-11 CCCH domain is UAUU 31 ., To investigate the RNA-binding specificity of ZC3H11 in more detail , we expressed a variety of different fusion proteins in E . coli and purified them ., The only proteins that could be obtained in reasonable quantity and purity were two N-terminal fragments of 104 and 119 residues ( Figure 2B , lanes 1 & 2 ) ., Both contain the zinc finger but the 119mer also includes additional conserved residues ( Supplementary Figure S1A ) ., Although the proteins formed single bands on denaturing gels ( Supplementary Figure S2A ) , on native gels , the pattern was very smeared and some protein remained in the well ( Supplementary Figure S2B ) ., This suggested that despite initial solubility , the proteins were not fully folded and some aggregation was occurring ., As controls , we expressed the same protein fragments with a C70S mutation in the zinc finger ., These were very poorly expressed and as a consequence , the purified samples were heavily contaminated ( Figure 2B , lanes 3 and 4 ) ., To test for RNA binding , we incubated the proteins with various radioactively-labelled oligo-ribonucleotides and examined migration in non-denaturing polyacrylamide gels ., The 104 residue protein ( ZC3H11-104 ) , interacted with both ( UAUU ) 5UAU ( classical ARE ) and U ( UAU ) 7U to give a clear band ( Figure 2C , lanes 4 & 5 , arrow s1 ) ., U ( UCU ) 7U gave a very faint band of slower mobility ( Figure 2C , lane 6 , arrow n ) which was also detected for both U ( UAU ) 7U and U ( UCU ) 7U using the zinc finger mutants ( Figure 2C , lanes 11 , 12 , 14 , 15 ) ., This band most likely represents binding of the probe by an E . coli contaminant , although zinc-finger-independent binding by ZC3H11 is also possible ., Using the 119-residue protein , ZC3H11-119 , a slower-mobility band was obtained using ( UAUU ) 5UAU and U ( UAU ) 7U ( Figure 2C , lanes 7 & 8 , band s2 ) , suggesting binding of additional copies of the protein: perhaps the extra 15 amino acids mediate protein-protein interactions ., In addition , there was strong accumulation of radioactivity in the well ( w ) ., This could be explained if the zinc finger were properly folded , but the remainder of the polypeptide were unfolded and formed aggregates ., Reducing the probe length to 14 residues did not affect the apparent aggregation ( not shown ) ., To investigate the binding in more detail , we used more probes ., Results for ZC3H11-104 are shown in Figure 2D and 2E ., The interactions with U ( UAU ) 7U and ( UAUU ) 5UAU were confirmed ( Figure 2D , lanes 2 & 4 , arrow s1 ) , and the faint band using U ( UCU ) 7U ( Figure 2D , lane 6 , arrow n ) was much stronger using U23 ( Figure 2D , lane 8 , arrow n ) ., A very faint shift was seen with A23 , but none with C23 ( Figure 2D , lanes 10 & 12 ) ., To further assess specificity , ZC3H11-104 was incubated with the labelled U ( UAU ) 7U probe in the presence of unlabelled competitors ., U ( UAU ) 7U competed effectively , most of the probe now remaining unbound ( Figure 2E , lanes 3 and 4 ) ., In contrast , addition of cold ( UAUU ) 5UAU ( Figure 2E , lanes 5 and 6 ) or of U23 shifted the radioactivity to the non-specific band ( Figure 2E , lanes 9 and 10 , band n ) ., U ( UCU ) 7U showed partial competition ( Figure 2E , lanes 7 and 8 ) whole A23 and C23 could not compete at all ., Results for ZC3H11-119 were similar except that as before , the specific complex showed slower migration and radioactivity accumulated in the well ( Supplementary Figure S2C–E ) ., The strong shift to the apparently less specific band ( “n” ) in the competition assays was not inhibited by heparin ( Supplementary Figure S2E ) ., We also attempted to assess the binding affinities of ZC3H11-119 to limiting amounts of U ( UAU ) 7U , ( UAUU ) 5UAU and U23 probes ., The only probe that bound at all under these conditions was U ( UAU ) 7U , but the results could not be interpreted quantitatively because at most protein concentrations , the only bound radioactivity was stuck in the well ( Supplementary Figure S2F ) ., We concluded that the zinc finger of ZC3H11 binds preferentially to ( UAU ) repeats , but is also able to bind to the classical ARE ., We examined the effect of ZC3H11 depletion by RNA interference ( RNAi ) ., Stable cell lines inducibly expressing a double stranded RNAi fragment were created in bloodstream- and procyclic-form trypanosomes ., In bloodstream forms , depletion of ZC3H11 was lethal ( Figure 3A ) , while no effect was observed in procyclic cells ( Figure 3B ) ., A similar result was obtained in a published high-throughput RNAi screen 41 ., To further investigate the reason why ZC3H11 was essential in bloodstream-form trypanosomes , the transcriptome of ZC3H11-depleted cells was compared with that of wild-type cells , initially using an oligonucleotide microarray ( not shown ) and later , using poly ( A ) + RNA , by high-throughput cDNA sequencing ( RNA-Seq ) ( Supplementary Table S1 , sheet 3 and Supplementary Figure S3A , B ) ., The RNA from ZC3H11-depleted cells was taken 24 h after induction of RNAi , before a growth defect was evident , and with no drug treatment apart from tetracycline , which is known not to affect the transcriptome at the level used 14 ., We compared the RNASeq results with a previous dataset for poly ( A ) + RNA from wild-type cells ., 452 transcripts were at least 2-fold increased after ZC3H11 depletion ( Supplementary Table S1 , sheet 3 ) ., The increased transcripts were significantly enriched in the categories of protein kinases and phosphatases , and also RNA-binding proteins , but have not yet been examined further ., There was no correlation between the effects of ZC3H11 RNAi in bloodstream forms and enrichment in the ZC3H11-bound fraction , suggesting that many of the effects seen were secondary ., Bound RNAs that increased included GPEET procyclin , but since procyclin-associated mRNAs , which are in the same transcription unit , also increased , an increase in procyclin locus transcription is possible ., Other increased ZC3H11-bound mRNAs included those encoding the putative RNA-binding protein RBP5 and a few proteins of unknown function ., RNASeq revealed 72 genes with at least 2-fold decreased mRNA expression after ZC3H11 RNAi ., The strong enrichment for genes encoding ribosomal proteins ( P\u200a=\u200a5×10−15 ) and translation factors ( P\u200a=\u200a0 . 05 ) suggests that some of the decreases could be indirect effects , secondary to the onset of growth arrest ., Looking at ZC3H11-bound mRNAs ( Table 1 ) , one encoding an FKBP-like petidyl-prolyl cis-trans isomerase was not decreased according to the RNASeq , but was decreased by Northern blotting ( 0 . 2× , Supplementary Figure 3C ) and microarray ( 0 . 4× , not shown ) ., HSP70 mRNA levels were reproducibly halved by RNASeq , microarray ( not shown ) and Northern blotting ( Figure 4A and Supplementary Figure 4B ) ., We therefore decided to investigate HSP70 regulation by ZC3H11 ., To confirm binding of HSP70 mRNA to ZC3H11 , we immunoprecipitated ZC3H11-myc from procyclic trypanosome extracts and subjected the resulting RNA to Northern blotting ( Figure 4B ) ., As controls , we used cells that expressed no myc-tagged protein , or cells expressing a myc-tagged version of ZC3H11 with the C70S mutation ., Since the immunoprecipitation is a lengthy procedure , some degradation of the mRNA occurred , but nevertheless , a band of HSP70 mRNA was visible in the preparation from cells expressing ZC3H11-myc , whereas no HSP70 mRNA was detected in the control pull-downs ., As a further control for non-specific RNA sticking to the beads we looked for the highly abundant tubulin mRNA ., As expected , some of this mRNA was found in all lanes , but with no specificity for pull-down by ZC3H11-myc ( Figure 4B ) ., The C70S mutant protein was rather poorly expressed relative to the wild-type ( not shown ) , so this experiment by itself allows no conclusions regarding a requirement for the C70 residue of the zinc finger in RNA binding ., To find out whether the effect on HSP70 mRNA abundance in bloodstream-form trypanosomes was caused by increased instability , we inhibited transcription and measured the amount of HSP70 mRNA left after 15 and 30 min ., In five independent measurements , the HSP70 mRNA half life was 23±7 min ( mean ± standard deviation ) ., The RNAi cell line yielded values of 21±9 min in the absence of tetracycline , and 15±6 min after one day of RNAi induction ., ZC3H11 RNAi decreased the half-life of HSP70 mRNA in every experiment , suggesting that ZC3H11 stabilises HSP70 mRNA ., It was already known that trypanosome HSP70 mRNA abundance is regulated by the 3′-UTR 17 , 18 , 22 ., To define the region that was targeted by ZC3H11 , we generated bloodstream-form cell lines that had inducible RNAi against ZC3H11 , and also constitutively expressed chloramphenicol acetyltransferase ( CAT ) reporter constructs flanked by different UTRs ( Figure 4C , 4D and Supplementary Figure S4A ) ., The constructs were integrated into the tubulin locus and expressed by read-through transcription by RNA polymerase II ., Reporter protein expression was measured by the CAT assay; CAT RNA levels and correct mRNA processing were assessed by Northern blotting ( Figure 4D and Supplementary Figure S4 , B & C ) ., The parental construct expressed an mRNA with the 5′-UTR from the EP mRNA , and a truncated actin 3′-UTR ( Figure 4D , control ) ., Introducing the HSP70 5′-UTR caused no significant change in expression levels compared to the parental constructs ( Figure 4D , HSP70 5′-UTR ) ., In contrast , when the HSP70 3′-UTR was included , either by itself or together with the HSP70 5′-UTR , the steady state levels of CAT mRNA and protein were approximately twice the control ., Induction of RNAi against ZC3H11 reduced this expression to the level of the control construct ., These results show that the HSP70 3′-UTR was sufficient for ZC3H11-mediated regulation ., Since all constructs were transcribed from the same locus , the mechanism must be post-transcriptional ., We attempted to compare the half-lives of the CAT-HSP70 reporter mRNAs but the low amounts present after ZC3H11 RNAi prevented accurate quantitation ., We wanted to see whether the AU-rich sequence was required for ZC3H11-mediated mRNA stabilisation ., A reporter with just the 5′ part of the HSP70 3′-UTR , which lacks the AU sequence element , was expressed at levels similar to the control , and showed no response to ZC3H11 RNAi ( Figure 4D , delAU ) ., In contrast , a construct containing only the 3′ part , with mainly just the AU element , behaved like the construct with the complete 3′-UTR ., We concluded that the part of HSP70 3′-UTR that contains ( AUU ) repeats is necessary and sufficient for regulation by ZC3H11 in bloodstream forms ., Finally , we inserted ( TAT ) 6 either at the beginning , or at the end , of the actin 3′-UTR in the reporter plasmid ., The insertion after the coding region and before the actin 3′-UTR had no effect on CAT RNA or protein ( Supplementary Figure S4D ) ., An insertion just before the usual poly ( A ) site resulted in a two-fold increase in both RNA and protein , but the effect was independent of the 18mer orientation and was not affected by ZC3H11 RNAi ( Supplementary Figure S4D ) ., This confirms our impression that in order to respond to ZC3H11 , the AU-rich sequence requires a particular context ., Other regulatory elements behave similarly in trypanosomes: for example , EP mRNA degradation in bloodstream-form trypanosomes is regulated by a 26mer 42 , 43 , but the 26mer alone does not work if placed at the start of the actin 3′-UTR 42 ., Since chaperones were strongly enriched among the possible ZC3H11 targets , we investigated ZC3H11 function in procyclic trypanosomes incubated above their normal culture temperature of 27°C ., At 37°C , wild-type cells stopped multiplying after 3–4 days , while cells with RNAi showed a slower cell number increase and were already starting to die after 2–3 days ( Figure 5A ) ., Since the latter result suggested that ZC3H11 was important in survival at elevated temperatures , we tested published heat-shock conditions ., After a one-hour incubation of our normal procyclic forms at 41°C , motility was strongly reduced , but , as previously observed 17 , the cells recovered rapidly after being returned to 27°C ( Figure 5B , WT ) ., If the cells were depleted of ZC3H11 , however , recovery was severely impaired ( Figure 5B , RNAi ) ., In another cell line , containing only one , V5-tagged copy of ZC3H11 , recovery kinetics were intermediate between the RNAi and wild-type ( not shown ) ., This suggests that V5-ZC3H11 is functional , but the presence of only a single copy of ZC3H11 causes haplo-insufficiency ., To look at the effect of the heat shock on cell cycle progression in more detail , we analysed cell shape and DNA content by FACS ( Figure 5C ) ., Normal cells before shock had identical patterns with a G1 peak of 1× diploid DNA content , a smaller G2/M peak with 2× diploid DNA content , and cells in S-phase in between ., One day after the heat shock , both populations showed relatively more G2/M cells , an accumulation of multinucleate cells with abnormally high DNA content , and some dead cells with less than 1× DNA content ., The wild-type population had returned to normal by day 2 , but for the population with ZC3H11 RNAi , dead cells and cells with abnormally high DNA content persisted and the G1/G2 ratio had not recovered ( Figure 5C ) ., As previously described 17 , a one-hour 41°C heat shock reproducibly decreased de novo synthesis of many proteins , as judged by 35S-methionine labelling ( Figure 5D ) ; among those spared were two migrating at about 90 kDa and 70 kDa , which are probably HSP83 and HSP70 ., This result was extremely similar to that previously seen for insect-stage Leishmania 44 , 45 ., Transcription initiation is shut down in trypanosomes after heat shock 15 and by preparing RNA , then analysing the amount of mRNA by Northern blotting with a spliced leader probe , we found that the global mRNA level was decreased after the 1 h-heat shock whether or not ZC3H11 RNAi had been induced ( Figure 5E ) ., The mRNA encoding alpha tubulin decreased by 20–30% and mRNA encoding glycerol-3-phosphate dehydrogenase by 50% ( Supplementary Figure S5 ) ., As expected , in heat-shocked cells without RNAi HSP70 mRNA persisted ( Figure 5F , lanes 1 & 2 ) ., In contrast , after ZC3H11 RNAi , stabilisation of HSP70 mRNA was no longer seen ( Figure 5F , lanes 5 & 6 ) ., Similar results were observed for HSP83 , HSP110 , FKBP , and the mRNA encoding the HSP40/DnaJ-like protein J2; moreover , HSP100 mRNA was induced by heat shock in wild-type cells but not induced after ZC3H11 RNAi ( Supplementary Figure S5 ) ., Cultivation of the parasites at 37°C for 1 h caused a 70% increase in HSP70 mRNA which was prevented by ZC3H11 RNAi ( Figure 5F , lanes 3 & 4 ) ., We transfected procyclic forms with the CAT reporters containing the full HSP70 3′-UTR , the HSP70 3′-UTR fragments or the actin 3′-UTR ( Supplementary Figure S4A ) and subjected the parasites to heat shock ., This revealed that the AU-rich segment from the distal portion of the HSP70 3′-UTR was sufficient to confer persistence of the reporter mRNA in heat shock conditions ( Figure 5G , lanes 7–9 ) whereas the mRNA with the 5′ portion ( Figure 5G , lanes 4–6 ) behaved similarly to the actin control ( Figure 5G , lanes 10–12 ) ., We concluded that ZC3H11 is required for the heat-shock response of procyclic trypanosomes , and that the heat-shock response is required for recovery of the parasites from incubation at 41°C ., A previous microarray analysis had identified mRNAs that escape degradation after heat shock of procyclic forms 17 ., We repeated this analysis by RNASeq , comparing the transcriptomes of procyclic trypanosomes after one hour at 41°C with those of parasites that remained at 27°C ., A large number of mRNAs was affected ( Supplementary Table S1 , sheet 4 and Supplementary Figure S3D ) ., In theory , the 41°C RNASeq data should be normalised to allow for the fact that the total amount of mRNA is four-fold diminished by heat shock ( Figure 5E ) , which means that the read count ratio ( heat shock/no heat shock ) should be divided by four ., In practice , however , the un-normalised RNASeq results agreed better with those from Northern blots ( Table 1 ) ., We do not understand why this is the case ., Of the 178 loci that showed at least 2-fold more expression in the published heat shock microarray , 88 were confirmed as at least 2-fold increased in the RNASeq analysis; examples are listed in Table, 2 . Intriguingly , the increased transcripts were significantly enriched for the class encoding RNA-binding proteins ( P\u200a=\u200a0 . 007 ) ., Several chaperones were actually decreased after heat shock , but these were preferentially those involved in vesicular transport ( Supplementary Table S1 , sheet 4 ) ., Some of the mRNAs that increase after heat shock are also normally preferentially expressed in bloodstream forms ( Table 2 ) ., Overall , there was no correlation between mRNA changes after heat shock and binding to ZC3H11 ( Supplementary Table S1 , sheet 4 ) , indicating that for most mRNAs , other regulatory mechanisms are involved in stabilisation after heat shock ., To find mRNAs that were dependent on ZC3H11 after heat shock , we compared the transcriptomes of wild-type heat-shocked parasites with those of heat-shocked parasites with ZC3H11 RNAi ., 27% of mRNAs were at least 2-fold less abundant in the RNAi cells; less than 1% were increased ., Although there was no transcriptome-wide correlation between ZC3H11 binding and the RNAi effect ( Supplementary Table S1 , sheet 5 ) , every single one of the ZC3H11-bound heat-shock chaperone mRNAs was decreased in the RNAi cells ( Table 1 ) : the enrichment of chaperones in the subset that was both ZC3H11-bound and reduced in heat-shock was highly significant ( P\u200a=\u200a1 . 6×10−13 ) ( Supplementary Table S1 , sheet 7 ) ., The RNASeq results therefore showed that ZC3H11 is required for the retention of mRNAs encoding refolding chaperones after heat shock ., The decreases in other mRNAs in the RNAi cells may be secondary to the loss of chaperones or other proteins encoded by ZC3H11-bound mRNAs ., There are two basic ways in which an RNA-binding protein can stabilise an mRNA ., One possibility is that it has a direct stabilising function , for example by binding to other proteins such as translation factors or poly ( A ) -binding protein ., Th

Introduction, Results, Discussion, Materials and Methods

In most organisms , the heat-shock response involves increased heat-shock gene transcription ., In Kinetoplastid protists , however , virtually all control of gene expression is post-transcriptional ., Correspondingly , Trypanosoma brucei heat-shock protein 70 ( HSP70 ) synthesis after heat shock depends on regulation of HSP70 mRNA turnover ., We here show that the T . brucei CCCH zinc finger protein ZC3H11 is a post-transcriptional regulator of trypanosome chaperone mRNAs ., ZC3H11 is essential in bloodstream-form trypanosomes and for recovery of insect-form trypanosomes from heat shock ., ZC3H11 binds to mRNAs encoding heat-shock protein homologues , with clear specificity for the subset of trypanosome chaperones that is required for protein refolding ., In procyclic forms , ZC3H11 was required for stabilisation of target chaperone-encoding mRNAs after heat shock , and the HSP70 mRNA was also decreased upon ZC3H11 depletion in bloodstream forms ., Many mRNAs bound to ZC3H11 have a consensus AUU repeat motif in the 3′-untranslated region ., ZC3H11 bound preferentially to AUU repeats in vitro , and ZC3H11 regulation of HSP70 mRNA in bloodstream forms depended on its AUU repeat region ., Tethering of ZC3H11 to a reporter mRNA increased reporter expression , showing that it is capable of actively stabilizing an mRNA ., These results show that expression of trypanosome heat-shock genes is controlled by a specific RNA-protein interaction ., They also show that heat-shock-induced chaperone expression in procyclic trypanosome enhances parasite survival at elevated temperatures .

When organisms are placed at a temperature that is higher than normal , their proteins start to unfold ., The organisms protect themselves by increasing the synthesis of “heat-shock” proteins which can re-fold other proteins when the temperature returns to normal ., In trypanosomes , the degradation of mRNAs that encode heat-shock proteins is slowed down at elevated temperatures ., Trypanosoma brucei multiplies as “bloodstream forms” in the blood of mammals , at temperatures between 37–39°C; and as “procyclic forms” in Tsetse flies , which are usually at 20–37°C but can survive at 41°C ., In this paper we show that in Trypanosoma brucei , a protein called ZC3H11 can bind to many heat-shock-protein mRNAs ., ZC3H11 is essential in bloodstream-form trypanosomes and for recovery of procyclic-form trypanosomes after heat shock ., ZC3H11 binds to an AUU repeat motif which is found in parts of the target mRNAs that do not encode protein ., Several heat-shock-protein RNAs were decreased when we decreased the amount of ZC3H11 in bloodstream-form trypanosomes ., These and other results show that expression of the specific subset of trypanosome heat-shock proteins is controlled by the interaction of ZC3H11 with the relevant mRNAs ., They also show that the heat-shock response could enhance survival of trypanosomes in over-heated Tsetse flies .

rna, cellular stress responses, molecular cell biology, cell biology, nucleic acids, biology, microbiology, rna stability, molecular biology, parasitology, parasite physiology

journal.pgen.1003222

Admixture Mapping in Lupus Identifies Multiple Functional Variants within IFIH1 Associated with Apoptosis, Inflammation, and Autoantibody Production

Systemic lupus erythematosus ( SLE , MIM 152700 ) is a clinically heterogeneous autoimmune disease with a strong genetic component , characterized by inflammation , dysregulation of type-1 interferon responses and autoantibodies directed towards nuclear components ., SLE overwhelmingly targets women , and its incidence and clinical course differ dramatically between ethnic populations ., In particular , SLE occurs with at least 3–5 times higher prevalence and more severe complications in African-Americans ( AA ) compared to Americans with European ancestry ( EA ) 1 ., However , the genetic basis of this increased risk is largely unknown ., The recently “admixed” AA population is likely to provide critical information necessary to identify chromosomal regions that harbor variants associated with SLE and provide insights about allele frequency differences among distinct ancestral populations ( i . e . , European and African ) ., Admixture mapping ( AM ) has proven to be a powerful method to leverage ancestry information to identify chromosomal segments linked to disease 2–9 ., For instance , AM has helped identify the risk gene MYH9 in idiopathic focal segmental glomerulosclerosis in AA 10 , and risk alleles in several genes associated with breast 11 and prostate cancer 12 ., In addition to the greater lupus incidence , studying AA populations offers a second advantage ., Africans have the smallest haplotype blocks of all human populations: African average population recombination distance is 6 kb , while it is 22 kb in Europeans and Asians 13 , 14 ., This 3-fold smaller haplotype size gives rise to correspondingly tighter genomic associations in admixed populations such as AA , making causal mutations easier to decipher ., Although several genes for SLE susceptibility have been found through candidate gene analysis and genome wide association scans ( GWAS ) , none or very few causal mutations have been identified in each gene ., In this study we employed AM in AA to identify admixture signals , and performed a follow-up association study on AA and EA to further identify and localize variants associated with SLE ., We experimentally validated predicted variants with biochemistry , cell culture experiments and sequencing of patient-isolated samples ., We showed distinct functions of two coding SNPs including changes in gene expression ., Through electrophoretic mobility shift assays ( EMSAs ) , protein identification and in vitro protein binding assays , we determined that the intronic SNP disrupts function of a transcriptional enhancer of the IFIH1 locus ., Taken together , these results explain the effects of three independent causal mutations on SLE , and begin to elucidate the disparity in disease prevalence between different human populations ., Since case-only analysis has greater statistical power than case-control , we first performed a case-only admixture scan 3 , 6 , 8 on 1032 AA SLE cases ( Figure 1 , Table S1 ) ., Individual admixture estimates and genome scans for admixture mapping were analyzed using STRUCTURE 15 and ANCESTRYMAP 7 and later verified with ADMIXMAP 6 ., As expected , a two-ancestral population model ( African and European ) best explained the population structure of these samples ., By applying the ANCESTRYMAP software , we identified seven potential admixture signals that exceeded our predefined LOD threshold of 2 ( Figure 2A , Table S2 ) ., Specifically , we identified a genomewide significant association 7 of SLE risk with European ancestry at 2q22–q24 ( highest LOD\u200a=\u200a6 . 28 was achieved between rs6733811 and rs4129786 ) using a weighted prior risk model; the strongest association at the same locus was observed at a fixed prior risk of 1 . 5 , which represents a 1 . 5-fold increased risk of SLE due to one European ancestral allele at this locus ., To evaluate how the prior risk model could influence the ANCESTRYMAP results , we also applied a uniform prior risk model and found consistent genomewide evidence for association at 2q22–q24 ( LOD\u200a=\u200a5 . 86 ) ., We also reassessed the strength of the admixture signal at 2q22–24 using alternate markers ( LODodd\u200a=\u200a3 . 65 , LODeven\u200a=\u200a4 . 32 ) , and computer simulation ( P\u200a=\u200a0 . 02 ) ., We also validated case-only admixture signals with a case-control admixture scan with 800 ancestry informative markers ( AIMs ) , using 1726 controls from the Dallas Heart Study ( DHS ) ( Table S2 ) ., We next repeated the admixture scan using ADMIXMAP , which uses a classical ( non-Bayesian ) hypothesis test ( i . e . , score tests for allelic associations with the trait , conditional on individual admixture and other covariates ) ., All the admixture signals identified by the case-only design using ANCESTRYMAP were strongly validated by ADMIXMAP ( Table S2 ) ., The strongest peak was identified at 2q22–24 through ADMIXMAP ( P\u200a=\u200a2 . 99×10−8 , Table S2 ) ., Both AM programs found that the strongest effect was on the AIM rs6733811 ( P\u200a=\u200a2 . 99×10−8 , LOD\u200a=\u200a5 . 65 ) ., Two weaker signals were also found on chromosome 2 ( LOD\u200a=\u200a3 . 61 , P\u200a=\u200a4 . 82×10−3; and LOD\u200a=\u200a3 . 52 , P\u200a=\u200a1 . 64×10−4 ) , as well as on chromosomes 7 ( LOD\u200a=\u200a3 . 26 , P\u200a=\u200a6 . 49×10−6 ) , 9 ( LOD\u200a=\u200a3 . 43 , P\u200a=\u200a1 . 41×10−5 ) , 14 ( LOD\u200a=\u200a2 . 44 , P\u200a=\u200a3 . 84×10−5 ) and 19 ( LOD\u200a=\u200a3 . 20 , P\u200a=\u200a1 . 15×10−5 ) ( Table S2 ) ., To identify SLE-susceptibility gene ( s ) within 2q22–24 , we performed a follow-up case-control ( CC ) association study in two ethnically diverse groups: CCAA ( 1525 cases , 1810 controls ) and CCEA ( 3968 cases , 3542 controls ) ( Figure 1 , Table S1 ) ., Individual ancestry was estimated using 216 highly informative AIMs ., Case-control association tests were performed using 284 SNPs from 20 plausible candidate genes spanning ∼21 megabases of 2q22–q24 ( 95% CI ( 142 . 4–163 . 6 ) , Table S3 ) ., Forty-two SNPs from 10 genes ( IFIH1 , CACNB4 , ACVR1C , KCNH7 , NEB , STAM2 , ZEB2 , NMI , ARHGAP15 , and ACVR2A ) showed significant association for the allelic test ( Puncorrected<0 . 05 ) in CCAA , whereas 23 SNPs ( in IFIH1 , CACNB4 , NEB , ARHGAP15 , and TNFA1P6 ) showed significant association in CCEA ( Table S4 ) ., The strongest associations occurred at IFIH1 ( interferon-induced helicase 1; PAA\u200a=\u200a3 . 52×10−5 , PEA\u200a=\u200a8 . 82×10−5 ) and CACNB4 ( voltage-gated calcium channel , beta subunit; PAA\u200a=\u200a9 . 07×10−5 , PEA\u200a=\u200a2 . 61×10−2 ) , which are separated by 10 . 2 Mb ., Among the 22 SNPs tested within IFIH1 , 13 were significantly ( P<0 . 05 ) associated in AA and 4 in EA ., Among 23 CACNB4 SNPs , 12 were significant in AA and 2 in EA ., Considering the number of associated SNPs , level of replication and involvement in autoimmune phenotypes 16–20 , we considered IFIH1 as the strongest candidate to explain the admixture-mapping signal ., Of the 13 SNPs significantly associated in AA , a preliminary imputation-based association analysis and comparing linkage disequilibrium ( LD ) determined that 11 were sufficient to tag the 13 associated SNPs ( Table S6 ) ., To increase the statistical power to detect variants associated with SLE , we genotyped these 11 IFIH1 SNPs in 949 healthy AA controls from the DHS , along with additional out-of-study controls ( Figure 1 , Table S1 ) ., Using single SNP analysis ( allelic and genotypic models ) , followed by conditional analysis and LD analysis across two populations , we detected three SNPs with potentially independent SLE association ( Table 1 , Figure 2B and 2E ) ., Based on an allelic model , intronic variant rs13023380 PAA\u200a=\u200a4 . 33×10−5 , PEA\u200a=\u200a9 . 52×10−11; Pmeta\u200a=\u200a5 . 20×10−14; OR\u200a=\u200a0 . 82 ( 0 . 78–0 . 87 ) , and a missense ( Ala946Thr ) variant rs1990760 PAA\u200a=\u200a2 . 02×10−4 , PEA\u200a=\u200a1 . 22×10−4; Pmeta\u200a=\u200a3 . 08×10−7; OR\u200a=\u200a0 . 88 ( 0 . 84–0 . 93 ) , were associated with SLE in both AA and EA ., Another non-synonymous missense variant ( Arg460His ) , rs10930046 , was initially associated only with SLE in AA ( PAA\u200a=\u200a1 . 81×10−7; OR\u200a=\u200a0 . 80 ( 0 . 73–0 . 87 ) ) , where the best fit genetic model was identified as dominant ( Pdom\u200a=\u200a1 . 16×10−8 , OR\u200a=\u200a0 . 70 ( 0 . 62–0 . 79 ) ) ( Table 1 , Table S5 ) ., This SNP is rare in EA , having a minor allele frequency ( MAF ) of only 1 . 3% in controls and 1 . 6% in cases ( PEA\u200a=\u200a0 . 086 ) ( Table 1 ) ., However , after conditioning on the other two associated SNPs ( rs13023380 and rs1990760 ) , the rs10930046 became marginally significant in EA ( PEA\u200a=\u200a0 . 017; OR\u200a=\u200a1 . 2 ) ., These SNPs remained significant in both AA and EA after adjusting for ancestry ( Table 1 ) ., Strikingly , the ancestral alleles at these three SNPs ( all ‘G’ ) are the minor alleles in at least one population: all three ancestral alleles are the minor alleles in EA; the ancestral rs10930046 allele is minor in AA as well ., We analytically estimated the joint population attributable risk ( PAR ) 21 using these three SNPs ( rs13023380 , rs10930046 and rs1990760 ) for AA ( 18 . 1% ) and EA ( 14 . 7% ) ., Most of the increased PAR ( % ) in AA was attributable to rs1090046 ( 12 . 5% PAR ) , whereas for EA very little was attributable to this SNP ( 0 . 3% PAR ) , likely due to the extremely low MAF ., For AA , we also sought to determine how much of the European ancestry risk ratio ( λ\u200a=\u200a1 . 5 , estimated by ANCESTRYMAP ) was attributable to the three SNPs at 2q22–24 ., Using the estimated ORs in AAs and the SNP allele frequencies of the two ancestral populations ( YRI , the Yoruba people of West Africa , was used as an African ancestral population; CEPH , Utah residents with Northern and Western European heritage , was used as an European ancestral population ( Table S8 ) , we estimated the locus-specific ancestry risk ratio ( λ; see Methods ) for each SNP ( λrs1990760\u200a=\u200a1 . 12 , λrs10930046\u200a=\u200a1 . 15 , λrs13023380\u200a=\u200a1 . 12 ) ., Assuming that each SNP contributes to the ancestry risk ratio independently , about 80% of the increased risk due to one copy of the European ancestry alleles estimated from ANCESTRYMAP ( ∼1 . 5 ) can be explained by the three SNPs at 2q22–24 , reinforcing our conclusion from admixture mapping that local European ancestry increases the disease risk at 2q22–24 ., We also repeated our admixture mapping by stratifying the cases by three genotype ( ‘AA’ , ‘AG’ and ‘GG’ ) at the most differentiated ( FstCEPH-YRI\u200a=\u200a0 . 38 ) SNP , rs10930046 ( NAA\u200a=\u200a279 , NAG\u200a=\u200a323 , NGG\u200a=\u200a114 ) ., Even with the small samples , we found a dramatically increased risk of European ancestry at rs10930046 ( LOD\u200a=\u200a10 . 78 ) in the homozygous ‘AA’ compared to the other genotypes , where ancestry association is insignificant ( LOD for ‘GG’\u200a=\u200a−6 . 46 and ‘AG’\u200a=\u200a−3 . 38 ) ., To identify additional SLE-associated variants , we performed an imputation-based association analysis in and around IFIH1 using MACH 22 with reference data from AA ( 207 controls ) and EA ( 594 controls ) using genotyping data from the ImmunoChip ( Figure 2B , 2E and Tables S6 , S7 ) ., Using stringent predefined criteria for imputation , there were 61 additional SNPs for AA , but only 1 for EA later used for conditional analysis ., Inefficiency of EA imputation was mainly due to presence of many low frequency ( <1% ) alleles and strikingly different LD structure ( Figure 2D and 2G , Table S7 ) ., A pair-wise logistic regression analysis conditioned on each SNP revealed that the three previously identified SNPs were each independently associated with SLE ., While in AA , rs13023380 , rs10930046 and rs1990760 accounted for the entire association spanning the whole gene ( Figure 2B , 2C ) , in EA , rs13023380 and rs10930046 were independently associated with SLE and accounted for the association ( Figure 2E , 2F ) ., Finally , comparing LD ( r2 ) between these three SNPs across nine datasets from seven ethnic populations , we concluded that these three SNPs are also physically independent ( Figure S2 ) ., Interestingly , using D′ we found that these SNPs are on the same haplotype in EA and AA , but most likely they are not in the ancestral populations ( Figure S2 ) ., In order to discover the ancestral origin of the risk ( ‘A’ in each of the 3 SNPs ) and protective ( ‘G’ in each case ) alleles for these three SNPs , we estimated local ancestry around the SNPs , then compared ( by allele frequency and fixation index ) AA individuals whose both haplotypes were European ( AAEUR , N\u200a=\u200a129 ) or African ( AAAFR , N\u200a=\u200a2124 ) , and to individuals from HapMap populations CEPH and YRI ( Table S8 ) ., Risk allele frequencies derived from the haplotypes were similar between AAEUR and CEPH , and between AAAFR and YRI ., Alignment of the human genome with other genomes strongly suggests that the protective alleles ( ‘G’ ) are ancestral , and that the risk ( ‘A’ ) alleles are derived ., For the two coding SNPs , the ‘G’ allele of rs1990760 ( and the resulting alanine amino acid ) is ∼100% conserved across 34 mammalian genomes ( Table S9 ) ; the ‘G’ allele of rs10930046 ( and the resulting arginine amino acid ) is ∼100% conserved across 50 vertebrate genomes ( Table S10 ) ., Introns are typically less conserved than protein-coding sequence , and accordingly the intronic sequence surrounding the rs13023380 SNP is only strongly conserved in primates; the base corresponding to rs13023380 is ‘G’ in each case ( Figure S6 ) ., In AA , only the rs10930046 risk allele is major; interestingly , all three ‘A’ risk , derived alleles are the major alleles in EA and the rs10930046 risk allele is almost fixed ( Table 1 ) ., This suggests a strong selective pressure against the SLE-protective alleles in humans 23 , which is not manifest in other animal species ., Given the strong association of these three SNPs in IFIH1 with SLE , we evaluated their effect on the function of the IFIH1 gene ., IFIH1 has been implicated in binding with dsRNA complexes generated as replication intermediates during RNA viral infections , leading to inflammation and apoptosis 24 , 25 ., The full length IFIH1 protein contains 1025aa in the following domains: caspase recruitment ( CARD ) ( aa115–200 ) , helicase ATP-binding ( aa305–493 ) , helicase C-terminal ( aa743–826 ) and RIG-I regulatory ( aa901–1022 ) ( Figure 5A ) ., Deletion of the ATP-binding domain , which includes rs10930046 , induces apoptosis in melanoma cells 26 ., The RIG-I regulatory domain , which includes rs1990760 , recognizes dsRNA , upon which the helicase domains are activated 27 ., Apoptosis has been associated with SLE pathogenesis in humans and mice 28 ., Furthermore , Ingenuity Pathway Analysis ( IPA ) indicates that IFIH1 interacts with several genes involved in apoptosis and inflammation ( Figure S3 ) ., To assess the effects of coding variants in apoptosis and inflammation , we mutagenized IFIH1 cDNA cloned in a mammalian expression vector with a poly-cistronic ( IRES ) GFP marker at the C-terminus ., We over-expressed IFIH1 in a K562 leukemia cell line and measured cell death for each risk SNP , comparing with the ancestral protective allele ., The rs10930046 risk allele ‘A’ significantly increased apoptosis over the protective allele ‘G’ ( 14 . 6% average increase at each time point between 44 and 92 hours , P\u200a=\u200a0 . 014 ) ( Figure 3A ) ., In contrast , the risk allele ‘A’ of rs1990760 had little impact on apoptosis ( P\u200a=\u200a1 . 0 ) , as expected since it is located in the RIG-1 regulatory domain , which is not involved in apoptosis ., To assess the effect of these polymorphisms on expression of downstream genes , additional transfected K562 cells were sorted for GFP+ cells by FACS and total RNAs were isolated from these cells ., These were subjected to RT-qPCR of 11 genes related to apoptosis , inflammation or viral response: NFκ-B1 , NFκ-B2 , RELA , CASP8 , CASP9 , TNFα , MAPK8 , MAVS , IFNA , IFIT1 and MX1 ., Gene expression analysis showed that over-expression of the ‘A’ allele of rs10930046 significantly increased expression of NFκ-B1 ( >2 . 8-fold , P\u200a=\u200a2 . 1×10−2 ) , CASP8 ( >1 . 8-fold , P\u200a=\u200a4 . 5×10−4 ) , CASP9 ( >3 . 5-fold , P\u200a=\u200a7 . 2×10−6 ) and MAVS ( >2-fold , P\u200a=\u200a9 . 6×10−3 ) compared to the ‘G’ allele ( Figure 3B , 3D , 3E , 3H ) but did not affect expression of NFκ-B2 ( P\u200a=\u200a0 . 14 ) , TNFα ( P\u200a=\u200a0 . 7 ) or MAPK8 ( P\u200a=\u200a0 . 9 ) ( Figure 3C , 3F , 3G ) ., While the ‘A’ allele of rs1990760 had no significant effect on expression of NFκ-B1 ( P\u200a=\u200a0 . 14 ) or CASP9 ( P\u200a=\u200a0 . 08 ) , it showed a significant decrease of TNFα ( >5-fold , P\u200a=\u200a1 . 8×10−7 ) , NFκ-B2 ( >3-fold , P\u200a=\u200a1 . 3×10−3 ) and CASP8 ( >1 . 5-fold , P\u200a=\u200a1 . 6×10−6 ) expression ( Figure 3B , 3C , 3D , 3E , 3F ) ., The rs1990760 risk allele also significantly increased expression of MAPK8 ( >2-fold , P\u200a=\u200a6 . 2×10−3 ) and MAVS ( >1 . 6-fold , P\u200a=\u200a8 . 5×10−3 ) ( Figure 3G , 3H ) ., Strikingly , interferon alpha ( IFNA ) expression was reduced for both risk alleles ( rs10930046 , >2-fold , P\u200a=\u200a4 . 2×10−3; rs1990760 , 2-fold , P\u200a=\u200a9 . 1×10−3 ) ( Figure 3I ) ., Reduced IFNA expression in SLE patients had been predicted for the risk allele of rs1990760 29 ., Our results not only confirmed this but also showed that expression of the risk allele of rs10930046 similarly reduced IFNA expression ( Figure 3I ) ., Similarly , expression of IFIT1 was also reduced ( rs10930046 , >0 . 75-fold , P\u200a=\u200a8 . 7×10−3; rs1990760 , >3 . 5-fold , P\u200a=\u200a9 . 1×10−7 ) ; and MX1 expression was decreased for rs10930046 ( >2 . 5-fold , P\u200a=\u200a2 . 1×10−3 ) but was increased for rs1990760 ( >1 . 5-fold , P\u200a=\u200a3 . 3×10−4 ) ( Figure 3J , 3K ) ., Following induction of transfected cells with Type-1 interferon IFN beta ( IFNB ) , IFIT1 and MX1 showed strong up-regulation ( Figure 3L , 3M ) by both SNPs ( IFIT1: rs10930046 , >2-fold , P\u200a=\u200a2 . 0×10−2; rs1990760 , >2-fold , P\u200a=\u200a2 . 7×10−2; MX1: rs10930046 , >1 . 3-fold , P\u200a=\u200a3 . 1×10−2; rs1990760 , >2 . 2-fold , P\u200a=\u200a4 . 7×10−5 ) ., RELA expression did not change significantly ( for rs10930046 , P\u200a=\u200a0 . 61 and for rs1990760 , P\u200a=\u200a0 . 77 ) for either risk allele ( not shown ) ., In our expression analysis , significant up-regulation of CASP8 , CASP9 and NFκ-B1 ( and unchanged NFκ-B2 and TNFα levels ) by the rs10930046 risk allele would be expected to dramatically increase apoptosis , as observed ., For rs1990760 , levels of these five pro-apoptotic factors are dramatically lowered , consistent with absence of an apoptosis phenotype ., MAVS ( mitochondrial antiviral-signaling protein ) expression was increased for both risk alleles ., MAVS is an antiviral protein in the host defense system whose virus-triggered cleavage is necessary to attenuate apoptosis 30 , 31 ., However , without viral attack MAVS induces apoptosis through caspase and NFKB activation 30 ., In our case , it could promote apoptosis , particularly for the risk allele of rs10930046 ., In terms of inflammation , the expression data shows some interesting effects ., For rs10930046 , neither TNFα nor MAPK8 changed , but for rs1990760 , TNFα decreased while MAPK8 increased leading to inflammation signaling through non-apoptotic pathways ., We next examined known transcriptional networks in the context of our expression data ., At the root , IFIH1 and type-1 interferons constitute a positive-feedback loop ( Figure 3P ) ., We verified this in our cellular model: indeed , in control cells , IFIH1 expression increased several hundred-fold upon IFNB treatment ( Figure 3N ) , and in IFIH1 over-expressing cells , IFNA expression increased ( Figure 3O ) ., Taken together , our results support the predicted IFIH1-Type1 interferon feedback loop through IRF7 32 and MAVS 33 ., IFNA and TNFα are known to drive IFIT1 expression 34 , and the IFIH1 SNP-driven decrease in IFIT1 may be mediated through decreased IFNA and/or TNFα ., MX1 ( interferon-induced GTP-binding protein ) is also driven by IFNA and TNFα 35 ., Surprisingly , although both IFNA and TNFα decreased in the presence of rs1990760 ‘A’ , MX1 was significantly up-regulated ., This result is similar to a recent paper 29 , which showed that when SLE patients cells were induced with IFNA , the rs1990760 ‘A’ risk allele displayed higher levels of MX1 than ‘G’ allele patients , even though these patients had lower circulating IFNA levels ., Our data suggest that IFIH1 risk alleles at these two coding SNPs down-regulate IFNA expression ( either through reduced expression or activity of IFIH1 ) and , in turn , IFNA down-regulates interferon regulatory antiviral genes , potentially conferring viral susceptibility ., It is also possible that the rs1990760 risk variant in IFIH1 may increase sensitivity of cells to IFNA pathway activation and subsequent IFN-induced gene transcription 36 ., The intronic variant rs13023380 could influence IFIH1 function either by producing a functional miRNA or by altering the binding efficiency to one or more nuclear regulatory proteins ., Through bioinformatic analyses ( miRBase: http://www . mirbase . org/ ) we confirmed that no reported or predicted miRNA-producing or binding sites were present in these sequences ., To address whether rs13023380 alters nuclear protein-DNA interaction , we performed EMSAs on nuclear protein extracts from K562 and JURKAT cell lines , using 150-bp PCR products amplified from genomic DNA of ‘AA’ and ‘GG’ homozygous patients ., Both PCR products containing the ‘A’ risk sequences or ancestral ‘G’ sequences bound to nuclear protein extract , but DNA containing ‘A’ sequences consistently showed ∼2-fold reduced binding efficiency to a protein complex compared to ‘G’ sequences ( Figure 4A , Figure S4F ) ., To identify any DNA-bound proteins , we performed mass spectrometric sequencing ( MALDI-TOF ) on the protein/DNA complexes isolated using two separate methods: 2D electrophoresis and protein pull-down ., In 2D electrophoresis , the visible DNA-bound protein complex in EMSA was excised from a native PAGE gel ( Figure S4A–S4C ) and sequenced directly , which identified lupus autoantigen Ku70/80 ( XRCC5/6 ) , nucleolin ( NCL ) and HSP90AA1/AB1 as the major constituents of the DNA–protein band ( Table S11 ) ., Using the second method , we performed EMSA with biotinylated PCR products and pulled down the DNA-bound proteins using immobilized streptavidin-coated agarose beads ., Subsequent fractionation by SDS-PAGE ( Figure S4D ) , and sequencing of two distinct visible protein bands ( not present in the control pull-down product ) , confirmed NCL and HSP90AB1 ( Table S11 ) ., We did not identify Ku70/80 in the streptavidin method , possibly because these two proteins were washed off or were present in insufficient quantities to detect and sequence ., However , when we performed “super-shift” assays with antibodies to these proteins , surprisingly , anti-NCL and anti-Ku70/80 antibodies released EMSA-bound DNA instead of super-shifting the complex ( Figure S4E ) ., It is possible that the antibodies either induce conformational changes in their targets to release DNA or compete with target proteins for DNA binding ., Autoantibodies against NCL and Ku70/Ku80 are characteristic features of SLE 37 , 38 and release of free DNA from EMSA-bound DNA in vitro implies that autoantibodies in vivo could impair the function of these proteins by disrupting the binding of bound proteins from target DNA , including the rs13023380 locus ., In light of the observed competition of added antibodies to protein-DNA binding , we determined whether purified recombinant proteins of NCL and Ku70/80 bound to these DNAs ., Both recombinant proteins , purified from insect cells , produced identical gel shifts as the nuclear extract ( Figure 4B , 4C ) , but again the risk ‘A’ allele bound to the recombinant proteins with ∼2-fold decreased efficiency relative to the protective ‘G’ allele ., These results prompted us to enquire whether DNA sequence containing rs13023380 and its surroundings could act as a transcriptional regulatory element ( TRE , e . g . enhancer/silencer ) in vivo , and if the risk allele has any effect on transcription ., The same sequences used for EMSA were cloned before a minimal TKmin promoter and a luciferase reporter gene , and luminescence assays were performed ., Both sequences increased reporter gene activity over the core vector , suggesting that the rs13023380 locus contains a transcriptional enhancer ., The risk allele-carrying sequences showed almost a 2-fold reduction ( Figure 4D ) in luciferase activity compared to those with the ancestral allele ., Taken together , these results suggest that the rs13023380 locus recruits transcriptional activity of IFIH1 through binding of Ku70/80 , NCL and HSP90AA1/AB1 ( and potentially more proteins ) , and that the risk allele at this base position interferes with this enhancer activity , potentially decreasing IFIH1 transcript levels ., The absolute conservation of Ala946 ( rs1990760 , Ala946Thr ) in all sequenced mammalian genomes , with diverse codons , strongly suggests selection at the amino acid level ( Table S9 ) ., We performed molecular modeling of IFIH1-Thr946 , based on the protein structure of the IFIH1 C-terminal domain ( PDB 2RQB ) , and the full-length structure of the homologous enzyme RIG-I , bound to dsRNA ( PDB 3TMI ) ., Ala946 is placed directly at the mouth of the helicase active site; in RIG-I this region makes contact with the helicase “cap” , which mediates dsRNA entry and processing 27 ., Mutation of alanine to the bulker threonine side-chain ( Figure 4E , 4F ) may alter the sterics and/or dynamics of this protein region , leading to loss-of-function ., Similarly , Arg460 ( rs10930046 , Arg460His ) is conserved in all vertebrate genomes sequenced , with diverse codons , again implying amino acid-level selection ( Table S10 ) ., Comparison with RIG-I ( PDB 3TMI ) suggests that in the ancestral protein , Arg460 may form hydrogen bonds with the 419–433 loop , most likely with the strictly conserved acidic side-chains of Glu425 and Glu428 , and the conserved Gln433 ( Figure 4E , 4G ) ., Intriguingly , the crystal structure of the human IFIH1 ATP-binding ( DECH ) domain ( PDB 3B6E ) incorporates the pervasive rs10930046 risk mutation ., In this structure the His460 side-chain does not make favorable contacts with the 419–433 loop and much of this loop is poorly structured ., Loss of stabilizing interactions of Arg460 might lead to weakened structural integrity of the helicase ATP-binding domain ( the 3B6E domain is internally shifted ∼1 . 5 Å relative to the RIG-I structure; Figure 4G ) , and subsequently with the helicase C-terminal and RIG-I regulatory domains ., DsRNA binding , which occurs at a site proximal to the rs10930046 mutation ( Figure 5A ) , leads to RIG-I dimerization 39 ., The disruptive nature of the rs10930046 risk allele on overall protein structural integrity apparently decreases dimerization , as the 3B6E structure was determined as a monomer ( all related structures are dimers ) ., Indeed it is likely that the rs10930046 risk allele structure “poisons” an ancestral binding partner , leading to a dominant negative phenotype , consistent with the genetically dominant model , especially in AA ., The intronic rs13023380 risk allele has no effect on the protein-coding sequence of IFIH1 ., The region directly surrounding rs13023380 is rich in strongly conserved C/G bases ( Figure S6A ) ., Given the binding of the locus to NCL and other nuclear regulatory proteins , we hypothesized that the site might play a role in mRNA processing ., Modeling of the region around rs13023380 predicts a highly structured pre-mRNA , with strongly favorable folding free energies ( CentroidFold , ncRNA . org ) ( Figure S6B ) ., In the ancestral pre-mRNA , the rs13023380 base is part of a highly structured 7-mer RNA stem with a 7-base loop ( Figure S6B ) ., In the risk allele pre-mRNA , mutation of the conserved rs13023380 base disrupts RNA stem formation , and likely perturbs structure and stability of the loop ( Figure S6C , S6D ) , which might disrupt the binding of RNA-binding proteins ( such as NCL 40 ) , impairing pre-mRNA trafficking and processing ., Our whole genome admixture scan identified 7 admixture peaks associated with SLE in AA , with the strongest at 2q22–24 , containing the IFIH1 gene ., Three SNPs ( two coding: rs1990760 and rs10930046 , and one intronic: rs13023380 ) accounted for the increased risk ., IFIH1 has been associated with Type 1 diabetes ( T1D ) 41 , IgA deficiency 18 , Graves disease 17 , and suggestively linked to SLE 20 , 42 ., The role of IFIH1 in apoptosis and inflammation makes it potentially critical for SLE progression ., Moreover , allele frequency differences in associated and non-associated SNPs ( high FST values ) , together with the differences in the number of rare variants between EA and AA , imply a strong positive selection in EA ( intriguingly , for the SLE-risk alleles at all three positions ) , as previously suggested 23 ., In AA , local European ancestry at these loci correlates with increased risk ., Variant rs1990760 has been recently reported to affect expression of viral resistance genes IFIT1 and MX1 in SLE patients 29 ., The risk allele of rs1990760 positively correlated with interferon-induced gene expression in SLE patients who were positive for anti-dsDNA antibodies 29 ., Another report on rs1990760 suggested that the risk allele correlated with increased expression of IFIH1 in T1D patients 32 ., The rs10930046 risk allele has been implicated in psoriasis susceptibility 43 ., Here we have systematically examined the effects of the two coding SNPs on immune cell biology , and demonstrated that the rs10930046 risk allele dramatically increases apoptosis , and that both significantly perturb inflammatory gene profiles ., The intronic risk allele disrupts a transcriptional enhancer that recruits nucleolin , lupus autoantigen Ku70/80 and HSP90 , potentially decreasing IFIH1 transcript levels ., Combined with molecular modeling , our results strongly suggest that these effects are due to several specific amino acid and nucleotide substitutions , rather than to indirect effects due to LD with other SNPs ., SLE is commonly identified with an up-regulation of the interferon pathway 44 ., Intriguingly , our results suggest that the two non-synonymous IFIH1 mutations down-regulate interferon signaling ., However , recent findings demonstrated that SLE patients with anti-DNA antibodies have lower serum IFNA levels 29 , and this dose-dependent decrease suggests that there exists a sub-population of SLE patients with lower serum IFNA levels with increased IFN sensitivity 36 ., Heterogeneity is also observed in clinical TNFα levels; rs1990760 would seem a likely candidate to be associated with low TNFα levels in this patient sub-population 45 ., The intronic SNP ( rs13023380 ) discovered in this study has not been previously implicated in SLE or any other medical condition ., The transcriptional enhancer uncovered in this genomic region , and the risk alleles disruption of its activity , opens up new avenues for investigation ., Nucleolin , in addition to contributing to RNA polymerase 1 function 46 , is known to be a principal component of the B-cell transcription factor complex LR1 47 , which binds the Ig heavy chain switch region and functions in Ig recombination ., Disruption of nucleolin binding to the rs13023380 risk allele may dysregulate polymerase binding , IFIH1 transcription , autoantibody production and interaction ., The region surrounding rs13023380 is rich in highly conserved C/G bases , which are pref

Introduction, Results, Discussion, Materials and Methods

Systemic lupus erythematosus ( SLE ) is an inflammatory autoimmune disease with a strong genetic component ., African-Americans ( AA ) are at increased risk of SLE , but the genetic basis of this risk is largely unknown ., To identify causal variants in SLE loci in AA , we performed admixture mapping followed by fine mapping in AA and European-Americans ( EA ) ., Through genome-wide admixture mapping in AA , we identified a strong SLE susceptibility locus at 2q22–24 ( LOD\u200a=\u200a6 . 28 ) , and the admixture signal is associated with the European ancestry ( ancestry risk ratio ∼1 . 5 ) ., Large-scale genotypic analysis on 19 , 726 individuals of African and European ancestry revealed three independently associated variants in the IFIH1 gene: an intronic variant , rs13023380 Pmeta\u200a=\u200a5 . 20×10−14; odds ratio , 95% confidence interval\u200a=\u200a0 . 82 ( 0 . 78–0 . 87 ) , and two missense variants , rs1990760 ( Ala946Thr ) Pmeta\u200a=\u200a3 . 08×10−7; 0 . 88 ( 0 . 84–0 . 93 ) and rs10930046 ( Arg460His ) Pdom\u200a=\u200a1 . 16×10−8; 0 . 70 ( 0 . 62–0 . 79 ) ., Both missense variants produced dramatic phenotypic changes in apoptosis and inflammation-related gene expression ., We experimentally validated function of the intronic SNP by DNA electrophoresis , protein identification , and in vitro protein binding assays ., DNA carrying the intronic risk allele rs13023380 showed reduced binding efficiency to a cellular protein complex including nucleolin and lupus autoantigen Ku70/80 , and showed reduced transcriptional activity in vivo ., Thus , in SLE patients , genetic susceptibility could create a biochemical imbalance that dysregulates nucleolin , Ku70/80 , or other nucleic acid regulatory proteins ., This could promote antibody hypermutation and auto-antibody generation , further destabilizing the cellular network ., Together with molecular modeling , our results establish a distinct role for IFIH1 in apoptosis , inflammation , and autoantibody production , and explain the molecular basis of these three risk alleles for SLE pathogenesis .

African-Americans ( AA ) are at increased risk of systemic lupus erythematosus ( SLE ) , but the genetic basis of this risk increase is largely unknown ., We used admixture mapping to localize disease-causing genetic variants that differ in frequency across populations ., This approach is advantageous for localizing susceptibility genes in recently admixed populations like AA ., Our genome-wide admixture scan identified seven admixture signals , and we followed the best signal at 2q22–24 with fine-mapping , imputation-based association analysis and experimental validation ., We identified two independent coding variants and a non-coding variant within the IFIH1 gene associated with SLE ., Together with molecular modeling , our results establish a distinct role for IFIH1 in apoptosis , inflammation , and autoantibody production , and explain the molecular basis of these three risk alleles for SLE pathogenesis .

molecular cell biology, gene expression, genetics, molecular genetics, biology, population genetics, genetics of disease, genetics and genomics

journal.pntd.0005747

Analysing published global Ebola Virus Disease research using social network analysis

The 2014/2015 West African Ebola Virus Disease ( EVD ) outbreak with more than 28 , 000 cases and 11 , 000 deaths , was a public health emergency of international concern 1 , 2 ., Although EVD was discovered in the former Zaire ( now: Democratic Republic of Congo ) more than 40 years ago , the absence of treatment generated global alarm and raised questions on the state of EVD research ., Studies analysing EVD transmission and clinical trials testing EVD treatments or vaccines have been difficult due to the small number of infected cases in previous outbreaks 3 , 4 ., Moreover , the pharmaceutical industry has been criticized for neglecting EVD research because it is not profitable enough as EVD occurred rarely and mostly in impoverished African communities 3 , 5–7 ., EVD outbreaks have attracted general public attention since the mid-90s , benefitting science funding , leading to increased publications , but EVD research funding is mostly spent outside of affected African countries and research capacity building there was neglected 8 ., The World Health Organization ( WHO ) called for greater transparency and better sharing of results from clinical trials as being a necessary contribution to facilitate research and development ( R&D ) for the benefit of science and patients 9 and published a research priority agenda 10 ., The necessity for increased transparency also applies to any existing EVD research and expertise to improve the value and efficiency of research efforts ., In order to enhance the understanding of on-going EVD research activities and its communities , social network analysis ( SNA ) of bibliometric data of EVD related scientific publications can be used ., Since co-authorships are the most visible and accessible indicator for collaborations , co-authorship-based SNA studies can be used to measure the presence of research collaborations and their evolution over time 11–13 ., SNA metrics can reveal network patterns and identify its most central and influential actors 14–16 ., The volume of publications , in combination with results from a co-authorship network analysis , can serve as a proxy indicator for R&D ., Besides mapping the research landscape 17 , especially co-authorship network analysis can provide insight into the degree of research governance and be relevant for strategic research planning 18 , 19 ., Moreover , information from collaboration networks can be used to identify potential collaborations in order to improve research communication and therefore maybe also influence research outcomes 12 , 20 ., The aim of this study is to identify EVD research activities and to analyse the structure of the evolving EVD research community network over time to map existing research collaborations and influential actors based on centrality network metrics ., Bibliometrics of 2 , 528 articles resulting for our WoS search were exported as tab-delimited data and imported into MS Excel as one bibliometric data set ( Fig 2 ) ., In the raw data set each entry referred to one publication ., We included data on title , authors , address of authors’ affiliated institution , publication year , source , language , document type , cited references , funding agency , publisher and subject category in further analysis ., Other columns were deleted from the data set ., Information on addresses of author’s affiliated institution , e . g . institution name , sub-departments and institution address including city and country , were split into separate columns ., Data processing and further cleaning was performed using the software AppleScript 22 and OpenRefine 23 ., Name disambiguation , e . g . Centers for Disease Control and Prevention was abbreviated as CDC , Ctr Dis Contr and Centers Dis Cont , orders within names , e . g . Univ Washington and Washington Univ or name spellings , e . g . , Univ Georgia , UNIV GEORGIA were identified and harmonised using OpenRefine algorithms or manually ., Missing data , e . g . missing country information of an affiliated institution , were substituted by manual web search ., If an institution name appeared with addresses in different locations in the data set , e . g . WHO with location Switzerland and location Copenhagen e . g . due to different regional offices , different locations were considered for construction of the network to account for institutions international representations ., Institutions duplicates originating from publications with multiple co-authors affiliated with the same institutions were eliminated to ensure a single weighting of institutions ., The free online application Table2net was used to extract network information from the refined data set to construct a Gephi readable file 24 ., Network nodes ( i . e . actors ) are institutions named as authors’ affiliations in original research publications ., Network edges are titles of joint publications from authors’ affiliated institutions ., The free software Gephi was used to calculate network metrics and visualise the networks 25 ., Network analysis provides various tools and metrics in order to assess different notions of importance of individual nodes and node groups ., As the simplest metric of centrality we calculated each nodes degree , as the sum of direct links to other nodes ., Nodes with more direct connections are considered more central ., The average node degree captures the number of actors that each actor is connected with on average ., The average weighted node degree also takes the weight of a connection between a pair of nodes into account 26 , 27 ., Betweenness centrality measures the frequency with which a particular node lays on the shortest paths between all other node pairs ., Therefore , nodes with a high betweenness are considered to have a broker position as they connect many other nodes and thus have a large influence on the transfer of items through the network , under the assumption that item transfer follows the shortest paths 26 , 28 ., We used a betweenness calculation algorithm for weighted graphs as developed by Opsahl 29 ., Besides positional properties of the nodes within the network , metrics are capturing topological aspects of the network as a whole ., This information can provide an insight on the evolution at network level ., Density measures were calculated to assess the connectivity of the network ., The density of a network is defined as the total number of existing edges divided by the total number of possible connections ., If edges exist between all nodes ( density = 1 ) a network is considered completely dense 26 , 28 ., Since density captures the probable feasible number of connections in a network , it is an indicator for possible community building 30 or innovation flow within a network 15 ., Communities within the network were detected using Gephi’s modularity algorithm ., Modularity measures the degree of separation of a network into modules or clusters ( communities ) ., While a modularity value of 1 indicates that the actors separate perfectly into self-contained clusters , a value of - . 5 suggest the opposite , a homogeneously connected network 27 , 31 ., Networks with a high modularity score employ dense connections between nodes within the modules but sparse connections between nodes from different modules ., For visual presentation of network metric calculations we used Gephis Force Atlas II algorithm in log-linear mode optimized towards hub dissuasion 32 ., Systematic search in WoS for publications containing “Ebola*” yielded a total of 4 , 587 publications between 1976 and 2015 , including original articles ( 2 , 531 ) , editorial material ( 659 ) , news items ( 437 ) , reviews ( 415 ) , letters ( 325 ) , meeting abstracts ( 157 ) , corrections ( 36 ) , notes ( 14 ) , reprints ( 7 ) , biographical items ( 4 ) and book reviews ( 2 ) ., Amongst the 2 , 531 original articles were 75 article proceedings and five article book chapters ., Three of those publications appeared with anonymous authors and were therefore deleted for social network analysis ( Figs 1 & 2 ) ., The first EVD research article was published in 1977 , shortly after the first noted EVD outbreak in 1976 ., Only few EVD publications were visible until the early nineties , whereas from 1994 onwards the number of yearly EVD publications increased continuously ( Fig 3 ) ., Since 1994 a higher frequency of EVD outbreaks were recorded and more EVD cases were being detected in almost every year ., Several localised EVD outbreaks in Africa have occurred with up to several hundred cases ., The initial EVD outbreak in 1976 , with a relatively high number of reported cases ( >600 ) , was followed by only a small number of publications on EVD research ., No EVD outbreaks were reported between 1979 and 1994 and hardly any publications were published on the topic ., The number of publications increased gradually and continuously after the second outbreak in 1994 , although compared to the 1976 outbreak only about one-tenth of cases were reported ( Fig 4 ) ., A substantial increase in EVD research publications occurred during the 2014/2015 West African outbreak ., An almost 10-fold increase from 2013 ( 171 ) , 2014 ( 772 ) to 2015 ( 1 , 621 ) was visible for almost all document types , but it was most pronounced for editorials ( 5 , 220 , 343 ) , letters ( 1 , 75 , 213 ) , news items ( 4 , 190 , 118 ) and meeting abstracts ( 9 , 5 , 66 ) respectively ., An increase in reprints , notes , biographical items and book reviews was not detected ., Bibliometrics of 2 , 528 original research articles were used for social network analysis ., Based on their co-authors’ affiliated institutions a global network including institutions from 101 different countries with 704 connections was constructed ( Figs 5 & 6 ) ., Research institutions in the United States ( US ) are among the most highly connected institutions in EVD research ( degree ( d ) = 80 ) ., They are mostly connected to institutions in Canada ( d = 40 ) with an edge weight ( ew ) of 130 and Europe , especially Germany ( d = 53 , ew = 110 ) , the United Kingdom ( UK ) ( d = 60 , ew = 90 ) and France ( d = 57 , ew = 51 ) , but also to Japan ( d = 32 , ew = 99 ) ., Connections between US institutions and institutions in EVD affected African countries are less frequent ( e . g . Guinea-USA ew = 14 , Sierra Leone-USA ew = 32 , Liberia-USA ew = 30 ) ., However , institutions in Sierra Leone and Guinea ( both d = 32 ) and other African countries , especially Nigeria , Uganda and Ghana , are embedded in the global research network with connections to UK , Germany , France and Switzerland ., The overall density of the global country-level EVD research network measures 0 . 15 , with an average degree of 14 . 65 and an average weighted degree of 61 . 01 ., Amongst all collaborations on country-level , nine research communities were identified using modularity-based community detection and visualised by different colours ( Fig 6 ) ., The largest community ( red ) is centred around the US with strong collaborations to Canada , Germany and the UK , representing 59 . 41% of the co-authorships collaborations ( weighted edges ) ., Another large community is a ( mostly francophone ) European–African community ( blue ) representing 31 . 68% of all co-authorships connections ., Among all published original research articles between 1976 and 2015 a total of 1 , 644 co-authors affiliated institutions were named , which yielded 9 , 907 co-authorship connections in the overall research network ( Fig 7 ) ., The main actors according to degree are the US government ( CDC USA , d = 353; NIH , d = 315; USAMRIID , d = 283 ) and WHO ( d = 256 ) ., Other prominent actors are from the US and European countries ., Most central institutions are publicly funded ( e . g . CDC USA , USAMRIID ) , government research institutions ( e . g . BNI , ISERM ) , ( mostly public ) universities ( e . g . Uni London , Univ Marburg ) or international institutions ( e . g . WHO ) or non-governmental institutions ( NGOs ) ( e . g . MSF ) ., Modularity analysis reveals 166 communities within the network ( Fig 7 ) , whereas the largest community ( blue ) represents 17 . 33% of the total network nodes and the second largest ( green ) represents 14 . 44% of the network nodes ., Numerous smaller and less connected communities exist in the periphery , with some being entirely disconnected from the main network ., The temporal development of the research network is visualised over four 10-year time periods ( Figs 8 , 9 , 10 & 11 ) ., In the first decade 1976–1985 , ( Fig 8 ) the network consists of only a few actors , with one large central cluster surrounded by four smaller clusters ., The German Bernhard-Nocht Institute ( BNI ) has the highest centrality degree ( d = 11 ) , closely followed by the Institut Pasteur , PHLS Center for Microbiology and Research ( Salisbury , UK ) and USAMRIID ., The CDC USA is a central institution ( d = 7 ) of a smaller cluster , publishing with African partners ( Kenyan Ministry of Health ) others ., Smaller research groups in Kenya ( Kemri Wellcome Trust , Institute of Primate Research , Kenya Trypanosomiasis Research Institute ) , UK and US published together , but had no connections with others ., In the second decade 1986–1995 , ( Fig, 9 ) two larger , but separate , research communities evolved ., One francophone French-Swiss-African community with a homogenous structure in which the Institut Pasteur published mainly with the University of Basel , Institut de recherche pour le développement ( IRD ) , Ecole national veterinaire Lyon and the Hospital Bichat Claude Bernard Paris ., The other community consists mostly of American and German institutions , with three main actors ( USAMRIID , CDC USA and the University of Marburg ) , where the USAMRIID and CDC USA connect this community ., During this period the WHO had its first appearance as a disconnected actor ., All institutions in the network of the second decade are public entities ., With the occurrence of new EVD outbreaks in 1994/1995 the EVD research network grew in the third decade 1996–2005 , ( Fig, 10 ) into a star-like structure with surrounding chains ., During this decade the CDC USA evolved as the most central actor ( d = 87 ) ., The University of Marburg ( d = 54 ) , USAMRIID ( d = 52 ) , WHO ( d = 46 ) and NIH ( d = 36 ) remain central but less prominent actors ., The network of the fourth decade 2006–2015 , ( Fig 11 ) is skewed by publications in 2014/2015 ., During this time only few public research institutions and university actors dominate the research collaborations but numerous new actors appeared ., Prominent cooperation exist between CDC USA and WHO and CDC , NIH and USAMRIID ., While the transnational WHO was well embedded in the network over these last two decades , all main network actors are public institutions , mostly from the US and European countries ., While the global EVD research network remains relatively consistent in the first two decades , the third and in particular the forth decade shows substantial overall increase in the number of institutions and the links between them ( Table 1 ) ., Simultaneously the average node degree and weighted node degree increased over time , which indicates a growing number of collaborations and research activity per institution ., The decreasing density of the network over all decades indicates a decreasing number of realised edges between nodes relative to the total number of possible edges ., The increasing average node degree implies a growing number of research connections per institution ., The number of communities increased in line with number of nodes ., The high modularity values show that the solutions of the community detection algorithm reflect the substructures of the graph well , i . e . the increase in communities is unlikely to represent a sheer increase in volume , but rather seems to capture an evolution of the field of EVD into several smaller communities ., A degree distribution analysis of the EVD research network in the fourth decade shows a skewed node-degree distribution ( Fig 12 ) ., While almost 100 nodes appear with a degree of zero ( d = 0 ) , indicating no collaboration at all , only few institutions have a very high degree above 160 ( mean 12 . 24; median 5 ) ., Most institutions had a degree of less than five ( d≤5 ) as they were named as affiliations by authors of few publications by authors that published with only few co-authors ., The few very well connected institutions , such as NIH and CDC USA , are the key actors in this period ., In fact the CDC USA has maintained a very central position in the network over all time periods ., The private NGO Médecins Sans Frontières ( MSF ) has only recently emerged within the network and is centrally embedded with a high degree ( d = 157 ) ., Since the first reported EVD outbreak in 1976 until today the total number of publications on EVD in WoS has exceeded more than 4500 publications , of which 2528 were original research articles ., Like in scientometric analyses we used joint publishing as a proxy indicator of scientific collaboration 17 and thus knowledge exchange for our SNA of the co-authorship network 11 , 13 , 30 ., Indeed for the EVD overall network we identified research contributions from 1 , 644 research institutions in 101 countries; most actors are indeed coming from the US 17 ., Since 1994 EVD research publications have increased continuously , steadily and independently of the major West African outbreak ., This growth in publications is mirrored by a growth in the number of institutions ( from 30 to 1 , 489 ) and edges ( from 60 to 9 , 176 ) and therefore on-going network growth accompanied by a decreasing network density ., The overall network is an extensive aggregation of 166 different communities with a clearly dominant anglophone and francophone community ., This same dominance is seen when analysing the most central actors by degree and betweenness centrality both confirming the dominance of 10 institutions in powerful , control or broker positions in the network 11 , 33 , 34 ., The pattern of a growing EVD network in size but with a reducing density is characterised by some outliers ( 106 institutions not connected ) , frequently less connected contributions from developing countries and the private sector , but with a strong and stable core of dominant or ‘central’ institutions ., These characteristics of the network are supported by many of the analyses we performed ., For example the relatively and increasingly poorly connected nature of the network ( network density ) , the heavily skewed node degree distribution with the median node degree remaining rather constant , the relatively compact nature of the network ( path lengths ) and the strong centralisation showing a dominance of a few very strongly connected actors and many poorly connected actors ., Although we acknowledge that our analysis is weakened by the absence of a comparator network ( a common challenge in emerging research fields ) , we also believe that our analysis bring some added value ., For example SNA metrics for the overall network shows a density of 0 . 007 and calculating network density for each decade individually showed progressively decreasing density from 0 . 138 in the first decade to 0 . 008 in the last decade ., While this is largely influenced by both the size ( the more actors a network includes , the more difficult it is for all actors to be connected ) and also the correspondingly rapid growth in the network ( connections take time to build ) , we still believe that these figures should raise questions about whether the network–and therefore research outputs–could benefit from greater connectivity and linkages and in doing so greater optimise knowledge transfer and the spread of innovation 15 ., The node degree distribution ( for the last decade from 2006–2015 ) further confirms both the observed increase in the average node degree is attributable to only a few central actors whereas the overall network was not well connected in this period ., Thus , the network growth during the 2014/2015 epidemic diluted connectivity , at a time when collaboration was arguably most needed ., These observations are built on when we look further at the node degree distribution for 2006–2015 ., This confirms that while most actors only had few connections during this time , some actors are extremely connected ., This distribution form has been described as “power law” or “scale free distribution” and is typically observed amongst poorly connected networks 35 , 36 ., This ‘concentrated core’ is corroborated by the high number of the average weighted node degree ( 17 . 89 ) , in contrast to the average node degree ( 12 . 05 ) , which is also an indicator that some actors in the EVD network are connected more strongly to each other than others due to repeated publishing 27 ., It shows that these actors have on-going collaborations , share research results intensely by jointly publishing—but focus sharing amongst their co-authors ., This latter finding is something confirmed by our SNA results , which show strong centralisation amongst six institutions ( CDC USA , NIH , USAMRIID , WHO , the University of Marburg and the University of Harvard ) , suggesting that knowledge is mostly exchanged within the network between and/or through these actors ., Centrality is a measure of power in SNA 37 , this is especially the case for our central actors whose knowledge broker status is confirmed with regard to EVD research due to their high degree and betweenness centralities ., Additionally , observation of the path lengths reveal further insight into the efficiency of information exchange , with the shorter the average path length of a network diameter , the more efficient is information exchanged within the network structure 26 , 35 ., We found that the average paths lengths ( 3 . 02 ) of the overall network is lower than the average node degree ( 12 . 05 ) , indicating both that some institutions have a lot of direct neighbours and that on average nodes can reach other nodes by crossing only two other nodes ., The network diameter ( 8 . 0 ) suggests that sub-graphs within the network do not span more than across a chain of eight nodes ., Taking both aspects into account this implies that the overall structure of the network is characterized by isolated and weakly connected components , i . e . localized small networks that have only few relations amongst each other ., Although our study cannot , unfortunately , reveal anything about the ‘type’ of research conducted , observations on the type of research institution maybe serve as a proxy for this insight ., Two notable observations here were both the relative underrepresentation and disconnectedness in the overall network of both research institutions from affected countries and the private sector ., Among the unconnected nodes appear some private industry actors ( e . g . , Novartis Vaccines , Biohelix Corp , Baxter Bioscience and Oravax Inc . ) , in addition to African universities such as the University of Benin and the University of Mbarara ., While there may be many good reasons that explain the disconnectedness , for example proprietary restrictions to collaboration ( in the case of industry ) , new entrants to the field or for resource-related barriers to International collaboration ., This observation remains significant for a number of reasons , presumably both of these actor types posses’ unique and distinct knowledge and capabilities that could diversify and strengthen the expertise within the network if better and more broadly integrated , this is likely even more the case during a public health emergency of international concern ., Also , this ability to identify disconnected but valuable nodes , demonstrates a great added value of tools such as SNA ., Finally the recent entry into the network of non-traditional research actors such as MSF should be welcomed , especially as endemic country capacity is being developed and integrated into international networks , due to their unique position as being close to patients in the field yet able to advocate–distant funders–on the need for a well-supported , needs-driven research agenda 5 , 38 ., We believe the structure , nature and evolution of the international EVD research network described in this paper presents some learnings for policy ., Looking positively , the network itself has maintained a similar structure–a relatively compact network with a few consistent actors at its core–over the four decades studied , implying it is a stable constellation ., This institutional memory provides a solid foundation for knowledge maintenance over time , indeed without central actors networks might be disrupted and knowledge exchange hampered 30 ., The growth in the network over time through the entry of new actors , particularly since 2014/2015 , is positive as it likely indicates the arrival of new ideas and approaches ., However although collaboration has increased over time , our analysis found that the network remains relatively poorly connected ., Hence there may be an additional role for the ‘central actors’ to expand their role beyond a hub for dissemination and exchange into a facilitator for integrating the newer actors and expertise into the network ., Additional opportunities presented by the network analysis include: a reflection on the , perhaps , over-reliance or vulnerability to the network of all of the ‘central actors’ being public government or university institutions ., The importance of predictable , sustainable , funding flows to their continued role as network ‘brokers’ feels more exposed in these current financially and politically turbulent times ., While the dominance of these institutions is not surprising , we assume that they have the infrastructure , capability and public-financing , it may represent a weakness in two respects: firstly , with respect to its insufficiently diverse expertise mix , particularly with respect to the translation of this research into the development of tangible , context-relevant tools and capacity building in affected countries 8 , 39; secondly , with respect to the risk of over-centralising expertise , resulting in the stifling or suppression of innovation and growth and development of new ideas ., Finally , in small research areas for diseases predominantly impacting the lives of those in low-income countries such as EVD , the inherent market failures indicate that this reliance of public-financing will likely continue Wölfel in: 3 , 5–7 ., Given this , we believe , that a valuable insight from our study is to observe ways in which the network efficiency could be enhanced to extract greater patient-impact from the public financing inputs ., For example: focused efforts on integrating new collaborators into the network , provision of tools to enhance the productivity and improved transparency and sharing of research data 9 , 40 the identification of expertise gaps and targeted filling of these gaps and lastly , but perhaps most importantly , National alignment , focus and financing coordination ( strategic research planning ) around the globally agreed prioritised research agenda 41 ., Although many of these calls have already been made by many actors , particularly since the 2014/2015 EVD outbreak we believe this study represents an important empirical tool to support these calls and inform National and global policy development as the global community works to avert the next EVD outbreak ., The use of bibliometric data has intrinsic limitations and restrictions related to any analysis of secondary data and where data ceases to provide information , in particular in relation to content or results of published research ., Two major limitations to our study were identified and previously highlighted ., The first being the absence of other publications with which to contextualise and compare our results ., This absence of relativity in our conclusions limits the comparative value of our findings although the absolute data remain valid ., Although SNA is increasingly being used as a tool to analyses research areas it remains a relatively new field so we are optimistic that this is a time-limited constraint ., Secondly , we acknowledge that our study would be greatly enriched by an ability to analyse the data by ‘type’ of research not only type of publication i . e . basic , applied , clinical , implementation research , translation , health systems etc ., However , at present , this is not a search field within WoS , so we were unable to attain the source data ., Should key , public , medical , search engines enable this in the future , SNA such as ours would be an even more powerful tool to provide insight into research focus and productivity ., This analysis we believe would have great value–supplementing existing financing and development pipeline analyses 42 , 43—in providing a more granular understanding of product development gaps and the persistent absence of tools for the prevention , diagnosis and treatment of EVD 6 , 44 ., Our analysis of decreasing network density over time could have been further triangulated with the use of an additional metric such as the percentage of the giant component or the clustering coefficient ., Other limitations include reporting delays and the possibility that some publications were not included in the WoS database , however sample testing of other databases , including PubMed . gov , did not reveal other publications on EVD ., Although the impact of missing publications was likely small future studies could aggregate studies from diverse databases and in particular try to assess contribution of private industries R&D ., Despite manual and automated attempts to resolve challenges with institution name cleaning and disambiguation it cannot be excluded that some actors and/or relationships were not captured or were captured incorrectly ., Although unlikely , errors of the software used cannot be completely excluded and different algorithms might lead to different presentations of results ., Therefore network visualisations should be critically assessed in context to minimise misinterpretations ., We further note that GeoLayout visualisation can be misleading since it locates the African continent in the map centre and visualised edges may overlap nodes ., For this reason a country distribution was processed additionally with Force Atlas 2 ., The use of only free available software and easy accessible bibliometric data from WoS both facilitate the easy reproducibility of our study ., We conducted the first systematic landscaping of published EVD global research bibliometrics using SNA tools for analysis and visualisation ., Since 1976 Ebola outbreak EVD research , numbers of authors and affiliated institutions and links between them are constantly increasing , mostly independent from outbreaks and in-particular in the past two decades ., The overall EVD research network is organised around a few co-authoring key actors , mostly publicly financed ., Low network density indicates room for increased cooperation between institutions , in-particular links to less connected and more peripheral institutions could foster knowledge exchange and innovation ., Key network actors , such as the CDC USA , maintained network coherence over time–and probably kept EVD research on-going ., Limited scientific collaboration of research organisations from LMIC and the private industry , and how they utilise their expertise and knowledge , is neglected ., However , the absence of effective treatments for EVD questions the existing EVD research network efficacy and efficiency and suggests the need for both direction and structure to optimize the network to focus on research relevant for treatments ., Since most institutions in the global network are publicly funded , guidance to direct and re-orientate research might be facilitated by funders ( through calls targeting knowledge and translation gaps ) and be offered by supranational policy setting entities such as WHO and its Global Observator

Introduction, Methods, Results, Discussion

The 2014/2015 West African Ebola Virus Disease ( EVD ) outbreak attracted global attention ., Numerous opinions claimed that the global response was impaired , in part because , the EVD research was neglected , although quantitative or qualitative studies did not exist ., Our objective was to analyse how the EVD research landscape evolved by exploring the existing research network and its communities before and during the outbreak in West Africa ., Social network analysis ( SNA ) was used to analyse collaborations between institutions named by co-authors as affiliations in publications on EVD ., Bibliometric data of publications on EVD between 1976 and 2015 was collected from Thomson Reuters’ Web of Science Core Collection ( WoS ) ., Freely available software was used for network analysis at a global-level and for 10-year periods ., The networks are presented as undirected-weighted graphs ., Rankings by degree and betweenness were calculated to identify central and powerful network positions; modularity function was used to identify research communities ., Overall 4 , 587 publications were identified , of which 2 , 528 were original research articles ., Those yielded 1 , 644 authors’ affiliated institutions and 9 , 907 connections for co-authorship network construction ., The majority of institutions were from the USA , Canada and Europe ., Collaborations with research partners on the African continent did exist , but less frequently ., Around six highly connected organisations in the network were identified with powerful and broker positions ., Network characteristics varied widely among the 10-year periods and evolved from 30 to 1 , 489 institutions and 60 to 9 , 176 connections respectively ., Most influential actors are from public or governmental institutions whereas private sector actors , in particular the pharmaceutical industry , are largely absent ., Research output on EVD has increased over time and surged during the 2014/2015 outbreak ., The overall EVD research network is organised around a few key actors , signalling a concentration of expertise but leaving room for increased cooperation with other institutions especially from affected countries ., Finding innovative ways to maintain support for these pivotal actors while steering the global EVD research network towards an agenda driven by agreed , prioritized needs and finding ways to better integrate currently peripheral and newer expertise may accelerate the translation of research into the development of necessary live saving products for EVD ahead of the next outbreak .

Ebola Virus Disease ( EVD ) research publications were used to analyse and visualise collaborations between institutions jointly publishing research results , using freely available social network analysis tools ., Constructed co-authorship networks between author affiliated institutions showed EVD research publications increased and networks evolved over time ., The global network is organised around a few co-authoring , mostly publicly financed key actors , highly connected with powerful and broker positions ., The results present an extensive narrative how modern empirical scientific methods for data processing and translation can supplement evidence-based arguments for public discussion on the status and focus of global EVD research ., Based on the network characteristics or concentration of expertise , we recommend a globally agreed and prioritized EVD research agenda may facilitate the translation of this research into new EVD tools ., Also , to analyse research networks regularly to enable public discussion on the direction in which research could be organized and optimised ., We would like to encourage others to utilize our methods with open access tools to enhance new methods to the field of NTD R&D .

united states, bibliometrics, sociology, geographical locations, social sciences, north america, network analysis, social networks, information technology, data processing, research and analysis methods, computer and information sciences, economics, research assessment, centrality, people and places, finance, social research

journal.pcbi.1003328

A Data-Driven Mathematical Model of CA-MRSA Transmission among Age Groups: Evaluating the Effect of Control Interventions

Staphylococcus aureus is one of the most common bacterial pathogens in humans and the most frequent cause of skin and soft tissue infections ( SSTIs ) 1 ., Strains of health care-associated methicillin-resistant S . aureus ( HA-MRSA ) were first identified among hospitalized patients in 1960 2 and dominated MRSA infections until late 1980s ., Since the early 1990s , there has been a dramatic increase in community-associated MRSA ( CA-MRSA ) , which is now endemic at unprecedented levels in many regions in the US 3–6 ., This increase in CA-MRSA appears to have regional variation and is more pronounced in children compared to adults 4 , 5 , 7 ., Cases of HA-MRSA and CA-MRSA are characterized by significantly different epidemiological and microbiological features 8 ., Evidence indicates that CA-MRSA infections result from physical contacts with MRSA carriers at home , community facilities such as gyms , nursing homes , or kindergartens 9 ., In addition to regular treatment of the actual infection of the infected individuals , typical intervention strategies against CA-MRSA related SSTIs focus on pharmaceutical treatment via decolonization using mupirocin 10–12 and reductions in contact rates between infected and non-infected individuals ., However , the role of these control interventions on CA-MRSA transmission dynamics remains poorly understood ., In particular , in this paper we asked if control strategies targeting symptomatic infected individuals were sufficient to achieve disease control in a population ., Although several mathematical models for MRSA transmission in hospital settings , nursing homes , and other inpatient facilities have been developed ( e . g . , 13–18 ) , there is a scarcity of transmission models of MRSA and relevant epidemiological data to parameterize them at the community level , but these could be useful to elucidate the transmission dynamics and the effect of control interventions on CA-MRSA ., For instance , in 13 , 14 and 19 , compartmental transmission models were developed to study the invasion of CA-MRSA into hospitals and the likelihood of coexistence between CA-MRSA and hospital-acquired ( HA- ) MRSA ., Moreover , HA-MRSA dynamics were exclusively studied in both 16 and 17 ., In these studies , health care workers were modeled as vectors transmitting disease among patients ( or residents ) with the goal of assessing the impact of quantifying the effect of targeted control measures ., Furthermore , most of these modeling studies have assumed homogenous mixing , but recent work has pointed to age-specific variation in MRSA infection risk 4 , 20 ., Hence , age-structured CA-MRSA transmission models at the community level and tailored to local epidemiological data could increase our understanding of the transmission dynamics of CA-MRSA and the impact of routine and novel intervention strategies ., Here we developed and parameterized an age-structured compartmental transmission model to study the transmission dynamics of CA-MRSA at the population level and evaluate the effect of various intervention strategies ., To calibrate the model , we employed a unique dataset Dataset S1 ) covering several years of SSTIs incidence during the period January 2004–December 2006 in Maricopa County , Arizona ., We also used additional incidence data for subsequent years 2007–2008 for validation purposes ., Based on our calibrated model , we estimated the reproduction number denoting the average number of secondary infections generated by primary infectious individual 21–23 and evaluated the effect of contact rate reductions aimed at infected individuals owing to awareness of infection as well as decolonization treatment strategies targeting symptomatic infected individuals or the general ( asymptomatic ) colonized subpopulation ., We obtained detailed data on CA-MRSA infections from the Center for Health Information Research ( CHIR ) , which is a university-community partnership between Arizona State University and several Arizona providers , insurers and employers ., The dataset ( Dataset S1 ) comprises records on hospitalization and outpatients visits by children and teenagers ( age≤19 years ) enrolled in the Medicaid program of Arizona from January 1 , 2004 to December 31 , 2008 ., We extracted records of all encounters diagnosed with skin or soft tissue infection ( SSTI ) based on ICD 9 codes ( 680 . xx-682 . 9x ) ., Each record contains information about the type of infection ( first-time infection or recurrent infection ) , age group , month and year of hospital or clinic visit , and whether the patient was treated with mupirocin ., Our data are based on SSTIs related infections , with a stationary fraction of MRSA-related infections during our study period 24 ., We also obtained population data by age groups for our study setting ., An extended data description is given in Text S1 ., We developed an SEIS ( Susceptible-Exposed-Infected-Susceptible ) transmission model that incorporates age heterogeneity in contact rates , infectiousness , and decolonization treatment rates ., Our model also keeps track of individuals with past infections because these individuals have been observed to have a higher rate of infection compared to those with no past infections 24 , and we are interested in assessing the effect of age-specific variation in infectiousness and the effect of targeted interventions ., Let , , be the respective number of susceptible , colonized ( asymptomatic ) , and infected ( symptomatic ) individuals in age group ( ) with no prior infections ., Similarly , let , , be the corresponding epidemiological states for individuals with prior infection history ., The schematic view of transitions among the 6 epidemiological states in our model for each age group is shown in Figure 1 ., Mupirocin is used for decolonization of CA-MRSA patients who have been treated by other antibiotics ., Hence , in our model only decolonized individuals coming from the infected compartments experience the additional decolonization rate ., Our transmission model is given by the following system of differential equations: ( 1a ) ( 1b ) ( 1c ) ( 1d ) ( 1e ) ( 1f ) for a\u200a=\u200a1 , 2 , … , n ., The spontaneous ( natural ) decolonization rate per unit of time ( month is denoted by d0 ) and is assumed to be the same for all colonized individuals independently of their infection history ., The progression rate from colonized to infected is τ0 for those colonized for the first time and τ1 for those with prior infections , where τ1>τ0 , with a relative risk factor ., Infected individuals in compartments I0 and I1 progress to the colonization ( with prior infections ) stage ( C1 ) following antibiotic treatment at a common cure rate γ ., Further , treatment for decolonization aimed at C1 transfers individuals to compartment S1 at a decolonization rate d1 ., Both colonized and infected individuals contribute to the force of infection , which is given by ( 1g ) where denotes the contact rate for the colonized and infected individuals with susceptible individuals in age group ., Moreover , and are the probabilities of transmission per contact ( contagiousness ) for colonized and infected individuals , respectively , which are assumed to be invariant across age groups ., Because infected individuals are assumed to be more infectious than colonized individuals , , with a relative contagiousness factor ., Further , is the population immigration/migration rate and is the total population size with new recruits into the susceptible population with no past CA-MRSA infection history ., We calibrated our model given by System 0 with time series data of first time and recurrent infections for age groups 0–4 , 5–9 , 10–14 , and 15–19 yrs ., from January 2004 to December 2006 and estimated the unknown epidemiological parameters ( Table 1 ) ., For this purpose , we assumed as initial conditions that all people in the population were free of past infections in January 2004 ( first time point ) ., That is , we set to 0 the initial value for , for ., The initial number of people with first-time infections in adult groups , and , as well as the percentage of completely susceptible people in each age group at the beginning of 2004 , ( common across the age groups ) , were taken as unknown parameters to be estimated ., We employed a delayed rejection adaptive Metropolis-Hastings ( DRAM ) algorithm in a Markov-Chain Monte-Carlo ( MCMC ) simulation framework 29 to estimate unknown model parameters ., We used the widely-used MCMC package coded in Matlab ( available from: http://helios . fmi . fi/~lainema/mcmc/ ) ( Text S3 ) ., For each estimated parameter we assumed uniform prior distributions with range values as given in Table 1 ., Posterior distributions for each parameter were obtained from the resulting Markov chains 30 ., Next , we selected a random sample of size 500 from the Markov chain of parameters to assess parameter uncertainty ., For model validation , we compared our calibrated model forecast for two subsequent years of time series data covering the period 2007–2008 ., Latin Hypercube Sampling ( LHS ) and Partial Rank Correlation Coefficient ( PRCC ) techniques 31 were used for sensitivity analysis to assess the impact of control strategies targeting different age groups on the reproduction number ., LHS provides remarkable efficiency in drawing a highly representative random sample of small size from a multi-dimensional distribution32 , 33 ., The magnitude of PRCC quantifies the importance of individual parameters , with the sign of the PRCC value indicating the specific qualitative relationship between the input and the output variable ., That is , positive values of PRCC implies that increasing values of the input variable lead to increasing values of the output variable ., Since a PRCC indicates the degree of the monotonicity between a specific input variable and a specific output variable , only input variables that are monotonically related to the output variable are included in this analysis 32 , 34 ., Hence , we examined the scatterplots between and each parameter to assess the monotonicity assumption ., We set out to analyze the effectiveness of the various types of control interventions including typical interventions targeting symptomatic infected individuals and novel decolonization strategies focused on the colonized reservoir pediatric population ., Specifically , we evaluated reductions in contact rates by infected individuals due to personal awareness of infection ., We also assessed the impact of decolonization treatment of infected people following regular antibiotic treatment of MRSA infections ., Finally , we examined the possibility of disease elimination by hypothetical decolonization treatment strategies targeting colonized people ( in compartment ) ., We modeled the first intervention by using an age-specific parameter , , , to denote the relative reduction in contact rates between different age groups ., Thus , the force of infection given by Equation ( 1g ) becomes ( 2 ) The decolonization treatment targeting symptomatic infected individuals was modeled by varying the decolonization treatment coverage ( in ) , for ., The effect of decolonization treatment for colonized people was modeled by adding an additional age-specific rate , , to the spontaneous decolonization rate , as the new transition rate from to ( Figure 1 ) ., Parameter is the product of decolonization coverage ( age-specific ) and reciprocal of duration of drug effect ( 2 weeks 10 , common across age groups ) by assuming perfect drug efficacy ., We estimated the associated reduction in and determined whether infections approached zero as we varied the age-specific decolonization coverage ., Model parameter estimates and their corresponding geweke indices are shown in Table 2 ., Overall the model yielded a good fit to the incidence curve covering the period 2005–2006 albeit three of the parameter estimates did not achieve high convergence based on their geweke indices ( , and ) ( Figure 3 ) ., Moreover , the model forecast for two subsequent years 2007–2008 tracked closely the additional incidence data ., We estimated the control reproduction number using Equation ( S2 ) ., which we derived using the next generation matrix method 35 , 36 ( Text S2 ) ., Parameters , , , , and satisfied the monotonicity assumption and the corresponding PRCCs are shown in Figure 4B ., We found that is most sensitive to the parameters ( progression rate from colonized to infected ) followed by ( reciprocal of the duration of being infected ) ., is also fairly sensitive to ( spontaneous decolonization rate ) ., All of them are significant from zero at the 0 . 001 level according to their p-values ., We have developed and parameterized the first age-structured model of the transmission dynamics of MRSA transmission at the community level using data on skin and soft tissue infections in children and teenagers who were enrolled in the Medicaid Program in Maricopa County , Arizona ., Our compartmental epidemic model includes age heterogeneity in contact rates , infectiousness , and decolonization treatment rates , and keeps track of individuals with past infection history ., We estimated the control reproduction number at 1 . 3 with 95% confidence interval 1 . 2 , 1 . 4 ., Sensitivity analysis of on the model parameters revealed that is most sensitive to the parameters ( progression rate from colonized to infected ) followed by ( reciprocal of the duration of being infected ) and ( spontaneous decolonization rate ) ., Using our calibrated model , we found that typical strategies focused on infected individuals were not capable of achieving disease control when implemented alone or in combination ., In contrast , our results suggest that novel decolonization strategies that target the general pediatric population colonized with CA-MRSA have the potential of achieving disease elimination ., We performed numerical simulations to explore the impact of various feasible intervention strategies such as reductions in contact rates by infected people owing to personal awareness of infection and effect of decolonization treatment targeting symptomatic infected individuals belonging to specific age groups or across the entire population ., We found that neither a single or combined strategy was able to achieve ., We also forecasted short-term disease dynamics in the presence of both types of interventions strategies starting in 2009 ., Our results suggest that reductions in contact rates by infected people has little effect on the disease prevalence , and that neither of the two strategies was capable of achieving disease control particularly among first-time infections ., Given that these two intervention strategies are the most widely practiced ( intuitively and clinically ) , our model-based results provide an explanation to the persistent levels of CA-MRSA in many US regions over 20 years since its first appearance 3–6 ., We also tested the effectiveness of some hypothetical decolonization treatment strategies targeting asymptomatic colonized people ., We found that could be reduced below 1 when the treatment was focused on 5–9 and 10–14 yrs ., age groups with coverage at 25% , and an even smaller coverage ( 5% ) was sufficient when all pediatric groups groups were targeted ., Long-term forecast of the infections showed that disease elimination is feasible within 5 years through decolonization treatment at 30% of colonized people in the pediatric population starting in 2009 ., Compared with published models on MRSA transmission , our model is novel in several ways ., First , our model accounts for transmission in a heterogeneous population based on age-specific contact rates calibrated using survey data on physical contacts 25 ., This enables our model to capture much more essential elements of the complex MRSA transmission problem and test the effect of different control strategies that target different age groups ., Second , our model was calibrated and provided a good fit to time series data using solid estimation techniques rather than solely relying on assumed parameter values from the literature ., It is worth mentioning that our estimate of the control reproduction number at 1 . 3 is in line with that assumed in prior studies ( e . g . , 37 ) ., Moreover , it important to note that while our model incorporates key features of the transmission process including a realistic population contact structure and dominant features of MRSA epidemiology , there are many limitations to policy models with respect to the incorporation of bio-medical , operational , political , and economic features ., No one model can claim to incorporate all assumptions and features given the limited data available to calibrate them ., We believe our transmission model could be useful to evaluate further control scenarios and formulate rational policy based on the best available evidence ., There are limitations to note about our model calibration ., Some of the Markov chains did not converge perfectly within 10000 steps in terms of the geweke index ., However , our objective here was to find reasonable values of parameters so that our model captured qualitatively the overall trend of the age-specific time series data ., In this sense , the MCMC algorithm worked much better than other sampling methods ( e . g . , Latin Hypercube Sampling or simple random sampling with repetition ) ., Moreover , we also calculated the R-squared ( ) , a widely-used index for goodness of fit for a general model , for each of the eight output variables , and it turned out that the s for first-time infections were much poorer ( less than 0 . 5 ) than those for the recurrent infections ( greater than 0 . 7 ) ., Hence , our model did a better job in modeling recurrent infections than first time infections ., Of note , our model described by System ( 1 ) does not account for seasonality patterns 38 , which may explain a significant fraction of variation in data that remained unexplained by the model ., Moreover , we did not account for the changing medical practice as our understanding of this pathogen has increased since 2004 ., The infection duration decreased over these years as the appropriate antibiotics changed from second line medications to first line medications 39–41 ., Therefore the model prediction might have overestimated the infection prevalence ., Our dataset itself has several limitations ., First , our data correspond to general SSTIs which may not be directly related to CA-MRSA ., However , it has been observed from microbiological analyses that the fraction of MRSA-related SSTIs remained stationary among the total SSTIs in our study population for our study period 24 ., Given that it will be difficult to truly determine the exact cases of MRSA-related SSTIs and their microbiological spectrum at the population level , our data set is the best among those available ., Second , the target population of the data set corresponds to the children and young adults in the Medicaid program in Maricopa County , AZ ., It is important to note that patients under Medicaid coverage belong to a lower income population , where the rate of infection has been described to be higher than that of the general population5 , 42 ., Hence , our model prediction likely overestimates the prevalence in the overall population in Maricopa County ., Further studies are needed to shed light on whether our findings can be generalized to other settings ., In summary , our model indicates that intervention strategies targeting only infected people , either by reducing their contact frequency with healthy people or by pharmaceutical decolonization are not capable of eliminating CA-MRSA infections at the population level ., By contrast , substantial reductions in the prevalence of HA-MRSA could be achieved via contact reductions via patient isolation , enhanced hand hygiene and screening and health-care worker cohorting strategies ( e . g . , 43 ) ., Our control scenarios based on decolonization treatment strategies that target asymptomatic colonized individuals indicate that finding a cost-effective method to locate colonized individuals and conducting decolonization treatment on a limited fraction of them could prove to be an effective intervention ., However , the possibility of resistance emerging from increased use of mupirocin cannot be overemphasized ., Future modeling studies could be carried out to to evaluate the effect of strategies aimed at reducing the impact of resistance emerging from large-scale use of mupirocin ., Moreover , further work should focus on the collection of morbidity data and quantification of the attack rates and transmission dynamics of CA-MRSA in other world settings with different socio-economic and climatic characteristics ., While our transmission model is based on key epidemiological features of CA-MRSA including age-specific heterogeneity in contact rates , infectiousness and decolonization treatment rates , predictions from more elaborate models that incorporate , for instance , seasonality in transmission efficiency or a more detailed picture of the transmission process ( e . g . , household level transmission models ) could be undertaken in future work as relevant epidemiological data become available .

Introduction, Methods, Results, Discussion

Community associated methicillin-resistant Staphylococcus aureus ( CA-MRSA ) has become a major cause of skin and soft tissue infections ( SSTIs ) in the US ., We developed an age-structured compartmental model to study the spread of CA-MRSA at the population level and assess the effect of control intervention strategies ., We used Monte-Carlo Markov Chain ( MCMC ) techniques to parameterize our model using monthly time series data on SSTIs incidence in children ( ≤19 years ) during January 2004 -December 2006 in Maricopa County , Arizona ., Our model-based forecast for the period January 2007–December 2008 also provided a good fit to data ., We also carried out an uncertainty and sensitivity analysis on the control reproduction number , which we estimated at 1 . 3 ( 95% CI 1 . 2 , 1 . 4 ) based on the model fit to data ., Using our calibrated model , we evaluated the effect of typical intervention strategies namely reducing the contact rate of infected individuals owing to awareness of infection and decolonization strategies targeting symptomatic infected individuals on both and the long-term disease dynamics ., We also evaluated the impact of hypothetical decolonization strategies targeting asymptomatic colonized individuals ., We found that strategies focused on infected individuals were not capable of achieving disease control when implemented alone or in combination ., In contrast , our results suggest that decolonization strategies targeting the pediatric population colonized with CA-MRSA have the potential of achieving disease elimination .

Community associated methicillin-resistant Staphylococcus aureus ( CA-MRSA ) is a bacteria that causes skin infections in the US ., We developed a mathematical model of CA-MRSA transmission among different age groups at the population level ., We parameterized the model using monthly time series data on number of SSTIs in children during the period January 2004–December 2006 in Maricopa County , Arizona ., Our model-based forecast to additional time series data covering the period 2007–2008 yielded a good fit to data ., Using our calibrated model , we calculated that an infected individual generates on average 1 . 3 infected people in a totally susceptible population in the study area ., We assessed the impact of intervention strategies including reductions in contact rates between infected and non-infected individuals and the effect of decolonization strategies aimed at infected individuals by drug treatment , and found that neither of the two strategies when implemented alone or in combination were able to control the disease ., In contrast , we found that decolonization strategies targeting the pediatric population colonized with CA-MRSA have the potential of achieving disease elimination .

journal.pgen.1001336

Population-Based Resequencing of Experimentally Evolved Populations Reveals the Genetic Basis of Body Size Variation in Drosophila melanogaster

Body size is a critical phenotype for all organisms ., In Drosophila species , considerable evidence from natural populations indicates size is under spatially varying selection , with larger size selected in higher latitudes 1–4 ., Body size is also sexually dimorphic in D . melanogaster , yet genetic variation for size is correlated between the sexes , and differences in optimal size have been shown to result in sexual conflict 5 ., As size can be correlated with development time 6 , 7 , life span 8 , 9 , sexual attractiveness 10 , fecundity 10–12 and other traits 13 , natural variation in this complex trait is of considerable interest ., Determining which genetic polymorphisms affect body size would elucidate the joint distributions of effect sizes , population frequencies , and network position—the key to understanding how biological systems accommodate change while maintaining performance ., It would also allow molecular dissection of trait correlations , sexual correlations , and interactions between genes and the environment , all of which may be crucial factors in maintaining genetic variation in natural populations ., However , genome-wide association studies of human variation , even when conducted with hundreds of thousands of individuals , have only explained a modest fraction of heritable variation for height 14 , probably due to small effect sizes and/or low population frequencies 14–16 ., In model organisms , where strict environmental controls and repeated measures of identical genotypes may reduce non-heritable variation , power is likely to increase substantially 17 , but adequate power to comprehensively map functional variation will likely remain challenging ., Because of the short generation time and ease of culture of D . melanogaster , a complementary approach may be to couple experimental evolution with population-based sequencing of evolved populations ., For over 100 years , experimentally evolved populations of D . melanogaster have been used to address fundamental questions in population genetics 18 , 19 , with experiments on body size from at least 1952 20 ., More recently , experimental evolution has been combined with complete genome sequencing to link genotype and phenotype in viral and microbial systems 21–23 ., In these systems , evolution generally begins with a single genotype , and adaptation occurs after mutation produces genetic variation ., This approach has yielded many insights into both the structure of molecular networks and the adaptive process ., By combining experimental evolution and population-based sequencing in D . melanogaster , it may be possible to investigate the genetic basis of natural population variation , which has been a primary goal of population genetics since molecular characterization revealed the extent to which genomes vary in natural populations ., Partial support for this proposal comes from a previous study which used microarray-based genotyping on individuals derived from populations selected for enhanced stress resistance 24 , and a recent study on selected lineages of domesticated chickens 25 ., Additionally , Burke et al . recently resequenced populations selected for divergent generation times 26 ., Though Burke et al . report little evidence for canonical “selective sweeps” on newly arising or rare causal variants , they do not attempt to estimate the locations or number of causal alleles ., Moreover , the history of the populations used potentially complicates these observations: before they were selected to have long and short generation time , the ancestors of all populations were selected to have long generation time , which may have biased later adaptive divergence towards alleles which have small enough effects to remain polymorphic during this initial period ., Here we further explore this approach , using populations of D . melanogaster derived from the outbred , lab-adapted LHM population ., This population was originally derived from a large collection of flies from California , and has been maintained under a precise and stable regime for over 400 generations 27 ., Although selection related to environmental variation has been minimized in this population , individuals compete for a limited amount of food and mates each generation , and variation in many traits , including fitness , is abundant 28 ., The factors that maintain variation in the face of drift and selection in this lab population are likely to be a subset of factors which maintain variation in populations in the wild ., To experimentally investigate the “Evolve and Resequence” ( hereafter E&R ) approach to genetic mapping , six populations were established: two were selected for large size , two for small size , and two were subjected to identical protocols , but not selected based on size ( controls ) ., A sieving apparatus was used to efficiently separate flies based on size: anesthetized flies were separated based on their ability to pass through a series of sequentially smaller sieves ( see Methods ) ., This allowed us to screen ∼1800 flies per population , each generation , for over 100 generations ., After 100 generations , the mean “sieve size” of the flies diverged substantially among the experimental populations ( Figure 1 , F9 , 32\u200a=\u200a89 . 52 , P\u200a=\u200a0 . 0001 ) ., Though a considerable response was seen in both directions , the response to selection was strongest among the small-selected lines ., Indeed , by the end of the selection experiment many of the male flies ( 79% and 35% ) in each small population passed through all 20 sequentially narrower sieves , whereas no flies pass this far through the sieve system in the control populations ., Anatomical measures of thorax , leg , and wing dimensions from each population verified considerable divergence in fly size ( Figure 1 and Table S1 ) ., All anatomical measurements agree that populations selected for small size evolved substantially , while populations selected for large size changed more modestly and were significantly different from controls for only some traits ., As large-selected populations evolved significantly in their ability to pass through the sieves , but have modest anatomical differences in the traits measured , this suggests that some of the response to selection is due to anatomical traits that were not directly measured , such as abdominal size ., To simultaneously determine the locations and frequencies of genetic polymorphisms , we extracted DNA from 75 pooled females ( 2n\u200a=\u200a150 chromosomes ) for each population , and sequenced these populations with the Illumina Genome Analyzer ., In total , we obtained 42 . 3 billion base pairs of sequence data , 99 . 8% of which aligned to the reference genome ., After excluding the 23% of alignments with low mapping qualities , which includes non-unique alignments , each population had between 17-fold and 23-fold median coverage , with 87% to 93% of the genome having more than 10-fold coverage in each population ( Table 1 , Figure S1; this excludes the 4th ( dot ) chromosome , Y chromosome , mitochondria , centromeric heterochromatin and unmapped regions , which are shown in supplementary data but will not be considered further ) ., Considering all 6 populations together ( ∼120-fold genome coverage ) , over 27 million base pairs were found to be variable ., However , the majority of these apparent polymorphisms are rare: 83 . 4% have overall frequencies less than 0 . 02 ., A considerable portion of these rare variants could be sequencing errors , which are difficult to completely exclude using pooled sequencing approaches ., Mean error rates from the UNC-CH sequencing facility , where sequencing was conducted , are 0 . 5–3 . 0% depending on read position ( C . D . Jones , pers . comm . ) , so apparent polymorphisms with >5% experiment-wide frequency should be true genetic variants ., Even when only considering apparent polymorphisms with population frequencies >10% across the entire data set , 1 . 68 million bases are variable , verifying that there is considerable genetic variation in these populations ., Although the large number of sequencing errors complicate some analyses by creating a large apparent excess of low frequency variation , these errors will be rare and randomly distributed , and are therefore not expected to be significantly differentiated between populations ., Differences in allele frequency between populations indicate that evolution has taken place , either due to stochastic forces ( drift ) , selection , or both: this evolution is quantified in Figure 2 ., As expected , evolution occurred between the two control populations after they were separated from a common ancestor for over 100 generations ., However , much more evolution has taken place between selection treatments than between control lines ., In the two independent comparisons between a large- and small-selected line , 41 , 399 and 48 , 645 variants are >95% differentiated , compared to only 1 , 260 variants between controls ( Figure 2 ) ., This considerable excess of highly differentiated variation indicates a substantial , genome-wide impact of artificial selection for body size ., Additionally , of the 5587 variants that achieved this extreme level of differentiation in both comparisons , the vast majority ( 5537 ) changed frequency in the same direction , clearly implicating selection for body size ., To investigate the genome-wide impact of selection further , we determined the genomic distribution of heterozygosity in each population ( only variants with an experiment-wide frequency >0 . 10 were used in order to avoid sequencing errors ) ., Median heterozygosity in 10-kb windows was 0 . 0031 and 0 . 0032 for control 1 and control 2 populations , respectively ., Average heterozygosity in each of the two large populations was only slightly , but significantly , less than the average of the controls: 0 . 0031 ( t-test P\u200a=\u200a0 . 048 ) and 0 . 0030 ( P≪0 . 001 ) ., The two small populations , however , have a 55% reduction in median heterozygosity compared to controls ( 0 . 00180 and 0 . 00175; P≪0 . 001 ) ., Fewer than 10 windows have near zero ( <0 . 0001 ) heterozygosity in each of the large and control populations , while 60 and 85 windows are depleted of variation in each of the small populations ., As this is still less than 1% of the genome , variation persists in most genomic regions ., The number of breeding adults of all six populations were identical and remained constant throughout the 100 generations of the study , so this reduction is clearly due to the action of selection ., To locate a set of variants that have evolved due to selection for body size , we considered data from the two large- and two small-selected lines together ., Only variants where the large-selected lines both had higher or lower allele frequencies than both the small-selected lines were considered high-confidence candidates for selection ., This immediately excludes 66% of allele frequency changes due to drift or selection unrelated to experimental treatment , as well as eliminating variants that were affected by linked selection in inconsistent directions ., For the 1 , 886 , 104 variants meeting this criterion , the minimum allele frequency difference ( diffStat ) between the four possible large-small comparisons was used as a composite statistic ( Figure 3 ) ., Most ( 74% ) variants had diffStats <0 . 10 , with just over 4% ( 76 , 719 ) of variants having diffStats >0 . 50 ., To determine a set of loci that are likely to have evolved under selection for body size , we compared the distribution of this statistic to simulations of drift alone ., The amount of allele frequency change expected due to drift depends on starting allele frequency , which we estimated for each variant as the average frequency of the two control populations ( see Methods ) ., For each allele frequency class , we then simulated drift for 110 generations , and simulated sampling error due to sequencing coverage ( see Methods ) ., Using these simulations , an expected distribution of diffStat was generated in the absence of selection ., By comparing the observed and expected distributions of the statistic , we estimated false discovery rate ( FDR ) thresholds for evolution due to selection for body size ( Tables S2 , S3 ) ., This FDR calculation implicitly assumes that the number of effective loci in the genome is approximated by the number of polymorphisms ., It does not assume that all polymorphisms are independent , however varying levels of linkage disequilibrium inflate the variance around the mean FDR ., For this and many other reasons , these calculations should be taken as approximations of the actual FDR ., As experimental populations are quite different from control comparisons ( Figure 2 ) , the observed and expected distributions are radically different ( Figure 3 ) , and differentiation occurs at many loci across the genome ( see below ) , this approximation seems acceptable ., Significantly differentiated variants are clearly distributed non-randomly across the genome: their distribution on chromosome arm 3L is shown in Figure 4 , and other chromosomes are shown in Figure S2 ., Differentiation exhibits a distinct peaked pattern in many regions , with a wide range of peak sizes ., Only a few variants are differentiated at some loci , providing precise hypotheses regarding functional targets ( Figure 5 ) ., In other regions , especially surrounding the chromosome 2 centromere , significant differentiation is spread across megabases ( Figure S2 ) ., A very large number of differentiated peaks are apparent when chromosomes are examined at a fine scale ., Regardless of the precise number of differentiated peaks , it is difficult to estimate the number of alleles that were selected ., For example , within each peak , it is possible that a single haplotype bearing the combination of multiple causal variants with the largest combined effect was selected ., However , to determine a very conservative minimum estimate , we counted the number of regions containing significant variants ( FDR<10% ) that were separated from all other significant variants by a minimum distance of 10-kb ., Based on this measure , 1236 distinct regions have evolved due to selection on body size—even if the minimum separation distance is extended to 50 kb , 304 distinct regions are observed ., As the true number of selected variants is likely much higher , this indicates that the response to selection was startlingly polygenic , with certainly hundreds , and likely thousands , of target loci ., Using the average frequency of each variant in the two control populations , we can obtain a rough estimate of the starting frequency of each differentiated polymorphism ., As shown in Figure 4 , the proportion of initially uncommon variants that became significantly differentiated varies greatly between peaks ( other chromosomes are shown in Figure S3 ) ., In some peaks , not even one differentiated variant has an average frequency <0 . 05 in the control populations , while other peaks have many such variants ., When using any correlative analysis ( e . g . genome-wide association studies , QTL mapping , or E&R ) , it is very challenging to differentiate causal alleles from linked variants ., For example , simulations of genome-wide association studies indicate that non-causal alleles can be more significant than causal alleles when the non-causal alleles are in linkage disequilibrium with multiple causal variants 29 ., Despite these caveats , we hypothesized that some variants with the local maximum diffStat values would be likely to either effect body size themselves , or be in close proximity to variants that do ., To delimit a set of such variants , we centered a 100-kb window on each significant variant ., As the structure of linkage disequilibrium is unknown in these populations , the choice of 100-kb is somewhat arbitrary , but is expected to be much larger than the normal extent of linkage disequilibrium across most of the genome , and is therefore conservative 30 ., If the diffStat value of the variant in question was larger than or equal to the maximum within this window , it was considered a “peak variant” ( that is , it was a local maximum; Figure 4 ) ., This method results in 5205 peak variants , 3572 of which lie in a 10-Mb region surrounding the chromosome 2 centromere ( 2L>18 Mb and 2R<5 Mb ) ., Of the 1633 peak variants outside this region , less than 10% have estimated starting frequencies less than 0 . 05; in contrast , 41% of the 3572 variants in the region surrounding the centromere started at frequencies below 0 . 05 ., Heterozygosity in this region is very low in the small-selected populations ( median<0 . 0001 ) , compared to the same region in the other four populations ( 0 . 0027–0 . 0030 ) , or the rest of the chromosome in the small-selected populations ( 0 . 0024–0 . 0025 ) ., Together , these results implicate one or more major selective sweeps in this region in the small-selected populations , which fixed a large number of rare variants and eliminated variation surrounding the centromere ., In regions with distinctly differentiated peaks , we hypothesize that peak variants are near the direct targets of selection ., As a partial test of this hypothesis , we assembled a list of genes at these loci ., The 10-Mb region surrounding the chromosome 2 centromere was excluded due to the large number of fixed differences throughout this region ., For the remaining 1633 peak variants , 632 genes either overlap the peak variant or are within 1-kb ., Functional annotations of these loci were compared to the complete genome using annotations from FlyBase 31 and the Database for Annotation , Visualization , and Discovery ( DAVID ) , which uses fuzzy clustering to group genes into functionally related classes based on the similarity of their annotations 32 , 33 ., The most over-represented cluster of biological processes ( GO terms ) includes genes affecting post-embryonic development and metamorphosis , with post-embryonic development also the most significantly over-represented biological process individually ( P\u200a=\u200a8 . 64E−7; Bonferroni-adjusted P\u200a=\u200a0 . 001; see Datasets S1 and S2 for full results ) ., As all anatomical features measured have changed between treatments , and the timing of metamorphosis is likely to alter adult size , these functions correspond precisely to phenotypic characterizations ., This functional cluster includes genes such as ecdysone-induced proteins ( l ( 3 ) 82Fd , Eip63E , Eip74EF , Eip75B ) , many genes involved in anatomical development ( vein , plexus , headcase , blistery , etc . ) and others ., The second most over-represented gene cluster was found to include the biological processes cell morphogenesis ( cell size and shape ) : cell number and cell size are both known to change with body size in Drosophila 7 , 34 , so genes with this annotation are excellent candidates for harboring natural variation for these traits ., Over 40 genes which are known to affect cell morphogenesis are near a peak variant ( P\u200a=\u200a3 . 46E−6; Bonferroni-adjusted P\u200a=\u200a0 . 006 ) , including genes involved in epidermal growth factor signaling ( including the epidermal growth factor receptor ) , the salvador/warts/hippo pathway ( salvador , crumbs ) 35–37 , and many others including E2F , dally , dally-like , knirps , and miniature ., When the 3572 peak variants surrounding the chromosome 2 centromere are included , these categories remain the two most over-represented categories of biological processes ., The large number of genes in this region precludes confident assessment of the targets of selection , but it is notable that the ecdysone receptor is found here , and contains fixed differences in multiple introns , exons , and both 5′ and 3′ UTRs ., Finally , as an additional control , we determined if any of the biological process mentioned above were over-represented near variants which had >90% allele frequency difference between controls ., Of the biological processes mentioned above , post-embryonic development was the most significant between control populations ( P\u200a=\u200a0 . 002 ) , and was not significant after Bonferroni correction ( P\u200a=\u200a0 . 74 ) ., This is in stark contrast to the overrepresentation between treatments: cell morphogenesis , post-embryonic development and metamorphosis all show greater than 2-fold enrichment , and are all significant after Bonferroni correction ., For a complete list of the genes near peak variants , and a list of the variants themselves , see Datasets S3 and S4 ., Some genes are notable for their absence from this list: although insulin regulation is crucial for body size determination , many of the canonical genes in the insulin pathway are not near a peak variant ., However , some of these genes , including Tor , slimfast , and glut1 38 overlap significant ( FDR<10% ) variants ., Because these variants are within 50 kb of a more significant variant , they are not included in the peak variants list ., This may indicate that natural variants at these loci have smaller effects than many other genes , and this hypothesis could be tested by including the significant variants in a follow-up association study ., Even more remarkably , the adaptor protein chico and the insulin receptor itself are far from any polymorphisms which responded to selection in this regime ., By resequencing experimentally evolved populations , the genomic impact of selection on allele frequency is shown to be considerable ., Nearly all variants which are >95% differentiated between the two independent comparisons of a large- and small-selected population are differentiated in the same direction , supporting the assertion that these variants have changed due to selection on body size ., Heterozygosity is reduced in the selected populations , especially in the small populations where the most phenotypic change has occurred , consistent with the expected effect of linked selection 39 ., It is clear that at least hundreds , and likely thousands , of polymorphisms affect body size in this long-term laboratory population ., Though some of these alleles may have arisen through mutation after the founding of this population from nature , most are expected to be variants that affect body size in natural populations ., In any case , it is clear that the response to selection was due evolution at many loci throughout the genome ., A recent meta-analysis of human height variation provides perspective on this number: by genotyping over 180 , 000 individuals , 180 loci affecting human height were located , but these loci together explain only 10% of the phenotypic variation 14 ., Body size in D . melanogaster would appear to be similarly polygenic , and our study demonstrates that these loci can be mapped with much less genotyping effort and expense than human GWA studies ., Of course , mapping these loci provides only the initial step towards comprehensive characterization of this variation ., For example , the relative effect sizes of these loci are not yet known , and it is possible that some loci are selected because they are compensating for the pleiotropic side effects of body size change ., These challenges can be addressed either using association studies or further artificial selection , but in either case the loci mapped here can be used as a priori candidates , tremendously increasing power while decreasing the effort required ., The proportion of trait variation due to common versus rare variants is of great interest ., Interestingly , differentiation of uncommon ( <5% frequency ) variants occurred at only some peaks ( Figure 5 , Figure S3 ) , which may indicate that these peaks are caused by selective sweeps on rare ( or at least uncommon ) variation ., At other peaks , no uncommon variation is differentiated , which is inconsistent with selective sweeps of haplotypes from low to high frequency ., It is tempting to conclude that selection response in these regions was due to common variants , but this conclusion is tenuous ., For example , if an uncommon combination of common mutations was selected in some regions , rare variants might hitchhike to high frequency on this uncommon haplotype ., Conversely , peaks where only common mutations are differentiated could theoretically be created by slight changes at a large number of linked rare variants ., We consider this second possibility to be unlikely , as the effective number of alleles per locus is likely modest in D . melanogaster , due to the short range of linkage disequilibrium in this species 30 ., Consequently , these data are consistent with much of the standing variation in body size being due to common variants , but definitive conclusions regarding the relative importance of common and rare variants await follow-up studies determining if the most differentiated variant in each peak is indeed causal ., Though the most differentiated variant in each peak is not assured to be under direct selection , it is the most probable variant in each peak ., Indeed , some peaks have only one or a few significant variants , allowing confident hypotheses to be formed about the specific causal variants , and in many others there is a single variant which is much more differentiated than all others nearby ( Figure 5 ) ., The correlation between significantly differentiated “peak variants” and gene functions expected to effect body size is striking , and supports the assertion that these variants may be near the direct targets of selection ., This correlation implies that many genes known to be involved in anatomical development , metamorphosis , cell number , and cell size harbor natural genetic variants affecting these same traits ., However , we were also able to exclude the involvement of several loci , including chico and insulin receptor ( InR ) , which have key roles in body size and development when experimentally manipulated 40 , 41 ., Consequently , not all loci which posses the capacity to affect traits actually harbor functional variants in any given population ., As body size varies clinally with latitude , and variation at InR is also clinal 42 , this gene is considered a candidate for adaptively affecting body size ., This is potentially compatible with our results: if variation in InR is maintained by spatially varying selection , this variation may have been lost when this selection was removed by founding the stable LHM population ., Other clinal genes , however , may have been selected in our experiment: recently , genotypic and expression variation at the dca gene ( Drosophila cold acclimation , a . k . a . Senescence marker protein-30 ) was shown to associate with wing size in Australian populations 43 ., In our experiment , several deletions in the 3′ UTR of dca changed under selection , providing a precise hypothesis for the location of functional variation at this locus ., The most differentiated of these deletions was present at frequencies of 1 . 0 in both large populations , and 0 . 0 and 0 . 5 in each small population , resulting in a diffStat of 0 . 50 ( with an estimated FDR of 12 . 9% , this variant was not considered significant in the genomic analysis ) ., The greatest strength of the E&R approach may be the possibility to refine annotations at genes expected to influence a phenotype by identifying specific sub-genic functional elements , as illustrated by the dca example above ., As an additional example , consider Ecdysone-induced protein 63E: this is a complex gene , with 13 alternative transcripts spanning nearly 95 kb ., Deletions at this locus are generally lethal , but larvae that survive to pupation form very small pupae 44 ., In response to selection for body size , only 4 SNPs and a 3-bp deletion became significantly differentiated at this locus , and all are within a 100-bp region in a single intron ( FDR<0 . 006; Figure 5 ) ., Functional characterization of this small region may lead to insights regarding ecdysone-regulated size determination ., Similarly , only 3 SNPs are differentiated in the gene dre4 ( FDR<0 . 00002; Figure 5 ) ., The product of this gene ( also known as SPT16 ) forms a heterodimer known as FACT ( with SSRP1 ) that is involved in chromatin remodeling in Drosophila and conserved throughout eukaryotes 45 , 46 ., Loss-of-function mutations at this gene dramatically reduce ecdysteroid production at ecdysone regulated developmental stages , preventing molting: this gene is therefore an excellent candidate for altering critical size at metamorphosis through ecdysone signaling 47 ., Finally , it should be noted that at many loci there is much less resolution to infer the causal mutations ., For example , significant variants span ∼25 kb at the epidermal growth factor receptor , and some differentiated regions are much larger and contain many genes ( Figure S4 ) ., The set of significant variants at these loci is still a minute fraction of genomic variation , so these variants can now be used as a priori functional candidates in an association study ., This will reduce the genotyping effort required , and greatly increase the statistical power ., For this reason , we consider our approach to be largely complementary to more traditional genome-wide association studies , with mapping resolution at some loci small enough to proceed directly to functional characterization , while at others additional mapping will be required ., Furthermore , this approach will increase the number of species where powerful genotype-phenotype mapping is possible , as it can be utilized in any species with a suitable life-history ., We therefore expect the E&R approach to be a major component of future efforts to identify and characterize the molecular polymorphisms responsible for the tremendous phenotypic diversity observed within populations ., To sort flies by size , approximately 1800 flies from each population were anesthetized with CO2 and placed into a shaking column ( Gilson Performer III , Gilson Company ) ., For generations , 1–30 , flies were separated using 6 U . S . A . Standard Test Sieves ( ATSM E-11 specification; #10 , 12 , 14 , 16 , 18 & 20 ) , in which the diameter of the openings in each sieve was approximately 20% larger than the openings of the sieve below ( average of top sieve openings\u200a=\u200a2000 µm , average of bottom sieve openings\u200a=\u200a850 µm ) ., In order to create a finer scale selection gradient after generation 31 , flies were sieved with a custom made set of 20 electroformed sieves , in which the diameter of the holes in each sieve were only 5% larger than the holes of the sieve below ( average of top sieve holes\u200a=\u200a1685 µm; average of bottom sieve holes\u200a=\u200a800 µm ) ., Each generation , the most extreme 160 males and 160 females in each selected line were allowed to reproduce ., For control populations , an equal number of random individuals was selected after they were sedated , sieved , and then mixed back together ., Throughout , fly populations were maintained under a standard culturing process , designed to match the rearing conditions of the LHM base population as closely as possible ., Rearing conditions are described in detail in 27 ., Leg , wing , and thorax measures were taken from each population to investigate the anatomical consequences of selection for sieve size ., Length measures were made from images taken with a 3 . 3 Megapixel IC D integrated digital camera on a Leica M205C stereomicroscope using Image-Pro Analyzer 7 . 0 software ., Fifteen individuals per sex were scored for twelve phenotypes ., Thorax length was recorded as the distance between the posterior tip of the scutellum to the anterior most point of the prescutal suture ., Distances between the posterior scutellar and upper humeral bristles and the posterior scutellar and anterior sternopleural bristles were recorded from the thorax as well ., All thoracic measurements were done on the left side of the thorax after the legs had been disarticulated ., Femur lengths for all

Introduction, Results, Discussion, Methods

Body size is a classic quantitative trait with evolutionarily significant variation within many species ., Locating the alleles responsible for this variation would help understand the maintenance of variation in body size in particular , as well as quantitative traits in general ., However , successful genome-wide association of genotype and phenotype may require very large sample sizes if alleles have low population frequencies or modest effects ., As a complementary approach , we propose that population-based resequencing of experimentally evolved populations allows for considerable power to map functional variation ., Here , we use this technique to investigate the genetic basis of natural variation in body size in Drosophila melanogaster ., Significant differentiation of hundreds of loci in replicate selection populations supports the hypothesis that the genetic basis of body size variation is very polygenic in D . melanogaster ., Significantly differentiated variants are limited to single genes at some loci , allowing precise hypotheses to be formed regarding causal polymorphisms , while other significant regions are large and contain many genes ., By using significantly associated polymorphisms as a priori candidates in follow-up studies , these data are expected to provide considerable power to determine the genetic basis of natural variation in body size .

Understanding the causes and consequences of natural genetic variation is crucial to the characterization of biological evolution ., Moreover , natural genetic variation is comprised of millions of perturbations , which are partially randomized across genotypes such that a small number of individuals can be used to combinatorially analyze a large number of differences , facilitating mechanistic understanding of biological systems ., Here we demonstrate a powerful technique to parse genomic variation using artificial selection ., By selecting replicate populations of Drosophila flies to become bigger and smaller , and then determining the evolutionary response at the genomic level , we have mapped hundreds of genes that respond to selection on body size ., As our approach is powerful and cost-effective compared to existing approaches , we expect it to be a major component of diverse future efforts .

genetics and genomics/genomics, genetics and genomics/functional genomics, evolutionary biology/genomics, evolutionary biology, genetics and genomics, genetics and genomics/population genetics

journal.pgen.1007394

Whole exome sequencing reveals HSPA1L as a genetic risk factor for spontaneous preterm birth

Preterm birth ( PTB ) , defined as birth before 37 completed weeks of gestation , is a major global public health concern ., Worldwide , over 15 million infants ( more than one in ten babies ) are born preterm and of those , more than one million die from complications related to preterm birth each year 1 ., Preterm birth and its complications are the leading cause of neonatal deaths and have become the major cause of death among children under five years old 2 ., Moreover , preterm infants are at increased risk , not only of short-term complications but also of life-long disabilities , such as respiratory and cognitive disorders 1 ., Preterm birth also increases the risk of adult-onset disorders , such as obesity , diabetes and cardiovascular diseases 3 , 4 ., Currently , there is no generally effective method for prevention of preterm delivery ., The majority ( ~70% ) of preterm births occur after spontaneous onset of labor , with or without preterm prelabor rupture of the membranes ( PPROM ) 5 ., Most spontaneous preterm births ( SPTBs ) are idiopathic 1 , 5; however , recurrence of preterm birth among mothers and within families indicates that genetic factors may be important ., Genetic factors are estimated to account for 25–40% of the variation in birth timing 6 , with the maternal genome playing the major , but not only , role in predisposition to preterm birth 7–11 ., Despite many studies of the genetics of SPTB 6 , 12 , 13 , only a few variants have been robustly associated with this outcome 14 , and their functional implications are unclear ., Previous genome-wide association studies ( GWAS ) of SPTB have involved common variants , but they explain only a small portion of the genetic risk ., The role of rare variants in SPTB has been essentially unexplored ., Whole exome sequencing ( WES ) in families offers a comprehensive method to identify rare variant associations with disease , including almost complete coverage of the protein coding regions of the genome ., Even though studies of rare variants underlying Mendelian disorders have revealed novel genes 15 , 16 , using WES to study complex multifactorial syndromes remains a challenge 17 ., Previous sequencing studies of PTB 18 or PPROM 19 , 20 have focused only on a set of candidate gene regions and , consequently , have missed the majority of the coding regions of the genome ., In contrast to whole genome sequencing , WES is more cost effective and has the advantage of providing more easily interpreted results ., We performed a WES study using families under the hypothesis that familial recurrence is influenced by rare variants with large individual effects on SPTB susceptibility ., Such an approach has the potential of identifying genes containing rare variants shared in these multiplex families , as well as genes in pathways common across families ., This method applies a hypothesis-free testing approach to identify potentially novel candidate genes for SPTB ., Seventeen mothers from seven northern Finnish multiplex families ( Discovery cohort ) and an additional 192 mothers from 95 Danish families ( Replication cohort ) were sequenced using WES ., The pedigrees of the multiplex Finnish families are shown in S1 Fig . For the Discovery cohort , all samples ( except one that was excluded from subsequent analyses ) passed the quality control parameters used for the clinical exome sequencing at the CMH; quality control cutoffs were 85% reads aligned , 80% aligned with alignment quality of 20 or greater ., For these samples , the mean and median heterozygous/homozygous variant ratios were 0 . 765 and 0 . 748 , respectively ., Prior to variant filtering in Ingenuity , the mean/median numbers of nucleotide variant calls per individual were 318 , 767/326 , 474 ( Discovery cohort ) and 221 , 682/184 , 381 ( Replication cohort ) ., This difference in variant calls between the Discovery and Replication populations is likely due the fact that populations were sequenced using different Next Generation Sequencing platforms , Illumina for Discovery cohort and Complete Genomics for Replication cohort , and their respective primary quality control measures and variant calling methods were thus different ., Mean transition/transversion ( Ti/Tv ) ratios were 2 . 2 and 2 . 0 for Discovery and Replication exomes , respectively ., An overview of the WES workflow is presented in Fig 1 ., Three software programs ( Ingenuity Variant Analysis , Varseq and the CMH Variant Warehouse ) were used to assess common shared ( by affected mothers per family ) rare variants ., Only those variants that passed the prioritizing steps with at least two of the annotating software tools were considered valid and are described below ( summarized in S4 Table ) ., The benefits of comparing data obtained from multiple software is that it minimized the possibility of picking up falsely called variants that passed quality control filters only by one software ., This approach resulted in a total of 844 variants in the Discovery population ., For the Replication population , we combined and compared the shared rare variants passing the annotation and prioritizing steps of Ingenuity Variant Analysis and Varseq; a total of 8431 variants passed the filters of both software tools ., The CMH Variant Warehouse was not available for the Replication set ., For both populations , variants were categorized as loss of function , moderate , or other , according to their predicted consequences , i . e . pathogenicity ( S5 Table ) ., We further compared the list of variants resulting from the family-based analyses ( as described above ) between the Discovery and Replication populations ., Numbers of common genes and variants for both populations are shown in S2 Fig . There were 72 rare variants that were found in both populations in 72 genes ( S6 Table ) ., Rare single nucleotide variants from HSPA1L heat shock protein family A ( Hsp70 ) member 1 like , identified by the Discovery Ingenuity pathway analysis , were further investigated using imputed GWAS data that also included variants with MAF <1% ., In the Discovery set , variants in AR , NCOA3 and NCOR2 were either CAG repeat length polymorphisms , in-frame deletions or insertions , respectively , and were , therefore , not investigated in the GWAS datasets ., Three independent GWAS datasets were used , one of general European ancestry containing more than 40 , 000 mothers of live births ( 23andMe dataset ) and two from Northern Europe containing 4 , 600 and 600 mothers ( Nordic and northern Finnish datasets , respectively ) ., In the large 23andMe preterm birth GWAS dataset , the minor allele of rs34620296 in HSPA1L , which is in the glucocorticoid receptor signaling pathway , was found to be more common in cases than in controls ( case frequency 0 . 0025 vs . control frequency 0 . 0010 , p = 0 . 002; Table 3 ) ., This association was also significant for gestational age as a continuous trait ( gestational age as weeks; p = 0 . 0016 , effect -0 . 8238 , standard error 0 . 2608 ) ., The HSPA1L variants from the Discovery ( rs34620296 and rs150472288 ) and the Replication ( rs482145 , rs139193421 ) analyses are listed in detail in Table 3 ., In the two smaller GWAS datasets , however , these four HSPA1L variants were absent or not significant ., Lack of significance may be due to smaller numbers of individuals , especially in cases ., Sanger sequencing confirmed the genotypes of the two rare HSPA1L missense variants ( rs34620296 and rs150472288 ) in the samples from the Discovery cohort ., The rare HSPA1L variants were observed in a total of six mothers from four unrelated families ., Additional family members with available DNA were sequenced for these variants ., Interestingly , in two of the families , female carriers of the maternally inherited rs34620296 minor T-allele were born preterm , whereas in the other two unrelated families the male carriers of maternally inherited rs150472288 minor T-allele were born preterm ., However , numbers of minor allele carriers are too small for any definite gender related conclusions ., Pathogenicity predictions for rs34620296 and rs150472288 derived from the Discovery cohort as well as for rs482145 and rs139193421 from the Replication cohort were assessed using in silico tools SIFT and PolyPhen-2 , and all these variants were predicted as damaging and probably/possibly damaging , respectively ( Table 4 ) ., In addition , MutationTaster and MutationAssessor predicted all four variants as disease causing and predicted functional ( high ) , respectively ., According to the Combined Annotation Dependent Depletion ( CADD ) score ( >20 ) , all of these variants , except for rs139193421 , are among the top 1% of deleterious variants in human genome ( Table 4 ) ., To assess potential consequences of these variants on transcriptional activity , we evaluated them for evidence of histone modification or DNase I hypersensitivity ., In silico tools HaploReg 4 . 1 and/or RegulomeDB showed that all four variants were in regions that had histone marks , as well as strong transcriptional regulatory signatures in various cells of the immune system , especially in T lymphocytes from peripheral blood ( Table 4 ) ., Evidence of an active transcription start site was predicted in the HeLa-S3 Cervical Carcinoma Cell Line for rs34620296 and in foreskin fibroblast primary cells for rs482145 ( Table 4 ) ., Further evidence of active DNA accessibility ( DNAse ) was found in ovarian tissue for rs150472288 , and in psoas muscle tissue for rs139193421 ( Table 4 ) ., There was also evidence of a transcriptional effect of rs34620296 and rs150472288 ( Discovery ) in ovary and fetal adrenal gland ., Together these results from HaploReg 4 . 1 and RegulomeDB provide evidence for the potential involvement of HSPA1L variants in the endocrine system , as well as in the adaptive immune cells ., These variants could , therefore , have a role in the etiology of SPTB ., We further investigated putative effects of HSPA1L rs34620296 on protein structure ., This variant was selected due to its association with SPTB in the large 23andMe GWAS dataset ., This variant causes an amino acid change from Alanine to Threonine at position 268 ( Ala268Thr ) ., According to the NetPhos 3 . 1 in silico prediction , Ala268Thr generates an additional phosphorylation site next to an existing phosphorylation site ( T267-p ) ( S3 Fig ) ., Furthermore , Ala268Thr is near an adenosine triphosphate ( ATP ) nucleotide-binding site located downstream at position 270−277 ( Fig 2A ) ., Gain of phosphorylation may cause changes in binding energy , modulate physio-chemical properties or stability kinetics and dynamics of the protein functions such as strength of protein-protein interactions 22 ., To investigate the possible effects of the missense variant on protein structure , the reference HSPA1L protein structure and a structure including the Ala268Thr variant were compared simultaneously using UCSF Chimera ., There was not a visible change in the overlaid protein structures ( Fig 2B ) ., Instead , there was a slight change in the chemical bond lengths ( ≥0 . 002Å ) of the adenosine diphosphate ( ADP ) -ligand binding amino acid side chains at positions Glu270 , Arg274 and Asp368 , shown in the 3D model of the HSPA1L ( Fig 2C ) ., This may be due to the change from a small size , and hydrophobic , ( Ala ) to medium size , and polar , ( Thr ) residue ., Such a change in the amino acid side chains could affect the binding efficiency of the ADP molecule ., To further explore possible underlying biological functionality , we investigated the tissue expression established via HSPA1L , along with AR , NCOA3 and NCOR2 , using HumanBase ( http://hb . flatironinstitute . org ) ., HSPA1L was expressed in placental tissue with reasonable confidence ( 0 . 65 ) , and in ovarian ( 0 . 57 ) and fetal tissues ( 0 . 48 ) as well as in uterus ( 0 . 29 ) ., For HSPA1L , AR , NCOA3 and NCOR2 together , the average expression confidence was high in placenta ( 0 . 74 ) , ovary ( 0 . 70 ) , fetus ( 0 . 65 ) , and moderate in uterus ( 0 . 29 ) , indicating high confidence for expression in female reproductive system overall ( S4 Fig ) ., To determine whether the HSPA1L Ala268Thr ( rs34620296 ) variant alters activity of the GR signaling pathway , we analyzed the consequences of glucocorticoid exposure during decidualization ., Human endometrial stromal fibroblasts were transfected with plasmids containing either WT or Ala268Thr cDNA , or with empty vector serving as control ., The cells were treated with decidualization media for 72h in a presence of glucocorticoids ( 100nM dexamethasone ) as a surrogate of stress ., Protein levels of HSPA1L and GR , as well as mRNA levels of Wnt Family Member 4 ( WNT4 ) were measured ., Cells transfected with the WT HSPA1L-pcDNA3 . 1 trended to greater increases in cytosolic HSPA1L protein content than those transfected with the Ala268Thr HSPA1L-pcDNA3 . 1 ( mean ± SEM; 1 . 272 ± 0 . 142 vs . 0 . 893 ± 0 . 146 , respectively , p = 0 . 09 ) ( Fig 3 ) ., Furthermore , the Western blot analysis showed that the relative cytosolic protein levels of GR differed significantly between the WT and Ala268Thr groups with more GR present in the WT group than in the Ala268Thr group ( mean ± SEM; 1 . 309 ± 0 . 099 vs . 0 . 993 ± 0 . 096 , respectively , p = 0 . 04 ) ( Fig 3; numerical data available in S7 Table ) ., Next , we determined the relative gene expression of WNT4 by qPCR ., WNT4 is a critical decidualization target found in the recent GWA study 14 associated with gestational length ., Increased expression of WNT4 was observed in the WT group , whereas , the Ala268Thr group was less able to activate the WNT4-signaling pathway leading to a lower expression of WNT4 ( p = 0 . 04 ) ., To move beyond traditional case-control GWAS and family-based linkage studies , we performed a case-only whole exome sequencing study designed to investigate the burden of rare variants in families with recurrent SPTB ., Whole exome sequencing enables the discovery of rare , putatively functional variants associated with the etiology of complex disease on a gene-by-gene or a pathway-by-pathway basis , and enrichment in multiplex families provides a means to filter large-scale sequencing data ., Comparisons of mothers with recurrent preterm deliveries identified the glucocorticoid receptor signaling pathway as a candidate for mediating the risk of SPTB ., Specifically , within this pathway , likely pathogenic missense variations in HSPA1L were found among four unrelated Finnish families ( rs34620296 and rs150472288 ) , and within Danish sister pairs ( rs482145 and rs139193421 ) ., Notably , the rs34620296 minor allele variant was observed at a higher frequency in cases than controls in a very large 23andMe GWAS set ., These variants were also identified via bioinformatics analyses as likely affecting either protein function or expression ., Further functional evidence linked HSPA1L activity and decidualization ., HSPA1L is a member of the Hsp70 superfamily and is near HSPA1A and HSPA1B within the major histocompatibility complex class III region on chromosome 6 ., The HSPA1L protein ( also known as Hsp70-hom ) is ~90% identical to HSPA1A and HSPA1B , also known as Hsp70-1 and Hsp70-2 , respectively 23 , 24 ., Heat shock proteins ( HSPs ) are highly conserved cellular defense mechanisms for cell survival and are present in all cell types in all organisms ., Some HSPs are expressed constitutively , while others are stress-induced ( e . g . heat , hypoxia , oxidative stress , infection and inflammation ) 25 , 26 ., Intracellular HSPs act as molecular chaperones and , together with co-chaperones , stabilize existing proteins against aggregation , mediate folding of newly translated proteins , and assist in protein translocation across intracellular membranes 25 , 27 ., HSPs are categorized into families according to their approximate molecular weight; of which Hsp70 ( a group of proteins sized approximately 70 kDa ) is the best characterized ., Potential involvement of stress-induced HSPA1A in adverse pregnancy outcomes , including preeclampsia and PTB , has previously been suggested 28 , 29 ., Although , studies of the role of HSPA1L and HSPA1L in pregnancy are lacking , there is some evidence of involvement in adverse pregnancy outcomes such as preeclampsia 30 ., The rare HSPA1L missense variants observed in our study , are in the nucleotide-binding domain ( NBD ) , except the rs482145 , which is in the substrate-binding domain ( SBD ) ( Fig 2 ) ., ATP binds to the NBD , which is followed by the exchange from low-binding affinity ATP state to high-binding affinity ADP state 29 , 31 , 32 ., We showed that the non-synonymous variant rs34620296 ( Ala268Thr ) generates an additional phosphorylation site near the nucleotide-binding site ., It showed a modest change in the binding efficiency at this site , which could affect the interaction with ADP or HSPA1L stability itself , as suggested by our transfection studies ., In agreement with our findings , a previous study of Caucasian patients with inflammatory bowel disease found that rare mutations in HSPA1L were significantly enriched in patients but absent in healthy controls 33 ., Interestingly , one of the associated rare variants was Ala268Thr , and further in vitro biochemical assays of the recombinant HSPA1L showed reduced chaperone activity with this variant 33 ., There is also evidence that possibly connects inflammatory bowel disease to adverse perinatal outcomes 34 ., Additionally , a previous SPTB study in African Americans found a common nonsynonymous HSPA1L variant , rs2075800 , to associate with SPTB 35 ., Furthermore , a meta-analysis of previously PTB associated genes linked HSPA1L and SPTB using Ingenuity Pathway Analysis 36 ., Due to a very low incidence of the rare HSPA1L variants associating with SPTB in our study , the anticipated attributable risk in the population level is probably small ., However , the identification of the damaging alleles may facilitate the identification of causative pathways ., For instance , interaction between Hsp70 and Hsp90 chaperones as well as their co-chaperones is essential in the maturation and inactivation of nuclear hormone receptors ( e . g . glucocorticoid , androgen , estrogen and progesterone receptors ) 37 , 38 ., In the absence of its ligand , glucocorticoid receptor ( GR ) is bound to a complex constituting of Hsp40 , Hsp70 and Hsp90 chaperones; this complex keeps the GR in a ligand-receptive conformation but remaining transcriptionally inactive until ligand binding 38 ., As shown previously 33 , rare HSPA1L variants can cause partial loss of HSPA1L chaperone activity , and therefore , altered function or expression ., Altered function of the chaperones can compromise the stability of the GR complex , leading to an accumulation of partially unfolded proteins that are prone for aggregation and degradation events 37 ., Glucocorticoids , steroid hormones that mainly signal through the GR , have anti-inflammatory and immunosuppressive actions ., Glucocorticoid signaling communicates with estrogen signaling pathways to tightly regulate the pro- and anti-inflammatory milieu in reproductive tissues 39 , and progesterone signaling , via nuclear GR , mediates anti-inflammatory and immunosuppressive effects in genital tract during pregnancy 40 , 41 ., Sustaining a pregnancy is a complex interplay and balance between the innate and adaptive immune cells in the reproductive tissues and at the maternal-fetal interface ., Imbalance between the inflammatory cells can cause a breakdown of maternal-fetal tolerance leading to activation of labor ( both term and preterm ) ., An untimely stimulus ( e . g . stress , infection or inflammation ) together with impairments in the glucocorticoid receptor signaling pathway could impose an inadequate response against inflammation or stress ., This can elicit a shift from an anti-inflammatory to pro-inflammatory microenvironment , causing a premature activation of labor initiating signals resulting in preterm birth 42 , 43 ., Possible limitations of our study are that the Discovery and Replication populations were sequenced using different Next Generation Sequencing platforms , and primary quality control measures and variant calling methods were thus different ., In addition , Next Generation Sequencing generates an enormous amount of data , which could lead to many sequencing artifacts that may be misidentified as variants ., We attempted to minimize these artifacts by applying a variety of quality control filters and using a large internal control population to detect potential sequencing or annotation errors ., We also compared the results of variant annotation and prioritizing filters from three different software tools to ensure reproducible results ., Furthermore , reported variants were confirmed by Sanger sequencing ., Another possible limitation was that our study did not include unrelated control samples ., This limitation has been partly overcome with the use of additional large GWAS datasets including control samples ., In conclusion , whole exome sequencing of families with recurrent occurrence of SPTB enables identification of rare alleles influencing the predisposition to SPTB ., Among the individual genes , two minor alleles of HSPA1L had a strong association to SPTB in multiplex Finnish families and the association of a specific minor allele was confirmed in a large GWAS set ., Furthermore , this variant was associated with altered modification and function of the protein ., Overall , our data suggest the need for precise regulation of steroid signaling in mediating birth timing ., Written informed consent was obtained from all individuals participating in this study , and the study was approved by the Ethics committees of the participating centers: Oulu University Hospital ( 78/2003 , 73/2013 ) , University of Southern Denmark ( NVK#1302824 ) , and University of Iowa ( IRB#200608748 ) ., Individuals in the large European American GWAS were research participants of 23andMe , Inc . , a personal genetics company ., All 23andMe participants provided informed consent and participated in the research online , under a protocol approved by the external AAHRPP-accredited IRB , Ethical & Independent Review Services ( E&I Review ) ., DNA samples of the 17 individuals from Discovery population were extracted from whole blood and saliva samples using standard methods 45 ., Although , using DNA from both blood and saliva samples , there were no major difference in the overall sequencing metrics ( alignment metrics or total number of variants ) between the sample types ., DNA samples were subjected to exon specific next generation sequencing performed at the Center for Pediatric Genomic Medicine , Children’s Mercy Hospital ( CMH; Kansas City , MO ) ., Exome samples were prepared with the Illumina Nextera Rapid Capture Exome kit according to the manufacturer’s protocols as described previously 47 ., Sequencing was performed on Illumina HiSeq 2500 instruments utilizing v4 chemistry with 2 x 125 nucleotide sequences ., Sequence data were generated with Illumina RTA 1 . 18 . 64 . 0 and bcl2fastq-1 . 8 . 4 , and aligned against the reference human genome ( GRCh37 . p5 ) using bwa-mem 48 , and variant calls were made using the Genome Analysis Toolkit ( GATK ) 49 version 3 . 2–2 using previously described methods 50 ., Duplicate reads were identified and flagged with the Picard MarkDuplicates tool ., Realignment of reads around known indels was performed with the RealignerTargetCreator and IndelRealigner , and variants were called on individual samples using the HaplotypeCaller modules of the GATK ., In addition , whole exome sequencing was performed on 192 affected individuals from 95 Danish families ( Replication set ) ., Exome capture of the samples were carried out with the BGI Exon Kit following manufacturer’s protocols ( BGI , Shenzhen , China ) ., DNA libraries were generated using combinatorial Probe Anchor Ligation ( cPAL ) technology , and 35 base paired end reads were generated from 500 bp genomic fragments ., Whole exome sequencing was performed using the Complete Genomics platform ( BGI ) and using the manufacturer’s pipeline ., Reads were aligned against the National Center for Biotechnology Information ( NCBI ) build 37 reference human genome ., The variant call files ( VCF ) , containing the variant call results , generated by CMH and BGI were analyzed using Ingenuity Variant Analysis Software ( Qiagen , Germany ) and Golden Helix VarSeq Software v . 1 . 2 . 1 ( Bozeman , MT ) for both the Discovery and Replication population sets ., Variants were filtered based on variant quality control measurements , frequency and predicted pathogenicity , as well as a dominant inheritance model ., In the Ingenuity Variant Analysis , low quality variants ( read depth <15 and call quality <20 ) were removed ., Furthermore , we only included rare variants ( i . e . MAF <1% in the 1000 Genomes Project , ExAC or in European American population in NHLBI ESP exomes ) and variants that would likely have functional effect ( i . e . variants that are predicted by SIFT or PolyPhen-2 as damaging or likely damaging , listed in Human Gene Mutation Database , or associated with gain or loss of function of a gene ) ., In VarSeq , variants with read depth <15 and genotype quality score <20 were excluded ., Only rare variants in Europeans ( MAF <1% or absent; 1kG Phase 3: Variant frequencies 5 , GHI Jan 2015 ) and missense or loss-of-function variants were included for analyses ., Further filtering was applied to the data obtained from VarSeq ., Variant quality was enhanced by applying range criteria ( 0 . 3–0 . 85 ) for alternative allele frequency ( i . e . ratio of alternate allele read depth / alternate allele read depth + reference allele read depth ) ; variants outside this range were excluded ., Allele frequencies were searched from the Sequencing Initiative Suomi ( SISu ) database ( www . sisuproject . fi ) for variants originating from the Finnish mother data , and from the Exome Aggregation Consortium ( ExAC ) database ( http://exac . broadinstitute . org/ ) for Danish sister pair data ., For these variants , a MAF cut-of value <1% in Finnish general population ( SISu ) or European “non-Finnish” ( ExAC ) population was used for Finnish or Danish mothers , respectively ., The SISu database was used for Finnish mothers to exclude rare variants that are enriched in Finnish general population compared to the rest of the Europeans ., For the Discovery samples , we also used allele frequency calculations derived from Center for Pediatric Genomic Medicine’s CMH Variant Warehouse database ( http://warehouse . cmh . edu ) including ~3900 individuals previously sequenced at the center 50 ., Pathogenicity was categorized according to the American College of Medical Genetics 21 as 1; previously reported to be disease-causing , 2; expected to be pathogenic ( loss of initiation , premature stop codon , disruption of stop codon , whole-gene deletion , frame shifting indel , and disruption of splicing ) , and 3; unknown significance but potentially disease-causing ( nonsynonymous substitution , in-frame indel , disruption of polypyrimidine tract , overlap with 5 exonic , 5 flank , or 3 exonic splice contexts ) ., Only variants that fit one of these criteria ( 1−3 ) were included for analyses ., Rare and novel variants with relatively high frequency in this internal control population were also excluded as they were thought to be technical artifacts ., In Ingenuity Variant Analysis , a dominant inheritance model ( including gain of function variants , and all heterozygous , compound heterozygous , haploinsufficient , hemizygous , and het-ambiguous variants ) was used to investigate predisposing variants that are inherited in the families ., When analyzing affected mothers as a group , rare variants in genes that were common for a proportion of all cases were investigated ., Whereas in family specific analyses , only variants that were shared by the affected individuals within each family were included ., Ingenuity Variant Analysis provides a list of most significant pathways calculated specifically for each filtering output ., P-values were calculated according to Fisher’s exact test assessing overlap enrichment of dataset-variant genes relative to known phenotype-implicated genes ., Here , only pathways with p<0 . 01 were included for further analyses ., To investigate rare variants in genes arising from the whole exome data in a larger population setting including controls , we used available sources of preterm birth GWAS data ., GWAS data from a large cohort , identified among 23andMe’s research participants , included 43 , 568 mothers of general European ancestry 14 and meta-analysis data including 4 , 632 mothers from three independent Nordic ( Finnish , Danish , and Norwegian ) birth cohorts of European ancestry 51 ., In addition , a set of GWAS data from a total of 608 mothers passing quality control measures was available ., This set included mothers with spontaneous preterm deliveries and mothers with term deliveries originating exclusively from northern Finland ., Genotyping was performed with Illumina Human CoreExome chip , followed by prephasing and imputation procedures with ShapeIT2 52 and IMPUTE2 53 ., Association analysis was performed using SNPTEST v . 2 . 5 . 2 54 ., Since WES methodologies are associated with significant false positive rates , the presence of interesting variant findings from WES analyses was confirmed using Sanger sequencing ., Samples were sequenced using capillary electrophoresis with ABI3500xL Genetic Analyzer ( Applied Biosystems , CA ) in Biocenter Oulu Sequencing Center , University of Oulu , Oulu , Finland ., Details of the PCR primers and reaction conditions are available upon request ., The possible functional effect of the rare HSPA1L variants ( rs34620296 and rs150472288 from Discovery analyses as well as rs482145 and rs139193421 from Replication analyses ) were investigated using in silico prediction tools such as SIFT , PolyPhen-2 , MutationTaster and MutationAssessor ., These pathogenicity predictions were annotated using Varseq , whereas CADD scores to identify pathogenic and deleterious variants were obtained from Ingenuity ., Variants with CADD score >20 are amongst top 1% of deleterious variants in human genome 55 ., RegulomeDB ( http://www . regulomedb . org/ ) was used to investigate variant locations for e . g . chromatin state activity ., In addition , we used HaploReg v4 . 1 ( http://archive . broadinstitute . org/mammals/haploreg/haploreg . php ) to assess whether variants are located within regions that show evidence for promoter or enhancer activity ( i . e . presence of histone modification marks H3K4me1 and H3K27ac that are associated with enhancer regions , or H3K4me3 and H3K9ca that are associated with promoter regions ) , as well as for DNase I hypersensitivity in human tissues and cell line samples ., We further investigated the potential effects that missense variation rs34620296 ( Ala268Thr ) could have on protein sequence or structure ., We used NetPhos 3 . 1 56 to investigate possible changes in phosphorylation events in HSPA1L sequence ., NetPhos 3 . 1 predicts serine , threonine or tyrosine phosphorylation sites in amino acid sequences of eukaryotic proteins ., Evidence of being a phosphorylation site is given when the score is above the threshold ( 0 . 5 ) ., To investigate the possible effects of missense variant in protein structure , the reference protein structure and the modified protein structure , including the missense variant , were compared ., Original ( UniProtKB: P34931 ) and modified ( Ala268Thr ) amino acid sequences were submitted to SWISS-MODEL ( https://swissmodel . expasy . org/ ) for protein modeling ., Resulting protein models were compared simultaneously using UCSF Chimera ., Molecular graphics and analyses were performed with the Chimera-1 . 11 . 2 .

Introduction, Results, Discussion, Materials and methods

Preterm birth is a leading cause of morbidity and mortality in infants ., Genetic and environmental factors play a role in the susceptibility to preterm birth , but despite many investigations , the genetic basis for preterm birth remain largely unknown ., Our objective was to identify rare , possibly damaging , nucleotide variants in mothers from families with recurrent spontaneous preterm births ( SPTB ) ., DNA samples from 17 Finnish mothers who delivered at least one infant preterm were subjected to whole exome sequencing ., All mothers were of northern Finnish origin and were from seven multiplex families ., Additional replication samples of European origin consisted of 93 Danish sister pairs ( and two sister triads ) , all with a history of a preterm delivery ., Rare exonic variants ( frequency <1% ) were analyzed to identify genes and pathways likely to affect SPTB susceptibility ., We identified rare , possibly damaging , variants in genes that were common to multiple affected individuals ., The glucocorticoid receptor signaling pathway was the most significant ( p<1 . 7e-8 ) with genes containing these variants in a subgroup of ten Finnish mothers , each having had 2–4 SPTBs ., This pathway was replicated among the Danish sister pairs ., A gene in this pathway , heat shock protein family A ( Hsp70 ) member 1 like ( HSPA1L ) , contains two likely damaging missense alleles that were found in four different Finnish families ., One of the variants ( rs34620296 ) had a higher frequency in cases compared to controls ( 0 . 0025 vs . 0 . 0010 , p = 0 . 002 ) in a large preterm birth genome-wide association study ( GWAS ) consisting of mothers of general European ancestry ., Sister pairs in replication samples also shared rare , likely damaging HSPA1L variants ., Furthermore , in silico analysis predicted an additional phosphorylation site generated by rs34620296 that could potentially affect chaperone activity or HSPA1L protein stability ., Finally , in vitro functional experiment showed a link between HSPA1L activity and decidualization ., In conclusion , rare , likely damaging , variants in HSPA1L were observed in multiple families with recurrent SPTB .

Preterm birth is the leading cause of infant mortality , and prematurity is further associated with serious morbidities in later life ., Genetic and environmental risk factors play a role in the susceptibility to preterm birth ., Despite numerous studies , the genetic basis for preterm birth remains poorly defined ., We investigated the presence of rare , possibly risk associated nucleotide variants in mothers with spontaneous preterm births ( SPTB ) ., The first set of mothers with family history of recurrent preterm births was of northern Finnish origin ., An additional set of mothers ( sister pairs , both giving birth preterm ) of European origin was also studied ., Whole exome sequencing identified multiple rare , likely damaging HSPA1L variants in several families affected by SPTB , and this gene was associated with the glucocorticoid receptor signaling pathway ., Potential involvement of one of the HSPA1L variants in SPTB was further supported by large GWAS dataset ., In addition , this variant alters protein post-translational modification potential , and thus may affect protein stability and its function as a chaperone .

phosphorylation, genome-wide association studies, medicine and health sciences, cellular stress responses, maternal health, obstetrics and gynecology, engineering and technology, cell processes, hormones, preterm birth, womens health, pregnancy, genome analysis, industrial engineering, quality control, protein structure, estrogens, pregnancy complications, birth, heat shock response, computer and information sciences, proteins, molecular biology, biochemistry, cell biology, post-translational modification, genetics, software engineering, biology and life sciences, genomics, software tools, computational biology, macromolecular structure analysis, human genetics

journal.pcbi.1000504

Evaluation of Objective Uncertainty in the Visual System

Every single human action happens in a context of uncertainty , being based on incomplete knowledge and undertaken despite unpredictable consequences ., When faced with uncertainty , humans employ heuristics 1 , 2 and show characteristic biases in their decision 3 ., The neural structures involved in some of these decisions are now being identified 4–6 ., Before one can make decisions that depend on uncertain information , the degree of uncertainty must be evaluated ., The basic question of how well humans do at evaluating their own uncertainty remains largely understudied ., Uncertainty is a familiar concept in cognitive science , in particular thanks to Signal Detection Theory ( SDT; Green and Swets 1966 ) ., In a typical psychophysical task , an observer has to detect small contrast increments near threshold ., The uncertainty in this task comes mostly from internal variability: because of fluctuations in her internal representation of contrast , the observer makes mistakes and is uncertain about the correctness of her decisions ., Unfortunately for the experimenter , this source of the uncertainty is internal to the observer and therefore only indirectly controllable ., Now consider another difficult perceptual task: listening to a speaker among cocktail-party chatter ., Here the difficulty depends not so much on variability in the brain , but rather on interactions between the different voice signals: the one emitted by the speaker you aim to listen to , and the sound of other voices ., Even with the volume of the other voices staying the same over time , difficulty will depend on the languages spoken , the gender of the speakers , and other sources of confusion ., More generally , background chatter plays the role of noise , and difficulty will vary based on how much signal and noise covary ., An analogous visual task can be obtained by adding visual noise to a signal –random perturbations to the stimuli shown to the observer ., Using visual noise , we are in a position to manipulate the objective uncertainty: objective uncertainty is inversely related to the amount of task-relevant information available in the stimulus ., Concurrently , we can measure the perceived uncertainty of the observer , the level of confidence she actually reports ., We introduce three experiments where we manipulate objective uncertainty and study its relationship with perceived uncertainty ., In the first two experiments , observers were presented with pairs of images of oriented objects embedded in high levels of noise , and had to report the orientation of the image of their choice ., Even though the two images contained the same level of noise , the particular noise structure made one image orientation more certain than the other ., We found that observers reliably chose the more certain of the two images , thereby providing evidence of a capacity to accurately evaluate objective uncertainty ., We confirmed this in another experiment , in which we held the objective uncertainty of one of two stimuli fixed while varying the other , and asked observers to pick the less uncertain one ., The greater the difference in uncertainty was , the greater the chance that observers picked the less uncertain stimulus , showing that uncertainty discrimination behaves similarly to normal psychophysical tasks ., In a third experiment , we extend our results to a letter discrimination task ., We discuss plausible computational mechanisms for achieving these results ., To determine whether observers did effectively pick the less uncertain stimuli , we contrasted two conditions ., In the so-called True Choice ( TC ) condition , the two stimuli presented resulted from independent draws from the same noise distribution ., Note that two stimuli with the same average noise level , as is the case here , can still vary in the objective uncertainty they induce , because different realizations of the same noise distribution can make the stimulus more or less ambiguous ., In that case there is a benefit to be had in choosing the less uncertain of the two: this gives observers a higher chance of responding correctly than if only one stimulus is available ., In the other condition , the False Choice ( FC ) condition , we removed that benefit: the first stimulus was computed the normal way , but the second was obtained by flipping the top one either once or twice ( Figure 2 ) ., We took advantage of the underlying symmetry of our templates: flipping the first template left-to-right yields the second , and flipping the second bottom-top yields back the first ., By applying these transformations to a noisy version of our template , we were able to create two stimuli that differed pixel-to-pixel , but were equivalent from the point of view of the classification task and thus carried equal objective uncertainty in that context ., In the False Choice case , there is therefore nothing to be gained by choosing one rather than the other ., At no point in the experiment were observers aware of the existence of the two conditions ., The two stimuli presented always had equal contrast , preventing observers from using a heuristic of selecting the lower-contrast stimulus as the most certain ., The False Choice condition therefore provides the performance baseline that will be used to determine whether or not observers are able to successfully compare objective uncertainties ., We measured observers performance , defined as proportion of correct classifications , in the two conditions across five different signal-to-noise ratios , chosen to span a range of performance between approximately 60 to 85% ., Both the signal-to-noise ratio and the condition each trial belonged to were randomized ., If observers are able to make accurate judgments of objective uncertainty , then we expect that measured performance will be higher in the TC than in the FC condition ., As expected given the nature of the task , mean performance for all observers grew with increased signal-to-noise ratio ., More interestingly , however , mean performance is higher in the TC condition than in the FC condition , which translates into lower performance thresholds in the TC condition ( Figure 3 a and b ) ., To establish that the effect is genuine we used a model comparison technique ., We used a likelihood-ratio test to evaluate the effect of True Choice versus False Choice ( details in Text S1 ) ., Using two psychometric functions , one per condition , rather than one psychometric function for both conditions provides a significantly better fit to performance data ( Nested hypotheses test 10: p\u200a=\u200a0 . 0004 , , d . f . =\u200a24 ) ., It appears then that observers were able to take advantage of the True Choice condition , by choosing the less uncertain stimulus a majority of the time ., It seems reasonable that , should the ability to pick the less uncertain stimulus be present , the probability of choosing the correct stimulus ought to be an increasing function of the magnitude of the difference: the more the two stimuli differ in their uncertainty , the more likely observers are to choose the right one ., We evaluate that by regressing observers choices of stimuli on the difference of log-entropies ( Text S1 ) ., We found a highly significant effect ( details in Text S1 ) of the difference in uncertainty on the probability of choosing the bottom stimulus: in other words , the more uncertain the bottom stimulus compared to the top one , the less likely observers were to choose the bottom one ., This last result hints at a more general property: in all psychophysical discrimination tasks , the larger the difference between two stimuli , the more reliable discrimination is ., For example , when asked to compare the length of two lines , an observers responses are likely to be better predictable when the two lines differ by 20 cm rather than 1 ., In a second experiment , we sought to confirm our findings by checking that discrimination of uncertainty behaves in the same way ., The task was identical to that of experiment 1 , but instead of introducing a False Choice condition , we manipulated the stimuli such that one – the standard – had always the same level of uncertainty and the other – the test – had lower uncertainty ., We show in the supplementary material that generating random stimuli with a controlled level of uncertainty can be achieved using a simple orthogonal projection ., Mathematically , the space of all possible stimuli of the kind used here can be described in terms of the contrast of individual pixels by having one dimension ( one axis ) for each pixel ., Then the two templates are two points u , v in that space , and stimuli obtained by adding white noise to a template are other points , forming Gaussian point clouds around the templates ., To decide whether a point is more likely to belong to the left-tilted template rather than the right-tilted one , a simple geometrical rule describes the ideal strategy ., Imagine drawing a line between u and v , as in figure 4 , where we illustrate the problem for stimuli with only 2 pixels ., Now draw the plane ( in higher dimensions; the hyperplane ) that is orthogonal to the line and cuts through it at the mid-point ., Then any stimuli falling on the same side of the plane as u we will call “left-tilted” and any falling on the side of v we will call “right-tilted”: the plane represents the decision boundary ., Stimuli falling right on the hyperplane are completely ambiguous: both categories are equally likely ., In fact , it is possible to show that the uncertainty of a stimulus is given by its ( unsigned ) distance to the decision boundary ., Then the set of stimuli of fixed uncertainty is the set of points that are of the same distance to the decision boundary , and that set is simply the union of two parallel planes ., We therefore generated our stimuli by constraining them to lie on a plane of distance d to the decision boundary ., Standard stimuli were always on a plane of distance dstandard and test simuli were on a plane of distance dtest ., The difference between dstandard and dtest was varied parametrically between 4 different levels: we expected the observers to more reliably choose the test stimulus as the difference increased ., The results appear in figure 5: the larger the difference in uncertainty between standard and test , the more likely observers were to choose the test stimulus ., We adapted the noise level to each observers performance , so the distances used varied between observers ., We normalise them with respect to the expected distribution of the distance to the hyperplane for the noise level chosen ( see Text S1 ) ., The effect of the difference is significant for every observer as modeled by logistic regression of stimulus choice on difference in uncertainty ( t-test for Generalised Linear Models coefficients , all p-values at 10−3 or below ) ., This confirms that uncertainty behaves in that respect just like other psychophysical quantities: the more dissimilar two stimuli are on that scale , the more predictable observers judgments are ., In experiments 1 and 2 , the underlying visual task is orientation discrimination under noise , with templates identical in every way except for one basic attribute – their orientation ., To check that our results were sufficiently general , we ran a variant of experiment 2 using a letter discrimination task ., Observers had to discriminate between the letters ‘T’ and ‘X’ ( shown on figure 5 ) , a pair chosen because the corresponding characters correlate very little ., Except for the nature of the templates , experiment 3 was identical to experiment 2 and we replicated its results ( figure 5 ) : observers were more likely to pick the less uncertain stimulus when the difference in uncertainty was larger ., Our results thus generalize to more sophisticated visual tasks ., Our results imply that observers had access to some estimate of the uncertainty in the orientation task ., How is that estimate computed ?, Do observers have effective access to a probability distribution over perceptual hypotheses , from which they can estimate their own uncertainty ?, Or do they rely on more limited information ?, To investigate that question we evaluated two distinct families of models that compute uncertainties globally over the full distribution for the first , and locally for the second ., We begin by defining the following quantities: let r and s be two stimuli , represented as vectors of pixel luminances ., Call u and v the left-tilted and right-tilted templates ., Then and are measures of how “different” r is to u and v , respectively ., If r is more like u than v ( i . e . , ) , then it is more likely to have been generated from u , and hence the observer should respond “left-tilted” for stimulus r ., In comparing the uncertainty between two stimuli - choosing between r and s - the following procedure is exactly equivalent to the strategy of the “ideal observer” ( i . e . , the strategy that maximizes performance , see Text S1 ) ., Compute as ( 4 ) and choose r if , r otherwise ., This corresponds to evaluating uncertainty based on the full posterior distribution ( see equation 1 ) : uncertainty is low if one hypothesis corresponds to the data much better than the other , and high otherwise ., We call this model the difference of responses model ., Another strategy , perhaps simpler for the observer , is to evaluate uncertainty based only on how well the best hypothesis fits the data ., We call this the maximum response model ., The same measures of distances are computed as in the first model , but only the maximum is retained for each stimulus ., The observer then compares the two maxima ( 5 ) Put into perceptual terms , this corresponds to a strategy of picking the stimulus that seems to have a more salient dominant orientation , when the templates were Gabor patches , or the stimulus that was more “letter-like” , when the templates were characters ., In statistical terms this is equivalent to evaluating uncertainty based on the magnitude of the likelihood of the maximum-likelihood hypothesis ( Methods ) , a strategy that is sub-optimal for our task but still gives an improvement over choosing between the two stimuli at random ., Both hypotheses are realistic from a neural-computation point of view ., Computing and is nothing more than a linear filtering of the neural input: although some important non-linearities have been identified in visual orientation discrimination , linear filtering remains the basic operation in all models 11 , 12 ., Computing the decision variables , whether dabs and dmax , is a simple non-linear step readily implementable in a neural system ., To test those models we make the same assumption we did for regressing choice on difference in log-entropy: the higher dabs and dmax , the more likely observers are to choose the bottom stimulus ., As above , we compute the decision variables for every trial and we fit a linear binomial regression model to the responses ( Text S1 ) ., Our models give for each trial a choice probability ., On figure 6 we plot the percentage prediction correct ( i . e . , the proportion of trials where the model predicted with p> . 5 the choice the observer actually made ) ., The two models have the same number of degrees of freedom , and can be directly compared ., Both predict the data significantly better than chance , but the maximum response has a significant lead ., Our data therefore point to a likelihood-based evaluation of visual uncertainty , rather than one based on the full posterior distribution ., In summary , we demonstrate here that humans display second-degree knowledge of a visual discrimination task: not only are they able to detect what signal is in the noise ( first-degree knowledge ) , but also to estimate how uncertain that knowledge is , at least comparatively ., Why humans should be so well calibrated to what is in essence a laboratory task rather than a natural one is a question that deserves attention ., It is possible that they learn the statistical properties of the task over time , although we find no conclusive evidence for that in our data ( see Text S1 ) ., Previous research lacked an objective standard to compare subjective judgements to , and relied on ratings 13 ., Various biases have been reported in human confidence judgments , including over- and under-confidence , global/local inconsistencies , as well as inter-cultural differences 14–17 ., The forced-choice method we outlined here allows one to test human observers objective capacity to detect differences in uncertainty contained in a task , and to evaluate possible computational mechanisms much more rigorously ., It is a potentially important methodology in the study of discrepancies between visual performance and confidence , a topic many believe to be connected to the wider issue of awareness 18 , 19 , but potentially also in investigations of metacognition in non-human species 20 , 21 ., Our work is in tune with a variety of current research that tries to understand visual function as a form of Bayesian inference 22–25 ., These theories posit that the visual system explicitly encodes probability distributions over perceptual hypotheses ., In that context , it makes intuitive sense that the system should be able to measure the uncertainty of such a distribution: comparing two uncertainties as we do here is rarely needed as such , but comes into play in more complicated decisions ., Just as a low feeling of confidence in an item to be memorized is a clue that further study is needed 26 , high visual uncertainty signals that more information is needed , making precise evaluation of visual uncertainty an essential aspect of exploration mechanisms 27 ., The results given here agree with other studies that have found unexpectedly accurate decision-making in perceptual 28 , 29 and motor systems 30 , 31 ., These results imply that uncertainty is dealt with at an implicit level: unlike them , we require observers to make explicit comparisons between levels of uncertainty ., The observers who took part in our experiment nevertheless found the task quite intuitive: indeed , we often make comparative judgments of visual uncertainty “in the wild” , as when we judge if we see better from one vantage point than another ., Generally , we expect that confidence measures have the potential to play a larger role in computational investigations of perceptual decision-making ., The evaluation of uncertainty is a necessary first step in any statistical decision-making system , and biases and approximations in evaluating uncertainty will cause sub-optimal decisions ., A systematic study of the evaluation of uncertainty in the visual system will help uncover the shortcuts taken by the brain in making perceptual decisions ., Our method can be generalized to other noise models , other sensory modalities , and other tasks ., But showing that fine-grained discrimination of uncertainty can be done is of course not an end in itself: uncovering how that essential operation is achieved in the brain is a natural next step ., This study was conducted according to French guidelines on research involving human participants ., All participants gave informed consent ., The experimental method was the same as in experiment one , unless indicated otherwise ., The experimental method was the same as in experiment 2 , unless indicated otherwise .

Introduction, Results, Discussion, Methods

The role of sensory systems is to provide an organism with information about its environment ., Because sensory information is noisy and insufficient to uniquely determine the environment , natural perceptual systems have to cope with systematic uncertainty ., The extent of that uncertainty is often crucial to the organism: for instance , in judging the potential threat in a stimulus ., Inducing uncertainty by using visual noise , we had human observers perform a task where they could improve their performance by choosing the less uncertain among pairs of visual stimuli ., Results show that observers had access to a reliable measure of visual uncertainty in their decision-making , showing that subjective uncertainty in this case is connected to objective uncertainty ., Based on a Bayesian model of the task , we discuss plausible computational schemes for that ability .

Most work in vision science focuses on the question of why we perceive what we do , and we now have many models explaining what physical properties of a stimulus make us see depth , colour , etc ., Here we ask instead what makes us feel confident in our visual perception: in the context of a visual task , what are the physical properties of the stimulus that will make us think we are doing the task well ?, The mathematical framework of Bayesian statistics provides an elegant way to frame the problem , by assuming that the visual system is trying to estimate physical properties of the world from incomplete , sometimes unreliable visual information ., Objective uncertainty will therefore depend on the quality of the information available in the stimulus ., In our experiments we compare objective uncertainty—as computed using the Bayesian framework—with subjective uncertainty , the confidence observers report about their visual percepts ., To this end , we use a visual task with well-defined statistical properties , discrimination under noise ., We report a surprising degree of agreement between objective and subjective uncertainty , and discuss possible computational models that could explain this ability of the visual system .

neuroscience/psychology, neuroscience/natural and synthetic vision

journal.pcbi.1002643

A Simple Histone Code Opens Many Paths to Epigenetics

The histone proteins that form the nucleosomes that package eukaryotic DNA are subject to various post-translational modifications of several of their exposed amino acid residues ., These chemical modifications are added and removed by a large number of specific enzymes , and create the potential for a vast number of different nucleosome types ., Specific modifications ( i . e . particular chemical modifications of particular histone residues ) affect the binding of other proteins to nucleosomes , which in turn can affect the packaging , replication , recombination , repair and expression of the underlying DNA ., Many different protein domains or modules have been shown to confer modification- sensitive nucleosome binding 1 ., The presence of multiple modification ‘reader’ domains within a single protein or protein complex inspired the histone code hypothesis - that specific combinations of different modifications can have distinct downstream consequences 2 ., More recent experimental observations support this idea ., First , high resolution ChIP analysis has shown that complex patterns of histone modifications are associated with specific sequences 3 ., Although ChIP does not prove co-existence of modifications on single nucleosomes , top-down mass spectrometry has revealed over 100 different modification patterns on individual histone proteins in vivo 4 , 5 ., Secondly , there is now a number of examples , at least in vitro , of multivalent reader proteins distinguishing different combinations of modifications at multiple histone positions 6–9 ., Combining a protein element that can recognize a specific nucleosome modification with a protein element that can catalyze ( ‘write’ ) the same modification creates the possibility for positive feedback and bistability 10–14 ., The idea is that specifically modified nucleosomes recruit enzymes that cause other nearby nucleosomes to become similarly modified ., This feedback and the distribution of parental nucleosomes to nearby locations on both daughter DNA molecules after DNA replication 15 , means that alternative modification states could be persistent and heritable within a cluster of nucleosomes ., Such states are believed to provide for epigenetic regulation of the underlying genes , allowing transient signals to produce the long-term and heritable expression states needed for cell differentiation and development ., Indeed , several histone modifying enzymes involved in epigenetic regulation are known to recognize and create the same modification 16–19 ., Theoretical analyses of such systems 11 , 20–25 have shown that in order to generate heritable bistability , the positive feedback recruitment reaction must be: ( 1 ) non-local on the DNA ( i . e . involving interactions between non-adjacent nucleosomes ) , ( 2 ) substantially more frequent than non-recruited changes in modification state ( such as by histone replacements , random modifications and DNA replication ) , and ( 3 ) cooperative ., We have previously examined two ways in which the cooperativity requirement can be met ., Direct or explicit cooperativity involves the modifying enzyme needing at least two nucleosomes of a particular type in order to create a new nucleosome of that type ( Fig . 1A ) , for example , if the enzyme is only efficiently recruited by simultaneous contact with two or more nucleosomes 24 , 25 ., Indirect , or two-step , cooperativity occurs when creation of a particular nucleosome type requires two successive modification steps , each of which can be catalysed by an enzyme recruited by just one nucleosome of that type 11 , 21 ., Two-step cooperativity requires at least three nucleosome types , which can be achieved using a single histone residue if there are at least three modification states of that residue e . g . a particular lysine may be unmodified , acetylated or methylated ( Fig . 1B ) ., Modeling has so far been confined to systems involving recognition and modification reactions at single histone residues ( e . g . Fig . 1AB ) ., Here , we analyze the possibilities for generating heritably bistable systems with a minimal histone code in which there are modifications at two separate histone positions ., In its simplest form , where each of the two histone positions is allowed only two possible states , unmodified or modified , this system produces four nucleosome types ( 0 , 1 , 2 and 3; Fig . 1C ) ., Applying the histone code concept , the enzymes that bind to and modify these nucleosomes should be capable of distinguishing the four combinations ., Thus , there are eight specific modification or de-modification reactions capable of causing inter-conversions between the four types ., ( It should be noted that because each nucleosome has two copies of each histone and has two-fold symmetry , two modification states at each of two histone positions in fact results in 10 distinct nucleosome types - see Fig . 1 legend . Our simplification is that all enzymes in the system are recognizing modifications on one half of a nucleosome - essentially that each nucleosome in our system is comprised of two independent half-nucleosomes ) ., Each of the four nucleosome types may or may not recruit an enzyme that catalyzes a particular reaction ., Thus there are possible combinations for each of the eight reactions , giving a total state space of circuits ., We reduced this state space by allowing each of the eight reactions to be catalyzed by only one or none of the nucleosome types , giving five possibilities for each reaction and thus possible circuits ., Effectively , we are examining those cases that have maximal discrimination between nucleosome types ., One particular circuit ( the ‘classical’ circuit , most similar to the 3-nucleosome type system ) is shown in Fig . 1C ., Each possible circuit is defined by a code assigning , for each reaction , the nucleosome that recruits the enzyme for that reaction ., All reactions can also occur by ‘noise’ , random transitions occurring irrespective of the status of other nucleosomes in the system ., Reactions not subject to recruited modification only occur through this process ., We also simplified the system by not adding cooperativity to the individual recruitment reactions and by making all recruitment reactions of equal strength and all noise reactions of equal strength ., In addition , the four nucleosome states should be distinguishable by the reader proteins that control the expression of the underlying genes ., A prevalence of a specific nucleosome type ( E1 ) would result in one epigenetic state ( e . g . gene activity ) while another type ( E2 ) would be associated with the alternative epigenetic state ( e . g . gene inactivity ) ., With a four-nucleosome code these E1 and E2 nucleosomes could either be opposite to each other in the circuit ( i . e . different at both histone positions ) or adjacent to each other ( i . e . different at only one histone position ) ., The final description needed for the circuit is to define the type of nucleosome that is inserted after replication , R ( Fig . 1 ) ., The R type can be one of the E1 or E2 types ( R-in circuit ) or one of the other nucleosome types ( R-out circuit ) , giving 4 different circuit types: E-opposite/R-out , E-opposite/R-in , E-adjacent/R-out and E-adjacent/R-in ., The ability of any specific circuit to generate stable and heritable alternative modification states was tested by iteration of recruitment reactions , non-recruitment ( noise ) reactions and DNA replication steps among a system of nucleosomes ( strictly , 30 half-nucleosomes ) as follows ., Given a particular circuit we simulate a system for 250 generations , starting with all nucleosomes of the E1 type ., Subsequently we re-initiate the system by making all nucleosome the alternative E2 type and continue the simulation for another 250 generations ., To minimize the effect of randomness of individual simulations , we repeat this procedure 4 times , thus simulating each circuit for a total of 2000 generations ., The simulations are done at low noise , , Just before each replication we record the state of the system by counting the number of E1 and E2 nucleosomes , and ., An example of such a time-series is shown in Fig . 1D ., The quality of our circuit is evaluated from this time-series ., The first measure , testing for a balanced bistability is ( 1 ) where and are the respective probabilities for the system as a whole being in one of the epigenetic states ., We characterize the system as being in state if the number of nucleosome of this type exceeds the alternate type by at least half of the nucleosomes in the system , i . e . and reversely for ., A value of 1 corresponds to ., We consider a value of 0 . 75 a threshold for a good balance of the two states; if either or is <0 . 25 or >0 . 75 , then ., This also allows the system some ‘undecided’ time , since for a system that is in state 43 . 3% of the time and in state 43 . 3% of the time ., The second score for bistability measures the stability of the epigenetic states , and is simply the frequency of switches between the states per generation ., Strongly bistable systems which remain in E1 for the first 250 generations of the simulation and in E2 for the second 250 generations , and so forth for all 4 simulations of each code are assigned a stability ., The third measure of bistability was the average number of generations the system spends in intermediate states when switching between the E1 and E2 states , ., We set a threshold value of , reflecting the decisiveness expected to be important in epigenetic regulatory systems ., In general these measures of bistability were strongly affected by noise , and thus the larger noise a code can sustain , the more robust ( to noise ) is its associated bistability ., Initial screening of circuits was done with a noise level ., Fig . 2 examines all possible circuits for each of the four possible assignments of the E1 , E2 and R nucleosome types ., The scatter plots show and scores for all circuits with ., Circuits falling in the lower right side of the plots are the most bistable; we defined ‘working’ circuits as those with , ( and ) ., Some of the bistable circuits are shown in Fig ., 2 . Although we used a low noise level for testing , many of the circuits remained bistable with noise increased four-fold ., We were surprised by the number and variety of circuits able to produce good heritable bistability , especially considering that our search was limited in several ways: 1 ) It disregarded the possibility that a given enzymatic reaction can be catalyzed by more than one of the 4 nucleosome types ., 2 ) It disregarded that enzymatic reactions and noise moves may be individually graded , and thereby fine tuned to balance an otherwise biased drift among the states ., 3 ) It disregarded the possibility for explicit cooperativity ., In addition , our criteria for deciding whether the system was in the E1 or E2 state and for defining reasonable balance between these states were quite stringent ., The number of working circuits can be taken as a measure of mutational robustness for the corresponding E1 , E2 and R arrangements ., The E-opposite/R-out arrangement was the most robust with 202 working circuits ., However the E-opposite/R-in and E-adjacent/R-out arrangements have 67 and 19 working circuits , respectively ., The E-adjacent/R-in arrangement gave no bistable circuits by our criteria ., These data allow us to conclude: In the following analysis we confine ourselves to the E-opposite/R-out arrangement ., The 202 working E-opposite/R-out circuits are examined in more detail in Fig ., 3 . The number of times that each type of reaction occurs in this group of circuits is shown in Fig . 3A ., The most frequent reactions are self-creation recruitment reactions by E1 or E2 , particularly from the R state ., These reactions provide direct positive feedback by E1 and E2 ., E1 and E2 also frequently ‘attack’ the opposite state , which not only weakens that state but is the first move towards self-creation ., Together , these common reactions generate a consensus circuit in which E1 and E2 each use both possible two-step positive feedback pathways ( Fig . 2A ) ., The consensus circuit is the most bistable of the two-modification circuits ., In fact , as already seen in Fig . 1D , the classical circuit which contains only half of these recruitment reactions provides strong bistability ., Thus these reactions can often be replaced; the attack on E1 or E2 can often be left to noise and , less frequently , is stimulated by the non-E nucleosomes ., Interestingly , the creation of E1 or E2 sometimes occurs as a result of non-E nucleosomes ‘destroying’ themselves - recruiting enzymes that act on their own type ., In contrast , E1 and E2 never recruit enzymes that destroy themselves , though they occasionally act to create the opposing E type ., Working circuits generally involve strong activity of the E1 and E2 nucleosomes , usually containing 4–6 reactions in which the E1 and E2 nucleosomes recruit enzymes that make modifications that move nucleosomes towards their own type , either creating themselves or attacking the opposing state ( Fig . 3B; the consensus circuit contains 8 of these reactions ) ., However , a reasonable number of circuits use only three reactions of this kind ., In contrast , few circuits have E1 and E2 stimulating moves ‘away’ from their own type ( Fig . 3C ) ., Fig . 3D shows that bistability requires at least 4 recruitment reactions in the circuit ., However , the only working circuit with so few recruitment reactions is the classical circuit ( Fig . 1C ) , while the consensus circuit ( Fig . 2A ) requires 8 such reactions ., The minimal number of specific recruited enzymes required by the circuits ( Fig . 3E ) is substantially less than the number of recruitment reactions because in many cases a specific nucleosome type ( e . g . 11 ) recruits enzymes that catalyse the same reaction ( e . g . 0 to 1 at the first position ) on two nucleosomes ( e . g . 00 and 01 ) ., In these cases only a single enzyme is required; one that is sensitive to modifications at both positions on the recruiting nucleosome but is insensitive to the modification at the other position on its target nucleosome ., Among the 202 motifs with balanced bistability , have at least one standard two-step positive feedback pathway , where E1 or E2 stimulate the reaction to create one of the intermediate ( non-E ) nucleosomes and also stimulate the reaction creating themselves from that intermediate nucleosome ( Fig . 3F ) ., Because this two-step pathway involves the successive action of TWO E1 ( or E2 ) nucleosomes in their self-creation , it can produce a positive feedback with a dependence on the square of the number of E1 ( or E2 ) nucleosomes , providing a more-than-linear response , or ultrasensitivity ., There are four possible standard two-step positive feedback pathways: two directions ( towards E1 or E2 ) and two paths ( over the R or non-R nucleosome ) ., Circuits with one two-step pathway in each direction are most abundant but nearly as many circuits have just one such pathway ( Fig . 3F ) ., Involvement of the R nucleosome in these pathways is preferred , presumably because there are large numbers of these nucleosomes that need to be rapidly converted after replication ., We noticed that the standard two-step recruitment pathway nearly always includes an extra recruited reaction converting the intermediate type into the attacked E nucleosome type , pushing the intermediate “back” , against the flow of the two-step reactions ., In fact , the standard two-step pathway with such a destabilized intermediate is seen in 185 of the 202 motifs that exhibit bistability ( Fig . 3F ) ., This additional reaction is critical for bistability ( Fig . 4 ) ., Simulations showed that a circuit with two standard two step pathways without intermediate destabilization gave poor bistability ( Fig . 4A; in these simulations , replication was omitted and circuits were symmetrical in order to provide balance ) ., Addition of destabilization by the intermediate converting itself to the opposing E type improved bistability somewhat ( Fig . 4C ) ., Strong bistability was obtained when the destabilization was catalyzed by the opposing E type ( Fig . 4D ) , the type of reaction seen in the classical circuit and the 3 nucleosome-type system ( Fig . 1 ) ., This need for destabilization of the intermediate type can be understood as introducing a “loss” term for the intermediate nucleosome type that is necessary for ultrasensitivity ., Considering the case where type nucleosomes create themselves in two steps from type through type ( see x02230x3 in Fig . 2 ) : ( 2 ) ( 3 ) giving a steady state occupation of the rare intermediate state ., When this “loss” is sizeable , occupation of the intermediate state is sensitive to and eq ., 2 then predicts ultra sensitive dependence of on itself ., Notice that the “loss” term only supports ultra-sensitivity if the loss is not due to enzymes recruited by type nucleosomes , as seen in the simulations in Fig ., 4 . Surprisingly , 12 of the E-opposite/R-out circuits work without any of these standard two-step positive feedback pathways ., This group of ‘exotic’ circuits is comprised of 6 unique circuits , each having a symmetrical twin ( Fig . 5A ) ., Five of these circuits contain one or more reaction motifs that provide a novel form of two-step cooperativity ( the exception is x10320x3 ) ., In these ‘pull-push’ reactions an E nucleosome recruits an enzyme that attacks the opposite E type , converting it to one of the intermediate types ( R or non-R ) , and the intermediate recruits an enzyme that converts its own type to the E type ( Fig . 5B ) ., Because the E nucleosome acts to create the recruiter AND the target for the reaction that creates its own type , this pair of reactions provides a positive feedback that can exhibit an ultrasensitive dependence on the number of E nucleosomes , provided that the intermediate state is again destabilized ., Simulation of a minimal circuit with two pull-push reactions ( without replication ) shows that this motif alone is not able to provide bistability ( Fig . 5B ) ., However , adding a loss of the intermediate , due to a self-creation reaction catalyzed by the attacked E type , provides robust bistability even at high noise levels ., This destabilized pull-push motif is present in 74 of the 84 bistable circuits that contain pull-push reactions ( Fig . 5C ) ., These reactions are therefore likely to play a role in strengthening the bistability in a large fraction of the 202 bistable E-opposite/R-out circuits ., even when standard two-step cooperativity is present ., Fig . 6 examines the frequencies of the destabilized pull-push and standard 2-step motifs in the 202 bistable E-opposite/R-out circuits ., Only 4 circuits have neither of these motifs ( 2 symmetrical pairs ) , one being the x10320x3 circuit ( Fig . 5A ) ., The x10320x3 circuit ( Fig . 5A ) and its symmetrical counterpart do not contain either a standard two-step motif or a pull-push motif ., However recruitment from 10 act to support the 01 type that attacks 10 , and thus recruitment reactions around 10 represent a variant destabilized pull-push motif ., At the same time , the 01 type is occupied simultaneously with type 11 , and together they provide a 2-step recruitment ., Thus the fact that this exotic motif is stable even up to noise level may well select that it integrates the two recruitment paths that both lead to ultrasensitivity ., The large number of different bistable circuits revealed in our screen suggests that different organisms , different cells and different genomic regions could utilize different variations of a 4-nucleosome-type modification system to achieve epigenetic regulation ., To examine how such differences could arise by evolution , we looked at the ‘connectedness’ of the different circuits ., Fig . 7 shows each of the 202 bistable circuits as nodes in a network , each linked to circuits that have only one reaction catalyzed differently ., The network tends to be clustered into two main groups , centered around the consensus circuit 00033033 and a more peripheral cluster centered around the classical circuit x003xx30 where the non-R paths are less catalyzed ., Remarkably , 191 of the circuits are connected into one network ., This connectedness is surprising , given the strong limitations we put on our search space for circuit motifs , and means that the large variety of bistable circuits can be explored through a succession of relatively small evolutionary changes , without loss of bistability ., Sequences of single enzymatic replacements can connect circuits that have no common recruitment , for example motif 0x2131x3 and x003xx3x shown explicitly in Fig . 7 ., Stepping through this network , a system may maintain basic epigenetic and regulatory properties while exploring subtle differences due to different combinations of recruitment processes ., This evolutionary network resembles evolution of RNA sequences , which traverses far ranging but connected neutral plateaus of primary sequences which fold in identical secondary structures 26 ., A degeneracy has also been suggested for regulatory networks that can digitalize morphogen gradients 27 ., By sampling histone code circuits for their ability to support heritable bistability we have addressed the minimal requirements for obtaining epigenetics , defined as the ability to remember one of several possible states across several cell generations ., Our findings are also important because histone code circuits with potential for bistability allow genetic regulation that can be ultrasensitive or graded , depending on the size of the system and the extent to which it couples to a noisy catalytic environment in the cell 21 ., Our search was constrained to a “circuit code” space which was limited in both the number of considered modifications and in the scope for regulating each transition ., In spite of these limitations , we found many solutions , including some unexpected new regulatory designs for bistable feedback systems ., Thus , having two histone positions that can be modified AND having reading and writing enzymes that can distinguish the four resulting nucleosome types allows a large variety of reaction circuits that can generate stable and heritable alternative modification states ., This increased number of circuits results from the increased number of pathways available for indirect cooperativity , where positive feedback involves two successive recruitment reaction steps ., This indirect cooperativity can be achieved by two kinds of two-step positive feedback ., In the standard pathway 11 , a nucleosome type recruits enzymes that catalyse a sequence of two steps:one step to create an intermediate nucleosome type that is different from it by one modification and a second step to create its own type ., We discovered a new cooperativity motif , the pull-push reaction , in which the second step is clarified out by the intermediate type recruiting the enzyme that converts itself ., To generate ultrasensitivity , both cooperativity motifs require that the intermediate nucleosome type is destabilized by a recruitment reaction that pushes it in the opposite direction ., Our analysis reveals a number of ‘rules’ for generation of such epigenetic circuits:, 1 ) The circuit should contain at least one two-step intermediate-destabilized pathway of either the standard or pull-push type ., There were only 2% exceptions to this rule among our accepted motifs ., 2 ) It is easier to produce working circuits if the alternative dominant nucleosome types are different at both modification positions , that is , are separated by 2 reaction steps ., This reflects the ease of producing two-step cooperativity ., 3 ) It is easier to produce working circuits if the new nucleosomes inserted after replication are not one of the alternative dominant nucleosome types ., This helps avoid biasing the system too heavily towards one dominant state ., 4 ) The dominant nucleosome types are highly active in recruiting enzymes that create their own type or destroy the opposing dominant type , and never self-destruct , that is , recruit enzymes that change their own type ., 5 ) Self-destruction is only seen for intermediate nucleosome types ., These rules could be relaxed with removal of some of the restrictions we placed on the circuits ., However , we believe that they are likely to be general features of nucleosome-based epigenetic systems ., Finally we found that conversion of one circuit into almost any other circuit can occur by a succession of small changes that retain heritable bistability , a feature that should facilitate circuit evolution ., This plasticity is consequence of the fact that the two motifs that generate cooperativity are only one “mutation” away from each other in the sense that only one recruitment separates them from each other ., Although our restricted 4-nucleosome-type system is capable of surprisingly complex behavior , it is extremely simple compared to real systems ., Nucleosomes are known to be modified in multiple ways at many positions , and modifying enzymes are likely to be sensitive to combinations of these modifications in complex ways ., Thus , our analysis indicates that real systems have huge potential for generating multiple stable and heritable nucleosome modification states .

Introduction, Methods, Results, Discussion

Nucleosomes can be covalently modified by addition of various chemical groups on several of their exposed histone amino acids ., These modifications are added and removed by enzymes ( writers ) and can be recognized by nucleosome-binding proteins ( readers ) ., Linking a reader domain and a writer domain that recognize and create the same modification state should allow nucleosomes in a particular modification state to recruit enzymes that create that modification state on nearby nucleosomes ., This positive feedback has the potential to provide the alternative stable and heritable states required for epigenetic memory ., However , analysis of simple histone codes involving interconversions between only two or three types of modified nucleosomes has revealed only a few circuit designs that allow heritable bistability ., Here we show by computer simulations that a histone code involving alternative modifications at two histone positions , producing four modification states , combined with reader-writer proteins able to distinguish these states , allows for hundreds of different circuits capable of heritable bistability ., These expanded possibilities result from multiple ways of generating two-step cooperativity in the positive feedback - through alternative pathways and an additional , novel cooperativity motif ., Our analysis reveals other properties of such epigenetic circuits ., They are most robust when the dominant nucleosome types are different at both modification positions and are not the type inserted after DNA replication ., The dominant nucleosome types often recruit enzymes that create their own type or destroy the opposing type , but never catalyze their own destruction ., The circuits appear to be evolutionary accessible; most circuits can be changed stepwise into almost any other circuit without losing heritable bistability ., Thus , our analysis indicates that systems that utilize an expanded histone code have huge potential for generating stable and heritable nucleosome modification states and identifies the critical features of such systems .

Specialized enzymes add and remove chemical modifications to the histone proteins that package DNA into nucleosomes ., These modifications act as labels to recruit various proteins to the DNA locations where they are needed to control DNA functions , such as gene expression ., The modifications are usually made and maintained in response to specific signals ., However , if a modifying enzyme is itself recruited by the modification it makes , then this positive feedback could cause the modification or its absence to be self-sustaining , and even heritable , once the signal has gone ., We used computer simulations to systematically explore the possibilities for such epigenetic states when there is an expanded modification ‘code’ - one that involves the presence or absence of two different modifications rather than just one ., We found that this small expansion of the histone code allows hundreds of different modification and enzyme recruitment schemes to give alternative stable and heritable states ., These worked best when the nucleosomes in alternative states were differently modified at both positions ., All working schemes involved positive feedback and cooperativity between nucleosomes ., Thus , even a simple histone code could be used in many ways to make stable and heritable , yet reversible , marks on DNA .

biology

journal.ppat.1006114

Experimental Estimation of the Effects of All Amino-Acid Mutations to HIV’s Envelope Protein on Viral Replication in Cell Culture

HIV evolves rapidly: the envelope ( Env ) proteins of two viral strains within a single infected host diverge as much in a year as the typical human and chimpanzee ortholog has diverged over ∼5-million years 1–4 ., This rapid evolution is central to HIV’s biology ., Most humans infected with HIV generate antibodies against Env that effectively neutralize viruses from early in the infection 5–7 ., However , Env evolves so rapidly that HIV is able to stay ahead of this antibody response , with new viral variants escaping from antibodies that neutralized their predecessors just months before 5–7 ., Env’s exceptional evolutionary capacity is therefore essential for the maintenance of HIV in the human population ., A protein’s evolutionary capacity depends on its ability to tolerate point mutations ., Detailed knowledge of how mutations affect Env is therefore key to understanding its evolution ., Many studies have estimated the effects of mutations to Env ., One strategy is experimental: numerous studies have used site-directed mutagenesis or alanine scanning to measure how specific mutations affect various aspects of Env’s function 8–17 ., However , these experiments have examined only a small fraction of the many possible mutations to Env ., Another strategy is computational: under certain assumptions , the fitness effects of mutations can be estimated from their frequencies in global or intra-patient HIV sequences 18–22 ., However , these computational strategies are of uncertain accuracy and cannot separate the contributions of inherent functional constraints from those of external selection pressures such as antibodies ., Therefore , a more complete and direct delineation of how every mutation affects Env’s function would be of great value ., It is now possible to make massively parallel experimental measurements of the effects of protein mutations using deep mutational scanning 23–25 ., These experiments involve creating large libraries of mutants of a gene , subjecting them to bulk functional selections , and quantifying the effect of each mutation by using deep sequencing to assess its frequency pre- and post-selection ., Over the last few years , deep mutational scanning has been used to estimate the effects of all single amino-acid mutations to a variety of proteins or protein domains 26–39 , as well as to estimate the effects of a fraction of the amino-acid mutations to many additional proteins ( e . g . , 40–42 ) ., When these experiments examine all amino-acid mutations , they can be used to compute the mutational tolerance of each protein site , thereby shedding light on a protein’s inherent evolutionary capacity ., Recently , deep mutational scanning has been used to examine the effects of amino-acid mutations on the binding of antibodies to Env protein displayed on mammalian or yeast cells 43 , 44 , or the effects of single-nucleotide mutations scattered across the HIV genome on viral replication in cell culture 45 ., However , none of these studies comprehensively measure the effects of all Env amino-acid mutations on viral replication ., Therefore , we currently lack comprehensive measurements of the site-specific mutational tolerance of Env ., Here we use deep mutational scanning to experimentally estimate how all amino-acid mutations to the ectodomain and transmembrane domain of Env affect viral replication in cell culture ., At most sites , our measurements correlate with the frequencies of amino acids in natural HIV sequences ., However , there are large deviations at sites where natural evolution is strongly shaped by factors ( e . g . , antibodies ) that are absent from our experiments ., Our results also show that site-to-site variation in Env’s inherent capacity to tolerate mutations helps explain why epitopes of broadly neutralizing antibodies are highly conserved in natural isolates ., Overall , our work helps elucidate how inherent functional constraints and external selective pressures combine to shape Env’s evolution , and demonstrates a powerful experimental approach for comprehensively mapping how mutations affect HIV phenotypes that can be selected for in the lab ., We used the deep mutational scanning approach in Fig 1A to estimate the effects of all single amino-acid mutations to Env ., We applied this approach to Env from the LAI strain of HIV 46 ., LAI is a CXCR4-tropic subtype B virus isolated from a chronically infected individual and then passaged in human T-lymphocytes ., We chose this strain because LAI and the closely related HXB2 strain have been widely used to study Env’s structure and function 8–11 , 47–49 , providing extensive biochemical data with which to benchmark our results ., LAI’s Env is 861 amino acids in length ., We mutagenized amino acids 31–702 ( throughout this paper , we use the HXB2 numbering scheme 50 ) ., We excluded the N-terminal signal peptide and the C-terminal cytoplasmic tail , since mutations in these regions can alter Env expression in ways that affect viral infectivity in cell culture 51–53 ., The region of Env that we mutagenized spanned 677 residues , meaning that there are 677 × 63 = 42 , 651 possible codon mutations , corresponding to 677 × 19 = 12 , 863 possible amino-acid mutations ., To create plasmid libraries containing all these mutations , we used a previously described PCR mutagenesis technique 31 that creates multi-nucleotide ( e . g , gca→CAT ) as well as single-nucleotide ( e . g , gca→gAa ) codon mutations ., We created three independent plasmid libraries , and carried each library through all subsequent steps independently , meaning that all our measurements were made in true biological triplicate ( Fig 1B ) ., We Sanger sequenced 26 clones to estimate the frequency of mutations in the plasmid mutant libraries ( S1 Fig ) ., There were an average of 1 . 4 codon mutations per clone , with the number of mutations per clone roughly following a Poisson distribution ., The deep sequencing described in the next section found that at least 79% of the ≈104 possible amino-acid mutations were observed at least three times in each of the triplicate libraries , and that 98% of mutations were observed at least three times across all three libraries combined ., The plasmid libraries therefore sampled most amino-acid mutations to Env ., We produced virus libraries by transfecting each plasmid library into 293T cells ., The viruses in the resulting transfection supernatant lack a genotype-phenotype link , since each cell is transfected by many plasmids ., We therefore passaged the transfection supernatants twice in SupT1 cells at an MOI of 0 . 005 to create a genotype-phenotype link and select for functional variants ., Importantly , neither 293T nor SupT1 cells express detectable levels of APOBEC3G 54 , 55 , which can hypermutate HIV genomes 56 , 57 ., This is a crucial point: although HIV encodes a protein that counteracts APOBEC3G , a fraction of viruses will lack a functional version of this protein and so have their genomes hypermutated in APOBEC3G-expressing cells ., For each library , we passaged 5 × 105 infectious particles in order to maintain library diversity ., We used Illumina deep sequencing to quantify the frequency of each mutation before and after passaging ., In order to increase the sequencing accuracy , we attached unique molecular barcodes or “Primer IDs” to each PCR amplicon 58–61 ., We sequenced the plasmids to assess the initial mutation frequencies , and sequenced non-integrated viral DNA 62 from infected SupT1 cells to assess the mutation frequencies in the viruses ., A concern is that errors from sequencing and viral replication ( e . g . , from viral reverse transcriptase ) would introduce bias ., To address this concern , we paired each mutant library with a control in which we generated wildtype virus from unmutated plasmid ., Sequencing the control plasmids and viruses enabled us to estimate and statistically correct for the rates of these errors ( S2 Fig ) ., Overall , these procedures allowed us to implement the deep mutational scanning workflow in Fig 1 ., Our deep mutational scanning experiments require that selection purge the virus libraries of non-functional variants ., As an initial gene-wide measure of selection , we analyzed how different types of codon mutations ( nonsynonymous , synonymous , and stop-codon mutations ) changed in frequency after selection ., In these analyses , we corrected for background errors from PCR , sequencing , and viral replication by subtracting the mutation frequencies measured in our wildtype controls from those measured in the mutant libraries ( S2 Fig ) ., Stop-codon mutations are expected to be uniformly deleterious ., Indeed , after correcting for background errors , stop codons were purged to <1% of their initial frequency in the twice-passaged viruses for each replicate , indicating strong purifying selection ( see the data for “all sites” in Fig 2A ) ., The second viral passage is important for complete selection , as stop codons remain at about ≈16% of their initial frequency in viruses that were only been passaged once ( S3 Fig ) ., Interpreting the frequencies of nonsynonymous mutations is more nuanced , as different amino-acid mutations have different functional effects ., However , a large fraction of amino-acid mutations are deleterious to any protein 63–65 ., Therefore , one might expect that the frequency of nonsynonymous mutations would decrease substantially in the twice-passaged mutant viruses ., But surprisingly , even after correcting for background errors , the average frequency of nonsynonymous mutations in the passaged viruses is ≈90% of its value in the mutant plasmids ( see the data for “all sites” in Fig 2A ) ., However , the average masks two disparate trends ., In each library , a few sites exhibit large increases in the frequency of nonsynonymous mutations , whereas this frequency decreases by nearly two-fold for all other sites ( see the data for the subgroups of sites in Fig 2A ) ., An obvious hypothesis is that at a few sites , amino-acid mutations are favored because they are adaptive for viral replication in cell culture ., Consistent with this hypothesis , the sites that experienced large increases in mutation frequencies are similar among the three replicates ( Fig 2B ) , suggestive of reproducible selection for mutations at these sites ., Moreover , these sites are spatially clustered in Env’s crystal structure in regions where mutations are likely to enhance viral replication in cell culture ( Fig 3 and S1 Table ) ., One cluster of mutations disrupts potential glycosylation sites at the trimer apex ( Fig 3A ) ., This result suggests that some of the glycans that help shield Env from antibodies in nature 6 , 66 actually decrease viral fitness in the absence of immune selection ., This idea is consistent with previous studies showing that that loss of glycosylation sites can enhance viral infectivity in cell culture 67–69 ., A second cluster overlaps sites where mutations influence Env’s conformational dynamics , which are commonly altered by cell-culture passage 70 , 71 ., It has been hypothesized that neutralization-resistant Envs primarily assume conformations that mask conserved antibody epitopes , while lab-adapted variants more efficiently sample different conformations associated with CD4 binding 72 ., Thus , the adaptive mutations we observe may enable Env to more efficiently use CD4 in cell culture , but would not be selected in nature because they expose conserved epitopes ., A third cluster is at the co-receptor binding interface ( Fig 3B ) , where mutations may enhance viral entry in cell culture ., Therefore , while most of Env is under purifying selection against changes to the protein sequence , a few sites are under selection for cell-culture adapting amino-acid mutations ., If our experiments are indeed identifying mutations to LAI that are beneficial in cell culture , then one expectation is that some of these mutations might fix after prolonged passage of LAI in cell culture ., Interestingly , almost exactly such an experiment was performed in the early study of HIV ., The LAI strain used in our study was initially isolated from a chronically infected individual and then passaged in cell culture for a short period of time before cloning 46 , 77 ., HXB2 , another common lab strain , is derived from a variant of LAI that was repeatedly passaged in a variety of cell lines , initially as a contaminant of other viral stocks 78 , 79 ., There are 23 amino-acid differences between the Env proteins of LAI and HXB2 ., Although the predecessor for HXB2 was not passaged in the same SupT1 cell line that we used , if its passage in other cell lines led to mutations that were generally adaptive to cell culture , then we would expect them to introduce amino acids in HXB2 that are also selected in our deep mutational scan of LAI ., Indeed , we found that most differences between LAI and HXB2 introduced mutations to amino acids that our experiments suggest are more preferred in cell culture than the wildtype LAI amino acid ( S2 Table ) ., Thus , our results are consistent with the expectation that HXB2 is more adapted to cell culture than LAI ., The average error-corrected frequency of synonymous mutations changes little after selection ( an average decrease to 96% of the original frequency; see the data for “all sites” in Fig 2A ) ., This overall trend is consistent with the fact that synonymous mutations usually have smaller functional effects than nonsynonymous mutations ., However , synonymous mutations can sometimes have substantial effects 21 , 80–82 , particularly in viruses like HIV that are under strong selection for RNA secondary structure and codon usage 83 , 84 ., To assess selection on synonymous mutations on a more site-specific level , we examined the change in frequency of multi-nucleotide codon mutations across env’s primary sequence ( Fig 4 ) ., The rationale behind examining only multi-nucleotide codon mutations is that they are not appreciably confounded by errors from PCR , deep sequencing , or de novo mutations from viral replication ( S2 and S4 Figs ) ., In a region roughly spanning codons 500 to 600 , selection strongly purged both synonymous and nonsynonymous multi-nucleotide codon mutations ( Fig 4 ) ., This region contains env’s Rev-response element ( RRE ) 85 , a highly structured region of RNA that is bound by the Rev protein to control the temporal export of unspliced HIV transcripts from the nucleus 86 , 87 ., The finding of strong selection on the nucleotide as well as the amino-acid sequence of the RRE region of Env therefore agrees with our biological expectations ., The previous section examined broad trends in selection averaged across many sites ., But our data also enable much more fine-grained estimates of the preference for every amino-acid at every position in Env ., We define a site’s preference for an amino acid to be proportional to the enrichment or depletion of that amino acid after selection ( correcting for the error rates determined using the wildtype controls ) , normalizing the preferences for each site so that they sum to one ., We denote the preference of site r for amino acid a as πr , a , and compute the preferences from the deep-sequencing data as described in 88 ., Since we mutagenized 677 residues in Env , there are 677 × 20 = 13 , 540 preferences ., If selection in our experiments exactly parallels selection in nature and there are no shifts in mutational effects as Env evolves , then these preferences are the expected frequencies of each amino acid at each site in an alignment of Env sequences that have reached evolutionary equilibrium under a mutation process that introduces each amino acid with equal probability 31 , 89 ., Fig 5 shows Env’s site-specific amino-acid preferences after averaging across replicates and re-scaling to account for the stringency of selection in our experiments ( details of this re-scaling are in the next section ) ., As is immediately obvious from Fig 5 , sites vary dramatically in their tolerance for mutations ., Some sites strongly prefer a single amino acid , while other sites can tolerate many amino acids ., For instance , site 457 , an important receptor-binding residue 8 , has a strong preference for aspartic acid ., However , this site is adjacent to a variable loop ( sites 460–469 ) where most sites tolerate many amino acids ., Another general observation is that when sites tolerate multiple amino acids , they often prefer ones with similar chemical properties ., For instance , sites 225 and 226 prefer hydrophobic amino acids , while sites 162 to 164 prefer positively charged amino acids ., To confirm that our experiments captured known constraints on Env’s function , we examined mutations that have been characterized to affect key functions of Env ., Table 1 lists mutations known to disrupt an essential disulfide bond , binding to receptor or co-receptor , or protease cleavage ., In almost all cases , the deleterious mutation introduces an amino-acid that our experiments report as having a markedly lower preference than the wildtype amino acid ., Therefore , our measurements largely concord with existing knowledge about mutations that affect key aspects of Env’s function ., A crucial aspect of any high-throughput experiment is assessing the reproducibility of independent replicates ., Fig 5 shows the average of the preferences measured in each replicate ., Fig 6A shows the correlations among the 13 , 540 site-specific amino-acid preferences estimated from each of the three replicates ., The correlations are modest , indicating substantial replicate-to-replicate noise ., In principle , this noise could arise from differences in the initial plasmid mutant libraries , bottlenecks during the generation of viruses by transfection , bottlenecks during viral passaging , or bottlenecks during the sequencing of proviral DNA from infected cells ., Analysis of technical replicates of the first or second round of viral passaging indicates that most of the noise arises from bottlenecks during the viral passaging or sequencing steps ., Specifically , measurements from replicate 3 are no more correlated to those from replicates 3b-1 or 3b-2 ( which are repeated passages of the same transfection supernatant , Fig 1B ) than they are to those from totally independent replicates ( compare Fig 6 and S6 Fig ) ., However , replicates 3b-1 and 3b-2 ( which shared the first of the two viral passages , Fig 1 ) do yield more correlated measurements than independent replicates ( S6 Fig ) ., The existence of bottlenecks during viral passage is also suggested by the data in S4 and S5 Figs ., Therefore , the experimental reproducibility could probably be increased by passaging more infectious viruses at each step ., If bottlenecks cause each replicate to sample slightly different mutations , then perhaps the total number of tolerated mutations per site will be similar between replicates , even if the exact mutations differ ., To test this hypothesis , we computed the effective number of amino acids tolerated at each site as the exponential of the Shannon entropy of the site’s amino-acid preferences ., Fig 6B shows that the effective number of amino acids tolerated at each site is more correlated between replicates than the preferences themselves ., We further reasoned that even if bottlenecking causes slight variations in the preferred amino acids between replicates , each site would still tend to prefer amino acids with similar chemical characteristics ., To test this hypothesis , we quantified the extent that each site preferred hydrophobic or hydrophilic amino acids by computing a site-specific hydrophobicity score from the amino-acid preferences ., Fig 6C shows that these preference-weighted hydrophobicities are more correlated between replicates than the preferences ., Therefore , even though there is replicate-to-replicate noise in the exact amino acids preferred at a site , the effective number of tolerated amino acids and the chemical properties of these amino acids are similar among replicates ., In the previous section , we showed that our experimentally measured amino-acid preferences captured the constraints on Env’s biological function for sites with known mutational effects ( Table 1 ) ., If this is true across the entire protein , then our measurements should correlate with the frequencies of amino acids in natural HIV sequences ., Table 2 shows that there is a modest correlation ( Pearson’s R ranging from 0 . 29 to 0 . 36 ) between the preferences from each experimental replicate and the frequencies in an alignment of HIV-1 group-M sequences ( a phylogenetic tree of these sequences is in Fig 7A; sites in Env variable loops that can not be reliably aligned are excluded as described in the Methods ) ., Since each replicate suffers from noise due to partial bottlenecking of the viral diversity , we hypothesized that averaging the preferences across replicates should make them more accurate ., Indeed , averaging the replicates increased the correlation to R = 0 . 4 ( Table 2 ) ., The concordance between deep mutational scanning measurements and natural sequence variation is improved by accounting for differences in the stringency of selection in the experiments compared to natural selection 89 , 91 ., Specifically , if the measured preference is πr , a and the stringency parameter is β , then the re-scaled preference is ( πr , a ) β/ ∑a′ ( πr , a′ ) β ., A stringency parameter of β > 1 means that natural evolution favors the same amino acids as the experiments , but with greater stringency ., Table 2 shows that for all replicates , the stringency parameter that maximizes the correlation is >1 ., Therefore , natural selection prefers the same amino acids as our experiments , but with greater stringency ., After averaging across replicates and re-scaling by the optimal stringency parameter , the Pearson correlation is 0 . 44 between our experimentally measured preferences and amino-acid frequencies in the alignment of naturally occurring HIV sequences ( Fig 7B ) ., Is this a good correlation ?, At first glance , a correlation of 0 . 44 seems unimpressive ., But we do not expect a perfect correlation even if the experiments perfectly concord with selection on Env in nature ., There are several factors that are expected to reduce the correlation between the experimentally measured preferences and amino-acid frequencies in natural sequences ., First , our experiments examine the effects of mutations to Env from the LAI strain ., However , it is well known that epistasis can cause the effects of mutations to differ among homologs of the same protein 92 , 93 , and many examples of this phenomenon have been documented in HIV Env 94–97 ., Therefore , our measurements for the LAI Env are probably not completely generalizable to all other strains ., In addition , natural HIV sequences are drawn from a phylogeny ( Fig 7A ) , not an ideal ensemble of all possible Env sequences ., The frequencies of amino acids in this phylogeny reflect evolutionary history as well as natural selection ., For instance , if several amino acids are equally preferred at a site , one is likely to be more frequent in the alignment due to historical contingency ., Additionally , natural evolution is influenced by the genetic code and mutation biases: a mutation from the tryptophan codon TGG to the valine codon GTT is extremely unlikely even if valine is more preferred than tryptophan ., Mutation biases inherent in reverse transcription 98 or APOBEC3G-induced hypermutation 54 could also bias some evolutionary outcomes over others ., Therefore , the correlation will be imperfect even if the preferences completely concord with natural selection—the question is how the actual correlation compares to what is expected given the phylogenetic history and mutation biases ., To determine the expected correlation if the experimentally measured amino-acid preferences reflect conserved constraints in Env , we simulated evolution along the phylogenetic tree in Fig 7A under the assumption that the experimentally measured preferences exactly match natural selection ., Specifically , we used pyvolve 99 to simulate evolution using the experimentally informed site-specific codon substitution models described in 91 , which define mutation-fixation probabilities in terms of the amino-acid preferences ., In addition to the preferences and the stringency parameter β = 2 . 1 from Table 2 , the substitution models in 91 require specification of parameters reflecting biases in the mutation process ., We estimated nucleotide mutation bias parameters of ϕA = 0 . 55 , ϕC = 0 . 15 , ϕG = 0 . 11 , and ϕT = 0 . 18 from the frequencies at the third-nucleotide codon position in sequences in the group-M alignment for sites where the most common amino acid had 4-fold codon degeneracy ., We used the transition-transversion ratio of κ = 4 . 4 estimated in 100 ., For these simulations , we scaled the branch lengths so that the average pairwise protein divergence was the same in the actual and simulated alignments ., The correlation between the preferences and amino-acid frequencies in a representative simulated alignment is shown in Fig 7C ., As this plot illustrates , the expected correlation is only about 0 . 46 if the experimentally measured preferences exactly describe natural selection on Env under our model ., The simulated frequencies in Fig 7C show the same pattern of bi-modality ( most values near zero or one ) as the actual frequencies in Fig 7B despite the fact that the preferences used in the simulations allow multiple amino acids at most sites ( see Fig 5 ) ., This fact illustrates that bi-modality in the amino-acid frequencies can arise from the historical contingency inherent in a phylogenetic tree even if multiple amino acids are tolerated at most sites ., As a control , we also simulated evolution using substitution models in which the preferences have been randomized among sites ( Fig 7D ) ; as should be the case , there is no correlation in these control simulations ., So the actual correlation is nearly as high as expected if natural selection concords with the preferences measured in our experiment ., We next investigated if there are parts of Env for which there is an especially low correlation between our experimentally measured preferences and natural amino-acid frequencies ., For instance , antibodies exert selection on the surface of Env in nature 6 , 7 , 101 , 102 ., We therefore examined the actual and simulated correlations between the preferences and frequencies as a function of solvent accessibility ( Fig 7E and 7F ) ., For all sites ( right side of Fig 7E , left side of Fig 7F ) , the actual correlation is only slightly lower than the range of correlations in 100 simulations ., For more buried sites , both the simulated and actual correlations increase ( Fig 7E ) , presumably because sites in the core of Env tend to have stronger preferences for specific amino acids ., But as sites become more surface-exposed , the actual correlation drops below the value expected from the simulations ( Fig 7F ) ., Therefore , our experiments provide a relatively worse description of natural selection on Env’s surface than its core—probably because the evolution of the protein’s core is shaped mostly by inherent functional constraints that are effectively captured by our experiments , whereas the surface is subject to selection pressures ( e . g . , antibodies ) that are not modeled in our experiments ., Comparing disulfide-bonded cysteines and glycosylation sites vividly illustrates this dichotomy between inherent functional constraints and external selection pressures ., Env has 10 highly conserved disulfide bonds , most of which are essential for the protein’s inherent function 49 ., Env also has numerous N-linked glycosylation sites , many of which are also highly conserved in nature , where they help shield the protein from antibodies 6 , 66 ., In contrast to the disulfides , only some glycosylation sites are important for Env’s function in the absence of immune selection 67 , 69 ., Fig 8 shows that our experimentally measured preferences are highly correlated with natural amino-acid frequencies at the sites of the disulfides , but not at the glycosylation sites ., This result can easily be rationalized: the disulfides are inherently necessary for Env’s function , whereas many glycosylation sites are important largely because of the external selection imposed by antibodies ., Our experiments therefore accurately reflect the natural constraints on the former but not the latter ., The fact that we found well-tolerated mutations at all of Env’s glycosylation sites ( S7A Fig ) might seem surprising given that other studies have shown that some glycosylation sites are important for Env’s function in certain HIV strains 67 , 69 ., However , these studies were all performed in HIV strains substantially diverged from LAI ., A study in HXB2 ( which is closely related to LAI ) found that individual mutations are at least partially tolerated at all glycosylation sites in Env’s gp120 subunit when assaying for viral infectivity in cell culture 103 ., Therefore , glycosylation sites may be especially expendable in the LAI strain used in our study ., Different sites in Env evolve at different rates in natural HIV sequences ., For instance , sites on the apical surface of Env evolve especially rapidly 104 ., These differences in evolutionary rate arise from two factors ., First , some sites are inherently better at tolerating mutations without disrupting Env’s essential functions ., Second , some sites are under stronger immune selection for rapid sequence change ., However , since Env in nature is under selection both to maintain its function and escape immunity , it is difficult to deconvolve these factors ., Our experiments estimate each site’s inherent tolerance for mutations under selection purely for Env’s function in cell culture , without the confounding effects of immune selection ( for the remainder of this section , we define a site’s mutational tolerance as the Shannon entropy of its amino-acid preferences shown in Fig 5 ) ., We can therefore assess whether regions of Env that evolve rapidly or slowly in nature also have unusually high or low inherent tolerance to mutations ., We focused on two regions of Env ., First , we analyzed portions of the protein classified as “variable loops” due to extensive variation in nature 105 , 106 ., These loops are frequently targeted by antibodies that drive rapid sequence evolution 102 , 107 ., Because these loops evolve rapidly , we hypothesized they would have a high inherent mutational tolerance ., But an alternative hypothesis is that their rapid evolution more attributable to strong selection from antibodies than an unusually high mutational tolerance ., Second , we focused on epitopes of antibodies that broadly neutralize many HIV strains ., Because these epitopes are highly conserved in nature and often overlap with regions of known functional constraint 108–113 , we hypothesized they would have a low mutational tolerance ., However , an alternative hypothesis is that these epitopes evolve slowly not because they are mutationally intolerant but simply because they are under weaker immune selection ., Indeed , broad immune responses targeting these epitopes only develop in 20% of infected individuals and generally only after multiple years of infection 114 ., In testing these hypotheses , it is important to control for other properties known to affect mutational tolerance ., This can be done by using multiple linear regression to simultaneously analyze how several independent variables affect the dependent variable of mutational tolerance ., Relative solvent accessibility ( RSA ) is the strongest determinant of mutational tolerance in proteins 115 , so we included RSA as a variable in the regression ., The region of env that contains the RRE is under strong nucleotide-level constraint 85–

Introduction, Results, Discussion, Materials and Methods

HIV is notorious for its capacity to evade immunity and anti-viral drugs through rapid sequence evolution ., Knowledge of the functional effects of mutations to HIV is critical for understanding this evolution ., HIV’s most rapidly evolving protein is its envelope ( Env ) ., Here we use deep mutational scanning to experimentally estimate the effects of all amino-acid mutations to Env on viral replication in cell culture ., Most mutations are under purifying selection in our experiments , although a few sites experience strong selection for mutations that enhance HIV’s replication in cell culture ., We compare our experimental measurements of each site’s preference for each amino acid to the actual frequencies of these amino acids in naturally occurring HIV sequences ., Our measured amino-acid preferences correlate with amino-acid frequencies in natural sequences for most sites ., However , our measured preferences are less concordant with natural amino-acid frequencies at surface-exposed sites that are subject to pressures absent from our experiments such as antibody selection ., Our data enable us to quantify the inherent mutational tolerance of each site in Env ., We show that the epitopes of broadly neutralizing antibodies have a significantly reduced inherent capacity to tolerate mutations , rigorously validating a pervasive idea in the field ., Overall , our results help disentangle the role of inherent functional constraints and external selection pressures in shaping Env’s evolution .

HIV is infamous for the rapid evolution of its surface protein , Env ., The ability to measure the effects of all mutations to Env under defined selection pressures in the lab would open the door to better understanding the factors that shape this evolution ., However , this is a daunting experimental task since there are over 104 different single-amino acid mutations to Env ., Here we leverage next-generation sequencing to perform a single massively parallel experiment that estimates the effects of all these mutations on viral replication in cell culture ., Our measurements are largely consistent with existing knowledge about the effects of mutations at functionally important sites , and show that inherent mutational tolerance varies widely across Env ., Our work provides new insight into Env’s evolution , and describes a powerful experimental approach for measuring the effects of mutations on HIV phenotypes that can be selected for in the lab .

organismal evolution, microbial mutation, medicine and health sciences, pathology and laboratory medicine, evolutionary biology, pathogens, biological cultures, microbiology, retroviruses, viruses, immunodeficiency viruses, rna viruses, microbial evolution, cell cultures, molecular biology techniques, research and analysis methods, sequence analysis, sequence alignment, bioinformatics, artificial gene amplification and extension, medical microbiology, hiv, microbial pathogens, viral replication, molecular biology, viral evolution, polymerase chain reaction, virology, viral pathogens, database and informatics methods, genetics, gene identification and analysis, biology and life sciences, mutation detection, lentivirus, organisms

journal.pbio.3000135

Uncovering and resolving challenges of quantitative modeling in a simplified community of interacting cells

Successful prediction of quantitative traits of a biological system can be tremendously useful ., For example , if we can quantitatively predict properties of microbial communities , then we will be empowered to design or manipulate communities to harness their activities 1–6 , ranging from fighting pathogens 7 to industrial production of vitamin C 8 , 9 ., An important community-level property is community dynamics , including how species concentrations change over time 6 ., Community dynamics can be predicted using statistical correlation models ., For example , community dynamics observed over a period of time can be used to construct a model that correlates the concentration of one species with the growth rate of another , and the model can then be used to predict future dynamics 10–12 ., However , even for two-species communities , statistical correlation models might generate false predictions on species coexistence 13 ., Alternatively , mathematical models can be constructed based on species interaction mechanisms , such as how metabolites released by one species might affect the growth of another species ., For example , genome-scale metabolic models use genome sequences , sometimes in conjunction with RNA and protein expression profiles , to predict metabolic fluxes within species as well as metabolic fluxes among species ( i . e . , metabolic interactions ) 14 , 15 ., However , these models face multiple challenges , including unknown protein functions or metabolic fluxes 16 ., When interaction mechanisms are known 14 , 17–21 , we can construct a model based on interaction mechanisms ., Ideally , we would use the model to first determine which parameters are critical for the phenomenon of interest , and then directly quantify those critical parameters ., However , parameter quantification can be time-consuming ., Thus , in many models , a fraction of model parameters are “free parameters” ( unmeasured parameters that can be chosen to fit data ) ., Sometimes , a free parameter is assigned a literature value measured in a different strain or even a different species ., This is a poor practice for quantitative modeling , because literature values can vary by orders of magnitude 22 ., Sometimes , a model is “calibrated” or “benchmarked” to fit experimental data 23 , and thus free parameters become “fitting parameters . ”, This type of model calibration can also be problematic , because wrong models can also be calibrated to fit empirical data 23 , and , not surprisingly , the resulting model predictions are likely wrong 23 , 24 ., Even when all parameters are directly measured , quantitative modeling can still be challenging ., First , a parameter measured from a cell population represents the population average and ignores cell-to-cell heterogeneity 25 , which can be problematic ., Second , parameter values may vary with the environment or time 26–29 ., For example , the rate of acetate excretion by Escherichia coli is sensitive to the growth environment 28 , 29 ., Third , during parameter measurements , cells may rapidly evolve , and thus parameters no longer correspond with the intended genotype ., Fourth , in a model with multiple parameters , measurement uncertainty in each parameter can accumulate such that prediction confidence interval is too broad to be useful ., Finally , the correctness or sufficiency of a particular model structure can be questionable ., It is unclear how severe each of these problems can be in empirical examples , nor how to overcome these problems ., As a result , it is not clear how feasible it is to perform quantitative modeling of living systems , including microbial communities ., Here , using a highly simplified community of engineered yeast cells , we stress test quantitative modeling of community dynamics ., Our community “Cooperation that is Synthetic and Mutually Obligatory” ( CoSMO ) 17 consists of two differentially fluorescent , non-mating haploid Saccharomyces cerevisiae strains ( Fig 1A; S1 Table ) ., One strain , designated A−L+ , cannot synthesize adenine ( A ) because of a deletion mutation in the ADE8 gene , and over-activates the lysine ( L ) biosynthetic pathway due to a feedback-resistant LYS21 mutation 30 ., The other strain , designated L−A+ , requires lysine because of a deletion mutation in the LYS2 gene , and over-activates the adenine biosynthetic pathway due to a feedback-resistant ADE4 mutation 31 ., Overproduced metabolites in both strains are released into the environment and are consumed by the partner ., In minimal medium lacking adenine and lysine supplements , the two strains engage in obligatory cooperation and stably coexist 17 , 32 ., The biological relevance of CoSMO is as follows ., First , simplified communities are useful for biotechnology applications 1 , 3 , 33 ., For example , mutualistic communities similar to CoSMO have been engineered to divide up the labor of synthesizing complex drugs 34 ., Second , cooperation and mutualisms modeled by CoSMO are widely observed in naturally occurring communities ( including those in the gut and oral microbiota 35 , 36 ) as microbes exchange essential metabolites such as amino acids and cofactors 37–42 ., Indeed , principles learned from CoSMO , including how fitness effects of interactions affect the spatial patterning of community members , mechanisms that protect cooperators from non-cooperators , and how to achieve stable species composition in two-species communities , have been found to operate in communities of non-engineered microbes 32 , 43–45 ., Because CoSMO has defined species interactions , and because all model parameters can be directly measured , we should be able to quantitatively predict community dynamics ., Our initial model predictions significantly deviated from experimental measurements ., In the process of resolving model–experiment discrepancies , we have uncovered and resolved multiple challenges in parameter quantification , a critical aspect of quantitative modeling ., Because these challenges are likely general , our work serves as a road map that can be applied to quantitative modeling of other cell communities where interaction mechanisms can be inferred from genetic determinants ( see Discussion ) ., Experimentally , CoSMO growth followed a reproducible pattern: after an initial lag marked by slow growth , the two populations and thus the entire community grew at a faster rate ( Fig 1B , “Experiment” ) ., Under optimized experimental conditions , post-lag growth rate reached a steady state ( Fig 7A ) ., We wanted to quantitatively predict CoSMO’s post-lag steady state growth rate ( “growth rate” ) gcomm , the rate of total population increase ., Community growth rate is a measure of how likely the community can survive periodic dilutions such as those in industrial fermenters 46 or during regular bowel movements ., By “quantitative prediction , ” we mean that model prediction should fall within experimental error bars ., We have formulated a differential equation model of the CoSMO dynamics as the following:, dL−A+dt= ( bL ( L ) −dL ) L−A+, ( 1 ), dA−L+dt= ( bA ( A ) −dA ) A−L+, ( 2 ), dLdt=rLA−L+−cLbL ( L ) L−A+, ( 3 ), dAdt=rAL−A+−cAbA ( A ) A−L+, ( 4 ), Eq 1 states that the L−A+ population density ( L−A+ ) increases at a birth rate ( bL ) dependent on the concentration of lysine ( L ) , and decreases at a fixed death rate ( dL ) ., Eq 2 describes how A−L+ population density ( A−L+ ) changes over time ., Eq 3 states that the concentration of lysine ( L ) increases due to releaser A−L+ releasing at a fixed rate ( rL ) , and decreases as the cL amount is consumed per birth of consumer L−A+ ., Eq 4 describes how the concentration A changes over time ., To predict community growth rate , we either simulated community dynamics ( Fig 1B , dotted lines ) or calculated it from an analytical formula ( Eq 5 ) derived from Eqs 1–4 ( see Methods , “Calculating steady state community growth rate” ) :, gcomm≈− ( dA+dL ) 2+rArLcAcL ., ( 5 ), Eq 5 suggests that community growth rate depends on metabolite release rates ( rA; rL ) and metabolite consumption per cell birth ( cA; cL ) in a square root fashion , and depends on death rates ( dA; dL ) in a linear fashion ., Simulations and analytical calculations yielded similar results ( e . g . , S1 Fig ) ., Because death rates are small ( Table 1 ) compared to community growth rate gcomm ( 0 . 11 ± 0 . 01/h in Fig 7B ) , release and consumption parameters are important and should be carefully measured ., Eq 5 also states that even if one parameter is free , its value can always be chosen such that the calculated community growth rate will perfectly match experiments , regardless of the accuracy of the remaining five parameters ., This is the well-known danger of free parameters ., Model parameters correspond to strain phenotypes and include metabolite release rate , metabolite consumption per birth , and cell birth and death rates ., Even though these phenotypes reflect strain interactions ( “interaction phenotypes” in Fig 1A ) , we measured them in monocultures to eliminate partner feedback ., In our earlier studies , we quantified some of these phenotypes and borrowed others from literature values 17 , 32 , 44 ., Our models correctly predicted various properties of CoSMO , including the steady-state strain ratio 17 as well as qualitative features of spatial patterning 32 , 44 ., Our first model ( Model, i ) underestimated community growth rate ., Unlike the published strains of A−L+ and L−A+ in the S288C background 17 , strains in this study were constructed in the RM11 background to reduce mitochondrial mutation rate 48 ., For each RM11 strain , we measured death rate during starvation using a microscopy batch-culture assay 27 ., We also quantified the amount of metabolite consumed per birth in batch cultures grown to saturation ( see Fig 4B for details; S13 Fig ) , similar to our earlier work 17 ., Because release rates were more tedious to measure , we initially borrowed published release rates of L−A+ and A−L+ in the S288C background in batch starved cultures 17 ., Predicted community growth rates were much slower than experimental measurements ( Fig 1B , “Model i”; Fig 7B , gray ) ., A revised model ( Model, ii ) without any borrowed parameters overestimated community growth rate ., For this model , we directly measured the release rates of RM11 L−A+ and A−L+ in batch starved cultures ( see Fig 5B and S15B Fig for details ) ., The release rates of both strains in the RM11 background were approximately 3-fold higher than those in the S288C background ( S2 Table ) ., Consequently , the predicted community growth rate greatly exceeded experiments ( Fig 1B , “Model ii”; Fig 7B , blue ) ., One possible cause for the model–experiment discrepancy could be that cells engineered to overproduce adenine or lysine 30 , 31 might instead release derivatives of adenine or lysine ., Consequently , when we quantified phenotypes such as metabolite consumption , we could have supplemented the wrong metabolite and been misled ., A genome-scale metabolic model of S . cerevisiae predicted that although A−L+ likely released lysine , L−A+ likely released hypoxanthine or adenosine- ( 3 , 5 ) -biphosphate instead of adenine 49 , 50 ., Nanospray desorption electrospray ionization mass spectrometry imaging ( nanoDESI MS ) 51 performed by the Julia Laskin lab revealed a lysine gradient emanating from A−L+ and hypoxanthine and inosine gradients emanating from L−A+ , although the signals were noisy ., We followed up this observation using high-pressure liquid chromatography ( HPLC ) ( Fig 2 ) ., Indeed , lysine mediates the interaction from A−L+ to L−A+ ., We subjected A−L+ supernatant to HPLC ( Methods , “HPLC” ) and a yield-based bioassay ( Methods , “Bioassays” ) ., In HPLC , a compound in A−L+ supernatant eluted at the same time as the lysine standards ( Fig 2A ) , and its concentration could be quantified by comparing the peak area against those of lysine standards ( Fig 2A inset ) ., In bioassay , we quantified the total lysine-equivalent compounds in an A−L+ supernatant by growing L−A+ in it and comparing the final turbidity with turbidities achieved in minimal medium supplemented with various known concentrations of lysine ., HPLC quantification agreed with the yield bioassay ( Fig 2B ) ., Thus , lysine-equivalent compounds released by A−L+ were primarily lysine ., Hypoxanthine mediates the interaction from L−A+ to A−L+ ., When we subjected L−A+ supernatants to HPLC , we found compounds at the elution times of hypoxanthine and inosine , but not of adenine ( Fig 2C ) ., Hypoxanthine but not inosine supported A−L+ growth , and inosine did not affect how hypoxanthine stimulated A−L+ growth ( S3 Fig ) ., Hypoxanthine concentration quantified by HPLC agreed with the concentration of purines consumable by A−L+ in the yield bioassay ( Fig 2D; Methods , “Bioassays” ) ., Thus , A−L+ primarily consumed hypoxanthine released by L−A+ ., Using phenotypes of A−L+ measured in hypoxanthine versus adenine happened to not affect model performance ., Death and release rates were not affected because they were measured in the absence of purine supplements ., Similar amounts of hypoxanthine and adenine were consumed to produce a new A−L+ cell ( S3 Fig ) ., Although the birth rate of A−L+ was slower in the presence of hypoxanthine compared with adenine , especially at low concentrations ( S4 Fig ) , this difference did not affect community growth rate ( Eq 5 ) ., Thus , distinguishing whether hypoxanthine or adenine was the interaction mediator did not make a difference in predicting community growth rate ( S1 Fig ) ., Here , we continue to use A to represent the adenine precursor hypoxanthine ., Model–experiment discrepancy ( Fig 1B ) could be caused by phenotypes being dependent on the environment ., So far , we had measured phenotypes in batch cultures containing zero or excess metabolite ., Thus , we set out to remeasure strain phenotypes in chemostats 52 that mimicked CoSMO environments ., Specifically , in a chemostat , fresh medium containing the required metabolite ( lysine or hypoxanthine ) was pumped into the culturing vessel at a fixed rate ( “dilution rate” ) , while culture medium containing cells exited the culturing vessel at the same rate ( Methods , “Chemostat culturing” ) ., After an initial adjustment stage , live population density reached a steady state ( Fig 3A ) , which meant that the population grew at the same rate as the dilution rate ( Eqs 6–11 in Methods , “Quantifying phenotypes in chemostats” ) 52 ., By setting the chemostat dilution rate to various growth rates experienced by CoSMO ( i . e . , 5 . 5-h to 8-h doubling ) , we could mimick the CoSMO growth environments ., From the population and chemical dynamics in the chemostat , we could then measure metabolite release rate , metabolite consumption per birth , and death rate ( Eqs 12–16 in Methods , “Quantifying phenotypes in chemostats” ) ., During chemostat measurements , ancestral L−A+ was rapidly overtaken by mutants with dramatically improved affinity for lysine ( Fig 3C; S7 Fig; Methods , “Detecting evolved clones” ) , consistent with our earlier work 43 ., These mutants , likely being present in the inoculum at a low ( on the order of 10−6 ) frequency , displayed a growth rate 3 . 6-fold that of the ancestor during lysine limitation ( S7 Fig ) ., Thus , to measure ancestral L−A+ phenotypes , we terminated measurements before mutants could take over ( <10% , before magenta dashed lines in Fig 3 ) ., In contrast , the evolutionary effects of A−L+ mutants on CoSMO growth were captured during phenotype measurements ., Unlike L−A+ mutants , A−L+ mutants were constantly generated from ancestral cells at an extremely high rate ( on the order of 0 . 01/cell/generation; Methods , “Evolutionary dynamics of mutant A−L+” ) , presumably via frequent chromosome duplication ( S8C Fig ) ., These mutants were present at a significant frequency ( 1%–10% ) , even before our measurements started , and slowly rose to 30%–40% during measurements due to their moderate fitness advantage over the ancestor under hypoxanthine limitation ( S8A Fig; S9 Fig; Methods , “Detecting evolved clones” ) ., Consequently , we measured the average phenotypes of an evolving mixture of ancestors and mutants ., Fortunately , these averaged phenotypes could be used to model CoSMO , because mutants accumulated in similar fashions during phenotype measurements and during CoSMO measurements so long as the two time windows were compatible ( S9B Fig; S11 Fig ) ., Metabolite consumption per birth depends on the growth environment ., Consistent with our previous work 17 , consumption during exponential growth in excess supplement was higher than that in a culture grown to saturation ( Fig 4; Methods , “Measuring consumption in batch cultures” ) , presumably due to exponential phase cells storing excess metabolites 54 ., Consumption in chemostats ( Methods , “Quantifying phenotypes in chemostats , ” Eq 12 ) was in between exponential and saturation consumption ( Fig 4C for L−A+ and S13 Fig for A−L+ ) ., For both strains , because consumption in chemostat was relatively constant across the range of doubling times encountered in CoSMO ( 5 . 5–8, h ) , we used the average value in Model iii ( dashed line in Fig 4C and S13 Fig; Table 1; S5 Table , S6 Table ) ., Metabolites can be released by live cells or leaked from dead cells ., We want to distinguish between live versus dead release for the following reasons ., First , if death rate were to evolve to be slower , then live release would predict increased metabolite supply , whereas dead release would predict the opposite ., Second , dead release would imply nonspecific release and , thus cell–cell interactions may be highly complex ., Finally , leakage from dead cells is thermodynamically inevitable , whereas active release of costly molecules would require an evolutionary explanation ., Hypoxanthine is likely released by live L−A+ ., In the absence of lysine ( Methods , “Starvation release assay” ) , we tracked the dynamics of live and dead L−A+ ( Fig 5A , magenta and gray ) and of hypoxanthine accumulation ( Fig 5A , lavender ) ., If live cells released hypoxanthine , then hypoxanthine should increase linearly with live cell density integrated over time ( i . e . , the sum of live cell density * h , Fig 5B , left ) , and the slope would represent the live release rate ( fmole/cell/h ) ., If cells released hypoxanthine upon death , then hypoxanthine should increase linearly with dead cell density , and the slope would represent the amount of metabolite released per cell death ( Fig 5B , right ) ., Because the live release model explained our data better than the dead release model ( Fig 5B ) , hypoxanthine was likely released by live cells during starvation ., In lysine-limited chemostats , we could not use dynamics to distinguish live from dead release ( note the mathematical equivalence between Eqs 9 and 10 in Methods , “Quantifying phenotypes in chemostats” ) ., Instead , we harvested cells and chemically extracted intracellular metabolites ( Methods , “Extraction of intracellular metabolites” ) ., Each L−A+ cell , on average , contained 0 . 12 ( ±0 . 02 , 95% CI ) fmole of hypoxanthine ( Methods , “HPLC” ) ., If hypoxanthine was released by dead cells ( about 105 dead cells/mL , Fig 3A ) , we should see 0 . 012 μM instead of the observed approximately 10 μM hypoxanthine in the supernatant ( Fig 3B ) ., Thus , hypoxanthine is likely released by live L−A+ in chemostats ., Hypoxanthine release rates of L−A+ are similar in lysine-limited chemostats mimicking the CoSMO environments ( Methods , “Quantifying phenotypes in chemostats , ” Eq 14 ) versus during starvation ( Fig 5C ) ., Thus , we used the average hypoxanthine release rate ( Fig 5C black dashed line; Table 1 ) in Model iii ., Note that release rates declined in faster-growing cultures ( ≤3-h doubling; Fig 5C ) , but we did not use these data because CoSMO did not grow that fast ., Lysine is likely released by live A−L+ ., When we measured lysine release from starving A−L+ cells ( S15A Fig ) , a model assuming live release and a model assuming dead release generated similar matches to experimental dynamics ( S15B and S15C Fig ) ., However , after measuring intracellular lysine content , we concluded that dead release was unlikely , because each dead cell would need to release significantly more lysine than that measured inside a cell to account for supernatant lysine concentration , especially during the early stage of starvation ( S16B Fig ) ., Lysine release rate of A−L+ is highly sensitive to the growth environment ( Fig 6B , details in S20 Fig ) and reaches a maximum at an intermediate growth rate ., Release rates in 7–8-h doubling chemostats were about 60% more than those during starvation ., Lysine release rate rapidly declined as hypoxanthine became more available ( i . e . , as growth rate increased , Fig 6B ) ., Variable release rate could be due to variable intracellular lysine content: lysine content per cell increased by severalfold upon removal of hypoxanthine ( from 2 . 9 fmole/cell to about 19 fmole/cell; Fig 6A black dotted line ) and leveled off at a higher level in 8-h chemostats than during starvation ( Fig 6A ) ., We incorporated a variable lysine release rate in Model iii ( Table 1 ) ., Death rates , which could affect CoSMO growth rate ( Methods , Eq 5 ) , are also sensitive to the environment ., We measured death rates in chemostats ( Methods , “Quantifying phenotypes in chemostats , ” Eq 13 or Eq 16 ) and found them to be distinct from the death rates in zero or excess metabolite ( S21 Fig ) ., Because death rates were relatively constant in chemostats mimicking the CoSMO environments ( S21 Fig , blue lines ) , we used the averaged values in Model iii ( Table 1; S7 Table; S8 Table ) ., Our chemostat-measured model parameters are internally consistent: mathematical models of L−A+ in lysine-limited chemostat ( S4 Code ) and of A−L+ in hypoxanthine-limited chemostat ( S5 Code ) captured experimental observations ( S12 Fig; S19 Fig ) ., Using parameters measured in chemostats ( Table 1 ) , model prediction on CoSMO growth rate quantitatively matches experimental results ., Experimentally , because L−A+ mutants quickly took over well-mixed CoSMO ( red in S22A Fig 43 ) , we grew CoSMO in a spatially-structured environment so that fast-growing mutants were spatially confined to their original locations and remained a minority ( red in S22B Fig ) ., Spatial CoSMO growth rates measured under a variety of experimental setups ( e . g . , agarose geometry and initial total cell density ) remained consistent ( 0 . 11 ± 0 . 01/h; Fig 7B purple; S24 Fig ) ., In Model iii , an analytical formula ( Eq 5; Methods , “Calculating steady-state community growth rate” ) and spatial CoSMO simulations based on chemostat-measured parameters ( Table 1 ) both predicted CoSMO growth rate to be 0 . 10 ± 0 . 01/h ( Fig 7B , green and brown ) ., Thus , chemostat parameters allowed our model to quantitatively explain experimental CoSMO growth rate ( Fig 7 green and brown versus purple ) ., This also suggests that our parameter measurements are valid ., Note that although Model iii captures the steady-state growth rate of CoSMO , it fails to recapitulate quantitative details of strain dynamics during and immediately after the initial lag phase ( S26 Fig ) ., This is not surprising , because strain phenotypes during starvation are complex ( e . g . , being time dependent , S18B Fig ) 27 and differ from those in chemostats ( Fig 4; Fig 6; S13 Fig; S21 Fig ) ., In summary , phenotypic parameters are often sensitive to the environment ., Thus , measuring phenotypes in a range of community-like environments may be required for quantitative modeling ., Rapid evolution may further interfere with parameter measurements and model testing ., Only after overcoming these challenges did we succeed in quantitatively predicting the steady-state growth rate of CoSMO ., Microbial communities are complex ., Thus , qualitative modeling has been deployed to yield biological insights 55 , 56 ., However , one would eventually like to understand how community-level properties quantitatively emerge from interactions among member species ., The simplicity of CoSMO has allowed us to directly measure all parameters , uncover some of the challenges to quantitative modeling , and devise means to overcome these challenges ., These challenges are likely general to other living systems ., Below , we discuss what we have learned from quantitative modeling of CoSMO steady-state growth rate ., Even when genetic determinants are known , interaction mediators can be nontrivial to identify ., In CoSMO , we previously thought that adenine was released by L−A+ , whereas in reality , hypoxanthine and inosine are released ( Fig 2 ) ., Fortuitously , hypoxanthine but not inosine affects A−L+ growth ( S3 Fig ) ., Otherwise , we might be forced to quantify how hypoxanthine and inosine , in isolation and in different combinations , might affect A−L+ ., A−L+ grows faster in adenine than in hypoxanthine ( S4 Fig ) , and although this does not affect our prediction of CoSMO growth rate ( S1 Fig ) , it could affect predictions on other community properties ., Many mathematical models have relied on free parameters , which can be problematic when predictions are sensitive to the values of free parameters ., In the case of CoSMO , release rates from two strain backgrounds differed by severalfold ( S2 Table ) , and not surprisingly , borrowing parameters affected prediction ( Fig 1B ) ., A major challenge we uncovered was environment-sensitive parameters ., A key assumption in modeling is invariant parameters ., As we have demonstrated here , phenotypes ( e . g . , metabolite consumption per birth , metabolite release rate , and death rate ) measured in zero or excess metabolite can differ dramatically from those measured in metabolite-limited chemostats ( Fig 4C; Fig 5C; Fig 6B; S13 Fig; S21 Fig ) ., Furthermore , even within the range of metabolite limitation experienced by CoSMO ( doubling times of 5 . 4–8 h ) , lysine release rate varied by as much as 2-fold ( Fig 6B ) , which could be caused by variable intracellular metabolite concentrations ( Fig 6A ) ., Based on parameters measured in chemostats ( including variable lysine release rate ) , Model iii quantitatively predicts experimental results ( Fig 7 ) ., Environment-sensitive parameters make quantitative modeling intrinsically difficult , because community environment often changes with time , and so will environment-sensitive model parameters ., Even if we are only interested in predicting the steady-state community property , we may not know in advance what that steady state is and thus which environment to measure parameters in ., For complex communities , multiple states could exist 57 ., Thus , we may need to measure parameters in a range of environments that are typically encountered in a community ., Another obstacle for model building and testing is rapid evolution ., If we quantify phenotypes in starved batch cultures , cells do not grow and thus evolution is slow , but the environment deviates significantly from the community environment ., In chemostat measurements , we can control the environment to mimic those encountered by the community ., However , in addition to the time-consuming nature of constructing and calibrating chemostats to ensure accurate flow rates 58 , rapid evolution occurs ., For L−A+ , mutants pre-exist at a low frequency but can grow severalfold faster than the ancestor ( S7 Fig ) ., Consequently , a population will remain largely ( >90% ) ancestral only for the first 24 h in the well-mixed chemostat environment ( Fig 3 ) ., A short measurement time window poses additional challenges if , for example , the released metabolite has not accumulated to a quantifiable level ., For A−L+ , mutants are generated from ancestral cells at an extremely high rate before and during phenotype quantification ( Methods , “Evolutionary dynamics of mutant A−L+”; S9 Fig ) ., Because mutants accumulated at a similar rate in CoSMO ( S9B Fig ) , we accounted for evolutionary effects by using similar quantification time windows for A−L+ phenotypes and for CoSMO growth rate ., Note that this approximation is valid here , because our model ( without any free parameters ) matches experiments quantitatively ( Fig 7B ) ., Rapid evolution also poses a problem for model testing ., For example , when quantifying CoSMO growth rate , which requires several days , we were forced to use a spatially structured environment so that fast-growing L−A+ mutants could not take over ( S22 Fig ) ., Thus , unless one is careful , one may not even know what one is measuring !, Rapid evolution need not be unique to our engineered community of “broken” strains ., Indeed , rapid evolution has been observed in phage–bacteria communities in aquatic environments and in acidophilic biofilms 59–61 ., Rapid evolution is not surprising: given the large population sizes of microbial populations , mutants can pre-exist 62 ., These pre-existing mutants can quickly take over in novel environments ( e . g . , exposure to evolving predators or to man-made pollutants and drugs ) where the ancestor is ill adapted ., Choosing the right level of abstraction is yet another important consideration during model building , because different levels of abstraction show trade-offs between generality , realism , and precision 63 ., When the level of abstraction is chosen properly , even complex biological phenomena can be described by simple and predictive equations ., For example , a simplified model considering negative feedback regulation of carbon intake in E . coli quantitatively predicted cell growth rate on two carbon sources based on growth rates on individual carbon sources using only one single parameter that is fixed by experiments 64 ., For CoSMO , one could construct a complex model that , for example , considers physiological and genetic networks of each cell to account for the dependence of phenotypes on the environment and on evolution ., However , this would require making numerous assumptions and measuring even more numerous parameters ., In the absence of free parameters , quantitative matching between model predictions and experimental results provides strong evidence that no additional complexity is required to explain the biological phenomenon of interest ., Once the right level of abstraction is chosen , a good model can serve multiple purposes 65–67 , especially when coupled with quantitative measurements ., First , a model suggests which parameters need to be carefully measured ., For example , for spatial CoSMO growth rate , parameters such as diffusion coefficients are not critical ( S23 Fig ) , but metabolite release and consumption parameters are ( Eq 5 ) ., Second , a useful model not only explains existing data but also makes extrapolative predictions accurately ., An example is the quantitative theory of the lac operon in E . coli ( 68–70 ) ., Finally , model–experiment discrepancy exposes knowledge gaps ., When predicting CoSMO growth rate , the missing piece was environment-sensitive phenotypes ., Our approach can be applied to communities where interaction mechanisms can be inferred from genetic analysis ., For example , we have applied this approach to understand an evolved metabolic interaction ., Specifically , we observed that a single yeast population evolutionarily diverged into two genetically distinct subpopulations 71 ., One subpopulation acquired a met− mutation that prevented the synthesis of organosulfurs and thus must rely on the MET+ subpopulation for organosulfurs ( which are essential for viability ) ., Similar to this work , we first identified the released organosulfurs to be mainly glutathione and glutathione conjugates , using liquid chromatography–mass spectrom

Introduction, Results, Discussion, Methods

Quantitative modeling is useful for predicting behaviors of a system and for rationally constructing or modifying the system ., The predictive power of a model relies on accurate quantification of model parameters ., Here , we illustrate challenges in parameter quantification and offer means to overcome these challenges , using a case example in which we quantitatively predict the growth rate of a cooperative community ., Specifically , the community consists of two Saccharomyces cerevisiae strains , each engineered to release a metabolite required and consumed by its partner ., The initial model , employing parameters measured in batch monocultures with zero or excess metabolite , failed to quantitatively predict experimental results ., To resolve the model–experiment discrepancy , we chemically identified the correct exchanged metabolites , but this did not improve model performance ., We then remeasured strain phenotypes in chemostats mimicking the metabolite-limited community environments , while mitigating or incorporating effects of rapid evolution ., Almost all phenotypes we measured , including death rate , metabolite release rate , and the amount of metabolite consumed per cell birth , varied significantly with the metabolite environment ., Once we used parameters measured in a range of community-like chemostat environments , prediction quantitatively agreed with experimental results ., In summary , using a simplified community , we uncovered and devised means to resolve modeling challenges that are likely general to living systems .

A crown jewel of any scientific investigation is to make accurate and quantitative predictions based on mechanistic understanding of a system ., Although quantitative prediction has been the norm and the expectation in physical sciences , living systems are notoriously difficult to predict quantitatively ., One of the major challenges is obtaining model parameters ., Choosing model parameters to fit data often results in a model that can explain the fitted data but not predict new data ., When modeling cells , “guessing” phenotype parameters or “borrowing” parameters from a different genetic background can be highly problematic ., In addition , phenotype parameters can vary significantly over time , as cell physiology changes with changing environments , or as cells evolve rapidly ., Thus , although parameters are assumed to be constant in most models , this is a far cry from reality and may interfere with quantitative prediction ., Here , using a simple engineered yeast community as our case example , we demonstrate that quantitative modeling is possible , but only after overcoming multiple difficulties in parameter measurements ., Our approach should be generalizable for modeling communities of interacting cells for which genetic or chemical information of interaction mechanisms is available .

death rates, flow cytometry, cell physiology, cell death, chemical compounds, cell processes, cloning, nucleotides, organic compounds, cell metabolism, metabolites, basic amino acids, amino acids, molecular biology techniques, population biology, research and analysis methods, proteins, chemistry, molecular biology, adenine, spectrophotometry, biochemistry, population metrics, cytophotometry, organic chemistry, cell biology, lysine, biology and life sciences, physical sciences, metabolism, spectrum analysis techniques

An attempt to model the behaviour of a cooperative community of two engineered yeast strains shows that quantitative modelling of even a simplified living system requires a Herculean effort for parameter quantification.

journal.pcbi.1003197

Transcriptional Regulation of Lineage Commitment - A Stochastic Model of Cell Fate Decisions

Understanding how primary stem and multipotent progenitor cells decide their fate is pivotal in studying mechanisms driving tissue development and maintenance in multicellular organisms ., Despite considerable advances in ascribing key genes and regulatory circuits to specific lineages , the diversity of molecular mechanisms employed by individual cells to commit to particular lineage fates remains largely uncharacterized ., Recent technical developments in quantitative measurements of single-cell gene expression 1 , 2 have revealed stem and progenitor cell populations to be highly heterogeneous , and suggest that individual cells can exhibit transient biases towards different lineages , even in clonal populations 3–10 ., This molecular heterogeneity may result from stochastic fluctuations caused by noisy gene expression 11 , leading to fluctuations in individual mRNA molecule transcription and degradation rates , and likewise for protein production in individual cells 12 , 13 ., Also , genes switch between active and inactive states , alternating between variable-length transcriptional bursts that can produce a large number of mRNA molecules , and refractory periods in which transcription is significantly reduced 14 , 15 ., Molecular mechanisms of commitment have been suggested to involve various degrees of gene expression coordination , from activation of a few genes 16 to gradual accumulation of a transcriptome-wide coordinated program 17 ., Finally , the role of external cues ( e . g . growth factors ) in commitment remains unresolved , with a long-standing debate on whether they can instruct cells to commit to a particular fate , or do merely act as survival factors of cells that have committed through intrinsic mechanisms 18 , 19 ., A considerable hurdle in elucidating these questions is the elusive nature of the lineage commitment transition , which confounds the experimental capture of cells undergoing commitment ., Recent advances in microscopy and imaging techniques enabled the tracking of single cells in time 20 ., However , the ability of such methods to simultaneously track expression of multiple genes at the single molecule level is still limited , more so for endogenous genes , which may have a role in effecting commitment decisions 2 ., Additionally , the molecular heterogeneity of individual committed cells poses a challenge for defining the relative contributions of single regulators , both individually and in combination , to transitions ., In this work we follow an integrative approach aiming at computationally modeling the stochastic dynamics of lineage commitment of individual multipotent progenitor cells ., We do so using static gene expression profiles of individual self-renewing ( SR ) , erythroid-committed progenitors ( CP ) and erythroid-differentiated ( Ediff ) cells , obtained from the bone marrow-derived multipotent hematopoietic cell line EML , for a panel of genes putatively relevant for erythroid and myeloid lineage development ( Methods ) 21 ., We first perform an exploratory analysis of the static gene expression data , which provides insight into relevant features of the multipotent and committed progenitor populations as well as the SR-to-CP transition ( Figure 1 - top panel ) : Based upon these results , we implement a novel expansion of the random telegraph model of transcriptional bursting 15 , 22 that provides a framework for stochastic commitment as a function of mechanistic aspects of gene expression dynamics ( Figure 1 - middle panel ) : This integrative approach is based on , and expands upon , recently published single cell expression data from the hematopoietic EML cell line for populations in the vicinity of the erythroid commitment boundary 21 ., We revisit the question of transcriptional program coordination at the outset of lineage specification through correlation analysis and infer putative regulators of the commitment transition ., Additionally , we explore the regimens of transcriptional regulation for these genes in the context of a stochastic model of transcriptional bursting and implement expression-dependent rates of commitment which allow the capture of simulated cells at the moment of transition and the assessment of how mechanistic parameters of gene expression regulation impact on the frequency of commitment ( Figure 1 - bottom panel ) ., The single-cell expression data in 21 is a valuable resource for studying the regulation of commitment transitions as it captures SR and CP cells in direct ontogenic relationship ., Of note , CP cells represent a uniquely early stage post-commitment but are also more molecularly heterogeneous ., In order to focus on molecular programs at the commitment transition boundary , we used a combination of hierarchical clustering and dimensionality reduction methods to identify sub-compartments amongst CP cells ( Figure S1 , Methods ) ., We isolated a minor subset of cells ( CP2 ) that are apparently late in their expression profiles and cluster with Ediff cells ., The remaining CP cells , denoted CP1 , are distinct from SR and Ediff and could not be further subdivided , and are thus used as early-committed CP cells in what follows ., We compared the frequency and level of expression of all 17 individual genes ( Text S1 ) in each of the compartments SR , CP1 , CP2 and Ediff ( Figure S2 ) ., A set of genes displays monotonic increase in frequency and/or average level of expression from SR through Ediff ( e . g . Gata1 ) ; the converse monotonic trend is observed for a smaller set of genes ( e . g . Mpo ) ., Interestingly , other genes have non-monotonic patterns of expression increasing at the SR to CP1 transition , to then decrease during differentiation ( e . g . Gata2 ) , or decreasing from SR to CP2 , to increase in the Ediff compartment ( e . g . Btg2 ) ., Pronounced changes between cell types can suggest functional relevance in commitment and/or differentiation ., We then calculated pairwise Spearman correlations for all genes within the SR and CP1 compartments to assess overall coordination of transcriptional programs at the commitment transition ( Figure S3 , Tables S1 , S2 , S3 , Methods ) ., Despite the choice of an inclusive correlation coefficient cutoff value , SR cells did not show broad gene-to-gene correlation ., Similarly , gene expression in the CP1 population is essentially uncorrelated , with a low number of weak correlations ., In contrast , a highly correlated and interconnected gene network could be observed for Ediff cells ., Of note , Gata1 and Epor , which are critical regulators of erythroid lineage development , are minimally or not at all correlated in SR or CP1 compartments ., Hence , this analysis shows no evidence of significant gene regulatory interactions around or at the point of erythroid commitment within our dataset , consistent with the findings in 21 ., We sought to identify the genes that best distinguish between the SR and CP1 populations , which we assume may function directly or indirectly in the commitment transition ., Using the single-cell expression data for all genes in both compartments , we first used a random forest classifier 23 ( Methods ) and evaluated the importance of each gene for the overall performance ( Figure 2A ) ., In this analysis , Gata2 and Mpo were by far the most important genes , with Gata1 ranking at the top of a second line of predictors ., Classifier performances are commonly measured by the Receiver Operating Characteristics curve ( ROC ) , which provides performance percentages for different discrimination thresholds ., The areas under the ROC curve ( AUC ) , which measure the ability of each gene on its own to discriminate between the two populations ( 1 being perfect and 0 . 5 no better than random ) , are shown in Figure 2B ., Again , Gata2 and Mpo ranked highest , with Gata1 following at the top of a second line of predictors ., The random forest classifier covers both linear and non-linear relations between the input variables ( in our case gene expressions ) and the output class , where linearity represents the weighted sum of the inputs and non-linearity encompasses more complicated relations ( e . g . combinations of products ) ., To investigate the presence of the latter we then explored an artificial neural networks ( ANN ) classifier using Gata1 , Gata2 and Mpo expressions as inputs varying the number of hidden nodes ( Methods ) ., We did not observe a difference in validation performance when comparing non-linear and linear methods , suggesting the absence of more complex relations between the genes ., In other words , for the genes in our dataset , the transition from SR to CP seems to be dominated by independent expression values adding up to a certain threshold with gene-specific weights set by the classifier ( Methods ) ., Furthermore , to confirm the dominance of Gata2 and Mpo when predicting the commitment probability , we trained ANN models with fixed complexity , using all possible combinations of one up to four genes as inputs ., Consistently with our observations , all combinations with the highest cross validation performance included Gata2 and Mpo ( data not shown ) ., Regarding the biological relevance of the three top performing genes , Gata2 is required for development of the blood system 24 , 25 , and regulates the adult stem cell compartment through effects on cell cycle 26 , 27 ., Mpo expression can be detected in multipotent as well as myeloid-restricted cells 28 , 29 ., It constitutes a regulatory hub on which transcription factors such as Runx1 , Pu . 1 and members of the Cebp family converge 30 , 31 ., Gata1 is a master regulator of erythropoiesis capable of reprogramming to the erythroid lineage 32 , 33 , although its requirement in the commitment decision remains unclear 34 , 35 ., In order to explore the stochastic dynamics of gene expression for the putative key commitment-associated genes , we have used a random telegraph stochastic model for transcriptional bursting 15 , 22 ( Methods ) , which provides a mechanistic framework for the non-Poissonian behavior observed in eukaryotic gene expression ( Figure 3A ) ., Considering our previous results , we followed a consensus approach and selected genes that consistently ranked high in all classification methods: Gata2 and Mpo were the two best predictors of the committed state and Gata1 , which also ranked consistently high , is well-described as a master regulator of erythroid differentiation capable of myeloid and lymphoid cell reprogramming to an erythroid fate , making it a likely candidate driver of erythroid commitment ., These three genes have distinct gene expression profiles in SR cells , providing an opportunity to assess how distinct modes of gene regulation can affect fate transitions ., We fitted model parameters for each of the three genes through simulated annealing , followed by grid search optimization , minimizing the error towards the experimentally observed distributions ( Figure 3B ) ., The mRNA decay parameter was fixed for each gene according to published data 36 , 37 ., The best parameter sets reproduce experimental distributions and provide insight into the gene-specific stochastic dynamics of expression , suggesting that the three genes have distinct modes of regulation ( Figure 3C , Table S4 ) ., Gata1 displays short infrequent bursts of transcriptional activity; Gata2 expression is set by short but frequent transcriptional bursts with high mRNA production rate; Mpo is expressed through very long bursts of promoter activity resulting in near-constitutive expression ., We tested the robustness of these parameter sets by exploring different combinations of parameters in the vicinity of the optimum solutions ( Figure 4 , Methods ) ., Given its low frequency of expression , the Gata1 distribution can be reconstituted by a fairly broad range of parameters and sensitivity is highest to parameters governing promoter activity ., In contrast , the parameter space for Gata2 is constrained to a smaller region around optimum values , with a clear positive correlation between mRNA production and promoter inactivation times ., Finally , for Mpo the most important parameter is mRNA production time , with a very narrow region of tolerance around the optimal value ., Overall , these results suggest that the observed gene expression distributions for the three genes may be governed by different regulatory mechanisms: Gata1 primarily by promoter activity , Mpo primarily by mRNA dynamics and Gata2 by both ., We selected the best set of parameters that describe the stochastic dynamics of expression for each of the three genes , and expanded upon the initial model to take into account the probability of a cell to commit as a function of gene expression ., Our stochastic model includes an expression-specific commitment rate , proportional to the probability of commitment ( Methods ) ., This probability is given by an expression-dependent logistic regression model trained with experimental data , that separates SR from CP populations ., The proportionality constant was set to reproduce the average commitment rate inferred from culture reconstitution assays ., The logistic regression model captures all relationships between genes , given that non-linear relationships seem to be absent ( see classifier analysis above ) ., This simple model for commitment focuses on the experimental data and abstracts the underlying complexity , weighing the importance of individual genes , as well as their combined effects ., Since we could not find significant correlations within the SR population suggesting regulatory interactions , we assumed complete independence in the stochastic dynamics of each gene ., For most gene expression combinations , the corresponding commitment probability is low , consistent with the fact that commitment is a rare event ( Figure 5A ) ., However , for a small subset of expression states , the probability increases sharply ., Due to the stochastic nature of the system , we can still observe instances where high probabilities do not lead to commitment , as well as others where commitment happens despite low probabilities ., Our modeling approach generated a population of in silico-committed cells , and we compared their expression of Gata1 , Gata2 and Mpo at the moment of transition against experimentally observed values in SR and CP1 cells ( Figure 5B ) ., In silico CP cells are located at the edge of the SR population and share some characteristics with experimental CP1 ., In particular , simulated CP cells can recapitulate expression patterns specific to experimental CP1 and absent from SR cells , such as absence of Mpo in the presence of Gata1 and Gata2 ., Events of in silico commitment occur more often with high values of Gata2 and Gata1 , and indeed , absence of Gata2 does not seem compatible with CP status ., Nevertheless , cells can commit both experimentally and in silico with low levels of Gata2 and in the presence of Mpo , if Gata1 is also present ., Given the stochastic nature of the commitment transition , it is possible for cells with commitment-permissive expression profiles not to effect commitment ( Figure 5C ) ., It is also possible for cells to commit as soon as they enter a commitment-permissive state , and to do so with different kinetics ( Figure S4 ) ., Overall , the data are compatible with the existence of multiple transcriptional routes into lineage commitment ., We assessed how graded changes in the parameters governing gene expression regimens affect the frequency of transition to the committed state ( Figure 6 ) ., Strongest effects are observed upon perturbation of mRNA processing parameters ( production and decay ) , particularly for Gata2 , whereas similar perturbations at the level of promoter activity state do not seem to cause major commitment frequency changes ., This suggests that for Gata2 as for other putative regulators of lineage commitment with similar expression profiles , mRNA dynamics may play a more important role than the regulation of promoter status ( e . g . through histone modifications ) in influencing the commitment transition ., Such subtle changes at this level of gene regulatory mechanisms are seldom feasible in a tightly controlled manner within experimental settings ., Instead , gain- or loss-of-function experiments are more often used to assess the functional relevance of a given gene , involving much more pronounced expression increase or decrease , respectively ., In this context , we used our stochastic model to predict the impact of pronounced Gata1 expression changes in the frequency of commitment in the EML model cell system ., Despite Gata1s capacity to reprogram cells to an erythroid fate through ectopic expression under a strong exogenous promoter 32 , 33 , our model suggested a less prominent though relevant role under its native expression regime , and we wished to test the consequences of enforcing its expression both in silico and in vitro ., To this end , we set the parameter for Gata1 to an infinite value , thus effectively keeping its promoter permanently in the active state ( Figure 7A ) ., The range of simulated values for Gata1 expression in this perturbation scenario is comparable to wild type , but the fraction of high-expressing cells is greatly increased ( Figure 7B ) ., The gene expression time-course reflects the permanent activity of the Gata1 promoter resulting in more frequent high commitment probability peaks as compared to wild type ( Figure 7C and Figure 5A ) ., These changes result in a 2-fold predicted increase in frequency of commitment from wild type to the Gata1 ON perturbation ( Figure 7D ) ., In order to test these results experimentally , we transduced EML SR cells with a GATA1-ERT fusion construct 32 , activated the resulting protein with a pulse of tamoxifen ( Methods ) , and assessed the status of the activated cells in clonal culture-reconstituting assays ( Figure 7E ) ., Importantly , we were able to recapitulate the 2-fold increase in commitment predicted by our model ( Figure 7F ) ., Overall , the data supports the in silico predictions of our stochastic model of commitment and attests to its utility in exploring alternative expression regimens at the transition between self-renewal and lineage commitment ., Our stochastic Monte Carlo model approach is to our knowledge novel ., It integrates the random telegraph model framework 15 , 22 with commitment probabilities obtained from single cell classifiers and cell culture properties ., Also , the robust conversion of static expression data , where each data point is considered a “snapshot” , into time series parameters is new in this context ., In 38 cell cycle FISH data were analyzed with the same goal using template matching ., Our approach , which can be expanded to a larger number of genes and extended to instances where regulatory interactions are present , provides insight into the mechanistic aspects underlying stochastic gene expression and , more importantly , establishes a link between such mechanisms and functional properties of individual cells , by assessing the relevance of promoter and mRNA regulation dynamics in the frequency of commitment ., The computational framework was designed and implemented using single cell expression data observations from different populations of the EML hematopoietic cell line 21 ., Clustering analyses distinguished cellular sub-compartmentalization from molecular heterogeneity within the CP population and identified subsets of early ( CP1 ) and late ( CP2 ) committed cells , with distinct molecular profiles ., Global characterization of CP1 cells revealed a heterogeneous population dispersed in their individual expression profiles , including absence of known erythroid regulators like Gata1 , Klf1 or Epor in a significant number of cells ., Importantly , we observed only few and weak pairwise correlations between genes in CP1 cells , a pattern that was even more evident amongst SR cells ., Hence , no significant level of gene expression coordination is discernible in the commitment transition , at least not within the gene signature analyzed ., We proceeded to infer potential key commitment regulators using machine learning methods to separate SR from CP1 cells across the commitment boundary ., We identified increase in Gata2 and decrease in Mpo expression as the best predictors of commitment , with changes in a second group of genes , including increase in Gata1 expression , also of some relevance ., Although we cannot directly equate predictors of the commitment event with commitment effectors , we have presumed it likely that those genes that best separate SR from CP1 states play a role in their identity or maintenance , and hence may directly effect or report the decision ., Also , in exploring mechanisms of commitment , we are aware that our data is exclusively transcriptional and , consequently , mechanistic approaches cannot consider the effects of translational mechanisms and protein quantities ., However , protein half-lives for Gata1 and Gata2 , for instance , are similar or even shorter than those of their respective mRNAs 36 suggesting that regulation is in fact dominated by transcriptional events ., Indeed , short half-lives of both mRNA and proteins seem to be a common feature of genes involved in regulatory mechanisms 39 and partially preclude the existence of buffering effects at the protein level , although they cannot account for all translational regulatory events ., A better understanding of the regimens of expression of Gata2 , Mpo and Gata1 and their consequences for the SR-to-CP transition could illuminate specific and global mechanisms of lineage commitment ., Thus , we explored the dynamics of these three genes by fitting the parameters of a stochastic gene expression model to experimentally observed distributions ., These solutions , validated by a local robustness analysis , were taken as strong indicators of the qualitative behavior of the system ., We found the genes to have different regulatory dynamics , compatible with global experimental observations in mammalian genes 40 ., In the case of Gata1 and to some degree Gata2 , the frequency of promoter activity bursts plays a fundamental role; Mpo , on the other hand , is most sensitive to variations in mRNA production times ., These patterns are consistent with measurements in yeast , in which transcriptional bursts were more important for larger variations , whereas smaller variations were mostly attributed to transcription-initiation mechanisms 41 ., We extended the stochastic model to account for commitment events by means of a logistic regression model that maximizes the separation between SR and CP cells; stochastic commitment events were thus the result of, ( i ) the inherent stochasticity resulting from the mechanistic parameters of Gata1 , Gata2 and Mpo regulation , and, ( ii ) the rate of commitment inferred from SR-seeded cell cultures , itself implemented as a random event ., Within this framework , the probability of commitment is very low for the vast majority of the time , with infrequent and short transient peaks at high values ., This behavior bears some resemblance to excitable systems of differentiation 42 ., The extended model allowed us to recreate and capture in silico the moment of commitment ., By analyzing the molecular patterns of simulated cells at the transition , we hypothesize that expression of Gata2 defines two states in SR cells: a commitment-impeded state with low Gata2 expression in which no commitment events were observed; and a commitment-permissive state with high Gata2 expression where multiple entry points into commitment can be reached ., Given the lack of correlations between the expression of Gata2 and other genes , we could not further explore specific molecular mechanisms by which Gata2 can drive cells into commitment ., Nevertheless , we systematically assessed how gradual changes in the stochastic dynamics of gene expression regulation for Gata2 , Mpo and Gata1 influence the frequency of commitment ., Again , changes in Gata2 regulation had the strongest impact , in particular when perturbing mRNA production and decay ., Additionally , we tested the impact of more drastic changes in regulatory parameters , by simulating permanent activity of the Gata1 promoter ., The predicted 2-fold increase in frequency of commitment is in agreement with experimental results measuring loss of culture-reconstitution capacity in clonal assays , and is compatible with the reported role of Gata1 in erythroid differentiation and reprogramming experiments 32 , 33 , 35 ., Taken together , these observations bridge mechanisms of gene regulation and functional impact on lineage commitment , and highlight the role of intrinsic noise in cell fate decisions 43 ., This integrative approach can also be applied to other differentiating systems , generating hypotheses on transcriptional regulation dynamics and its impact on commitment ., The gene expression data ( see Text S1 ) were originally expressed as for each gene i to reference Atp5a1 and linearly transformed to the variable ( 1 ) where 30 is the experimental detection limit ., The variable grows with multiplicity in contrast to ., To confront modeled distributions of multiplicities with measured -distributions , we assumed ( 2 ) where is a gene specific parameter ., This represents an ideal experiment , where abundances double in every amplification cycle , and a single molecule is eventually detected after cycles ., The threshold may be gene specific , depending on properties of the reference reporter used ., Thus , we get ( 3 ) where is a gene specific shift parameter to be fitted together with the model rates ( Table S4 ) ., We should stress that single-cell RT-qPCR data is a relative measure of mRNA abundance for each individual gene analyzed ., Quantification is obtained by measuring the number of amplification cycles needed to detect individual mRNA species above an experimental threshold ., This detection threshold may represent a different number of mRNA molecules for each gene , since the measured relative level depends on gene-specific parameters ( such as amplification efficiency from the initial mRNA molecule number ) as well as on the interrogating primers/probe ., As a consequence , comparisons of single-cell expression levels are internally consistent and can be made between populations for a given gene ( such as presented in Figure S2 ) but do not reliably measure differences between genes in a given population ., The shift parameter , , takes into account gene-specific detection thresholds and unique amplification efficiencies , mapping the number of mRNA molecules in our Monte Carlo simulations onto the experimentally-observed gene expression scale ., Time evolution is performed using the Gillespie MC algorithm 49 on the random telegraph model for transcriptional bursting 14 , 22 ., A given gene i is defined by its promoter state ( 6 ) and multiplicity ., Different actions a can take place: We pick times for potential actions a for each gene i from exponential distributions ( 7 ) where the -parameters are for turning the promotor on , for turning it off , and for production and decay of mRNA respectively ., With representing the different components , we use ( Eq ., 3 ) and the trained logistic regression classifier ( Eqs . 4 and 5 ) to calculate the state-dependent commitment rate as explained with Eq ., 16 and pick a time for potential commitment , with the parameter ., Optimized parameter values are found in Table S4 ., The action with the shortest time is selected and the time spent in the current state is recorded ., Then the state is updated and new times are selected ., After completed simulation , the fraction of time spent in a state is our resulting probability for finding a cell in that state ., The system is thermalized for each new cell by requiring the promotor to turn ON and OFF at least once for each gene ., The dimensionless time scale of the Monte Carlo procedure is related to physical time by inferring the characteristic time of commitment events as a function of gene expression ., This is accomplished in two steps:, i ) the overall commitment rate is inferred through the implementation of a compartmental model describing the dynamics of SR and CP cultures in time , where parameters of cell division and death are fitted to experimental SR and CP cell culture dynamics data ; and, ii ) the expression-specific commitment rate is obtained by combining overall commitment rate with the commitment probabilities given by the logistic regression classifier for a finite set of genes ( Eqs . 4 and 5 ) ., An in-house implementation of the simulated annealing algorithm 50 was used to optimize parameters for the stochastic gene expression model by minimizing the sum squared error between experimental and observed single-cell gene expression distributions ., Optimization was further refined by subsequently performing a local grid search in the vicinity of the best parameter sets .

Introduction, Results, Discussion, Methods

Molecular mechanisms employed by individual multipotent cells at the point of lineage commitment remain largely uncharacterized ., Current paradigms span from instructive to noise-driven mechanisms ., Of considerable interest is also whether commitment involves a limited set of genes or the entire transcriptional program , and to what extent gene expression configures multiple trajectories into commitment ., Importantly , the transient nature of the commitment transition confounds the experimental capture of committing cells ., We develop a computational framework that simulates stochastic commitment events , and affords mechanistic exploration of the fate transition ., We use a combined modeling approach guided by gene expression classifier methods that infers a time-series of stochastic commitment events from experimental growth characteristics and gene expression profiling of individual hematopoietic cells captured immediately before and after commitment ., We define putative regulators of commitment and probabilistic rules of transition through machine learning methods , and employ clustering and correlation analyses to interrogate gene regulatory interactions in multipotent cells ., Against this background , we develop a Monte Carlo time-series stochastic model of transcription where the parameters governing promoter status , mRNA production and mRNA decay in multipotent cells are fitted to experimental static gene expression distributions ., Monte Carlo time is converted to physical time using cell culture kinetic data ., Probability of commitment in time is a function of gene expression as defined by a logistic regression model obtained from experimental single-cell expression data ., Our approach should be applicable to similar differentiating systems where single cell data is available ., Within our system , we identify robust model solutions for the multipotent population within physiologically reasonable values and explore model predictions with regard to molecular scenarios of entry into commitment ., The model suggests distinct dependencies of different commitment-associated genes on mRNA dynamics and promoter activity , which globally influence the probability of lineage commitment .

Stem cells have the capacity to both self-renew and differentiate into specialized cell lineages , thus sustaining tissue formation during embryonic development and permitting tissue homeostasis throughout adult life ., Previous studies have suggested that stem cell commitment to a specific lineage may constitute a discrete event of stochastic activation of a small number of key regulator genes ., Experimental exploration of this question is challenging , in face of the elusive nature of the commitment transition and due to considerable gene expression heterogeneity between cells ., Here , we implement a computational model that simulates gene expression variation through time and affords the capture of in silico commitment events ., This model integrates statistical analysis of experimental single-cell gene expression data with dynamical modeling methods to implement a mechanistic framework for stochastic regulation of gene transcription and a probabilistic approach for the commitment rules ., Applied to blood cells , our method identifies potential commitment-associated genes , explores how their expression patterns can define alternative commitment regimes , and suggests how differences in regulation of gene expression dynamics can impact the frequency of commitment .

systems biology, developmental biology, biology, computational biology, molecular cell biology

journal.pcbi.1002576

Deceleration of Fusion–Fission Cycles Improves Mitochondrial Quality Control during Aging

Mitochondria are double-membrane enclosed organelles that fulfill a number of essential cellular roles including oxidative phosphorylation , thermogenesis , iron-sulfur cluster biogenesis , biosynthesis of heme , certain lipids and amino acids , and regulation of apoptosis ., They are semiautonomous organelles which are depending on the expression of both , nuclear as well as mitochondrially encoded genes ., For example , in humans the mitochondrial DNA ( mtDNA ) which exists in several 100 s to 1000 s copies per cell encodes two rRNAs , a set of tRNAs and 13 subunits of the electron transport chain ( ETC ) ., About of all mitochondrial proteins are encoded by the nucleus , synthesized in the cytoplasm and imported into mitochondria ., Mitochondrial dysfunction is linked to a number of human disorders including neurodegenerative diseases , myopathies , obesity , diabetes , and cancer ( for review see 1 , 2 ) ., Moreover , several theories aiming to explain aging in eukaryotes ascribe a crucial role to mitochondria ., One hypothesis dominating the last decades , known as the ‘mitochondrial free radical theory’ ( MFRT ) of aging proposed by Harman 3 states that reactive oxygen species ( ROS ) , predominantly generated within the mitochondrial ETC , cause molecular damage in a cumulative manner ., Later refinements of this theory suggest a vicious cycle to occur since mtDNA encoding essential subunits of the ETC are damaged by ROS 4 ., Although the aforementioned ‘refined’ MFRT is currently hotly debated and specifically the existence of a vicious cycle is unclear 5–9 , it is undisputed that oxidative stress contributes to and mitochondrial dysfunction is involved in aging processes ( for reviews see 6 , 10–12 ) ., However , the vast majority of studies undertaken so far have addressed mainly how oxidative stress is generated within mitochondria , what are the cellular and molecular targets being damaged , and how these processes contribute to aging ., A different point-of-view has gained increasing attention recently , namely , what are the molecular mechanisms that reduce the amount of mitochondrial dysfunction and ROS formation ., Thus , rather than focusing on the formation of molecular damage our main interest has now shifted to the mechanisms ensuring its removal which appears equally important to the aging process ., Mitochondrial quality control has recently been linked by several studies to the astonishing dynamic organization of the mitochondrial network 13 , 14 and to the selective removal of dysfunctional mitochondria by mitophagy 15–21 ., Mitochondria constantly undergo fusion and fission events 13 , 14 ., Impairment of the dynamic behavior is linked to a range of neurodegenerative diseases and aging 22–34 ., Mitochondrial fusion was proposed as a mechanism primarily mediating content mixing and by that allowing inter-mitochondrial complementation compensating for missing or dysfunctional gene products of individual mitochondria 35 , 36 ., Impairment of mitochondrial function was shown to strongly inhibit mitochondrial fusion both in yeast as well as in mammalian cells 16 , 17 , 37 ., Consequently , under bioenergetically compromised conditions , the tubular network of mitochondria transforms into individual , spatially separated , mitochondria exhibiting a rather fragmented appearance ., The selective loss of the fusion capacity of damaged mitochondria , but not of functional ones , could act as a mechanism to distinguish non-functional from functional mitochondria on a morphological basis 16–18 ., This in turn might prevent or minimize further molecular damage as , first , dysfunctional mitochondria are spatially separated from the intact mitochondrial network , and , second , the smaller size itself might be prerequisite for their selective removal by autophagy a process known as mitophagy ., Indeed , impairing mitochondrial fission in mammalian cells inhibited mitophagy and , conversely , inhibiting autophagy increased the levels of depolarized mitochondria and led to reduced maximal respiration 19 ., Taken together , these measurements indicate that fusion occurs selectively between functional mitochondria and that fission acts as to continuously separate mitochondria from each other facilitating the isolation and the subsequent removal of damaged , non-fusogenic mitochondria ., The mitochondrial life cycle thus appears to represent an efficient mechanism ensuring the molecular quality of the cellular ensemble of mitochondria 38 ., A possible link of such a quality control system to aging was also discussed in more detail recently 39 ., Although the general view that mitochondrial dynamics and mitophagy are beneficial for maintaining mitochondrial integrity is supported by numerous studies , it should be noted that basically all studies in mammalian cells rely on tumor cell lines and/or on conditions in which mitochondrial functions are affected by non-physiological conditions ( e . g . , overexpression of PARKIN and the addition of the uncoupler CCCP ) ., It is by far not clear whether all conclusions derived from such systems can indeed be transferred to primary tissues or entire organisms ., This concern is further strengthened by findings demonstrating that mitochondrial dynamics is tightly linked to the cell cycle as e . g . hyperfusion of mitochondria was found to be required for progression from G1 to S phase and mitochondrial fission is promoted during mitosis 40–42 ., Unfortunately , the entire field is lacking quantitative data on the rates of mitochondrial fusion and fission as well as of mitophagy in primary tissues under normal and pathological conditions ., Moreover , there are a number of observations that clearly appear counterintuitive considering such a mitochondrial quality control system ., For example , the pathogenesis of certain mitochondrial diseases caused by mtDNA mutations ( e . g . , MELAS , mitochondrial encephalomyopathy , lactic acidosis , and stroke-like episodes syndrome; MERRF , myoclonic epilepsy with ragged red fibers syndrome ) is not well understood ., These diseases develop progressively and the severity of the symptoms is positively correlated to age and degree of heteroplasmatic prevalence of mutated mtDNA molecules ( for review see 1 ) ., However , the mutational load appears to be highly variable between different cells , tissues , or even between mother and affected offspring ., Clonal expansion of various mtDNA haplotypes in higher eukaryotes is frequently observed 12 , 43–46 ., Another counterintuitive observation is that in a cellular model of aging the rate of both fusion and fission events become reduced by more than in aged cells compared to young cells 47 ., As aging is well known to be associated with the accumulation of molecular damage , an increase in mitochondrial dynamics would be expected to better cope with this situation ., Moreover , it was shown that ablation of mitochondrial fission extends the life span of the two fungal species , Podospora anserina and Saccharomyces cerevisiae 34 ., These experiments reveal that a reduced fission rate retards aging without impairing fitness and fertility of the mutants which also appears incompatible with a fission-dependent mitochondrial quality control mechanism ., Taken together , there are a number of apparent discrepancies suggesting that mitochondrial quality control and its age-dependent regulation might be more complex than initially assumed ., This prompted us to apply a systems biology approach which aims to integrate and to challenge current views in mitochondrial biology during aging into a probabilistic model ., This model considers alterations of mitochondrial dynamics as an adaption to the accumulation of molecular damage in aging cells and thus differs from the Monte Carlo approach 48 or the ‘organelle control’ theory 49 reported previously ., We investigated the hypothesis that mitochondrial dynamics itself could be harmful , for example in situations when molecular damage has already accumulated to some degree and further content mixing would lead to an infection-like phenomenon causing molecular damage to spread across the entire mitochondrial population ., Thus , the fission processes itself could , in addition to its beneficial effect on mitochondrial quality control under certain conditions impose molecular damage to mitochondria ., This is supported by the observation that mitochondrial membrane potential was reduced in a significant fraction of mitochondria subsequent to a fission event in mammalian cells 19 ., Our simulations suggest that a reduction in mitochondrial dynamics rather than being merely a sign or even putative cause of aging , may actually reflect a systemic adaptation to slow down the molecular damage and its propagation by cycles of fusion and fission to prolong cellular function and organismic life span ., Here we present and discuss this concept as the ‘mitochondrial infectious damage adaptation’ ( MIDA ) model ., We derive the master equation for the time evolution of the probability distribtuion by considering all possible processes that contribute to the probability of finding mitochondria in quality state at time : ( 1 ) The first term on the right-hand side accounts for the case where the system is in quality state at time with probability and remains in this state during the time-interval ., The probability refers to the transition from state at time to during the time interval ., For , denotes the probability of making no transition during ., The second term on the right-hand side in Eq ., ( 1 ) refers to all possible transitions during from quality states to with ., Representing by its counter transition probability , , we can rewrite Eq ., ( 1 ) as: ( 2 ) After division of both sides by the time interval , we obtain in the limit the master equation for the time evolution of in the generic gain-loss structure: ( 3 ) Here , denotes transition rates that are to be determined for each process of the mitochondrial dynamics ., In what follows , we determine the contributions for each of these processes separately and then add them up to numerically solve the master equation for ., Starting from a normalized probability distribution at , the norm ( 4 ) is conserved at all times ., Mitochondria undergo cycles of fusion and fission by which they exchange their molecular content such that the mitochondrial distribution in quality state-space is altered ., As the time between fusion and fission of two mitochondria is typically much shorter than the time that the next fusion event follows a fission event , we consider fusion and fission as paired consecutive events consistent with reported experimental evidence 19 ., This approximation is justified for large time scales on which we focus in this study ., Furthermore , as shown in Fig . 1B , we impose the condition that the total functional quality of the involved mitochondria is conserved ., Thus , two mitochondria in quality state and are fusing and subsequently undergo fission into two mitochondria of functional quality and , such that ., In practice , we start by computing the list of all possible fusion and fission events , , that fulfill the quality conservation condition ., In order to avoid double counting of processes , we consider the quality states in the ranges and ., Changes of the mitochondrial distribution in quality state are now easily identified by searching the list ., As depicted in Fig . 1B , for or , the fusion–fission process ends up with one mitochondrion in quality state from two fusing mitochondria in quality states and ., These processes give rise to the gain term in the master equation for fusion–fission events under quality conservation: ( 5 ) Here , we introduced the fusion–fission rate that may depend on time and on the involved quality states ., Furthermore , we introduced the Kronecker delta , with for and for , to distinguish the different kinds of transitions ., The second term on the right-hand side in Eq ., ( 5 ) refers to the case where mitochondria of quality state , either or , fuse and undergo fission ., These are loss processes of quality state and , therefore , enter the master equation with a minus sign ., In addition , we account for permanently , yet modestly , ongoing damage to molecular constituents of mitochondria over time ., This represents the permanent accumulation of e . g . oxidative damage and hydrolytic degradation of proteins , lipids and nucleic acids , but also includes partial misfolding and aggregation of proteins , which overall cause a gradual decline in the amount of active molecules ., This process is schematically shown in Fig . 1C and the corresponding master equation is given by: ( 6 ) Here , the first term on the right-hand side represents the gain in quality state due to mitochondrial quality stepping down the quality ladder from ., This process requires , which is taken care of by the Kronecker delta ., Similarly , the second term on the right-hand side of Eq ., ( 6 ) describes the process of mitochondria quality stepping down the quality ladder from to , a process which can only occur for ., Thus , natural decline of molecular integrity gives rise to a functional quality-shift of the probability distribution from higher to lower quality states with quality decay rate ., Mitophagy leads to the removal of mitochondria in cells ., At the same time , however , mitochondrial mass becomes renewed by biogenesis ., Depending on the balance between the mitophagy rate and the renewal rate , these process can add up to both a net loss or a net gain in quality state ., The combination of these processes is schematically shown in Fig . 1D and the corresponding master equation reads: ( 7 ) Here , we introduced the dimensionless quantity ( 8 ) that represents a homeostatic rheostat: The rate of mitochondrial biogenesis is continuously adapted until the net change of the total mitochondrial mass vanishes ., This ensures that the norm Eq ., ( 4 ) remains a constant in time , i . e . the total number of mitochondria in the cell is unchanged ., Combining Eqs ., ( 5 ) , ( 6 ) and ( 7 ) yields the master equation for the reference simulation of the mitochondrial life cycle: ( 9 ) This simulation accounts for fusion–fission events , functional quality decay and mitophagy in the presence of mitochondrial biogenesis under homeostatic conditions ., It allows studying the impact of cycles of fusion and fission on the quality maintenance in the ensemble of mitochondria ., Furthermore , it can be extended to account for molecular damage in aging cells ., We extend the reference simulation Eq ., ( 9 ) by accounting for the impact of molecular damage in mitochondria that give rise to a decrease in their functional quality ., As depicted in Fig . 2 , molecular damage is assumed to be the result of two conceptually different types of damage:, ( i ) random molecular damage and, ( ii ) infectious molecular damage ., Random molecular damage is caused by external or internal sources , e . g . by ROS , and occurs randomly in time and to a random degree concerning the severity of the damage ., We use the term infectious molecular damage for impairments induced and/or propagated by cycles of fusion and fission ., The inclusion of this type of damage was simulated for the following reasons ., It was reported that a partial dissipation of the membrane potential occurred upon mitochondrial fission in mammalian cells 19 ., This may also be accompanied by loss of mitochondrial ion homeostasis ., In addition , certain types of molecular damage may exert a dominant-negative mode of function and thus any additional propagation by ongoing content mixing would impair mitochondrial function resembling an infection-like process ., One example of such a dominant-negative-like behavior could relate to the possibility that certain mutant mtDNA molecules have a replicative advantage over wild type mtDNA molecules - a hypothesis that was proposed in a number of studies 43 , 51 , 52 ., Consequently , mitochondria harboring wild type mtDNA molecules can only be outcompeted by the replicative advantage of mutant mtDNA molecules when they were infected by a fusion event with mitochondrial population harboring mutant mtDNA molecule ., In this way a pre-existing mutation is spread over the entire network in a fusion-dependent manner ., In our simulation , infectious molecular damage implies a violation of the quality conservation in the molecular exchange between mitochondria during fusion–fission events ., The mitochondrial distribution in the quality state-space is affected by molecular damage induced by ROS production ., As depicted in Fig . 2A , we consider this damage to occur with random impact on the mitochondrial distribution and the corresponding master equation may be written as: ( 10 ) Here , represents the random rate for leakage of the mitochondrial distribution from quality state to ., Note that this random process is restricted to values of and enters Eq ., ( 10 ) as the gain term ., Similarly , the second term on the right-hand side of Eq ., ( 10 ) represents the loss of mitochondria with functional quality to quality states ., Random molecular damage is considered to reshuffle the probability distribution while conserving its norm , which imposes constraints on the involved rates ., In practice , we perform the computation in a way that ensures the norm conservation by construction ., At each time step , we choose random pairs of quality states with from a uniform distribution with the occurrence rate of damage ., Next , we reduce the probability of the higher-quality state by a random fraction with and simultaneously increase the probability of lower-quality state by this fraction: ( 11 ) ( 12 ) Within the same time step , random pairs reshuffle one after the other under conservation of the norm Eq ., ( 4 ) ., This type of molecular damage is associated with the violation of the strict quality conservation condition during fusion–fission events ., As shown in Fig . 2B , two mitochondria in quality states are fusing and subsequently undergoing fission into two mitochondria in quality states ., Thus , infectious molecular damage gives rise to a net loss of functional quality after fission , with the two mitochondria in quality states and , respectively ., Here , and represent random values of a particular molecular exchange event and are drawn from a uniform distribution in the range and ., In the probabilistic modeling approach , damage in the exchange of molecules between mitochondria is directly associated with fusion–fission events and occurs with rate ., Note that this rate is limited by the fusion–fission rate that corresponds to the maximal rate by which infectious molecular damage can occur ., In practice , the list of all possible fusion–fission events is replaced by , which has to be re-computed at each time step ., Once has been generated , we continue with the calculation of the corresponding master equation Eq ., ( 5 ) on the basis of this list ., The processes of mitochondrial dynamics occur with characteristic transition rates that are specified within a product ansatz ., For example , the fusion–fission rate involves the interaction of two mitochondria whose interaction may depend on the involved quality states and may change with time: ( 13 ) Here , the rate accounts for time-dependent changes of the fusion–fission events , while the selectivity function with accounts for the involved quality state , such that the occurrence rate of fusion–fission events is modulated by the functional quality of the involved mitochondria ., Similarly , we represent the rates depending on a single mitochondrial quality level by ( 14 ) where the index refers to the process of functional quality decay ( ) , mitophagy ( ) , mitochondrial biogenesis ( ) , occurrence of random molecular damage ( ) , and occurrence of infectious molecular damage ( ) ., The product ansatz provides full flexibility with regard to the dependence of the rates on the involved quality state and the current time ., For example , the time-dependence may be specified by a Hill function: ( 15 ) Here , and denote the initial ( ) and final ( ) rate value , respectively ., The monotonic transition between these two values is determined by the Hill exponent and the Hill coefficient , which corresponds to the time point at which the rate attains its half-value ., For the rate is constant in time ., Note that the occurrence rate for infectious molecular damage , , poses a special case ., This rate is limited by the rate for the occurrence of fusion–fission events ., Thus , the upper limit of the Hill function for is set by the time-dependent fusion–fission rate , , such that is not necessarily a monotonic function of time ., As is clear from Eqs ., ( 13 ) and ( 14 ) , the selectivity function modulates the time-dependence of the rate to account for the impact of quality state on the occurrence of the process ., In analogy to the rates , the dependence of the selectivity functions on the quality state may as well be represented by Hill functions , ( 16 ) where , and and denote again the Hill coefficient and Hill exponent , respectively ., We developed an algorithm for the time integration of the master equation with time step ., In all simulations the time step is chosen to be much smaller than the inverse of the largest rate and the stability of the time integration is checked by monitoring the conservation of the norm Eq ., ( 4 ) at each time step ., In all presented simulations , we obtained that is preserved upto at least seven post decimal positions ., Furthermore , time-dependent changes in the probability distribution are monitored by computing the deviation factor ( 17 ) at each time step ., If the system reaches equilibrium , , the deviation factor vanishes , , and the simulation may be stopped ., All presented simulation results in the absence of random processes reached values for below ., It should be noted that the system reaches a flow equilibrium , where the involved processes – such as fusion–fission events , quality decay , mitophagy and biogenesis of mitochondrial mass – are continuously taking place but are balancing each other to yield ., In the presence of random processes , does not vanish but fluctuates around a constant average value that depends on the random strength of these processes ., Thus , monitoring is still useful in order to check that a constant average value is reached in time ., For all presented simulation results in the presence of random processes we typically observe values for well below , where averages of the characteristic quantities do not change anymore , indicating that the system has reached its quasi-equilibrium state ., The readout of the computer simulations includes the time-dependent fraction of mitochondria accumulating in quality state and being non-active: ( 18 ) The fraction of mitochondria remaining active is distributed over quality states and calculated by summation: ( 19 ) While and are quantitive measures of the relative size of active and non-active mitochondria populations in the cell , a qualitative measure is given by the time-dependent average quality ., We compute the average quality of all mitochondria by summation over all quality states: ( 20 ) However , since it turns out that the profile of is typically characterized by a peak at state and a peak at high-quality states , the average quality over all states does not adequately represent this distribution ., Therefore , we also compute the average quality of active mitochondria , ( 21 ) involving only quality states of active mitochondria ., To simulate mitochondrial quality control under standard conditions in silico we applied a probabilistic modeling approach taking into account well-known as well as novel aspects in mitochondrial biology ., This simulation of the mitochondrial life cycle is defined by Eq ., ( 9 ) and accounts for distinct quality states of mitochondria , fusion and fission events , for progressive decline of molecular integrity , and for mitochondrial turn-over ( Fig . 1 ) ., All parameters of the reference simulation were either taken from the literature or were estimated ( Table 1 ) ., The size of the quality state-space was chosen to be and the rates of all processes were set to constant values in time ., In accordance with experimental findings 19 , we set the time between two fusion–fission events to minutes , while the time scales on which mitochondrial quality decay and mitophagy take place were chosen to be orders of magnitude larger ., This is in line with reports showing that turn-over of mitochondrial proteins in mammalian cells occurs with a half-life in the order of days 53–55 ., The selectivity functions are represented by Hill functions according to Eq ., ( 16 ) with the corresponding parameters given in Table 1 ., Although these parameters cannot directly be derived from experimental data , it is reasonable to expect that the selectivity functions resemble the qualitative behavior shown in Fig . 3A ., Quality decay and mitophagy are gradually decreasing for increasing quality states , whereas mitochondrial biogenesis and fusion–fission events are both increasing functions of ., Note that the selectivity functions for mitochondrial mass renewal and fusion–fission events vanish at , representing the fact that dysfunctional mitochondria are incapable of fusing with the intact mitochondrial network ., Further , these mitochondria will undergo mitophagy with a higher probability than functional mitochondria ., The simulation is started from a random distribution that is generated from a uniform distribution over and is shown in Fig . 3B ., At time min it has evolved into the equilibrium distribution ( Fig . 3C ) ., Starting from different initial random distributions , we found that the reached equilibrium distribution in Fig . 3C is robust against these changes ., Similarly , we confirmed that the qualitative simulation results do not depend on the choice of the size of the quality state-space and do not depend on the precise choice of the rate values as well as on the profile of the Hill functions Eq ( 15 ) and Eq ( 16 ) , respectively , for the time-dependence and quality-dependence of the rates ., Changes in these parameters merely affect kinetic aspects of the system , e . g . regarding the time scale on which the system equilibrates , and a discussion on this issue as well as on the impact of each process is provided by the Supporting Information ., The simulations confirm that cycles of fusion and fission represent a reliable mechanism for mitochondrial quality enhancement and maintenance ., The equilibrium distribution revealed that of all mitochondria in the cell reach the highest possible quality state , whereas only populate the state of fully dysfunctional mitochondria at ., It should be noted that the equilibrium distribution represents a steady-state , where all processes are continuously taking place but are balancing each other ., This is shown in Fig . 3D where we plot the rates of all processes normalized to their maximum value ., All rates are constant in time , except for the renewal rate that is dynamically adapting under the imposed homeostatic conditions and attains the equilibrium value min−1 ., The time point at which the renewal rate becomes independent of time is an adequate measure for deciding that the system has reached its flow equilibrium ., In Fig . 3E , we plot the average quality of mitochondria as a function of time , both over all quality states and over active states ( ) yielding ., This number characterizes the equilibrium distribution of active mitochondria with the error bars reaching the size corresponding to the standard deviation of the distribution in Fig . 3C for ., Note that , since the typical profile of the equilibrium distribution is double-peaked at state and at high-quality states , interpretation of the average quality over all states may be misleading , which is why we also present the average quality restricted to active mitochondria ., The proper functioning of the cell will require mitochondria of sufficiently high quality to be present at sufficiently high quantities ., As can be seen from Fig . 3F , in addition to the fact that the average quality of active mitochondria is high , the fraction of active mitochondria reaches the high value of ., Taken together , it can be concluded that cycles of fusion and fission combined with selective removal of dysfunctional mitochondria give rise to large amounts of active mitochondria with high average quality ., All subsequent simulations are started from the equilibrium distribution of the reference simulation ( see Fig . 3C ) as initial distribution ., Next we decided to address whether and how such a system can handle the appearance of random molecular damage e . g . by ROS ., In the presence of random molecular damage , as depicted in Fig . 2A , the master equation reads ( 22 ) where the individual terms are given by Eqs ., ( 9 ) and ( 10 ) , respectively ., Starting computer simulations from the equilibrium configuration of the reference simulation ( see Fig . 3C ) , all parameters of the reference simulation were left unchanged ( see Table 1 ) ., In addition , the occurrence rate of random molecular damage follows a Hill function with parameter values , , , , and the random fraction is set to ( see Eqs ., ( 11 ) and ( 12 ) ) ., In Fig . 4A the normalized rates are plotted as a function of time ., The renewal rate was again observed to dynamically adapt under the imposed homeostatic conditions ., However , due to the presence of random molecular damage , its value was fluctuating and attained the constant average value min−1 ., As is observed in Fig . 4B , the quality of active mitochondria was decreasing to the average value with increased standard deviation of ., Thus , compared to the reference simulation , in the presence of random molecular damage the quasi-equilibrium distribution became broader and had a lower average quality of active mitochondria ., However , in contrast to this modest change in the average quality of active mitochondria , a significant difference was observed for the fraction of active mitochondria in the cell under steady-state conditions ( Fig . 4C ) ., This value was decreased by compared to the reference simulation indicating that random molecular damage can give rise to an inversion: While in the reference simulation of all mitochondria were active and occupied high-quality states , random molecular damage caused a reduction to about of only of all mitochondria to accumulate in the highest quality state ., Conversely , the number of dysfunctional mitochondria with the lowest quality state ( ) increased drastically from about in the reference simulation to here ( Fig . 4C ) ., As dysfunctional mitochondria are modeled to be non-fusogenic , which is consistent with experimental data obtained for mammalian cells 16 , 56 , our simulations imply that the mitochondrial network becomes fragmented ., Taken together , upon random molecular damage we observe the emergence of two major classes of mitochondria: one class characterized by high average quality actively undergoing fusion and fission cycles , and one class of non-active and non-fusogenic mitochondria ., Thus , mitochondrial morphology is a good parameter for the relative abundance of these two major classes ., Intrigued by the observation that following mitochondrial fission the mitochondrial membrane potential was diminished in a fraction of the resulting daughter mitochondria 19 we decided to test what implications any fusion–fission dependent damage would have on mitochondrial quality control ., We termed this ‘infectious molecular damage’ as it may not only inflict damage directly caused by fission events but , in addition , it may also represent loss of mitochondrial ion homeostasis , or a propagation of dominant-negative properties ( e . g . , spreading of mutant mtDNA with a replicative advantage over wild type

Introduction, Models, Results, Discussion

Mitochondrial dynamics and mitophagy play a key role in ensuring mitochondrial quality control ., Impairment thereof was proposed to be causative to neurodegenerative diseases , diabetes , and cancer ., Accumulation of mitochondrial dysfunction was further linked to aging ., Here we applied a probabilistic modeling approach integrating our current knowledge on mitochondrial biology allowing us to simulate mitochondrial function and quality control during aging in silico ., We demonstrate that cycles of fusion and fission and mitophagy indeed are essential for ensuring a high average quality of mitochondria , even under conditions in which random molecular damage is present ., Prompted by earlier observations that mitochondrial fission itself can cause a partial drop in mitochondrial membrane potential , we tested the consequences of mitochondrial dynamics being harmful on its own ., Next to directly impairing mitochondrial function , pre-existing molecular damage may be propagated and enhanced across the mitochondrial population by content mixing ., In this situation , such an infection-like phenomenon impairs mitochondrial quality control progressively ., However , when imposing an age-dependent deceleration of cycles of fusion and fission , we observe a delay in the loss of average quality of mitochondria ., This provides a rational why fusion and fission rates are reduced during aging and why loss of a mitochondrial fission factor can extend life span in fungi ., We propose the ‘mitochondrial infectious damage adaptation’ ( MIDA ) model according to which a deceleration of fusion–fission cycles reflects a systemic adaptation increasing life span .

Mitochondria are organelles that play a central role as ‘cellular power plants’ ., The cellular organization of these organelles involves a dynamic spatial network where mitochondria constantly undergo fusion and fission associated with the mixing of their molecular content ., Together with the processes of mitophagy and biogenesis of mitochondrial mass , this results into a cellular surveillance system for maintaining their bioenergetic quality ., The accumulation of molecular damage in mitochondria is associated with various human disorders and with aging ., However , how these processes affect aging and how they can be reconciled with existing aging theories is just at the beginning to be considered ., Mathematical modeling allows simulating the dynamics of mitochondrial quality control during aging in silico and leads to the ‘mitochondrial infectious damage adaptation’ ( MIDA ) model of aging ., It reconciles a number of counterintuitive observations obtained during the last decade including infection-like processes of molecular damage spread , the reduction of fusion and fission rates during cellular aging , and observed life span extension for reduced mitochondrial fission ., Interestingly , the MIDA model suggests that a reduction in mitochondrial dynamics rather than being merely a sign or even cause of aging , may actually reflect a systemic adaptation to prolong organismic life span .

aging, systems biology, developmental biology, organism development, biology, computational biology

journal.pbio.1002112

Natural Selection Constrains Neutral Diversity across A Wide Range of Species

The level of neutral genetic diversity within populations is a central parameter for understanding the demographic histories of populations 1 , selective constraints 2 , the molecular basis of adaptive evolution 3 , genome-wide associations with disease 4 , and conservation genetics 5 ., Consequentially , numerous empirical surveys have sought to quantify the levels of neutral nucleotide diversity within species , and considerable theory has focused on understanding and predicting the distribution of genetic variation among species ., All else being equal , under simple neutral models of evolution , levels of neutral genetic diversity within species are expected to increase proportionally with the number of breeding individuals ( the census population size , Nc ) ., Although this prediction is firmly established , surveys of levels of genetic variation across species have revealed little or no correlation between levels of genetic diversity and population size 6–9 ., This discrepancy—first pointed out by Richard Lewontin in 1974 6—remains among the longest standing paradoxes of population genetics ., One possible explanation for this disagreement is an inverse correlation between mutation rate and population size ., This is expected if there is relatively weak selection against alleles that cause higher mutation rates 8 , 10 ., Alternatively , this paradox could result from greater impact in large populations of nonequilibrium demographic perturbations such as higher variance in reproductive success 11 or population size fluctuations 12 ., Indeed , one recent empirical study suggests that demographic factors play an important role in shaping levels of genetic diversity within animal populations 13 ., However , none of these potential explanations is sufficient to fully account for the observed patterns of neutral diversity across species 8 ., Another potential cause of this paradox is the operation of natural selection on the genome 7 , 14 , 15 ., Natural selection can impact levels of neutral diversity via the adaptive fixation of beneficial mutations ( hitchhiking; HH ) 7 , 15 , 16 and/or selection against deleterious mutations ( background selection; BGS ) 17 , 18 ., Both processes purge neutral variants that are linked to selected mutations , implying that if natural selection is sufficiently common in the genome , it can reduce observed levels of neutral polymorphism ., Furthermore , theoretical arguments 7 , 14 , 19 suggest that , when the impact of natural selection is substantial , the dependence of neutral diversity on population size is weak or even nonexistent ., Although many authors have demonstrated that natural selection could , in principle , be sufficiently common to explain Lewontin’s paradox 7 , 8 , 14–16 , 20 , few direct empirical tests of this explanation exist ., One unique prediction of the hypothesis that natural selection is a primary contributor to disparity between Nc and levels of neutral genetic variation within species is that natural selection will play a greater role in shaping the distribution of neutral genetic variation in species with large Nc ., To test this prediction , we relied on the fact that the impact of natural selection on linked neutral diversity depends on the local recombinational environment ., In regions of relatively low recombination , selected variants affect more neutral sites through linkage , and vice versa , in regions of relatively high recombination ., The resulting correlation between recombination and polymorphism 21–26 ( reviewed in depth in 27 ) allows a quantitative assessment of the magnitude of the impact of selection on linked neutral diversity ( e . g . , 22 , 23 , 26 , 28 ) ., Specifically , if the effects of linked selection can explain the lack of correlation between neutral diversity and population size , we expect that species with larger population sizes will display stronger correlations between recombination and polymorphism than those with smaller population sizes and show a concurrently larger impact of natural selection on levels of neutral diversity across the genome ., Although empirical studies that explore the relationship between neutral diversity and population size are relatively infrequent compared to theoretical studies on this topic , there are two interesting patterns that merit consideration here ., First , the proportion of nonsynonymous substitutions that have been driven to fixation by positive selection varies widely across taxa ., In humans 29 , yeast 30 , and many plant species 31 , estimates of this proportion are close to zero ., In contrast , in Drosophila 32 , 33 , mice 34 , and Capsella grandiflora 35 , as well as other taxa ( reviewed in 8 ) , a large fraction of nonsynonymous substitutions are inferred to have been driven to fixation by positive selection , implying that natural selection is common in the genomes of these organisms ( which generally have large Nc ) ., Second , the strength of the correlation between polymorphism and recombination varies widely among the limited number of taxa 8 , 27 that have been studied in depth ., Here again , Drosophila 21 , 25 , 36 is among the taxa that shows the strongest correlation and thus the clearest evidence for natural selection , and the correlation in Drosophila is substantially larger than , for example , in humans 28 ., In a related study to the work presented here , Bazin et al . 37 showed that there is no correlation between nucleotide diversity in nonrecombining mtDNA and nucleotide diversity in the nuclear genome ., While this is consistent with some predictions of theoretical work on this subject , the mitochondrion has unusual patterns of replication and inheritance , and it is therefore challenging to disentangle the processes that generate diversity from those that shape its distribution across the genome ., Although suggestive , the evidence accrued thus far is fragmentary , has not been analyzed in aggregate , and varies widely in quality of samples , data collection , and analyses performed 8 , 27 ., It is therefore difficult to draw firm conclusions about the relative importance and prevalence of natural selection in shaping patterns of genetic variation in the genome based on existing studies ., Due to rapid advances in genome sequencing technologies , whole genome polymorphism data are now available for a wide variety of species ( e . g . , 36 , 38 ) , and these data enable us to conduct a quantitative test of the natural selection hypothesis as an explanation for Lewontins paradox ., Towards this , we identified 40 species with sufficiently high quality reference genomes , linkage maps , and polymorphism data to enable a broad-scale , robust comparison of the relative strength of correlation between polymorphism and recombination rate within a single unified alignment , assembly , and analysis framework ., Using these data , and reasonable proxies for Nc , we show that the effect of selection on linked nucleotide diversity is indeed strongly correlated with population size ., In other words , natural selection plays a disproportionately large role in shaping patterns of genetic variation in species with large Nc , confirming the idea that natural selection is an important contributor to Lewontin’s paradox ., We identified 40 species ( 15 plants , 6 insects , 2 nematodes , 3 birds , 5 fishes , and 9 mammals ) for which a high-quality reference genome , a high-density , pedigree-based linkage map , and genome-wide resequencing data from at least two unrelated chromosomes within a population were available ( Table 1 , S1 Table , S2 Table ) ., Because our model ( below ) requires that recombination has been sufficiently frequent to uncouple genealogies across large tracts of DNA on chromosomes , we required that each species have an obligatory sexual portion of its life cycle ., This requirement necessarily excludes clades such as bacteria , which are predominantly clonally propagated ., Nonetheless , extending this framework to bacterial taxa will be an important step towards understanding the mechanisms by which natural selection shapes patterns of variation across the tree of life ., Additionally , our sampling is biased towards more commonly studied clades ( e . g . , mammals ) , but this is unavoidable in this type of analysis , and there is no reason in principle why this taxonomic bias would affect the basic conclusions we describe here , as the sampled taxa likely span a large range of census population sizes ., After acquiring sequence data , we developed and implemented a bioinformatic pipeline to align , curate , and call genotype data for each species ( see S1 Fig and methods for a full description of the bioinformatics pipeline ) ., We further used the available genetic maps to estimate recombination rates across the genomes ., Across all species , we analyzed recombination across nearly 385 , 000 markers and aligned more than 63 , 000 , 000 , 000 short reads ., This is therefore one of the largest comparative population genomics dataset that has been assembled to date ., We used both simple nonparametric correlations and explicit coalescent models to test for a relationship between the impact of selection on linked neutral diversity and census size ., Although correlations between recombination rate and neutral diversity are informative , the extensive literature in theoretical population genetics provides an opportunity to develop a robust modeling approach ., Two primary types of selection can introduce a correlation between recombination rate and levels of nucleotide diversity: background selection ( BGS ) and hitchhiking ( HH ) ., Here , we are not primarily concerned with distinguishing between the two models , and so focus on their joint effects ., In addition to combining BGS and HH , we would also like to relax the assumption that these processes act uniformly across the genome ., All else being equal , regions of the genome with a higher density of potential targets of selection should experience a greater reduction in neutral diversity ., Starting from considerable prior theoretical work 14 , 17 , 18 , 32 , 39–41 , we develop an explicit model relating polymorphism , recombination rate , and density of functional elements in the genome ., We fit both a joint model that allows for both HH and BGS , as well as models of BGS only , HH only , and a purely neutral model ( in which there is no predicted correlation between recombination or functional density and neutral diversity ) ., Using these models , we estimate the proportion of neutral diversity removed by linked selection for beneficial alleles and/or against deleterious alleles ( Fig . 1 ) for each species , as well as the relative likelihood of each model ., In practice , it is not feasible to determine Nc for the majority of species we studied ., Instead , we used the species’ geographic range and individual body size as proxies for Nc ., Size has been previously validated as a proxy for individual density in a wide variety of taxa and ecosystems ( e . g . , 42–44 ) ., Under some simplifying assumptions , the product of geographic range and local density should be sufficient to roughly estimate a species census population size , and each factor is expected to independently capture some information related to species’ Nc ., Specifically , we expect that range will be positively correlated with Nc , size will be negatively correlated with Nc , and Nc will be positively correlated with the impact of selection ., For many of the species that we studied , it is clear that selection plays a central , even dominant , role in shaping patterns of neutral genetic diversity ., Specifically , both our correlation analysis and our explicit modeling support the hypothesis that natural selection on linked sites eliminates disproportionately more neutral polymorphism in species with large Nc , and in this way , natural selection truncates the distribution of neutral genetic diversity ., At a coarse scale , there is a stronger correlation between polymorphism and recombination in invertebrates ( mean partial τ after correcting for gene density = 0 . 247 ) , which likely have a large Nc on average , than in vertebrates ( median partial τ = 0 . 118 ) , which likely have a smaller Nc on average ( two-tailed permutation p = 0 . 021 ) ., We observe similar patterns for herbaceous plants ( mean partial τ = 0 . 106 ) versus woody plants ( mean partial τ = −0 . 020; two-tailed permutation p = 0 . 058 ) and for medians as opposed to means ( Fig . 2 ) ., When we repeat the analysis with alternate window sizes , we observe consistent effect sizes , albeit occasionally with reduced statistical support ( S3 Table ) ., More generally , we tested the hypothesis that Nc is positively correlated with the impact of selection by fitting a linear model that includes body size , geographic range , kingdom , and the significant interactions among them as predictors , and uses the impact of selection estimated from our coalescent model as the response variable ( Table 2; Fig . 3 ) ., Both size and range are significant predictors of the impact of selection in the expected directions ( Table 2; log10 ( size ) : coefficient = −0 . 092 , p = 0 . 0005; log10 ( range ) : coefficient = 0 . 112 , p = 0 . 0002 ) , and model as a whole explains 63 . 88% of the variation in impact of selection across species ( Table 2; overall p = 3 . 518 x 10−8 ) ., This is clear evidence that more variation is removed by linked selection from the genomes of species with smaller body size and larger ranges than from the genomes of species with larger body size and smaller ranges ., A number of confounding factors could potentially influence our conclusions , including variation in map or assembly quality across species , differences in overall recombination rate , and differences in genome size ., To test whether these factors can explain our results , we fit a confounder-only model including two measures of genetic map quality ( density of useable markers and proportion of total markers scored as useable ) ; two measures of assembly quality ( proportion of assembly that is not gaps and proportion of total assembly assembled into chromosomes ) ; overall recombination rate; and genome size ., We then compare this confounder-only model to a model that includes all confounding parameters and , in addition , includes our population size proxies ( kingdom , size , and range ) ., The model with proxies for Nc both explains substantially more total variation in impact of selection ( adjusted R2 of 0 . 6359 compared to 0 . 3388 for the confounder-only model ) and is a significantly better fit to the data ( F = 7 . 7322 , df = 4 , p = 0 . 0002 ) ., In order to ensure that variable sampling of chromosomes is not a source of bias ( given that the number of chromosomes sampled ranges from a minimum of 2 to a maximum of 517; S1 Table ) , we tested whether sampling depth is correlated with either size or range ., In neither case do we find a correlation ( size versus sampling depth: Kendall’s τ = 0 . 022 , p = 0 . 84; range versus sampling depth: Kendall’s τ = 0 . 044 , p = 0 . 699 ) ., We also find no evidence that species with only two chromosomes sampled are atypical with respect to range ( Wilcoxon Rank Sum Test , p = 0 . 944 ) or size ( Wilcoxon Rank Sum Test , p = 0 . 423 ) ., Finally , we find no evidence that mean depth per individual is correlated with either size ( Kendalls τ = −0 . 044 , p = 0 . 683 ) or range ( Kendalls τ = −0 . 02 , p = 0 . 862 ) ., Taken together , these results strongly suggest that the variable sampling across species , both in terms of sequencing depth and in terms of number of chromosomes sequenced , does not bias our conclusions ., To get a lower bound on the proportion of variation in impact of selection explained by our parameters of interest ( range , size , kingdom , and the kingdom–size interaction ) , we fit a linear model with these parameters as predictors and the residuals of the confounder-only model as the response variable ( S4 Table , S5 Table ) ., This is a conservative test , as genome size is strongly correlated with body size ( Kendalls τ = 0 . 296 , p = 0 . 007 in our dataset ) ., Nonetheless , our proxies for Nc explain 34 . 05% of the remaining variation in impact of selection after accounting for all confounding parameters ( overall model p = 0 . 0008 , S4 Table ) , and 47 . 36% of the variation after accounting for all confounding parameters except genome size ( overall model p = 2 . 042 x 10−5 , S5 Table ) ., For five species , our polymorphism data included individuals from domesticated populations , which could potentially affect our conclusions if selection has a different signature during domestication events than it leaves in natural populations ., However , removing these five species has virtually no impact on our model fit ( overall adjusted R2 = 0 . 6281 , overall p = 6 . 094 x 10−7 , S6 Table ) , suggesting that their inclusion has not biased our results ., Additionally , we obtain similar results if we fit our model ( excluding the kingdom term and its interaction with size ) to animals and plants independently ( S7 Table , S8 Table ) ., Finally , varying the filtering criteria , window size , assumed deleterious mutation rate ( U ) , or population genetic modeling approach produces nearly identical results ( Fig . 3C ) , implying our primary conclusion is robust to a wide range of analysis choices ., Taken together , our analysis demonstrates that the central pattern—natural selection reduces neutral diversity more strongly in species with large Nc than in species with small Nc—is consistently observed with both nonparametric model free approaches ( Fig . 2; S3 Table ) and with explicit population genetic models ( Fig . 3A , B , Table 2 ) across a wide range of possible analysis and filtering choices ( Fig . 3C , S4–S8 Tables ) ., If the process of recombination is itself mutagenic , neutral processes could produce a correlation between recombination and polymorphism 21 , 25 , 27 ., However , no or very weak correlations between divergence and recombination have been found in most species that have been closely studied 21 , 25 ( reviewed in 27 ) ., Moreover , for those species in which a positive correlation between divergence and polymorphism has been found ( e . g . , 45 , 46 ) , it is likely at least partially the result of linked selection acting on polymorphisms present in the ancestral population 27 , 32 ., Furthermore , the two species that showed the strongest correlation between polymorphism and recombination ( partial τ = 0 . 5196 for D . melanogaster , partial τ = 0 . 4637 for Drosophila pseudoobscura ) have no such correlation between recombination rate and divergence either on broad scales 21 or fine scales 25 ., Finally , many authors have found strong evidence that recombination is not mutagenic in a number of other animal species ( e . g . , 28 , 47 , 48 ) , and it therefore appears a general consensus has emerged that recombination-associated mutagenesis is unlikely to influence the overall patterns we report in this work 27 ., As an alternative approach to estimating the impact of natural selection on linked neutral diversity , we considered whether our proxies for Nc correlate with the strength of evidence that selection shapes patterns of neutral diversity , derived from our population genetic modeling approach ., To do this , we focus on the relative likelihoods ( Akaike weights ) of four models: the BGS+HH model , the BGS-only model , the HH-only model , and the neutral model ., These relative likelihoods can be interpreted as the probability that a particular model is the best model according to Akaike Information Criteria ( AIC ) , given the set of models tested and the underlying data ., We initially focus on the relative likelihood of the support for a purely neutral model ., Species with weak or no support for neutrality ( relative likelihood of the neutral model < 0 . 05 ) have significantly larger ranges ( p = 0 . 006 , Wilcoxon Rank Sum Test , Fig . 4A ) and significantly smaller sizes ( p = 0 . 0001 , Wilcoxon Rank Sum Test , Fig . 4B ) than species with moderate ( relative likelihood of neutral model ≥ 0 . 05 and < 0 . 90 ) or strong ( relative likelihood of neutral model ≥ 0 . 90 ) support ., This pattern also holds if we compare the species with strong support for neutrality or species with moderate support for neutrality individually to species with weak or no support ( moderate versus weak: p = 0 . 0005 for size and 0 . 02 for range; strong versus weak: p = 0 . 02 for size and 0 . 02 for range , all p-values from Wilcoxcon Rank Sum Tests ) ., This suggests that the evidence for non-neutral processes ( BGS and/or HH ) is significantly stronger in species with larger ranges and/or smaller sizes , consistent with our results above and with the hypothesis that natural selection explains Lewontins paradox ., Given the extensive debate on the relative importance of HH versus BGS in shaping patterns of diversity across the genome 17 , 21 , we also attempt to disentangle the relative roles of these two processes in reducing neutral diversity ., This is potentially relevant to the resolution of Lewontins paradox , as models of frequent , recurrent HH ( i . e . , genetic draft 7 ) demonstrate that recurrent HH can remove the dependence of neutral diversity on population size entirely ., Thus , evidence that HH specifically is more likely to occur in species with large census sizes would be compelling evidence for a role of selection in resolving the discrepancy between population sizes and neutral diversity ., However , it is crucial to note that our test does not take into account features , such as patterns of polymorphism around amino acid fixations 23 , 49 , that are particularly powerful for distinguishing HH and BGS , and thus suffers from many of the limitations of previous work relying purely on patterns of neutral diversity across the genome ( e . g . , 26 , 28 , 40 , 41 ) ., With that caveat , we begin by noting that , consistent with recent work in Drosophila 49 , 50 and other organisms 26 , 28 , 48 , background selection is ubiquitous ., Either the BGS-only model or the BGS+HH model has at least some support ( relative likelihood ≥ 0 . 05 ) for 95% ( 38 of 40 ) of the species we analyzed , and for 90% ( 36 of 40 ) of species one of the BGS-containing models was the best fit , as measured by AIC ., Thus , it seems clear that , in most cases , BGS is a more appropriate null model for tests of natural selection than strict neutrality ., To test whether species with moderate ( relative likelihood of HH or BGS+HH ≥ 0 . 05 and < 0 . 9 ) or strong ( relative likelihood of HH or BGS+HH ≥ 0 . 9 ) evidence for HH differ from species with little or no evidence for HH ( relative likelihood of HH or BGS+HH < 0 . 05 ) , we examined our proxies for Nc among these evidence classes ., Species with moderate or strong evidence for HH have significantly larger ranges than species with weak or no evidence for HH ( p = 0 . 03 , Wilcoxon Rank Sum Test , median range ( weak ) = 2 , 681 , 693 sq km , median range ( moderate/strong ) = 5 , 592 , 037 sq km ) , and these species tend to have smaller sizes as well ( p = 0 . 15 , Wilcoxon Rank Sum Test , median size ( weak ) = 0 . 91 m , median size ( moderate/strong ) = 0 . 54 m ) ., As a second test of this pattern , we compared whether the relative likelihood of HH was greater for species estimated to have particularly high Nc compared to species estimated to have particularly low Nc ., We define the high-Nc class as those species with ranges greater than the median range , and sizes below the median size , and we define the low-Nc class as those species with ranges below the median range and sizes above the median size ., The relative likelihood of HH models is greater for species in the high-Nc class than the low-Nc class ( p = 0 . 023 , Wilcoxon Rank Sum Test ) , and the proportion of species with moderate or strong evidence for HH ( either alone or in combination with BGS ) is higher in the high-Nc class than the low-Nc class ( 4/10 in high-Nc class , 0/10 in low-Nc class , p = 0 . 086 , Fishers Exact Test ) ., Despite the fact that our test is unlikely to have substantial power to distinguish BGS and HH models , we suggest that these results imply that HH in particular is a stronger force shaping genomic diversity in species with large Nc , while BGS appears to be much more pervasive ., The observation that pervasive HH may predominantly occur in species with large Nc suggests that genetic draft may play a substantial role in limiting neutral diversity among the species with the largest population sizes ., More data on species with very large Nc , and the application of tests specifically designed to detect HH to a wider taxonomic range , will be necessary to fully disentangle the relative roles of HH and BGS in shaping levels of neutral diversity ., On the strength of early allozyme polymorphism data , Lewontin 6 observed that in contrast with theoretical predictions of the neutral theory 51–53 , the range of neutral genetic variation among species is substantially smaller than the range of Nc among species ., Because both positive selection via HH and negative selection via BGS purge linked neutral mutations , the operation of natural selection affects patterns of neutral genetic variation at linked sites across the genome ., Although many authors have suggested that natural selection may play a role in truncating the distribution of genetic variation and may play a greater role than neutral genetic drift in shaping patterns of neutral nucleotide polymorphism 7 , 8 , 14 , 15 , few empirical tests of this hypothesis have been proposed or conducted ., Here , we show that species with larger Nc display a stronger correlation between neutral polymorphism and recombination rate , and that natural selection removes disproportionately more linked neutral variation from species with larger populations ., This indicates that natural selection plays a disproportionately large role in shaping patterns of polymorphism in the genome of species with large Nc ., One important consideration when interpreting our results is that cryptic population structure can influence patterns of variation across the genome in a way that obscures the effects of selection ., In the extreme case , where populations do not exchange any migrants for an extended period of time , genetic divergence is expected to accumulate at equivalent rates across the genome and would obscure the effects of linked selection ., Elucidating the complex relationship between population structure and patterns of natural selection is an important and longstanding question in population genetics ( for recent work see 54 , 55 ) ., Nonetheless , especially given the scope of our analysis , it is not feasible to simultaneously estimate the effects of linked selection and population structure , and there are many reasons to believe that the results presented here will be robust to potential cryptic population structure ., So long as the population subdivision is not especially ancient ( in the timescale of coalescence , on the order of Ne generations ) , a correlation between recombination and polymorphism is expected to remain due to the effects of selection on linked sites in the ancestral population 27 , 32 ., Additionally , if migration is sufficiently common , it is reasonable to treat data derived from samples from separate localities as a single population 56 ., One straightforward assumption is that species with larger geographic ranges will have greater opportunity on average to accumulate cryptic population structure than species with small ranges , which would imply we should preferentially underestimate the effects of linked selection in species with larger ranges ., If population structure is a primary determinant of patterns of nucleotide diversity in taxa that we studied , we could reasonably expect a negative correlation between species range and the effects of selection on linked sites ., Given that we instead obtain the opposite effect—one consistent with the effect of selection on linked neutral sites—it is reasonable to conclude that cryptic population structure has not drastically influenced the basic results presented herein ., Understanding the proximate and ultimate factors that affect the distribution of genetic variation in the genome is a central and enduring goal of population genetics and it carries important implications for a number of evolutionary processes ., One implication of this work is that in species with large Nc , such as D . melanogaster , selection plays a dominant role in shaping the distribution of molecular variation in the genome ., Among other things , this can affect the interpretation of demographic inferences because it indicates that even putatively neutral variants are affected by natural selection at linked sites ., Furthermore , to whatever degree standing functional variation is also affected by selection on linked sites ( e . g . , 40 ) , local recombination rate in organisms with large Nc may also predict what regions of the genome will contribute the greatest adaptive responses when a population is subjected to novel selective pressures ., More broadly , this work provides direct empirical evidence that the standard neutral theory may be violated across a wide range of species ., Indeed , it is clear from this work that in many taxa , natural selection plays a dominant role in shaping patterns of neutral molecular variation in the genome ., It is therefore essential to consider selective processes when studying the distribution of genetic diversity within and between species ., Incorporating selection into standard population genetic models of evolution will be a central and important challenge for evolutionary geneticists going forward ., Reference genome versions , annotation versions , map references , and other basic information about the genetic and genomic data for species we included in our analysis is summarized in S1 Table and S2 Table , and described in more detail below ., Our approach to estimating recombination rates is to first obtain sequence information and genetic map positions for markers from the literature , map markers to the genome sequence where necessary , filter duplicate and incongruent markers , and finally estimate recombination rates from the relationship between physical position and genetic position ., Specific details of map construction for each species are described in S1 Text ., We begin with the very general selective sweep model derived by Coop and Ralph 41 , which captures a broad variety of HH dynamics ., To include the effects of BGS , we rely on the fact that to a first approximation , BGS can be thought of as reducing the effective population size and therefore increasing the rate of coalescence ., This effect can be incorporated by a relatively simple modification to equation 16 of 41 ., Specifically , we scale N by a BGS parameter , exp ( -G ) , in equation 16 , which then leads to a new expectation of average pairwise genetic diversity ( π ) :, Eπ=θ1/exp ( −G ) +α/rbp, ( 1 ), where α = 2N * Vbp * J2 , 2 ( per 41 ) and rbp is the recombination rate per base pair ., This is very similar to previously published models of the joint effects of background selection and HH ( e . g . , 39 ) ., To account for variation in the density of targets of selection , we build upon the approach of Rockman et al . 40 and Flowers et al . 26 , which derives from the work Hudson , Kaplan , Charlesworth , and others that originally described models of background selection in recombining genomes 17 , 18 ., Specifically , we fit the following model to estimate G for each window i:, Gi=ΣkU*fdi*sh2* ( sh+P|Mi−Mk| ) * ( sh+P|Mi−Mk+1| ), ( 2 ), where U is the total genomic deleterious mutation rate , fdi is the functional density of window i , sh is a compound parameter capturing both dominance and the strength of selection against deleterious mutations , Mk and Mi are the genetic pos

Introduction, Results, Discussion, Materials and Methods

The neutral theory of molecular evolution predicts that the amount of neutral polymorphisms within a species will increase proportionally with the census population size ( Nc ) ., However , this prediction has not been borne out in practice: while the range of Nc spans many orders of magnitude , levels of genetic diversity within species fall in a comparatively narrow range ., Although theoretical arguments have invoked the increased efficacy of natural selection in larger populations to explain this discrepancy , few direct empirical tests of this hypothesis have been conducted ., In this work , we provide a direct test of this hypothesis using population genomic data from a wide range of taxonomically diverse species ., To do this , we relied on the fact that the impact of natural selection on linked neutral diversity depends on the local recombinational environment ., In regions of relatively low recombination , selected variants affect more neutral sites through linkage , and the resulting correlation between recombination and polymorphism allows a quantitative assessment of the magnitude of the impact of selection on linked neutral diversity ., By comparing whole genome polymorphism data and genetic maps using a coalescent modeling framework , we estimate the degree to which natural selection reduces linked neutral diversity for 40 species of obligately sexual eukaryotes ., We then show that the magnitude of the impact of natural selection is positively correlated with Nc , based on body size and species range as proxies for census population size ., These results demonstrate that natural selection removes more variation at linked neutral sites in species with large Nc than those with small Nc and provides direct empirical evidence that natural selection constrains levels of neutral genetic diversity across many species ., This implies that natural selection may provide an explanation for this longstanding paradox of population genetics .

A fundamental goal of population genetics is to understand why levels of genetic diversity vary among species and populations ., Under the assumptions of the neutral model of molecular evolution , the amount of variation present in a population should be directly proportional to the size of the population ., However , this prediction does not tally with real-life observations: levels of genetic diversity are found to be substantially more uniform , even among species with widely differing population sizes , than expected ., Because natural selection—which removes genetically linked neutral variation—is more efficient in larger populations , selection on novel mutations offers a potential reconciliation of this paradox ., In this work , we align and jointly analyze whole genome genetic variation data from a wide variety of species ., Using this dataset and population genetic models of the impact of selection on neutral variation , we test the prediction that selection will disproportionally remove neutral variation in species with large population sizes ., We show that genomic signature of natural selection is pervasive across most species , and that the amount of linked neutral variation removed by selection correlates with proxies for population size ., We propose that pervasive natural selection constrains neutral diversity and provides an explanation for why neutral diversity does not scale as expected with population size .

Analysis of whole genome genetic variation data from 40 species shows that natural selection disproportionately depletes linked neutral variation in species with large population sizes, explaining why levels of neutral diversity do not scale with population size.

journal.ppat.1007899

STING is required for host defense against neuropathological West Nile virus infection

Encephalitic Flavivirus infections , including West Nile virus ( WNV ) , are ongoing or emerging threats to global health 1–4 ., In particular , WNV continues to re-emerge in the Americas , causing neuropathology and death in the most severe cases 3 , 5–7 ., Since its emergence in the USA in 1999 , annual outbreaks of WNV are impacted with fluctuations in neurovirulence attributed to the circulating strain 4–6 , 8 , 9 ., Morbidity and mortality are dramatically increased in years where the circulating strain has enhanced neurovirulence , highlighting the significance of understanding host-pathogen interactions that control neurotropism 5 , 10 ., An analysis of CDC reports reveals that of all cases reported between 1999–2014 , 9% of neurovirulent cases result in death , in contrast to 0 . 5% of non-neurovirulent WNV cases ., Factors that limit WNV neurovirulence are not well understood but are critical to restrict pathology associated with WNV infections 5 ., WNV infection in humans most commonly manifests as an asymptomatic or mild febrile illness known as West Nile Fever ( WNF ) with symptoms that include headache , generalized weakness , rash , fever or myalgia , and in some cases vomiting , diarrhea , joint or eye pain 3 , 5–7 , 11–13 ., While most patients displaying WNF generally display symptoms for days to weeks , in some cases persistent symptoms continue to impact quality of life and cognitive abilities rendering a chronic disease outcome to WNV infection 11 ., More serious disease occurs if the virus crosses the blood brain barrier and progress to West Nile Neuroinvasive Disease ( WNND ) 7 ., WNND disease symptoms include meningitis , encephalitis , myelitis marked with acute flaccid paralysis , gastric complications , tremors and Parkinson-like symptoms 7 , 11 , 14–18 ., Patients with WNND can maintain symptoms for weeks to months , with persistent symptoms including chronic fatigue , functional cognitive disorders or neuropsychiatric disabilities and physiological complications , particularly those who exhibited acute flaccid paralysis symptoms during acute infection 7 , 11 , 18 ., Currently no therapeutics or vaccines are available for treatment of WNV infection or neuropathogenesis ., Thus , there remains a critical need to understand the virus-host interactions of WNV neurovirulence ., Both the innate and adaptive immune response are required to clear WNV infection and restrict immune mediated pathology 19 ., In humans , infection with WNV typically occurs through subcutaneous inoculation from the bite of an infected mosquito ., A parallel form of infection using sub-cutaneous challenge of WNV in a mouse model has been shown to replicate the progression , tissue involvement , and pathology of WNV infection that occurs in humans 19–22 ., In the mouse model , viral replication occurs at the subcutaneous site of entry followed by infection of the draining lymph node and splenic infection 19 ., These processes first trigger innate immune activation in peripheral tissues outside of the central nervous system ( CNS ) through viral recognition by the RIG-I-like receptors to induce IRF3 activation and the production of types I and III interferon ( IFN ) 23–26 ., Innate ( RLR ) immune defenses triggered by RLR signaling and IFN actions serve to restrict the tissue tropism of WNV and are essential for protection against neuroinvasion 19 , 23 , 24 , 27–34 ., Type I and III IFN are essential to inform the innate and adaptive immune interface to balance development of effective immunity , protect the blood-brain barrier , and limit immune-related pathology in the CNS 19 , 23 , 24 , 35–39 ., In particular , type I IFN-dependent cytokine and chemokine signaling cascades are essential for functional development of the cytotoxic CD8+ T cell response , as well as its regulatory T cell ( Tregs; FoxP3+ CD4+ T cells ) counterpart 24 , 36 , 37 , 39–42 ., While CD8+ T cells are required for controlling both peripheral and CNS viral load , CD4+ T cells , specifically Tregs , are essential for preventing symptomatic disease in the CNS 40–43 ., The adaptor protein , Stimulator of Interferon Genes ( STING ) , has also been implicated in host defense against WNV 44–46 ., STING was first described as an essential defense mechanism against both RNA and DNA viruses 47 , 48 ., Since then , STING has been recognized for its role in responding to cytoplasmic DNA and mediating subsequent innate immune activation and IFN production ., However its role in the defense against RNA viruses is poorly understood 47–54 ., Intriguingly , multiple RNA viruses , including dengue virus , yellow fever virus , hepatitis C virus and coronaviruses , direct viral evasion strategies to disrupt the STING signaling pathway , reflecting a likely role for STING in host defense against RNA viruses 52 ., STING was found to be required for host defense during infection with influenza A virus , as well as dengue virus , a closely related flavivirus to WNV 55–57 ., Additionally , during infection with related flavivuses including Japanese encephalitis virus ( JEV ) and Zika virus , STING deficiency led to increased neuropathology in vivo and in vitro , suggesting a critical role for STING in CNS defense 58 , 59 ., The role for STING in the CNS has been implicated in multiple other neurodegenerative diseases including Aicardi-Goutières syndrome , sterile immune mediated CNS pathology and during chronic CNS diseases 14 , 16 , 60–66 ., In this study , we investigated the hypothesis that STING plays a regulatory role in the immune response against WNV , thereby restricting viral neurotropism and neuropathology ., We show that STING is essential for host defense against WNV in a mouse in vivo model of infection ., Clinical and pathological analyses demonstrate a novel role for STING in conferring CNS defense against WNV in vivo ., We found that tonic levels of type I IFN were decreased in STING-/- bone marrow derived macrophages ( BMDM ) and linked with increased susceptibility to WNV infection ., Following infection , we observed heightened immune responses in vitro and in vivo concomitant with increased viral load ., STING deficiency led to the development of an aberrant adaptive immune response , with decreased activation of CD8+ cells and T regulatory cells ( Tregs ) in the spleen , and decreased CD4+ T cell numbers resulting in an altered CD4/CD8 T cell ratio in the CNS coupled with CNS disease ., Our observations imply an essential role for STING within the interface between the innate and adaptive immune responses for effective immune programming in the control of WNV infection and CNS disease ., Previous studies demonstrated that mice defective in STING signaling experienced increased mortality during WNV infection , yet the linkage of STING to immune response programming for defense against WNV has not been defined 46 ., Using genetically knocked-out Tmem173 ( STING-/- ) mice 67 , we first performed a survival analysis to confirm the role of STING in host survival during WNV infection ( Fig 1A ) ., C57B/6J ( B6 , WT ) and STING-/- mice were infected through subcutaneous virus challenge via foot-pad injection and monitored for 18 days post infection ( dpi ) ., Mice were scored daily for morbidity , marked as loss in body weight ( Fig 1B ) and overall increased clinical score ( Fig 1C ) ., Consistently , between 8–12 dpi , mice either met euthanasia criteria ( Terminal; T ) or went on to survive ( Survivors; S ) through 18 dpi ( study end-point ) ( Fig 1D ) ., Using this model , we confirmed the occurrence of increased susceptibility to WNV infection in the complete absence of STING ( Figs 1A and S1A ) , similar to what was previously described in STINGgt/gt mice 46 ., We also observed significantly increased clinical severity scores in the STING-/- mice that persisted until the study-endpoint , when WT mice had returned to a base-line clinical score ( Fig 1B and 1C ) ., Additionally , we monitored mice daily for the duration of the experiment until they either met euthanasia criteria or at the study end-point , day 18 post infection ., Results from each mouse were analyzed to determine if there were differences in clinical signs between WT and STING-/- mice ., Notably , STING-/- mice displayed increased neurological signs of disease , characterized by loss of balance , reduced muscle tone and reflexes predominantly in the pelvic limbs and increased paresis and paralysis , implicating more severe damage to the hind-brain and spinal cord ( Fig 1E and 1F ) ., In order to determine if there was a survivor bias in the clinical data , we retrospectively stratified the data into cohorts of mice that met euthanasia criteria ( Terminal; T ) or ones that survived until day 18 post-infection ( Survivors; S ) , the pre-determined study end-point ( Fig 1D , 1G and 1H ) ., By doing so , we found that significant differences in body weight loss and clinical scores between WT and STING-/- mice were only observed in the Survivor cohort and not in the Terminal cohort ., While there is an essential role for STING in host survival during acute infection ( Figs 1A and S1A ) , these data implicate an additional prolonged requirement for STING in both prevention and recovery from neurological pathology ., When we examined CNS pathology , we found that in both WT and STING-/- mice , pathological scores were significantly increased in the spines of the Survivor cohort , with a trend toward increased scores in the brains and spines of the Terminal cohort ( Fig 1D , 1I and 1J ) ., Intriguingly , while STING-/- Terminal mice displayed increased CNS pathology , WT mice that met Terminal criteria had unexpectedly low clinical scores , suggesting that they met euthanasia criteria for reasons independent of severe encephalitis ., During necropsy , we observed that the gastro-intestinal ( GI ) tract of Terminal mice exhibited gross distension or other aberrant phenotypes including stool compaction , disintegration and in some cases severe reduction in size or collapse of the GI tract ( S1B Fig ) ., Pathologic analysis confirmed that Terminal mice display increased GI pathology that included microbiome overgrowth and neuronal degeneration and loss in the myenteric ganglia , particularly in STING-/- ( S1C and S1D Fig ) ., Previous studies have indicated that GI manifestations during WNV infections exist in both mice and humans , and are positively correlated to increased neurotropism and mortality 15–17 , 22 ., This outcome may imply that WT mice are meeting euthanasia criteria following WNV infection due to severe GI disease rather than severe CNS involvement as previously thought ., Further , these results demonstrate that STING plays a systemic role in host defense against WNV , with increased frequency of mortality and pathology occurring in the CNS and GI tract in STING-/ mice ., Together , these results show an essential role for STING in host survival and neuropathological defense in the CNS during WNV infection ., To determine if STING is required for viral control in the CNS , we challenged mice with WNV via footpad injection and examined tissue viral load at 4 dpi ( peak of peripheral viremia ) and 8 dpi ( peak of detectible virus in the CNS ) ( Fig 2A ) ., Viral titer of macrodissected brains and extracted spinal cords were examined by plaque assay individually for each mouse in the cohort ( Fig 2A ) ., As expected , virus was not detected at 4 dpi in the CNS but by 8 dpi virus was clearly detected in different CNS regions ., Virus was not consistently found in the CNS of all mice nor in every tissue examined ., There was however , a consistent trend toward increased numbers of infected mice with detectible virus in the CNS as well as increased viral titers in the CNS of STING-/- mice compared to WT ., To determine if there was detectible virus in the brains of Terminal vs Survivor mice , tissues from retrospectively sorted mice utilized for pathological analysis ( Fig 1I and 1J ) were immunostained for the presence of WNV antigen ( Fig 2B ) ., WNV foci were found in the brains of WT and STING-/- Terminal mice but were not apparent in WT or STING-/- Survivors , suggesting that either the virus had cleared or that surviving mice did not have CNS infection ., Neuronal death was assessed by TUNEL stain in both WT and STING-/ Survivors ., Here we observed enhanced neuronal apoptotic death in the STING-/- cohort , suggesting STING may have a direct or indirect role in neuronal defense in the CNS ( Fig 2C ) ., In order to determine if STING is required for neuronal defense against WNV , primary cortical neurons were isolated and cultured , followed by infection with WNV to determine viral growth kinetics under conditions of single and multi-step growth ( Fig 2D ) ., Surprisingly , no difference was detected between in WNV replication in WT and STING-/- primary cortical neurons ( Fig 2D ) ., To determine if the actions of STING might be restricted to the CNS for WNV protection , we performed an intracranial virus inoculation bypassing the role of the peripheral immune response and physical barriers such as the blood-brain barrier to directly infect the brain with WNV ( Fig 2E ) ., At 4 dpi , there was no difference in CNS viral load found in WT vs STING-/- mice nor was viral load different between STING-/- and WT mice ., Taken together , our observations imply that the role of STING is not limited to mediating viral control in the CNS ., It is possible that STING is therefore required in the development of a protective immune response in the periphery such that in the absence of STING the immune response is aberrantly programmed , leading to CNS immunopathology ., Given that STING deficiency was associated with enhanced mortality ( see Fig 1 ) without a significant increase in CNS viral burden ( Fig 2 ) , we considered that STING deficiency could result in defective antiviral innate immune signaling and lead to loss of viral control in the periphery , thereby leading to enhanced morbidity and mortality ., We first tested the role of STING in BMDMs , as macrophages are a tropic cell and key modulator of peripheral viral control during WNV infection ( Fig 3A ) 19 ., As expected , WNV levels were significantly increased by 24 and 48 hours post inoculation ( hpi ) ., Unexpectedly however , STING-/- BMDM had increased innate immune and inflammatory gene expression , including enhanced level of type I IFN expression during WNV infection ( Fig 3B ) ., We then examined the spleens of infected mice to determine if there was an overall loss of viral control manifested as increased viral load over WT ., As expected , virus was detected at 4 dpi in both WT and STING-/- ., Surprisingly however , there was no difference in 4 dpi viral titers between WT and STING-/- , nor was there a sustained virologic response in STING-/- mice ( Fig 3C ) ., These data indicate that peripheral loss of viral control does not occur in the absence of STING ( Fig 3C ) ., Similarly , viral RNA was detected equally in spleens of infected WT and STING-/- mice at 4 dpi , but the virus was largely cleared from the spleen by 8 dpi ( Fig 3D ) ., In the CNS however , we observed a trend toward increased viral RNA and innate immune gene expression at 8 dpi in WNV-infected STING-/- mice , similar to that observed in BMDM ( Fig 3A and 3D ) ., These data were unexpected as we initially predicted that STING deficiency would reduce innate immune activation based on the known role of STING signaling in IFN induction ., These data demonstrate that innate immune activation and the inflammatory response are exacerbated in both in vitro and in vivo STING deficient models , possibly culminating in enhanced immunopathology in STING-/- mice ., The canonical STING sensing pathway is dependent on upstream recognition of DNA danger- or pathogen-associated molecular patterns ( DAMP , PAMP ) such as DNA viruses , cell-free or mitochondrial DNA , by cyclic GMP-AMP synthase ( cGAS ) ., In mammals , cGAS binding to dsDNA activates its synthase activity to produce a cyclic di-nucleotide , cGAMP ( cyclic guanosine monophosphate-adenosine monophosphate ) , which binds to STING , initiating downstream activation of STING by phosphorylation , STING relocalization from diffuse cytosolic to punctate pattern , and subsequent induction of innate immune signaling and IFN production 47 , 48 , 53 , 68 , 69 ., During RNA virus infections however , the role for STING defense has not been well-characterized ., To evaluate the activation of STING during WNV infection , we utilized a recently described telomerase reverse transcriptase human foreskin fibroblasts ( HFF ) model to assess activation of endogenous STING by phosphorylation and relocalization from the cytoplasm to the perinuclear space during WNV infection 70 ., Transfection of interferon-stimulated DNA ( ISD; calf-thymus DNA ) into HFFs initiated re-localization of STING as previously reported by 3hpi 48 , 70 ., Intriguingly however , STING was not relocalized in WNV infected cells ( Fig 4A ) ., It is possible that the kinetics of STING activation are different from ISD activation of STING as compared to WNV infection , so we performed a time course experiment to detect STING activation by phosphorylation status 71 , assessing a range of 1–24 hpi at MOI = 1 ( Fig 4B ) ., Similar to what was observed by IFA , STING phosphorylation was not observed at any time point during WNV infection , although phosphorylated STAT1 and WNV protein was detected at 24 hpi , suggesting virus replication and innate immune signaling were occurring normally ( Fig 4B ) ., To determine if activation was dependent on viral load , we infected HFF with a MOI = 1 and MOI = 10 of WNV , but also observed no STING activation as measured by phosphorylation ( Fig 4C ) ., These data suggest that STING is not canonically activated during WNV infection in HFF cultures and reveals a potential non-canonical role for STING in host defense during infection with WNV ., In order to determine if there was a systemic change in the innate immune profile in STING-/- mice , we examined the cytokine and chemokine profile in the serum of WT and STING-/- mice at the peak of peripheral viremia ( 4 dpi ) and CNS viral burden ( 8 dpi ) ., We found that mock infected STING-/- mice had an increased basal production of multiple cytokines and chemokines at 4 dpi ., We also observed significant increases in IL33 , IL4 , IL6 , IL15 , MCSF , Gro-alpha , while at 4 dpi IP-10 ( CXCL10 ) was decreased in STING-/- compared to WT mice ( S2 Fig ) ., While these cytokines have multiple roles in immune modulation , a common role among them is in activation and recruitment of T cells ., These data suggest that STING is required for regulation of immune cytokine and chemokines that program immune cell trafficking and actions during WNV , as has been shown for STING in cancer immunity and autoimmune signaling 53 ., To determine if STING is required for proper programming of the T cell response during WNV infection , we examined splenic T cells from WT and STING-/- mice at 8 dpi , a time point when the adaptive immune response is established in WT mice 24 ., We observed a reduction in the frequency of CD8+ T cells , along with a trend toward decreased numbers of T cells in the spleens of STING-/- mice compared to WT during WNV infection ( Fig 5B ) ., Additionally , within the CD8+ T cell subset ( Fig 5C ) , there was a significant decrease in frequency of activated ( CD44+ ) and CXCR3+ T cells , and we observed a consistent trend of decrease in the frequency of WNV-specific CD8+ T cells in the spleens of STING-/- mice compared to WT , suggesting that STING is required for optimal anti-WNV CD8+ T cell responses ., We also observed a significant increase in the frequency of CD4+ T cells in STING-/- mice ( Fig 5B ) , with a corresponding trend toward increased absolute cell numbers ., While we observed a trend toward differences in the absolute number of most cell populations examined between WT and STING-/- mice , we found that significant differences most typically occurred in cell frequencies , suggesting that the balance of T cells subsets may be skewed in the absence of STING ., In particular , we found skewing within the T regulatory cell ( FoxP3+ ) populations ( Fig 5E–5G ) , with significant deficits in Ki67+ , CD44+ and CD73+ Tregs , CD44 and CD73+ Tregs ., These data suggest that STING is required for modulating T cell responses and T cell frequencies during WNV infection that lead to a protective rather than pathogenic outcome ., Because of the heightened innate immune profile and aberrant programming of the T cell responses in spleens of STING-/- mice , we examined the CNS-specific T cell profile across mouse lines ., Histological analyses revealed trends toward increases in CNS immune cellularity , both in the form of perivascular and parenchymal mononuclear infiltrate , suggesting the CNS pathology may be immune-mediated ( Fig 6A ) ., We then performed a CD3 IHC stain in the brains of Survivors , we found increased clusters of CD3 infiltrate in the hind and mid-brain regions ( Fig 6B ) co-localized with robust lesions ., In serial slices of the same tissues , we did not observe WNV staining by IHC in STING-/- Survivors ( Fig 2 ) , however we did observe continued gliosis , suggesting that a potential immunopathology may occur in the brain of STING-/- mice infected with WNV ., Previous studies indicated that cellular infiltrate in the brain is predominantly comprised of CD3+ T cells during WNV infection 72 ., Therefore , we characterized T cell responses of WT and STING-/- mice in the CNS on 4 dpi to examine baseline differences at 8 dpi when WNV and leukocytes are both present in the CNS ( Fig 6J ) ., Lymphocyte and T cell responses in both mock and WNV-infected mice were comparable at 4 dpi , indicating that there was no gross difference in the CNS between WT and STING-/- mice ( Fig 6C and 6D ) ., By 8 dpi however , we found statistically significant decreases in the frequency and numbers of CD4+ T cells in STING -/- mice ( Fig 6F ) ., Although there was no difference in the total numbers of CD8+ T cells , there was a statistically significant increase in the frequency of CD8+ T cells in the CNS of STING-/- mice , likely due to overall trend of decreased numbers of lymphocytes in the brain ( Fig 6C–6E ) ., By 8 dpi , these changes resulted in a significantly decreased CD4/CD8 ratio of T cells , indicating an imbalanced T cell response to WNV in the CNS of STING-/- mice ( Fig 6I ) ., Of cells that made it to the brain by 8 dpi , no differences were found in the absolute number of activated ( CD44+ ) or WNV-specific ( NS4b Tetramer+ ) CD8+ T cells ( Fig 6G and 6H ) , FoxP3+CD25+CD4+ T cells ( Fig 6K ) in the brain ., These data suggest that STING is not essential for recruitment of WNV-specific cytotoxic T cells in the CNS , however it may be required for balancing the cytotoxic vs immunosuppressive adaptive response ., Furthermore , it is also possible that the enhanced recruitment of cells to the CNS is in response to damage caused by the virus , aberrant immune signaling , or both ., This outcome would suggest that STING plays an essential role in modulating the balance between immunopathogenic and immunoprotective response in the CNS during WNV infection ., The increase in clinical disease and pathological damage observed in the STING-/- versus WT mice , particularly in Survivors , could be due to an aberrant immune response resulting in CNS damage after initial viral insult ., We found that CNS pathology in WT mice is largely restricted to the cortex and meninges , while STING-/- mice display increased pathology in the cerebellum and hind/mid brain regions in addition to the cortex and meninges ( Fig 7A and 7C ) ., These data correlate with the increased CD3 staining observed by IHC in STING-/- mice ( Fig 6B ) , also noted as the same brain regions where WNV is often detected by IHC ( Fig 2B ) ., These observations suggest that STING plays a role in directing or maintaining the T cell response to specific loci within the CNS or that initial viral infection led to increased recruitment of a localized adaptive immune response that resulted in immunopathology ., Furthermore , pathology in the spine was more diffuse , suggesting that STING has a widespread protective role in the CNS during WNV infection ( Fig 7B and 7D ) ., These observations led us to investigate if there was a localized polarization of microglia or infiltrating macrophages in CNS regions toward an M1 or M2 phenotype ( Fig 7E ) ., Microglia have the highest levels of STING ( Tmem173 ) expression observed in any cell within the adult mouse 73 , 74 and it is possible that in the absence of STING , microglia are aberrantly polarized , enhancing immune-mediated pathology ., To examine this possibility , we assessed the expression of M1 ( CXC1 and IL6 ) and M2 ( Pparg , Arg1 , Chil3 and Retnla 1 ) associated genes by RT-qPCR in different regions of the CNS ., In WT mice , we found that CXCl1 ( marking an M1 phenotype ) was present in the brain stem by day 8 post infection , and Retnla 1 expression ( marking an M2 phenotype ) occurred in both the mock and 4 dpi tissues within the brain stem and sub-cortex ( containing the thalamus ) regions of the brain ( Fig 7E ) ., This profile suggests that CNS homeostasis includes a localized M2 phenotype that is induced to a M1 phenotype in WT mice following WNV infection ., In STING-/- mice however , we found a widespread increase in the M1 response gene expression ( marked by CXCL1 and IL6 ) with the highest expression observed in the brain stem and spinal cord ., Simultaneously , there was also a corresponding increase in Pparg and Chil3 ( marking the M2 phenotype ) , with no clear difference in Arg1 expression and an overall trend toward decreased expression of Retnla ., These observations reveal a widespread increase in both M1 associated genes , with altered regulation of the M2 associated genes in STING-/- mice , potentially resulting in aberrant balance of the M1 and M2 polarization in the CNS ., To determine where in the CNS STING is actually localized and if this tissue localization overlaps with the location of the cellular infiltrate noted histopathologically or with expression of innate immune genes , we utilized the Allen Brain Institute database to search for STING ( Tmem173 ) localization in the mouse brain 75 ., Within the brain , STING expression is found within the olfactory bulb , thalamus/midbrain , brainstem and cerebellum , as well as low levels throughout the cortex , overlapping areas that are affected most severely by WNV infection ( S3 ) 14 , 75 ., These regions of brain affected correlate with the clinical signs we observed including loss of balance , tremors , and loss of motor function ( Figs 1E and 7C–7E ) ., Furthermore , these areas of STING expression overlap with the brain regions where altered regulation of M1 or M2 gene expression were most readily observed , implicating a role for STING in polarization of either or both microglia and macrophages in the CNS ., Cumulatively , these data suggest that STING has an essential role in maintaining immune response homeostasis and immune programming in initial defense against WNV infection ., Without STING , immunopathology occurs , leading to exacerbated CNS disease and clinical sequelae ., Recent years have seen a marked increase in the global health threat presented by emerging and re-emerging encephalitic viruses , particularly those with increased neurotropism and neuropathology such as WNV 1 , 3 , 10 , 76 , 77 ., Previous studies indicated an important role for STING in host survival during WNV infection 46 , however it is unclear what role STING plays in conferring host defense against RNA viruses 52 , 54 ., Here , we demonstrate that STING is essential to prevent host morbidity and mortality during WNV infection where it plays a role in immune homeostasis and programming ., However , STING is not canonically activated in vitro upon infection with WNV , revealing a novel function for STING during infection with RNA viruses ., Furthermore , we show that STING is essential for host neuropathological defense against WNV through regulation of the innate-adaptive immune interface in vivo ., We found that STING deficient mice exhibit increased mortality and morbidity including increased and sustained neurological clinical signs , particularly in mice that survive infection ( Fig 1 ) ., These data were corroborated by pathological analysis , which also revealed distinct differences in CNS pathology ., Intriguingly , there seems to be a stratification in clinical and pathological findings between the STING-/- mice that meet euthanasia criteria and those that go on to survive ., Survivorship bias has been previously reported in the WNV model , with these data further implicating this bias as a critical factor to consider when performing time course vs . end-point experiments 78 ., Unexpectedly , these studies also revealed that there was minimal CNS pathology in WT mice that met euthanasia criteria ., It is typically assumed that mice meeting euthanasia criteria do so because of neuroinvasion and subsequent encephalitis ., Our data instead indicates that both WT and STING-/- Terminal mice have severe gross GI abnormalities , with corroborating abnormalities by histopathology , which may be the proximate cause of morbundity and meeting euthanasia criteria ( S1 ) ., GI complications during WNV have been previously described , however further study is necessary to understand the implications of GI pathology on WNV induced morbidity and mortality 15–17 , 79 ., Recently it has been shown that during WNV infection causes delayed GI transit , dependent on infiltrating antiviral CD8+ T cells 80 ., Furthermore , both in this model and in a lung model where STING exhibits a gain-of-function mutation , T cell-dependent chronic tissue damage occurs , supporting our findings that STING may play a broad and significant role in communicating between the innate and adaptive immune responses 80 , 81 ., Together , these data demonstrate an essential neuroprotective role for STING during WNV infection , potentially through a cellular mediated mechanism instead of the canonical interferon antiviral function typically attributed to STING ., WNV typically is cleared through development of an innate immune response and effective T cell immunity 19 ., To prevent progression to neuroinvasion , both the innate and adaptive immune response are critical to control WNV viremia and prevent viral induced pathology 19–21 , 24 , 82 , 83 ., Because the known function of STING is to initiate a type I IFN response to both PAMPs and DAMPs , we anticipated that the type I IFN response would be diminished both in vivo and in vitro explaining the increased viral loads ., Surprisingly , we actually observed an increased inflammatory and antiviral innate immune response in STING-/- mice in the CNS during WNV infection ., This same increase in the cytokine-chemokine response was also observed in BMDM ( Fig 3 ) and in serum of infected mice ( Fig 5 ) ., These outcomes were highly unexpected as the most commonly described role for STING is known as initiating a type I IFN response 46–48 , 53 , 54 ., In particular , STING was shown previously to facilitate the actions of the ELF4 transcription factor to promote type I IFN expression from WNV-infected cells wherein loss of STING associated with reduced IFN and ISG expression ( 49 ) ., While we observed significant increases in IFN and ISG expression in BMDM lacking STING , it is likely that STING imparts cell type-specific actions for regulation of innate immune signaling , similar to other pathogen recognition receptors that govern innate immune signaling against WNV , likely explaining this discrepancy between studies 19 ., It is also important to note that our studies employed STING-/- mice produced through classical gene targeting approach 48 while the previous study used STINGgt/gt mutant mice produced from N-ethyl-N-Nitrosourea mutagenesis and encoding a T596A point

Introduction, Results, Discussion, Methods

West Nile Virus ( WNV ) , an emerging and re-emerging RNA virus , is the leading source of arboviral encephalitic morbidity and mortality in the United States ., WNV infections are acutely controlled by innate immunity in peripheral tissues outside of the central nervous system ( CNS ) but WNV can evade the actions of interferon ( IFN ) to facilitate CNS invasion , causing encephalitis , encephalomyelitis , and death ., Recent studies indicate that STimulator of INterferon Gene ( STING ) , canonically known for initiating a type I IFN production and innate immune response to cytosolic DNA , is required for host defense against neurotropic RNA viruses ., We evaluated the role of STING in host defense to control WNV infection and pathology in a murine model of infection ., When challenged with WNV , STING knock out ( -/- ) mice displayed increased morbidity and mortality compared to wild type ( WT ) mice ., Virologic analysis and assessment of STING activation revealed that STING signaling was not required for control of WNV in the spleen nor was WNV sufficient to mediate canonical STING activation in vitro ., However , STING-/- mice exhibited a clear trend of increased viral load and virus dissemination in the CNS ., We found that STING-/- mice exhibited increased and prolonged neurological signs compared to WT mice ., Pathological examination revealed increased lesions , mononuclear cellular infiltration and neuronal death in the CNS of STING-/- mice , with sustained pathology after viral clearance ., We found that STING was required in bone marrow derived macrophages for early control of WNV replication and innate immune activation ., In vivo , STING-/- mice developed an aberrant T cell response in both the spleen and brain during WNV infection that linked with increased and sustained CNS pathology compared to WT mice ., Our findings demonstrate that STING plays a critical role in immune programming for the control of neurotropic WNV infection and CNS disease .

In recent years , outbreaks of emerging and re-emerging neuroinvasive West Nile virus ( WNV ) infection have brought about a critical need to understand host factors that restrict neuropathology and disease ., WNV infection in humans typically is either asymptomatic or results in a mild febrile illness , but in some cases virus spreads to the central nervous ( CNS ) causing a more severe form of neuropathological disease ., Previous studies established that both innate and adaptive immune responses are essential for controlling WNV disease and restricting the virus from the CNS ., In this study , we examined the role of Stimulator of Interferon Genes ( STING ) in conferring host defense during WNV infection in a murine model ., Our studies revealed that STING is essential for restricting pathology in the CNS during WNV infection ., Further , STING is required for effective programming of the innate and adaptive immune response to WNV ., In the absence of STING , aberrant immune development leads to ineffective viral clearance and immunopathology in the CNS ., These studies uncover a critical and previously unidentified role for STING in the restriction of WNV that may have broader implications for a role in conferring host defense against RNA viruses .

blood cells, medicine and health sciences, immune cells, pathology and laboratory medicine, viral transmission and infection, nervous system, pathogens, immunology, microbiology, viruses, euthanasia, rna viruses, cytotoxic t cells, viral load, white blood cells, animal cells, proteins, medical microbiology, microbial pathogens, t cells, immune response, biochemistry, west nile virus, anatomy, flaviviruses, central nervous system, cell biology, virology, viral pathogens, interferons, biology and life sciences, cellular types, organisms

journal.pgen.1000781

Increased Expression and Protein Divergence in Duplicate Genes Is Associated with Morphological Diversification

Duplicate genes rarely exhibit de novo functions ( neofunctionalization ) ; more usually , the functions of the original gene are split into multiple functions among the duplicate genes ( subfunctionalization ) 1–5 ., Such functionalization through gene duplication is considered to be an important source of diversification in complex organisms 6 ., As a mechanism of functionalization in duplicate genes , differentiation of both gene expression and protein function are thought to be important ., In particular , differential patterns of gene expression among paralogs are widely believed to play a prominent role in morphological diversification , because such differences are essential for development 7–10 ., However , substantial amounts of data support morphological diversification through divergence of protein function 11 ., Many researchers have studied divergence of either expression or protein function in duplicate genes at the genome scale 12–24 ., Although divergence of either expression or protein sequence tends to increase as a duplication ages , it is unclear whether either expression or protein divergence in duplicate genes has been elevated by functionalization ., Therefore , it is of interest to compare the divergence rate of either expression pattern or protein sequence of duplicate genes of the same age that have and have not undergone functionalization ., If divergence of both expression and protein function are important sources for functionalization , the divergence rate of both should be higher in duplicate genes that have undergone functionalization compared with those that have not ., A . thaliana is an excellent model organism for addressing the above issue because it has a highly duplicated genome and many knock-out mutants have been generated ., Here , to address how duplicate genes have contributed to morphological evolution , we classified Arabidopsis duplicate genes into high , low and no morphological diversification groups based on knock-out data , and examined the divergence rates of both expression pattern and protein sequence among the three morphological diversification groups ., From the literature and from our earlier work ( see Materials and Methods ) 25 , 26 we identified 398 pairs of duplicate genes in which the knock-out mutant of either gene in a pair induced abnormal morphological changes relative to wild type ., Abnormal morphological changes were classified into seed , vegetative and reproductive phenotypes on the basis of the definition of Meinke et al 27 ., When the knock-out phenotype is totally different between genes in a paralogous gene pair , it is reasonable to assume that functionalization occurred after gene duplication ( Figure 1A ) ., For example , the knock-out mutant of AT4G09820 and AT5G41315 genes induced a yellow seed coat in the reproductive stage and a reduction of trichomes in the vegetative stage , respectively ., Therefore , the knock-out phenotype is completely different between AT4G09820 and AT5G41315 because two abnormal phenotypes appeared in different developmental stages ., Thus , paralogous genes with different phenotypes ( morphological differences between phenotypes ) are defined to have high morphological diversification ., It is more common , however , to observe knock-out phenotypes that are similar or identical between paralogous genes ( Figure 1B ) ., For example , the knock-out mutants of AT1G62830 and AT3G10390 genes both induced late flowering ., Although the knock-out phenotype of the two genes is similar , there would appear to be functionalization in such paralogous genes because a morphological change resulting from the deletion of one gene occurs when there is no or little functional redundancy between the paralogous genes ., We , therefore , thought that such paralogous genes had some degree of functionalization after gene duplication ., However , it is likely that similar or identical phenotypes indicate paralogous genes that have lower functionalization compared with paralogous genes with different phenotypes ., Therefore , paralogous genes with either similar or identical phenotypes ( morphological changes within phenotypes ) were defined to have low morphological diversification ., In this study , we identified 163 and 235 paralogous gene pairs associated with high and low morphological diversification , respectively ., As a control set , we focused on paralogous gene pairs in which abnormal morphological changes are observed only upon the deletion of multiple paralogous genes but deletion of each gene separately did not induce abnormal morphological changes ( Figure 1C ) ., For example , the double knock-out mutant of AT3G58780 and AT2G42830 exhibits fruit dehiscence but knock-out of each gene alone did not induce abnormal morphological changes ., Such paralogous gene pairs are likely to have some degree of functional redundancy ., We , therefore , defined these paralogous gene pairs as having no morphological diversification ., The number of paralogous gene pairs identified without morphological diversification was 94 ., Thus , we identified a total of 492 paralogous gene pairs associated with the three kinds of morphological diversification ( Table S1 ) ., To examine the expression pattern divergence for a paralogous gene pair , we obtained intensities of gene expression by microarray analysis under 634 conditions ., Expression divergence in a pair of genes is usually inferred by 1 minus R ( Pearsons coefficient of correlation ) of the expression intensities among experimental conditions ., Here , we transformed the value as log ( ( 1−R ) / ( 1+R ) ) , because the transformation is more sensitive for examining expression differences 19 ., When we applied the log ( ( 1−R ) / ( 1+R ) ) values to paralogous gene pairs among the three morphological diversification groups , the log ( ( 1−R ) / ( 1+R ) ) values increased as morphological diversification increased ( Figure S1 ) ., However , the relationship may be strongly influenced by duplication age ( sequence divergence ) in the case that morphological diversification increases as sequence divergence increases ., We , therefore , investigated sequence divergence in paralogous gene pairs by examining synonymous ( Ks ) and nonsynonymous ( Ka ) distance among morphological diversification groups 28 ., Consequently , both synonymous and nonsynonymous distances increased as morphological diversification increased ( P<0 . 01 by Wilcoxons test; Figure S1 and Table S2 ) ., To minimize the effect of duplication age , log ( ( 1−R ) / ( 1+R ) ) was divided by Ks ., This is because expression divergence is expected to increase as duplication timing becomes earlier and Ks increases in a nearly linear fashion with duplication age 17 , 19 , 24 ., Ed ( log ( ( 1−R ) / ( 1+R ) ) /Ks ) is an indicator of the expression divergence rate between a paralogous gene pair: high and low Ed indicates high and low expression divergence at the same duplication age , respectively ., When we calculated Ed between a paralogous gene pair in the three morphological diversification groups , Ed increased as morphological diversification increased ( Figure 2A ) ., Ed differed significantly between each pair of morphological diversification groups ( P<0 . 01 by Wilcoxons test; Table S2 ) , suggesting that expression divergence is an important source for morphological diversification of duplicate genes ., There are genetic and epigenetic factors that are the source of expression divergence ., Since the differentiation of cis-regulatory elements can be a major genetic effect , we examined the proportion of known cis-regulatory elements that overlap in the promoter regions of paralogous gene pairs 29 ., The proportion of cis-regulatory elements that overlap decreased as morphological diversification increased ( Figure S2 ) ., The proportion of overlapping cis-regulatory elements differed significantly between each pair of morphological diversification groups ( P<0 . 05 by Wilcoxons test; Table S2 and Figure S2 ) , indicating that the divergence of cis-regulatory elements contributes to morphological diversification ., With respect to epigenetic factors , we investigated the proportion of methylated cytosines to non-methylated cytosines in the promoter regions of paralogous genes 30 ., The proportional difference in paralogous gene pairs did not significantly differ between each pair of morphological diversification groups ( Table S2 and Figure S2 ) , indicating that an epigenetic effect through methylation is unlikely to contribute to morphological diversification ., Taken together , expression divergence led by the differentiation of cis-regulatory elements is an important source for morphological diversification in duplicate genes ., Because duplication age ( sequence divergence ) between paralogous gene pairs increased as morphological diversification increased ( Figure S1 ) , we examined divergence rates of protein sequences of the same duplication age ., Divergence rates of protein sequences are commonly inferred from selection pressure in coding sequences , i . e . the ratio of the non-synonymous substitution rate ( Ka ) to Ks ., High and low Ka/Ks ratios indicate high and low protein divergence rates at the same duplication age , respectively 28 ., When we applied the Ka/Ks ratio to paralogous gene pairs within the three morphological diversification groups , the Ka/Ks ratio increased as the morphological diversification increased ( Figure 2B ) ., The Ka/Ks ratio differed significantly between each pair of morphological diversification groups ( P<0 . 01 by Wilcoxons test; Table S2 ) , suggesting that protein divergence is an important source for morphological diversification of duplicate genes ., To analyze the kinds of amino acid replacements that have occurred during morphological diversification , we classified all amino acid replacements as either ‘chemical radical’ or ‘conservative’ on the basis of an amino acid classification generated in an earlier report 31 ., We examined the ratio of the radical nonsynonymous substitution rate ( Kr ) to the conservative nonsynonymous substitution rate ( Kc ) ., Interestingly , the Kr/Kc ratios of all types of paralogous gene pairs were similar ( Figure 2C and Table S2 ) , indicating that paralogous gene pairs with either high , low or no morphological diversification tend to have the same level of radical protein divergence ., The Kr/Kc ratio based on this amino acid classification is significantly correlated with the Ka/Ks ratio at the whole genome level 31 ., Therefore , radical changes become restricted in paralogous gene pairs with higher morphological diversification ., One explanation for this restriction is that radical changes do not affect morphological diversification ., However , some reports have shown that radical changes significantly influence functional divergence 23 , 32 ., Therefore , it does not seem to be a reasonable explanation ., Another explanation is that radical changes may induce serious functional errors ., To maintain duplicate genes that encode functional proteins , radical changes may be too deleterious ., Therefore , paralogous gene pairs involved in higher morphological diversification may be subject to purifying selection against radical amino acid changes ., To compare the divergence rate of expression pattern with that of protein sequence in paralogous gene pairs associated with morphological diversification , we focused on paralogous gene pairs without morphological diversification because the divergence rate of expression pattern and/or protein sequence in these duplicate genes has little effect on morphological diversification ., Therefore , the top 5% of Ed and Ka/Ks ratios for paralogous gene pairs without morphological diversification were defined to be the threshold of higher divergence rate of expression pattern and protein sequences , respectively ., We then counted the numbers of paralogous gene pairs with a higher divergence rate in each of the high and low morphological diversification groups ( Table 1 ) ., To make the relative roles clear , we simply compared the observed ratio between paralogous gene pairs with only higher expression divergence and those with only higher protein divergence , assuming no bias between expression and protein divergence in either high or low morphological diversification groups ., Interestingly , the number of paralogous gene pairs ( 37 in either high or low morphological diversification groups ) with a protein divergence but no expression divergence was significantly higher than the number of paralogous gene pairs ( 62 in either high or low morphological diversification groups ) with a higher expression divergence but no protein divergence , as determined by the chi-square test ( P<0 . 05 ) ., These results indicate that paralogous gene pairs with a higher divergence rate of protein sequence contribute to morphological diversification more effectively than those with a higher divergence rate of expression ., The inference from these results is that protein sequence plays the major role ( 59–67% ) and expression plays the minor role ( 33–41% ) in morphological diversification ., We performed the same analysis using the top 10% of Ed and Ka/Ks ratios of paralogous gene pairs without morphological diversification as the threshold of higher divergence rate of expression pattern and protein sequences , and obtained essentially the same results ( Table S3 ) ., Therefore , we believed that the relative rates of expression and protein divergence are stringent in morphological diversification ., Finally , we addressed to what extent duplicate genes were associated with expression or protein divergence exerting morphological diversification at the whole genome level ., To examine this question , we randomly chose 1000 pairs of paralogous gene pairs ., We then compared Ed and Ka/Ks ratios among the 1000 random paralogous gene pairs and among paralogous gene pairs with high , low or no morphological diversification ( Figure 2 ) ., Both Ed and Ka/Ks ratios for the random paralogous gene pairs were significantly lower compared with that for the paralogous gene pairs with high or low morphological diversification but were significantly higher compared with that for the paralogous gene pairs without morphological diversification ( P<0 . 01 by Wilcoxons test , ( Figure 2A and 2B and Table S2 ) ., However , the Kr/Kc ratio was not different between any pair in the four categories ( P>0 . 05 by Wilcoxons test , Figure 2C and Table S2 ) ., As discussed earlier , the Kr/Kc ratio is not an indicator for functionalization , therefore , no difference is reasonable ., These results suggest that duplicate genes have not experienced divergence of expression or protein sequence exerting morphological diversification on a genome-wide scale ., It is , therefore , likely that most duplicate genes have experienced only minor functionalization , at least in A . thaliana ., To understand to what extent molecular changes in duplicate genes have contributed to morphological diversification in A . thaliana , we examined the divergence rate of either expression pattern or protein sequence in duplicate genes associated with morphological diversification and found that both divergences are important sources in morphological diversification ., Although both mechanisms are not mutually exclusive , our analysis suggested that changes of protein sequence play the major role and changes of expression pattern play the minor role in morphological diversification ., However , randomly chosen duplicate genes have not experienced divergence of expression or protein sequence exerting morphological diversification ., These results indicate that most duplicate genes have experienced minor functionalization and only a few duplicate genes are likely to be crucial to morphological evolution ., We used data from the available literature and from our bank of previously generated T-DNA insertional mutants 25 , 26 , to identify 1203 duplicate genes whose knock-out induced abnormal morphological changes relative to wild type ., The nucleotide sequences of A . thaliana ( TAIR7 ) were obtained from TAIR ( www . arabidopsis . org ) ., Duplicate genes were defined as proteins that matched other proteins in a BLAST search with E<1×10−4 33 ., We then classified the 1203 duplicate genes into 786 gene families by the Markov clustering algorithm ( http://micans . org/mcl/ ) ., In every pair of each family , we examined the amino acid identity and the coverage ( percentage of alignable regions ) ., We found 405 paralogous gene pairs with amino acid identity >0 . 3 and coverage >0 . 5 ., Since tandem duplicates have a higher chance of exhibiting similar expression due to leaky expression or conserved sequences by gene conversion than non-tandem duplicates 34–36 , we removed tandem duplicates from the 405 paralogous gene pairs ., As reported earlier 37 , tandem duplicates were defined as genes in any gene pair , T1 and T2 , that ( 1 ) belong to the same gene family , ( 2 ) are located within 100 kb of each other , and ( 3 ) are separated by at most 10 nonhomologous ( not in the same gene family as T1 and T2 ) genes ., In this definition , we identified 7 tandem paralogous gene pairs ., After removing these tandem paralogous gene pairs , we used 398 non-tandem paralogous gene pairs in this study ., Note that each knock-out mutant of paralogous genes induced abnormal phenotypic changes ., To examine the degree of morphological diversification between the genes of the paralogous gene pairs , we classified morphological changes into seed , vegetative and reproductive phenotypes , according to the definition of Meinke et al 27; the changes were defined as high ( morphological changes between phenotypes ) and low ( morphological changes within phenotypes ) morphological diversification ., Briefly , seed , reproductive and vegetative phenotypes show visible changes in development ., We identified 163 paralogous gene pairs associated with high morphological diversification and 235 associated with low divergence ( Table S1 ) ., As a control set , we identified from the literature165 duplicate genes that did not show morphological diversification ., Absence of morphological diversification was defined as the observation of morphological change only upon the deletion of multiple paralogs; deletion of each gene separately did not induce morphological change ., After removing tandem paralogous gene pairs , we found 95 paralogous gene pairs with amino acid identity >0 . 3 and coverage >0 . 5 ( Table S1 ) ., We obtained Affymetrix ATH1 data from the AtGenExpress expression atlas at TAIR ( http://www . arabidopsis . org/ ) ., We compiled 1280 microarray datasets under 634 conditions , consisting of 82 different developmental stages , 72 biotic treatments , 285 abiotic treatments , 11 nutrient treatments , 81 hormone treatments , 40 chemical treatments , 21 cell cycle stages and 42 different genotypes ., The array intensities were processed with the Bioconductor ( http://www . bioconductor . org ) affy package in the R software environment ( http://www . r-project . org ) ., Specifically , the array intensities were adjusted to reduce background with the mas5 function , and the normalize quantiles function was used for between-array normalization ., The background-corrected and background-normalized intensities were used for further analysis ., We obtained the mapping data of known cis-regulatory elements in 1 kb promoter regions of all A . thaliana genes at ATCOECIS ( http://bioinformatics . psb . ugent . be/ATCOECIS/ ) 29 ., To examine the divergence of cis-regulatory elements in each paralogous gene pair , we used the proportion of overlapping cis-regulatory elements ( the number of overlapping cis-regulatory elements over the number of observed cis-regulatory elements ) ., To examine divergence of methylation in paralogous gene pairs , we obtained the mapping data of bisulfite-treated DNA sequences in the TAIR7 genome at NCBI Gene Expression Omnibus ( GSM276809 ) 30 ., The bisulfate-treatment converts cytosine to uracil in unmethylated cytosine sites but does not affect cytosine in methylated cytosine sites ., Since the methylation of each cytosine site was determined multiple times , a methylated cytosine site was defined when that site is more often methylated than not ., We calculated the proportion of methylated cytosine sites ( the number of methylated cytosine sites over the number of observed cytosine sites ) in promoter regions ( 500 bp upstream from either start codon or transcriptional start site ) of all A . thaliana genes because the methylation of 500 bp upstream regions is considered to be sensitive for gene expression 30 ., The proportional difference of methylated cytosine sites in a paralogous gene pair was used to represent the methylation divergence in a paralogous gene pair ., Nucleotide sequences of A . thaliana ( TAIR7 ) were obtained from TAIR ( www . arabidopsis . org ) ., Pairwise alignment was performed with the program CLUSTALW to align coding regions 38 ., Ks and Ka between paralogous genes were estimated by the modified Nei–Gojobori method 28 ., The transition/transversion ratio was estimated for each paralogous gene pair , and the ratio was then used to estimate Ka and Ks ., To infer the ratio of the radical non-synonymous substitution rate ( Kr ) to the conservative non-synonymous substitution rate ( Kc ) , we classified amino acids according to Hanada et al . 2007 31 ., Radical and conservative changes were defined as amino acid replacements between and within groups , respectively ., The ratio of Kr to Kc for each paralogous gene pair was estimated by the Zhang method 39 ., We randomly chose genes from the total set of annotated A . thaliana genes ( TAIR7 ) ., For a chosen gene , similarity searches were conducted against all annotated A . thaliana genes using BLASTP 33 ., We aligned the chosen gene and all homologous genes identified in the BLASTP search using CLUSTALW and estimated the amino acid similarity among them 38 ., We calculated the amino acid identity and the coverage ( percentage of alignable regions ) between the chosen gene and the matched gene with the highest identity ., If the paralogous gene pair had amino acid identity >0 . 3 and coverage >0 . 5 , we added the pair to a random set ., We repeated this procedure until we obtained 1000 paralogous gene pairs .

Introduction, Results/Discussion, Materials and Methods

The differentiation of both gene expression and protein function is thought to be important as a mechanism of the functionalization of duplicate genes ., However , it has not been addressed whether expression or protein divergence of duplicate genes is greater in those genes that have undergone functionalization compared with those that have not ., We examined a total of 492 paralogous gene pairs associated with morphological diversification in a plant model organism ( Arabidopsis thaliana ) ., Classifying these paralogous gene pairs into high , low , and no morphological diversification groups , based on knock-out data , we found that the divergence rate of both gene expression and protein sequences were significantly higher in either high or low morphological diversification groups compared with those in the no morphological diversification group ., These results strongly suggest that the divergence of both expression and protein sequence are important sources for morphological diversification of duplicate genes ., Although both mechanisms are not mutually exclusive , our analysis suggested that changes of expression pattern play the minor role ( 33%–41% ) and that changes of protein sequence play the major role ( 59%–67% ) in morphological diversification ., Finally , we examined to what extent duplicate genes are associated with expression or protein divergence exerting morphological diversification at the whole-genome level ., Interestingly , duplicate genes randomly chosen from A . thaliana had not experienced expression or protein divergence that resulted in morphological diversification ., These results indicate that most duplicate genes have experienced minor functionalization .

The relationship between morphological and molecular evolution is a central issue to the understanding of eukaryote evolution ., In particular , there is much interest in how duplicate genes have contributed to morphological diversification during evolution ., As a mechanism of functionalization of duplicate genes , differentiation of both gene expression and protein function are believed to be important ., Although it has been reported that both expression and protein divergence tend to increase as a duplication ages , it is unclear whether expression or protein divergence in duplicate genes is greater in those genes that have undergone functionalization compared with those that have not ., Here , we studied 492 duplicate gene pairs associated with various degrees of morphological diversification in Arabidopsis thaliana ., Using these data , we found that the divergence of both expression and protein sequence were important sources for morphological diversification of duplicate genes ., Although both mechanisms are not mutually exclusive , our analysis suggested that expression divergence is the minor contributor and protein divergence is the major contributor to morphological diversification ., However , the expression or protein sequence of randomly chosen duplicate genes did not show significant divergence that resulted in morphological diversification ., These results indicate that most duplicate genes experienced minor functionalization in the genome .

computational biology/comparative sequence analysis, computational biology/genomics, evolutionary biology/evolutionary and comparative genetics

journal.pcbi.1002405

Cortical Surround Interactions and Perceptual Salience via Natural Scene Statistics

Contextual influences collectively denote a variety of phenomena associated with the way information is integrated and segmented across the visual field ., Spatial context strongly modulates the perceptual salience of even simple visual stimuli , as well as influencing cortical responses , as early as in V1 1–12 ( for a review , see 13 ) ., At least two main lines of theoretical inquiry have addressed these influences from different perspectives ., First , computational models have related perceptual salience to low-level image features 14–18 ., Of particular note for us , 14 , 19–21 proposed that V1 builds a visual saliency map , performing segmentation where the spatial homogeneity of the input breaks down ( e . g . , at the border between textures ) ., A model of the dynamical , recurrent , interactions among nearby cortical neurons induced by long-range horizontal connections realized this theory , accounting for cortical and perceptual contextual data , including popout , visual search asymmetries , and contour integration 14 , 19–24 ., Second , the hypothesis that sensory processing is optimized to the statistics of the natural environment 25–27 , has led to successful models of the linear and non-linear properties of V1 receptive fields ( RFs ) 21 , 28–35 ., However , although collectively covering a huge range of computational , psychophysical , and neural data , these two theoretical approaches have not been unified ., To this end , we introduce a computational model of the statistical dependencies of neighboring regions in images , rooted in recent advances in computer vision 36–39 ., This model provides a formal treatment of the idea of statistical homogeneity vs heterogeneity of visual inputs , to which V1 has been proposed to be sensitive 20 ., Correct inference in the model involves a novel , generalized , form of divisive surround normalization 30 , 40–43 , that is engaged by stimuli comprising extended single objects ( e . g . in Fig . 1 the image patch inside the region with uniform vertical texture ) , but not by stimuli involving independent visual features ( e . g . in Fig . 1 the patch across the zebra and the background ) ., We focused in particular on the dependencies between orientations across space , optimizing model parameters based on natural scenes ., We used the resulting model to simulate neural and perceptual responses to stimuli used in physiological and perceptual experiments to test orientation-based surround modulation ., Note that we did not fit the model to experimental data from individual cells or subjects , but rather compared the qualitative behavior of a model trained on ecologically relevant stimuli with general properties of early visual surround modulation ., There is a wealth of experimental results on surround modulation , some classical , and some still subject to debate ., We chose to model a set of findings that we intend to be canonical examples of both sorts of results ., We included data that have been the subject of previous theoretical treatments , but paid particular attention to phenomena that lie at the boundary between integration and segmentation ., Indeed , one claim from our approach is that subtleties at this border might help explain some of the complexities of the experimental findings ., We make predictions for regions of the stimulus space that have not yet been fully tested in experiments ., In sum , we show that the statistical principles introduced can account for a range of neural phenomena that demand tuned surround suppression as well as facilitation , and encompass V1 as a salience map 20 ., We concentrated on the statistical dependencies between V1-like filters ( or receptive fields , RFs; see also Text S1 ) across space in natural images ( the images are shown in Fig . S1 ) ., We adopted RFs derived from the first level of a steerable pyramid 44; more details are provided below ., For conciseness , we will refer to the projection of a visual stimulus onto an RF as the RF output ., Fig . 2A–D show the joint conditional histograms of the output of one vertical RF given the output of another vertical RF in a different spatial position ., In the case of white noise image patches , outputs are linearly correlated for RFs that overlap in space ( Fig . 2A ) , but not for RFs that are farther apart and so non-overlapping ( Fig . 2B ) ., Natural scenes differ from white noise ., The characteristic bowtie shape of Fig . 2C , D indicates statistical coordination in the form of a higher-order variance dependency 30: i . e . , the variance of one RF depends on the magnitude of the output of the other ., Further , elongated structures , such as edges and contours , cause strong co-activation of RFs with particular geometrical configurations such as collinearity , leading to linear correlations between non-overlapping collinear RFs ( the tilted bowtie shape of Fig . 2D ) , but not parallel RFs ( Fig . 2C ) ., Natural images are also spatially heterogeneous: different image regions can elicit different levels of dependence between RFs outputs 45 ., Extreme examples are regions involving single objects with a uniform texture ( homogeneous patches ) which show a strong variance dependence ( Fig . 2E ) , as opposed to regions spanning multiple objects or objects and background ( heterogeneous patches ) for which the dependence is weaker ( Fig . 2F ) ., We extended a well-known probabilistic model of the variance dependence of Fig . 2C , D ( the Gaussian Scale Mixture , or GSM; 46 ) to capture the variability across image regions exemplified in Fig . 2E , F ., The GSM describes an RF output , k , as a random variable obtained by multiplying a Gaussian variable ( i . e . a random variable that is Gaussian distributed ) , , and a second random variable that takes only positive values , , also called the mixer ., ( 1 ) The mixer in the model can be shared between multiple RFs ( we then describe these RFs as being co-assigned to the same mixer ) , and can therefore generate statistical coordination , and can be intuitively thought of as representing a relatively global image property , such as contrast , that changes smoothly across space ., In contrast , the Gaussian variable in the model is local to each RF ., Consider the case of two RFs , ( k1 , k2 ) , whose respective Gaussian variables are multiplied by a common mixer ., Then the coordination is generated in the following way: in a specific instance where has a large value , both Gaussian variables are multiplied by a large value and therefore k1 and k2 are more likely to be large together ., Conversely , when takes a small value , k1 and k2 will likely be small together ., This generates precisely the type of dependency observed in Fig . 2C , D ., The tilt of the bowtie evident in Fig . 2C comes from linear correlation between the Gaussian variables , captured by their covariance matrix C . We then introduced a variant of the GSM ( see also 37–39 , 47 ) that could capture both dependent and independent RFs outputs ., Since we were interested in center-surround effects , we designed the model to approximate variance dependencies between a center group of RFs and a surround group , using a combination of:, 1 ) a standard GSM for the case in which the center and surround groups are dependent ( they are co-assigned to a common mixer , and can be linearly correlated ) ; and, 2 ) an independent GSM model , whereby the center and surround do not share a common mixer and lack linear correlation ., This model , which combines the two extreme cases , is technically a Mixture of GSMs ( MGSM 39 , 47 ) ., We implemented the model with a bank of 72 bandpass oriented filters or RFs taken from the first level of a steerable pyramid 44 , 48 ( see Text S1 ) ., We used 8 RFs in a central position ( center group , whose outputs are denoted in the following by k , and the corresponding Gaussian variables by ) comprising 4 orientations , 0 , 45 , 90 and 135 degrees from vertical , each at 2 phases ., The inclusion of multiple orientations in the center group was motivated by the strong variance dependence typically observed across oriented RFs at a fixed position; it also guaranteed local contrast normalization of the model responses ( see below ) as commonly assumed in divisive normalization 41 ., The 2 phases correspond to a pair of even- and odd-symmetric RFs ( quadrature pair ) ., For the surround we used 64 RFs ( whose outputs are denoted in the following by S , and the corresponding Gaussian variables by ) comprising 4 orientations , 2 phases , and 8 positions on a circle surrounding the center RFs with radius 6 pixels ., We organized the surround RFs into 4 separate groups , each including all RFs of a given orientation ., RFs had peak spatial frequency of 1/6 cycles/pixel , and a diameter of 9 pixels and were partly overlapping ., The orientation bandwidth was chosen to approximate the median value found in V1 by Ringach et al . ( 49 23 . 5 degrees half-width at 70% height of the tuning curve ) ., The qualitative character of the results did not change when using different bandwidths ., We included multiple surround groups to guarantee that all orientations were treated equally ., Fig . 3 illustrates the spatial layout of the RFs , and their structural dependencies in the different mixture components ., The leftmost panel depicts the component ( denoted by ) in which none of the surround groups is co-assigned with the center ( each RF group in the surround has a different color ) ., The remaining four components ( denoted by for ) comprise the cases in which all the center RFs , and just those surround RFs that favor orientation , are co-assigned to a common mixer ., For example , under the same mixer multiplies all the center Gaussian variables and the Gaussian variables for the vertical surround , and the corresponding RFs are co-assigned ( indicated by black color ) ., Note that even when the center and surround are independent , we still assume that the RFs in the surround share the same mixer , an approximation that was needed for computational tractability ( see Discussion ) ., The parameters governing RF interactions – i . e . the covariance matrices and the prior probability of each component of the MGSM – were optimized by maximizing the likelihood of an ensemble of natural scenes ( downloadable from http://neuroscience . aecom . yu . edu/labs/schwartzlab/code/standard . zip ) ., Mathematical details are provided at the end of Materials and Methods ., Fig . 4A depicts the resulting covariance matrices ., Notice that in the co-assigned components ( top row ) but not in the independent component ( bottom row ) , the model found larger variances for each central RF and its collinear neighbors , and larger covariance between them , than for its parallel neighbors ., For all the results in the paper except where noted , and in Text S4 , we renormalized the covariance matrices to make them identical under rotation of the spatial configuration of RFs: e . g . , the covariance between the vertical central RF and its collinear neighbors ( vertical , above and below the center ) under , was forced to be the same as the covariance between the horizontal central RF and its collinear neighbors ( horizontal , left and right of the center ) under ., In practice , the effect of the renormalization is to guarantee that model responses corresponding to different RF orientations are identical when the respective input stimuli are rotated copies of each other ., However , to explore the effects of cardinal axis biases and variability across different scenes , we also trained the model separately on each of 40 natural images from the Berkeley Segmentation Dataset ( http://www . eecs . berkeley . edu/Research/Projects/CS/vision/bsds ) , and did not renormalize the covariances ., The training was repeated 3 times for each image from random starting points ., Training on individual images converged to similar values each time , except for one image on which it did not converge ., This image was therefore excluded from further analysis ., To relate the model to contextual modulation in the visual cortex we assume that firing rates in V1 represent information about the Gaussian variables associated with the center RFs ., We choose because it represents the local structure of the RFs , which is what contains specific information about the local input image patch ., By contrast , encodes more general information , i . e . the average level of activity in a neighborhood of RFs across orientations and spatial locations ., The MGSM is a statistical model of images , which is sometimes called a generative or graphics model 50 ., The task as a whole for vision is to invert this model – using Bayes rule to find the posterior distribution over and ., However , here , we compute just the expected value of the Gaussian variables , which we assume to be related to V1 firing rates ( via Equation 3 , below ) : in the rest of the paper we also call such estimates the model responses ., In practice , we obtained model responses in two steps , i . e . we first collected linear RFs outputs with a given input stimulus , and then we used them to compute the Gaussian estimates; this is an abstract schematization of how V1 firing rates are produced , conceptually similar to the canonical linear-nonlinear scheme 51 , and does not imply that the two steps are actually preformed by separate V1 mechanisms ., We will address in the following section the relation with divisive normalization and the possible neural mechanisms underlying the computations presented in the remainder of this section ., For a given input stimulus , we first collected the outputs of center RFs k and surround RFs S . We then computed the expected value of the Gaussian variable corresponding to a center RF ( e . g . , tuned to vertical ) : ( 2 ) Equation 2 comprises the sum of the estimates under each of the mixture components ( respectively denoted and ) , weighted by the posterior probability of the components , and ., In the first term the surround is not co-assigned ., The second term sums over each of the components corresponding to the co-assigned surround orientations ., We derived an analytical form for both the mean estimates and the posterior probability , as described below ., Eventually , we combined the estimates ( Equation, 2 ) for the two phases of the RF ( denoted by and ) , to obtain orientation tuned , phase invariant responses: ( 3 ) We obtained the mean estimates under each of the components as follows ., In the GSM , given knowledge of the value of the mixer , one can obtain by definition as ., However , the visual system does not know the value of , and so can only perform statistical Bayesian inference using information about the prior distribution of ., We chose a Rayleigh prior for the mixer variables which gave analytically tractable solutions; however , the qualitative behavior of our simulations would be the same for a range of priors ( similar to 37 , 47; and see also priors in 46 ) ., This choice resulted in the following estimates for the co-assigned components: ( 4 ) and similarly , we obtained the estimate for the component in which the surround is not co-assigned: ( 5 ) where nk represents the number of RFs in the center group , and n the total number of center and surround RFs ., The terms denoted by are generically defined as follows: ( 6 ) where , in , x is to be replaced with the center filter outputs and C is to be replaced with ; and in , x is to be replaced with the center filters and the surround filters with orientation , and C is to be replaced with ., We introduced a small constant ε which sets a minimal gain when the filter activations are zero , to prevent infinities in Equations 4 , 5 when the RF outputs are zero ., In all the simulations , was set to a value ( 10−10 ) several orders of magnitude smaller than the smallest value observed for the other term under the square root ., is the modified Bessel function of the second kind , and the ratio of Bessel functions diverges at and asymptotes to 1 at infinity , at a rate that depends on n; it remains approximately constant over the range of values of in all our simulations ., Eventually , we inferred the posterior probability , , that the center group shares a common mixer with the surround group labeled , using Bayes rule: ( 7 ) and substituting the first term on the r . h . s . with the learned co-assignment prior , and the second term with its analytic form ( see below , Equation 10 ) ., The inferred model responses ( Equations 2 , 4 , 5 ) thus constitute a form of divisive normalization 40–42 , where the terms in the denominator represent the normalization signal , and the RFs that contribute to form the normalization pool ., It has been shown that divisive normalization has the effect of reducing the higher-order dependencies illustrated in Fig . 2C , D , and accounts for some data on V1 surround modulation 30 ., However , our model generalizes standard divisive normalization in two substantial ways ., First , consider the effect of covariance ., Normalization allows cells to discount a global stimulus property that is shared across RFs , i . e . contrast in simple divisive normalization , or the mixer value in the GSM ., The mixer corresponds to RFs that are statistically coordinated and tend to be high or low together in their absolute value ., However , in the generative model , large outputs from RFs that often respond together ( i . e . RFs with large covariance or linear correlations ) could be generated either by a large value of the mixer or by similar draws from the correlated Gaussians; whereas similar , large outputs from RFs that rarely respond together ( small covariance ) are more likely to have been generated by a large value of the mixer ., Therefore linearly correlated RFs should contribute less to the estimate of the mixer ( which is loosely proportional to the normalization signal ) ., This arises in the model since the covariance matrices learned from scenes weight the contribution of the RFs to the normalization signal ( Equation 6 ) ., For instance , a pair of RFs with large variances and large covariance between them ( often leading to negative values in the corresponding off-diagonal term of C−1 ) exerts less normalization than a pair of RFs with the same variances but small covariance ., In addition , an RF with large variance ( corresponding to small diagonal terms in the inverse covariance ) weights less than one with small variance ., In the visual cortex , the effect of such weights may be represented indirectly in the strength of excitatory connections that are set with development; indeed the higher correlations , in the model , between overlapping RFs , as well as non-overlapping collinear RFs , are qualitatively in agreement with the known specificity and anisotropy of horizontal and feedback connections 52–58 ( see also Discussion ) ., Second , Equation 2 uses a stimulus-dependent normalization pool , since , for any given input stimulus , only RFs that are inferred as being statistically coordinated and thus to share a common mixer , are jointly normalized ( the normalization pool comprises different RFs for each mixture component in Equations 4 , 5 ) ., The same RFs can be coordinated for some stimuli and not for others ., The computation involved , which in the model is distinct from the evaluation of the corresponding normalization signals ( Equation 6 ) , does not necessarily have to be segregated in the biological system , and may be achieved by a number of neural mechanisms ., One possibility may be that the normalization signals are computed by inhibitory interneurons that pool the outputs of distinct subpopulations: the different firing thresholds or diversity across types of interneurons 59 , 60 have been previously indicated as mechanisms that may control whether or not surround inhibition is active on a given input ., A complementary view is that surround modulation is an emergent property of the cortical network; in this scenario , the strength of the surround influence may be determined by stimulus-dependent switching between cortical network states 61 or changes in functional connectivity 62 , or by the exact balance between excitatory and inhibitory conductances , which are known to change in parallel with surround stimulation 63 ., For a given input stimulus , we inferred the posterior co-assignment probabilities of each mixture component , , using Bayes rule ( see Equation 7 ) ., These probabilities measure how well each component explains the input data ., Intuitively , the probability is large when the stimuli in the center and surround are similar ( e . g . for gratings of similar orientation and contrast ) , and for a given stimulus , it tends to be larger at high contrast as illustrated in Fig ., 5 . More precisely , is a function of the center and surround RFs outputs , combined using the corresponding covariance matrices as in Equation, 6 . In our implementation , for a given input stimulus the probability of co-assignment does not vary across center orientations ( because we grouped together all center RFs with different orientations ) ., The surround RFs are grouped together according to their orientation , and so all RFs within one surround group have the same probability of being co-assigned with the center , but each surround group has a different probability ( e . g . for a large vertical grating , the vertical surround has high probability , while the horizontal surround has low probability ) ., On each given stimulus , the probabilities across the 4 surround groups , plus the probability of no assignment ( i . e . that no one of the surrounds is co-assigned with the center ) , jointly add up to 1 ., Note also that , differently from standard divisive normalization which is inherently suppressive , our model encompasses both surround suppression and facilitation as summarized in Fig . 6: a strongly driven surround suppresses center responses , whereas a weakly driven surround facilitates , and the relative modulation is larger when the center RF is weakly driven ., The full distribution of the RFs variables under the MGSM is given by the mixture model: ( 8 ) in which the terms represent the joint distribution of the center and surround RF outputs k and S under each of the five possible mixture components , weighted by their prior probabilities ., The terms can be derived analytically as follows ., First , we assumed a Rayleigh prior on the mixer variables: ( 9 ) Under the configuration in which the surround of orientation is co-assigned with the center , we can exploit the independence among groups that do not share a mixer , and obtain: ( 10 ) where , in the first line , denotes the surround RFs with orientation , and similarly for ., In the second line the factors are derived as in a standard GSM , as detailed in 37 , 47; nk and nS represent the number of RFs in the center and surround groups respectively , and n\u200a=\u200ank+nS; is the covariance matrix among the center group and the surround group with orientation ; is the covariance of the surround with orientation ; is the modified Bessel function of the second kind ., The terms denoted by are defined as in Equation, 6 . Similarly , we can derive the distribution conditional on , namely in the case that all groups are mutually independent: ( 11 ) where is the covariance matrix of the center Gaussian variables , and is defined as in Equation, 6 . The parameters governing RF interactions – i . e . the covariance matrices and the prior probability of each component of the MGSM – need to be optimized for an ensemble of natural scenes ., The training data were obtained by randomly sampling 25000 patches from an ensemble of 5 natural images from a database of standard images used in image compression benchmarks ( known as Einstein , boats , goldhill , mountain , crowd; see Fig . S1; the images are downloadable from http://neuroscience . aecom . yu . edu/labs/schwartzlab/code/standard . zip ) ., The parameters to be estimated were the covariance matrices associated to the different components of the model , collectively denoted by , as well as the prior probability of each component , denoted by and ., The model learned the full covariances , including the terms coupling opposite RFs phases which were typically close to zero ., To find the optimal parameters we maximized the likelihood of the data under the model; we implemented a Generalized Expectation-Maximization ( GEM ) algorithm , where a full EM cycle is divided into several sub-cycles , each one involving a full E-step and a partial M-step performed only on one covariance matrix ., In the E-step we computed an estimate , , of the posterior distribution over the assignment variable , given the RF responses and the previous estimates of the parameters ( denoted by the superscript old ) ., This was obtained via Bayes rule: ( 12 ) and similarly for ., Then in the M-step we adopted conjugate gradient descent , to maximize the complete–data Log Likelihood , namely: ( 13 ) The training was unsupervised , i . e . we did not pre-specify the co-assignments of the training set , but rather let the model infer them at each E-step ., The EM algorithm is not guaranteed to find a global maximum; however , repeated training runs with different randomly chosen starting points produced convergence to similar parameter values ., More details are provided in Text S2 ( see also 47 ) ., We learned the parameters of the model ( i . e . , the covariance matrices and the prior co-assignment probabilities ) entirely from the natural scenes , and then fixed them ., We first evaluated the model by comparing its response ( Equation, 3 ) to presentations of different forms and sizes of sinusoidal gratings with those described in previous neurophysiological studies of spatial contextual modulation ., Physiology experiments have made extensive use of gratings to show that stimuli presented in regions of visual space that do not drive the neuron ( i . e . the surround ) can still strongly modulate the responses to a stimulus presented within the RF ., First , we tested gratings of variable size and contrast whose orientation and spatial frequency match the chosen RF ., Experiments in cat and macaque monkey V1 8 , 64 , 65 show that at fixed contrast , neural responses typically increase as a function of stimulus size up to a peak value , and then decrease for larger stimuli that recruit the suppressive surround ., The peak responses correspond to larger diameters at low than high contrasts ( Fig . 7A-left; the studies cited above reported an average expansion factor across the population in the range 2 . 3 to 4 ) ., Fig . 7A-right shows the similar behavior of our model , including a contrast-related expansion of similar magnitude from high to low contrast in our model ., For intermediate contrasts we observed a gradual peak shift ., The contrast-dependence of size tuning has previously been ascribed to divisive normalization 30 , 35 ., Cavanaugh et al . 8 showed that a divisive model with variable gains for the center ( numerator ) and surround ( denominator ) accounted well for the contrast-related expansion; model parameters obtained from data fitting showed that the relative influence of the surround increased with contrast , suggesting that the inhibitory surround is less sensitive than the center at lower contrasts ( but see also 62 , that supported the alternative hypothesis that low contrast enhances recurrent excitatory interactions ) ., In our model , the flexible assignment process provided a related , but statistically motivated explanation of the expansion ., For gratings of intermediate sizes that covered the surround only partially , the co-assignment probability was larger when the contrast was high than when it was low ( Fig . 7B; see also Fig . 5 ) ., Therefore larger stimuli were necessary to recruit surround modulation at low than high contrast ., This pattern of assignments was obtained directly from statistical inference rather than from data-fitting constraints , and did not depend on any asymmetry between center and surround ., We next addressed three sets of data related to the orientation tuning of surround modulation ( Fig . 8 ) , for which the involvement of multiple surround normalization pools in the model provided a novel , potentially unifying explanation ., We first verified that the model produces contrast-invariant orientation tuning curves ( Fig . S2 ) ., We then assessed how a surround annular grating modulated the response to a fixed , optimally-oriented central grating , as a function of their orientation difference ., Without loss of generality , we refer to the central grating as being vertical , both for the data and the model ., Several studies , e . g . 7 , 9 , 10 , 66–68 , reported that the strongest suppression ( relative to the response to the center grating alone ) occurred when the center and surround were similar in orientation , while large orientation differences led to little or no suppression , as exemplified by the cell in Fig . 8A-top-left 10 ., Our model reproduced these features of surround orientation tuning ( Fig . 8A-top-right ) ., We also explored the model behavior as a function of the stimulus contrast and the sizes of the central grating patch and surround annulus ( see Text S3 ) ., Generally speaking , in the model we observed surround facilitation when the center was not optimally stimulated , and the surround was weakly driven and co-assigned ( Fig . 6 ) ., For instance , when we introduced a gap to make the center patch smaller than the center RFs , and with the surround annulus covering only part of the surround RFs , we found facilitation for large orientation differences between center and surround ., The response was maximized with annuli tilted less than 90 degrees relative to the center ( Fig . 8A-bottom-right ) ., These conditions have not been explored systematically in experiments , but some evidence for the above effects was reported by studies that tested surround orientation finely ( e . g . , Fig . 8A-bottom-left , reported in 10 using mid contrast; 67 using low contrast in the center and high in the surround ) , and recent models of the perceptual tilt illusion have implicated facilitation from non-orthogonal surrounds 69 , 70 ., On the other hand , other experimental studies have not reported facilitation at all 7 , 68 , and 8 reported that an iso-oriented surround stimulus is always suppressive , regardless of the contrast of the center and surround stimulus ., Our model partly failed to reproduce this observation , since it produced facilitation when we combined small center patches with large gap sizes at low contrasts ( see Discussion; see also Text S3 ) ., The dependence of surround tuning on contrast , and more generally on how strongly the RF is driven , has yet to be explored more systematically ., The responses of the model ( Fig . 8A-right ) depended on the interaction between different surround components as follows ( Fig . 8B; see also Text S3 for more examples ) ., In the top row , the center grating patch partly activated surround RFs , leading to a high co-assignment probability for the vertical surround , regardless of the annulus orientation: the strength of the suppression then simply depended on how much the surround annulus increased the outputs of the vertical surround RFs ., In the bottom row , we reduced the size of the center stimulus ., In this case , for stimuli with small orientation differences ( less than 45 degrees ) between the ( vertical ) center and the surround we observed large co-assignment probability for the vertical surround group: in this regime , surround modulation was suppressive ., Modulation switched to facilitation as the orientation difference increased and surround strength decreased ( corresponding to reducing for fixed in Fig . 6 ) ., At much larger orientation differences , the co-assignment probability decreased for the vertical surround group , but in

Introduction, Materials and Methods, Results, Discussion

Spatial context in images induces perceptual phenomena associated with salience and modulates the responses of neurons in primary visual cortex ( V1 ) ., However , the computational and ecological principles underlying contextual effects are incompletely understood ., We introduce a model of natural images that includes grouping and segmentation of neighboring features based on their joint statistics , and we interpret the firing rates of V1 neurons as performing optimal recognition in this model ., We show that this leads to a substantial generalization of divisive normalization , a computation that is ubiquitous in many neural areas and systems ., A main novelty in our model is that the influence of the context on a target stimulus is determined by their degree of statistical dependence ., We optimized the parameters of the model on natural image patches , and then simulated neural and perceptual responses on stimuli used in classical experiments ., The model reproduces some rich and complex response patterns observed in V1 , such as the contrast dependence , orientation tuning and spatial asymmetry of surround suppression , while also allowing for surround facilitation under conditions of weak stimulation ., It also mimics the perceptual salience produced by simple displays , and leads to readily testable predictions ., Our results provide a principled account of orientation-based contextual modulation in early vision and its sensitivity to the homogeneity and spatial arrangement of inputs , and lends statistical support to the theory that V1 computes visual salience .

One of the most important and enduring hypotheses about the way that mammalian brains process sensory information is that they are exquisitely attuned to the statistical structure of the natural world ., This allows them to come , over the course of development , to represent inputs in a way that reflects the facets of the environment that were responsible ., We focus on the case of information about the local orientation of visual input , a basic level feature for which a wealth of phenomenological observations are available to constrain and validate computational models ., We suggest a new account which focuses on the statistics of orientations at nearby locations in visual space , and captures data on how such contextual information modulates both the responses of neurons in the primary visual cortex , and the corresponding psychophysical percepts ., Our approach thus helps elucidate the computational and ecological principles underlying contextual processing in early vision; provides a number of predictions that are readily testable with existing experimental approaches; and indicates a possible route for examining whether similar computational principles and operations also support higher-level visual functions .

computer science, computational neuroscience, synthetic vision systems, biology, sensory systems, neuroscience

journal.pcbi.1004953

The Computational Development of Reinforcement Learning during Adolescence

Adolescence is defined as the period of life that starts with the biological changes of puberty and ends with the individual attainment of a stable , independent role in society1 ., During this period , significant changes in value-based decision-making are observed2 ., Adolescents are often characterised as prone to engage in suboptimal decision-making , which although probably adaptive in many circumstances3–6 , can sometimes result in negative real life outcomes7 , 8 ., The computational framework of reinforcement learning formally captures value-based decision-making9 , 10 ., Reinforcement learning ( RL ) refers to the ability to learn to improve one’s future choices in order to maximise the expected value ., The simplest RL algorithm ( Q-learning ) learns action-outcome associations directly from experienced rewards on a trial and error basis11 , 12 ., However , more complex behaviours , such as counterfactual learning and punishment- avoidance learning cannot be explained using the basic RL algorithm , due to its computational simplicity ., Counterfactual learning refers to the ability to learn not only from direct experience , but also from hypothetical outcomes ( the outcomes of the option ( s ) that were not chosen ) 13 , 14 ., Punishment avoidance , compared to reward seeking , requires an additional computational step in which outcomes are considered relative to a reference point ( i . e . outcome valuation is contextualised ) 15 , 16 ., Thus , compared to simple reward seeking , counterfactual and avoidance learning are more computationally demanding ., Accordingly , whereas simple reward learning has been largely and robustly associated with the striatum17–19 , punishment and counterfactual processing have been consistently associated with the dorsal prefrontal system and the insula , areas that are classically associated with cognitive control13 , 20–23 ., Theories of adolescent brain development have pointed to differential functional and anatomical development of limbic regions , such as the striatum , and cognitive control regions and there is some evidence to support this notion 1 , 2 , 6 , 24–26 ., We hypothesise that this asymmetrical development might be translated into a difference in the computational strategies used by adolescents compared with adults ., Differences in reinforcement learning strategies may in turn contribute to an explanation of features of adolescent value-directed behaviour ., More precisely , we hypothesise that , while the basic RL algorithm successfully encapsulates value-based decision-making in adolescence , adults integrate more sophisticated computations , such as counterfactual learning and value contextualisation ., To test this hypothesis , adults and adolescents performed an instrumental probabilistic learning task in which they had to learn which stimuli had the greatest likelihood of resulting in an advantageous outcome through trial and error ., Both outcome valence ( Reward vs . Punishment ) and feedback type ( Partial vs . Complete ) were manipulated using a within-subjects factorial design ( Fig 1A ) ., This allowed us to investigate both punishment avoidance learning and counterfactual learning within the same paradigm ., In a previous study , model comparison showed that adult behaviour in this task is best explained by a computational model in which basic RL is augmented by a counterfactual learning module ( to account for learning from outcomes of unchosen options ) and a value contextualisation module ( to account for learning efficiently to avoid punishments ) ( Fig 2A ) 15 ., Our computational and behavioural results are consistent with our hypothesis and show that adolescents utilise a different , simpler computational strategy to perform the task ., During the learning task , participants made choices between two options , presented within different choice contexts ( Fig 1 ) ., In each context , one option had a higher probability of resulting in an advantageous outcome ( the ‘correct’ option; gaining a point or not losing a point ) than the other ., We submitted participants’ correct choice rate to computational analyses , based on an algorithm that has been shown to provide a good account for both behavioural and neural data within the same task in adults ( Fig 2A ) 15 ., In short , the model includes a factual learning module ( Q-learning ) , which updates the value of the chosen option ( governed by a first free parameter: α1 ) , a counterfactual learning module , which updates the value of the unchosen option ( governed by a second free parameter: α2 ) and , finally , a contextual learning module , which learns the average value of the choice context and uses this to move from an absolute to a relative encoding of option value ( governed by a third free parameter: α3 ) ., The counterfactual learning module has been shown to underlie the enhanced learning induced by the presence of complete feedback information , whereas the contextual learning model has been proposed to underpin the ability to perform similarly in both punishment and reward contexts ., Thus , our model space included three nested and increasingly sophisticated models ., Model 1 was a simple , option-value learning model ( Q-learning ) , with no counterfactual or contextual learning modules ( α2 = α3 = 0 ) ., Model 2 also included counterfactual , but no contextual learning ( α3 = 0 ) ., Finally , Model 3 was the “complete” model ., Model 3 can be seen as the most parsimonious translation into the reinforcement-learning framework of the fictive learning models and relative value-based decision-making models proposed in economics27 , 28 ., To describe the properties of the three models and illustrate how their performances differ across the different choice contexts ( states , ‘s’ ) , we ran ex-ante model simulations and analysed the model estimates of option values ( Q ( s , : ) ) and decision values ( ΔQ ( s ) ) ( Fig 2B ) ., Decision value is defined for each context as the difference in value between the correct and incorrect option ., Decision values ultimately determine the percentage of correct choices during the learning task ., Model 1 ( basic Q-learning ) predicts higher performance in the Reward compared to the Punishment contexts , a learning asymmetry predicted by the punishment avoidance learning paradox29 , and similar performance in the Partial and Complete feedback contexts ., Model 2 ( Model 1 plus the counterfactual learning module ) permits an improvement in performance in the Punishment/Complete context , however still predicts a learning asymmetry in the Partial feedback contexts ., Finally , Model 3 ( Model 2 plus the value contextualisation module ) predicts similar performance in the Reward and Punishment contexts and increased performance in the Complete compared to the Partial feedback contexts: this is the behavioural pattern that we expected for the adult group , based on our previous study15 ., We fitted the three models to individual histories of choices and outcomes , in order to obtain , for each participant and each model , the parameters that maximised the negative log-likelihood of participants’ choices during the learning task ( see S1 Table ) ., To assess whether baseline model fitting differed between adolescents and adults , we submitted the negative log-likelihood and the inverse temperature parameter ( β ) to mixed-design ANOVA with group ( Adolescents vs . Adults ) as the between-subjects factor and model as the within-subjects factor ., For negative log-likelihood ( a measure of model quality of fit ) , there was no main effect of group ( F ( 1 , 36 ) = 1 . 3 , P>0 . 2 ) and the group x model interaction did not reach significance ( F ( 2 , 72 ) = 2 . 7 , P<0 . 08 ) ., Note that the main effect of model cannot be tested since the models are nested and therefore the negative log-likelihood can only decrease ., Analysis of the inverse temperature ( β ) parameter supported these results ., This parameter can be taken as a measure of how well choices are predicted by the model and strongly correlates with the model likelihood ( for all models: R>0 . 93; P<0 . 001 ) ., There was no main effect of group ( F ( 1 , 36 ) = 2 . 3 , P>0 . 1 ) but there was a significant group x model interaction ( F ( 2 , 72 ) = 5 . 0 , P<0 . 01 ) ( Fig 3A ) ., Post-hoc comparisons showed that this interaction was driven by adults showing increases in inverse temperature when comparing Model 1 to Model 2 ( T ( 19 ) = 3 . 2 , P<0 . 01 ) and Model 2 to Model 3 ( T ( 19 ) = 2 . 2 , P<0 . 05 ) ., Baseline ( Model 1 ) inverse temperature did not differ between adults and adolescents ( T ( 36 ) = 0 . 4 , P>0 . 70 ) ., The absence of main effects of group indicates that baseline quality of fit was not different between age groups , thus allowing further model comparison analyses ., Posterior probability ( PP ) was calculated for each of the models , using the log of the Laplace approximation of the model evidence ., Similar to other model comparison criteria , quality of fit is penalised by the model complexity30 ., As above , we submitted the PP of the models to a mixed-design ANOVA with group as the between-subjects factor and model as the within-subjects factor ( Fig 3B ) ., This analysis indicated a significant group x model interaction ( F ( 2 , 72 ) = 38 . 9 , P<0 . 001 ) ., The effect of model was not quite significant ( F ( 2 , 72 ) = 3 . 0 , P<0 . 06 ) ., Note that the main effect of group cannot be tested , since the model posterior probabilities by definition must sum to one , thus creating equal group means ., Post-hoc comparisons showed that in the adolescent group , the posterior probability of Model 1 was significantly greater than chance level ( T ( 17 ) = 3 . 0 , P<0 . 01; exceedance probability = 0 . 77 ) and greater than that of the adult group ( T ( 36 ) = 8 . 0 , P<0 . 001 ) ., Conversely , in adults , the posterior probability of Model 3 was significantly greater than chance level ( T ( 19 ) = 5 . 2 , P<0 . 001; exceedance probability = 0 . 80 ) and greater than that of the adolescents ( T ( 36 ) = 7 . 8 , P<0 . 001 ) ( see also Tables 1 , S1 , S2 and S3 ) ., This result indicates that different computational models explain learning behaviour in the two groups ., More precisely , a simple RL model better describes adolescents’ behaviour , whereas a more complex model , which integrates counterfactual and contextual learning processes , better accounts for adults’ behaviour ., Our model comparison analyses suggest that adults and adolescents do not use the same computational strategy ( Fig 3B ) ., If this is the case , this computational result should be reflected in behavioural differences between the two groups ., To verify this , we analysed the correct choice rate learning curves using a mixed-design ANOVA with group ( Adolescents vs . Adults ) as the between-subjects factor and trial ( 1:20 ) , valence ( Reward vs . Punishment ) and feedback information ( Partial vs . Complete ) as within-subjects factors ( Fig 4A ) ., There was a significant main effect of trial on correct choice rate ( F ( 19 , 684 ) = 26 . 8 , P<0 . 001 ) , in which the rate of correct choices increased over the course of the learning task ., There was also a significant interaction between group and trial ( F ( 19 , 684 ) = 5 . 7 P<0 . 001 ) , which was further moderated by valence ( F ( 19 , 684 ) = 2 . 0 , P<0 . 01 ) ., This suggests that adults and adolescents differed in the way their correct choice rate evolved during learning and that this difference interacted with outcome valence ( Reward vs . Punishment ) ., Post-hoc comparisons performed on the correct choice rate improvement ( the difference between the first and last trials ) indicated that , compared to adults , adolescents showed lower correct choice rate improvement in the Punishment/Partial context ( T ( 36 ) = -2 . 9 , P<0 . 01 ) ( Fig 4B ) ., Post-hoc comparisons performed on the correct choice rate in the final trial ( trial 20 ) indicated that , compared to adults , adolescents had lower rates of correct choice in the Punishment/Complete context ( T ( 36 ) = -2 . 1 , P<0 . 05 ) ( Fig 4B ) ., Finally , while there was no significant interaction between feedback information and group , exploratory analyses indicated that whereas adults performed better in Complete feedback contexts ( final correct choice rate: T ( 19 ) = 2 . 7 , P<0 . 05 ) , adolescents showed no such positive effect of counterfactual information on correct choice rate ( T ( 17 ) = 0 . 9 , P>0 . 4 ) ., To summarise , adolescents displayed reduced punishment learning compared to adults ., Also consistent with our computational analyses , adolescent performance did not benefit from counterfactual feedback , although the interaction with group did not reach statistical significance ( see Table 2 ) ., The behavioural analyses support the model comparison analyses , suggesting that adolescents implement a simpler computational model than adults ( Fig 4A and 4B ) ., To further verify the ability of the models to reproduce the observed behaviour , we used the optimised model parameter values to simulate correct choice rate ( ex-post model simulations; see Methods ) ., Trial-by-trial model estimates of the probability of choosing the correct response in the learning task were generated for each participant using the best fitting model for their age group ( i . e . Model 1 for adolescents; Model 3 for adults ) ., Model-simulated data were submitted to the same analyses as the behavioural data , which indicated significant group x valence x trial ( F ( 19 , 684 ) = 2 . 8 , P<0 . 001 ) , and group x feedback information x trial ( F ( 19 , 684 ) = 8 . 7 , P<0 . 001 ) interactions , consistent with the reduced capacity to learn from counterfactual information and to efficiently avoid punishments observed in adolescents ( Fig 4A ) ., Although reinforcement learning models and paradigms are primarily concerned with choice data , RTs are also supposed to carry relevant information concerning both option and decision values31 , 32 ., RTs were analysed in the same way as correct choice rate ., We analysed the RT curves with a mixed-design ANOVA with group ( Adolescents vs . Adults ) as between-subjects factor and trial ( 1:20 ) , valence ( Reward vs . Punishment ) and feedback information ( Partial vs . Complete ) as within-subject factors ( Fig 5 ) ., There was a significant main effect of trial on RT ( F ( 19 , 684 ) = 12 . 1 , P<0 . 001 ) , reflecting a learning-induced RT reduction ., There was also a significant main effect of valence ( F ( 1 , 36 ) = 9 . 6 , P<0 . 01 ) , and a significant interaction between valence and trial ( F ( 19 , 684 ) = 5 . 9 , P<0 . 001 ) , which reflected shorter RTs in the Reward compared to the Punishment contexts ., Post-hoc comparisons performed on the final RT reduction ( RTs at trial 20 ) indicated that both adults and adolescents showed higher RT ( i . e . slower responses ) in the Punishment compared to the Reward contexts ( adults: T ( 19 ) = 2 . 1 , P<0 . 05; adolescents: T ( 17 ) = 2 . 9 , P<0 . 05 ) ., We also found a significant interaction between feedback information and trial , indicating that RT reduction differed in Partial and Complete feedback contexts ( F ( 19 , 684 ) = 2 . 3 , P<0 . 001 ) ., There was no main effect of group on RT ( F ( 1 , 36 ) = 1 . 6 , P>0 . 2 ) , however there was a significant interaction between group and feedback information ( F ( 1 , 36 ) = 12 . 2 , P<0 . 01 ) , which was further moderated by trial ( F ( 19 , 684 ) = 4 . 1 , P<0 . 001 ) , indicating that RT reduction in the two groups was differentially influenced by the presence of counterfactual information ., Post-hoc comparisons performed on the RT reduction ( i . e . RTs at trial 1 minus RTs at trial 20 ) indicated that , compared to adults , adolescents showed less of a reduction in RT in the Reward/Complete context , which was not quite significant ( T ( 36 ) = 1 . 9 , P<0 . 06 ) and the Punishment/Complete context , which was significant ( T ( 36 ) = 2 . 2 , P<0 . 05 ) ( T ( 36 ) = 2 . 4 , P<0 . 05; when collapsed across the two Complete contexts ) ( Fig 5B ) ., Accordingly , whereas adult RT was reduced in the Complete compared to the Partial context ( -89 . 8ms: T ( 19 ) = 2 . 4 , P<0 . 05 ) , adolescents increased their speed ( +10 . 7ms; T ( 17 ) = 1 . 8 , P<0 . 09 ) ., To summarise , in both age groups RTs are slower in the Punishment compared to the Reward contexts , which is consistent with an implicit Pavlovian inhibition effect32 ., Consistent with the model comparison analyses and choice , the influence of counterfactual information on RT over the course of the learning task was reduced in adolescents compared to adults ( see Table 2 ) ., The post-learning test measured the ability to retrieve and transfer the value of the cues , as learnt by trial and error during the learning task ., Post-learning choice rate was extracted for each of the eight cues and analysed using a mixed-design ANOVA with group ( Adolescents vs . Adults ) as a between-subjects factor , and cue valence ( Reward vs . Punishment ) , feedback information ( Partial vs . Complete ) , and cue correctness ( Correct vs . Incorrect ) as within-subject factors ., There was a significant effect of valence ( F ( 1 , 36 ) = 92 . 2 , P<0 . 001 ) on post-learning choice rate , indicating that cues associated with Reward ( G75 and G25 ) were preferred over those associated with Punishment ( L25 and L75 ) ., Similarly , Correct cues ( G75 and L25 ) were preferred over Incorrect ones ( G25 and L75; F ( 1 , 36 ) = 38 . 1 , P<0 . 001 ) ( Fig 6 ) ., These effects indicate that , overall , participants were able to retrieve the value of the cues during the post-learning test ., Crucially , the analysis also revealed a significant interaction between feedback information and cue correctness ( F ( 1 , 36 = 11 . 6 , P<0 . 01 ) , which was further moderated by group ( F ( 1 , 36 = 6 . 0 , P<0 . 05 ) ., Post-hoc between-groups comparisons of these difference scores ( Fig 6 and Table 3 ) indicated that cue discrimination was significantly lower in the adolescents than in the adults in both the Complete contexts ( Reward/Complete: T ( 36 ) = -2 . 4 , P<0 . 05; Punishment/Complete: T ( 36 ) = -2 . 6 , P<0 . 05 ) ., While adults showed improved cue discrimination in Complete contexts compared to Partial contexts ( T ( 19 ) = 4 . 1 , P<0 . 001 ) , adolescents did not ( T ( 17 ) = 0 . 6 , P>0 . 5 ) ., To summarise , in adults , cue value retrieval in the post-learning test was enhanced for cues associated with counterfactual feedback during the learning task ., Adolescents did not show this effect ., We also tested the model’s ability to account for choices made in the post-learning test ., Under the assumptions that choices in the post-learning test were dependent on the final option values in the learning task , and that there was no significant memory decay between the two tasks , the post-learning test , as in previous studies , can be used as an out-of-sample measure to compare the predictions of the different models33 , 34 We calculated the probability of choice in the post-learning test using a softmax function , using the same individual choice inverse temperature optimized during the learning task ( note that similar results have been obtained by optimising a beta specific to the post-learning test ) ., Again , we submitted the model-simulated post-learning choice rates to the same statistical analyses as the behavioural data ( Fig 6 ) ., Analysis of the model-simulated choices in the post-learning test also showed a significant group x feedback information x correctness interaction ( F ( 1 , 36 ) = 13 . 0 , P<0 . 001 ) , consistent with the behavioural finding of enhanced cue value retrieval in adults for cues associated with counterfactual information that was not observed in adolescents , and the model comparison analyses ., As indicated by the ex-ante model-simulated option values , higher cue discrimination in both the Reward/Complete and Punishment/Complete contexts and inverted preferences for intermediate value cues ( i . e . small gains and small losses ) requires both counterfactual learning and value contextualisation ( Fig 2B ) ., Within the factorial design of our task , the Reward/Partial context represented a “baseline” learning context ., From a computational perspective , this context is the simplest as participants can efficiently maximise rewards by directly tracking outcome values using a basic model of reinforcement learning ( RL ) ., Neuroimaging and pharmacological studies have demonstrated the importance of subcortical structures , particularly the ventral striatum , in this basic reward-value learning10 , 35 ., The striatum shows earlier anatomical maturation compared with the more protracted development of the prefrontal cortex24–26 ., Basic reward seeking has also been associated with the dopaminergic modulation of the striatum33 , 36 , 37 , and animal studies show that striatal dopamine peaks during adolescence38 , 39 ., A previous task using a simple reward maximisation task , comparable to our Reward/Partial condition , showed stronger encoding of reward learning signals in the striatum in adolescents compared to adults , with no negative behavioural consequences 2 ., Consistent with these data , we observed no differences between age groups in basic reward learning in the Reward/Partial context ., The similar performance between groups in the Reward/Partial context provides evidence that the group differences we observed concerning punishment and reward learning cannot be explained by a generalised lack of motivation or attention , but rather are likely to be associated with specific computational differences ., While less extensively studied than simple action-value learning , previous neuroimaging and computational studies of counterfactual learning suggest that learning from the outcome of the unchosen option recruits dorsolateral and polar prefrontal structures13 , 14 , 21 ., We hypothesised that , since these regions are still developing in adolescence 40–43 , adolescents would display a reduced ability to learn from counterfactual feedback ., Both our computational and behavioural analyses ( specifically the reaction times and post-learning test ) supported this prediction ., This reduced integration of counterfactual outcomes in adolescent behaviour is also consistent with a previous study showing limited feedback use as a possible source of higher risky decision-making during adolescents44 ., Counterfactual learning can also be understood within the framework of “model-based” ( as opposite to “model-free” ) RL45 , 46 ., Algorithms that operate without using a representation ( model ) of the environment , such as basic Q-learning , are termed model-free ., Conversely , algorithms that build option values by simulating different possible courses of action ( i . e . planning ) , based on an explicit model of the environment ( the task ) , are termed model-based ., Counterfactual learning can be conceptualised as a “model-based” process , as it involves the updating of option values according to mental simulations of what the outcome could have been if we had chosen an alternative course of action21 ., Like counterfactual learning , model-based learning has been theoretically and experimentally associated with prefrontal systems47–49 ., A key area for future research will be to examine whether or not the developmental changes in counterfactual learning observed here generalise to and interact with other forms of computation implicated in model-based learning , such as state transition learning ., In our task , symmetrical performance in the reward seeking and punishment avoidance learning conditions depends on the ability to contextualise outcome values ., Value contextualisation consists of updating option value as a function of the difference between the experienced outcome and an approximation of the average value of the two options ( i . e . the context value ) ., Thus , in punishment contexts , where the overall context value is negative , an intrinsically neutral outcome ( neither gaining nor losing points: 0pt; Figs 2A and 4 ) acquires a positive value and can therefore reinforce selection of the options that lead to successful avoidance of punishment ., In the absence of value contextualisation , the neutral outcome , which represents the best possible outcome in the punishment contexts will inevitably be considered as less attractive than a positive outcome ( the best possible outcome in the reward contexts: +1pt ) , and consequently the participant will perform less optimally in punishment contexts ., Previous studies of punishment avoidance learning , using the same or similar tasks as ours , have implicated the dorsomedial prefrontal cortex and dorsal anterior cingulate cortex in the representation of negative values and negative prediction errors20 , 22 ., Similarly to counterfactual learning , we predicted that adolescents would show reduced punishment avoidance learning based on the continuing development of prefrontal “control” regions ., Indeed , our results demonstrated that adolescents were less likely to engage in value contextualisation computation and thus showed less effective punishment avoidance learning and different cue evaluation in the post-learning test ., Thus , our results provide a computational substrate to neurobiological theories pointing to a reward/punishment imbalance as a driving force of adolescent risk- and novelty-seeking behaviour6 , 24 , 26 , 50 ., Previous studies of punishment avoidance learning in adolescents have elicited somewhat inconsistent results ., While some studies showed a reduction of punishment learning in adolescents51–53 , others reported no effect of valence54 , or even higher performance in punishment than reward contexts55 , 56 ., One possible way to reconcile these discrepancies is to consider the modular nature of computational RL ., In addition to value contextualisation , at least one other learning process , the Pavlovian inhibitory system , has been implicated in punishment avoidance learning32 ., According to this theory , and supported by experimental findings , Pavlovian expectations may influence choice behaviour via Pavlovian-Instrumental Transfer ( PIT ) 57 ., In instrumental tasks , PIT is observed in the form of increased motor inertia for actions leading to potential harm ( losses ) ., Since Pavlovian learning has been shown to be underpinned by subcortical structures , such as the amygdala , which mature relatively early in adolescence 40 , 58 , it is possible that PIT occurs similarly in adolescents and adults ., We would predict that , for avoidance tasks that rely only on PIT , adolescents and adults would display similar performance , whereas in tasks that require value contextualisation ( such as multi-armed bandit tasks , with probabilistic outcomes ) , adolescents and adults would not behave similarly ., To investigate the Pavlovian inhibitory system in adolescent we considered the reaction time from stimulus onset to the decision point ., We found that in both adolescents and adults , RTs were longer in Punishment than in Reward contexts ., Interpreted within the framework of Pavlovian-Instrumental Transfer learning , this effect may reflect an increase in motor inertia of actions associated with potential losses ., In other words , punishment avoidance actions require more time to be performed , compared to reward seeking actions , because avoidance is more naturally linked to “nogo” responses ., It is possible that in adolescents the Pavlovian inhibitory system is fully responsive and can mediate successful punishment avoidance in tasks that do not require value contextualisation55 ., Finally , RT profiles differed between adolescents and adults in the Reward/Complete context , which may provide supplementary evidence of reduced counterfactual learning in adolescents ., This “multiple systems” account of avoidance learning is also consistent with the proposal that reward/punishment imbalance in pathology , development and aging , could be underpinned by different neurophysiological mechanisms59 , 60 ., From a methodological perspective our study underlines the importance of using computational approaches to study the development of learning and decision-making61 , 62 ., Few studies have used computational models to interpret adolescent behaviour63–65 , and fewer still have implemented model comparison techniques51 , 54 ., Behavioural measures provide a relatively rough measure of performance in learning tasks for the following reasons ., First , in probabilistic learning tasks an incorrect response , as defined by the experimenter with knowledge of the task design , may locally be a “correct” response , according to the actual history of choices and outcomes experienced by the participant , as a function of misleading trials ., Second , the final estimation of learning performance may be affected by differences in initial choice rate ., For example , a participant who starts choosing the correct option by chance is favoured compared to a participant who would need to “explore” the options in order to find out the correct option ., Third , aggregate model-free analyses are not able to formally tease apart the possible computational processes underlying performance differences , which could be characterised either by differences in free parameter values within the same model , or by differences in the computational architecture itself ., By incorporating into the analysis the individual history of choices and outcomes , and formalising different learning mechanisms in discrete algorithmic modules , computational model-based analyses offer an elegant solution to these issues ., As such , our study , together with others , has be seen as part of a broader agenda aiming at moving from an “heuristic” to an “mechanistic” modelisation of human cognitive development66 ., Our results suggest that adolescents show heightened reward seeking compared to punishment avoidance learning and a reduced ability to take into account the outcomes of alternative courses of action ., Together , these processes may contribute to the adolescent propensity to engage in value-based decision-making ., Atypical value processing and learning are also implicated in multiple mental health disorders , at both the behavioural and neural level67 ., Increasing our understanding of normative changes in learning and decision-making during adolescence may thus provide insight into why adolescence is a period of increased risk for risky behaviours and mental health difficulties such as substance abuse and depression68 ., Finally , our results might also have implications for education , since they suggest that adolescents might benefit more from positive than from negative feedback when improving behavioural performance69 ., We recruited 50 volunteers aged between 12 and 32 years ., Adolescents ( N = 26; 12–17 years ) were recruited from a local Community Theatre and UCL volunteer databases; adults ( N = 24; 18–32 years ) were recruited from UCL volunteer databases ., The study was approved by the UCL Research Ethics Committee , and participants , or their legal guardians ( adolescents ) , gave written informed consent ., All participants were native English speakers and non-verbal IQ was assessed using the matrix reasoning subset of the Wechsler’s Abbreviated Scale of Intelligence ( WASI ) 70 ., Due to group differences in non-verbal IQ scores ( T ( 48 ) = 4 . 59 , P<0 . 001 ) , we restricted our analysis to those participants with scores falling within the range shared by both groups ., The lower level of the range was determined by the lower IQ of the initial adult group , the higher IQ level of the range was determined by the higher IQ of the initial adolescent group ., This gave a final sample of 38 participants , in which age groups ( 20 adults; 18 adolescents ) were matched in non-verbal IQ and gender composition ( see Table 4 ) ., All participants received a fixed a

Introduction, Results, Discussion, Materials and Methods

Adolescence is a period of life characterised by changes in learning and decision-making ., Learning and decision-making do not rely on a unitary system , but instead require the coordination of different cognitive processes that can be mathematically formalised as dissociable computational modules ., Here , we aimed to trace the developmental time-course of the computational modules responsible for learning from reward or punishment , and learning from counterfactual feedback ., Adolescents and adults carried out a novel reinforcement learning paradigm in which participants learned the association between cues and probabilistic outcomes , where the outcomes differed in valence ( reward versus punishment ) and feedback was either partial or complete ( either the outcome of the chosen option only , or the outcomes of both the chosen and unchosen option , were displayed ) ., Computational strategies changed during development: whereas adolescents’ behaviour was better explained by a basic reinforcement learning algorithm , adults’ behaviour integrated increasingly complex computational features , namely a counterfactual learning module ( enabling enhanced performance in the presence of complete feedback ) and a value contextualisation module ( enabling symmetrical reward and punishment learning ) ., Unlike adults , adolescent performance did not benefit from counterfactual ( complete ) feedback ., In addition , while adults learned symmetrically from both reward and punishment , adolescents learned from reward but were less likely to learn from punishment ., This tendency to rely on rewards and not to consider alternative consequences of actions might contribute to our understanding of decision-making in adolescence .

We employed a novel learning task to investigate how adolescents and adults learn from reward versus punishment , and to counterfactual feedback about decisions ., Computational analyses revealed that adults and adolescents did not implement the same algorithm to solve the learning task ., In contrast to adults , adolescents’ performance did not take into account counterfactual information; adolescents also learned preferentially to seek rewards rather than to avoid punishments , whereas adults learned to seek and avoid both equally ., Increasing our understanding of computational changes in reinforcement learning during adolescence may provide insights into adolescent value-based decision-making ., Our results might also have implications for education , since they suggest that adolescents benefit more from positive feedback than from negative feedback in learning tasks .

learning, decision making, social sciences, neuroscience, learning and memory, simulation and modeling, age groups, optimization, adults, cognitive psychology, mathematics, cognition, research and analysis methods, behavior, adolescents, people and places, psychology, biology and life sciences, population groupings, physical sciences, cognitive science

journal.pntd.0006284

The seasonal influence of climate and environment on yellow fever transmission across Africa

Yellow fever ( YF ) is a viral disease caused by the mosquito transmitted yellow fever virus ( YFV ) of the genus Flavivirus 1 ., Though endemic to both Africa and South America it is thought that 90% of cases occur in Africa 2 , where the virus causes between 51 , 000 and 380 , 000 severe cases per year resulting in around 19 , 000 to 180 , 000 deaths 3 ., Roughly half of those infected experience an asymptomatic infection and a further third a mild illness , however , about 12% of YF infections progress to severe disease , defined as fever with jaundice and/or haemorrhage , with a mortality rate of 30–60% 4 ., In Africa the disease is found in three transmission cycles:, ( i ) a sylvatic cycle where transmission is maintained between non-human primates via sylvatic Aedes mosquitoes , such as Ae ., africanus , that gives way to, ( ii ) an intermediate cycle where peri-domestic Aedes species act as a bridging point between humans and non-human primates ., This can lead to the establishment of, ( iii ) an urban cycle propagated by Aedes aegypti 2 , 5 ., In this urban cycle the disease is spread between humans and can establish itself in areas without a sylvatic host ., Historically , this has led to devastating epidemics which reached as far as The United States of America , England and Italy during the 16th-19th century 6 ., The primary vector species for YFV transmission in humans is the mosquito Ae ., aegypti 7 , which carries several other human viruses such as dengue and chikungunya 8 , 9 ., Ae ., aegypti , like all insects , is a poikilothermic ectotherm which means that its internal temperature is linked to the ambient temperature ., As temperature increases , so does metabolism ., This has a wide variety of implications for the transmission of YFV ., Higher temperatures , below temperature induced mortality , increase the speed of pupation 10 , raise the competence of the immune system 11 , the frequency of blood meals 12 and reduce the extrinsic incubation period ( EIP ) of YFV , defined as the time from a mosquito acquiring infection to becoming infectious 13 ., However , temperature is neither the sole determinant of Ae ., aegypti natural history , nor a standalone factor ., The interplay between precipitation and temperature may be more important than either variable alone: both optimal temperatures with insufficient rainfall and insufficient temperatures with optimal rainfall are unfavourable for Ae ., aegypti , and viral replication within the mosquito , therefore limiting the potential for transmission ., Previously , several models have used climatic factors in order to map the distribution and transmission of YF ., While Rogers et al . , ( 2006 ) 14 modelled the global distribution of YF using only satellite derived environmental data , Garske et al . , ( 2014 ) 3 used human population sizes , surveillance quality and geographical coordinates , along with environmental data to predict the distribution of YF in Africa ., This was then used in conjunction with serological surveys in order to generate the force of infection across the endemic zone and , taking into account vaccination coverage , the burden of disease ., In addition to YF , the distribution and transmission of several other vector-borne diseases such as malaria 15–19 and dengue 20–23 have been modelled using climatic factors ., Not only have static variables such as annual temperature and rainfall been shown to have substantial predictive ability , but the importance of seasonality has also been highlighted 16 , 19 , 24 , suggesting that taking into account seasonality in disease control and prevention interventions could translate to substantial public health gains 16 ., Characterising the relationship between YF occurrence and environmental factors , in combination with future predictions of these factors , could allow for forecasting of periods of “heightened risk” ., These predictions would then allow proactive intensification of surveillance to detect initial cases and respond with appropriate measures such as vaccination or vector-control ., Despite previous success in utilising climate and seasonality to evaluate disease dissemination and intensity , this approach has not previously been applied to YF ., The importance of further quantifying the epidemiology behind YFV transmission has been highlighted in recent years by large scale outbreaks in Angola , the Democratic Republic of the Congo and Brazil 25 , 26 ., The aim of this study was to investigate the effect of climate and the role of seasonality in YFV transmission using a limited number of temporally varying covariates , in contrast to the initial stage of the Garske et al . , ( 2014 ) 3 which used a wide array of static environmental and non-environmental covariates ., To this end we developed a temperature suitability index for the transmission of YFV by Ae ., aegypti which captures our understanding of the temperature dependence of the mosquito life cycle and viral replication mechanistically , following similar approaches used for malaria and dengue 8 , 17 ., The predictive power of this index was assessed alongside other variables by fitting a series of multilevel logistic regression models to YFV occurrence across Africa to capture both geographical and seasonal variation ., Disease occurrence in a province was defined as at least a single laboratory confirmed YF case ., An occurrence dataset was compiled from various sources , including YF reports between 1971 and 2015 from the Weekly Epidemiological Record 27 and the Disease Outbreak News 28 , as well as individual reports of laboratory confirmed cases between 2007 and 2013 from the Yellow Fever Surveillance Database , see Garske et al . , ( 2014 ) 3 ., Each of the reports were geo-located to the first sub-national administrative level , here termed province , and provinces across Africa were classified into the presence or absence of any confirmed YF reports from any data source over the study period ., Furthermore , we generated a seasonal dataset , where we similarly classified each province and calendar month into YF presence or absence , based on the start month of the outbreak or individual reported case ., Surveillance quality was informed by the volume of reporting suspected YF cases into the Yellow Fever Surveillance Database ( YFSD ) ., This collects data on suspected cases of YF using a very broad case definition based on the syndrome fever and jaundice , of which only 1–2% are laboratory confirmed while the remaining suspected cases , which form the vast majority , are due to other causes ., Assuming an approximately constant incidence of the syndrome “fever and jaundice” across the region covered , we used the per-capita rate of reporting suspected cases as a proxy for the surveillance quality in participating countries ., With non-participating countries assigned the mean value ., Country and year specific population sizes were obtained from the UN World population prospects ( UN WPP 2014 ) 29 and then averaged over the study period ( 1971–2015 ) to obtain average population sizes ., Population sizes were disaggregated to the first sub-national administrative unit , termed province , by aggregating the LandScan 2011 30 population estimates from a 1km grid 3 to the province level to calculate the proportion of the national population within each province ., This proportion was assumed to be constant through time to calculate the year-specific population sizes by province from the UN WPP national population sizes ., Datasets for seasonal air temperature 17 , the enhanced vegetation index ( EVI ) , an optimised satellite derived measure of vegetation , 31 and rainfall 32 , from 2003 to 2006 , the period for which all this data was available , were used as covariates ., These were chosen a priori based on the ability to plausibly explain their influence on transmission mechanistically and adoption in models from literature ., This data was available in grids at a resolution of between 1 and 10km , which we aggregated to the province level , by calculating the population-weighted mean value , based on the spatial population distribution from the LandScan dataset 30 , in order to generate environmental values that are representative of human habitation ., While the timescale of this data is temporally mismatched with the case data , previous work by Garske et al . , ( 2014 ) has found the approach taken here is a valid simplification of fitting to datasets that vary annually , not just seasonally ., Given the uncertainty and sparsity of underlying case data the model complexity is appropriate considering the richness of data ., All environmental datasets were stratified geographically to the province level and we used overall means through time as well as monthly and weekly values averaged across years to describe the typical seasonal patterns ., The monthly and weekly datasets were obtained through Fourier transforms 17 to produce the smoothed and averaged outputs desired ., The Ross-Macdonald model for mosquito borne disease transmission 33 defines the basic reproductive number , R0 , as, R0=ma2bce−μEIPγμ ., 1, Here , the vector to host ratio m , the probability of transmission between host and vector and vector and host in a single infectious bite , b and c , respectively and the recovery rate of the human host γ , are assumed to be temperature independent , while the mosquito mortality rate μ , the extrinsic incubation period ( EIP ) and the biting rate a , are assumed to be temperature dependent ., The temperature suitability index z , is then defined as the temperature dependent factors of the basic reproduction number 8 , 17 , 18 ,, z ( T ) =a ( T ) 2e−μ ( T ) EIP ( T ) μ ( T ) ., 2, This provides a single value parameterising the suitability for transmission at a given temperature T . The vector to host ratio , m , is likely to vary between locations and seasonally due to , among other factors , the abundance of breeding sites made available by rainfall ., While the transmission probabilities may depend on the vector competence of local mosquito species , the human recovery rate is likely less variable geographically ., Humans are only viraemic during the initial period of infection and not during the severe stage of disease 34 ., The temperature dependent death rate μ ( T ) , was parameterised by fitting a piecewise linear relationship to mosquito mortality data for Ae ., aegypti at different temperatures , with a cut point at 41°C 35 , 36 ., We fitted a linear regression to the data on the temperature dependence of mortality and biting rates ( Fig 1 ) ., For mortality rates we fitted these separately for low ( below 41°C ) and high ( above 41°C ) temperature regimes , following exploratory analysis ., Assuming only one bite per gonotrophic cycle ( i . e . the biting rate equals the oviposition rate ) and that biting occurs solely on human hosts we fitted a linear relationship between temperature and biting rate to data on the temperature dependence of Aedes feeding using 37 a polynomial relationship ., The EIP shortens with increasing temperature , and the rate of turning infectious can be described by a certain number of degree-days required for viral replication 15 , 38 , 39 ., For this we used the temperature dependent relationship described in Johansson et al . , ( 2010 ) 13 ., Suitable temperatures alone are not sufficient for mosquito survival and disease transmission; they need to coincide with rainfall in order to produce an optimal climate for transmission , as such the interaction between temperature suitability and rainfall was also considered as a potential covariate ., In the seasonal version this was defined as the product of the monthly temperature suitability index and the rainfall during the previous month ., This delay was implemented following exploratory analyses of the data , and may reflect the delayed impact rainfall has on the availability of mosquito breeding sites ., To account for the seasonal synchronicity between temperature and rainfall we averaged the seasonal temperature suitability and rainfall interaction across the year for use in the annual model ., Seasonality was investigated using a series of multivariable models 40 fitted to the YF report dataset using a complementary log-log link function , with a variety of sets of candidate environmental covariates ., This is conceptually similar to running a logistic regression with each data point referring to a province and calendar month ., However , unlike in a logistic regression , multilevel models allow for parameters to vary by group , here the province , in order to avoid biases introduced by treating monthly covariates within a province as independent 41 ., These models were fit to the whole continent of Africa ., All models included the log10 of population size and the surveillance quality proxy for countries included within the YFSD 3 , and the mean value of participating countries given to countries not included ., Furthermore we included each possible subset out of the four considered environmental covariates , the temperature suitability index , rainfall , the interaction of temperature suitability and rainfall , and the EVI , resulting in 15 different models that each included at least one of the climatic covariates ., Using these covariates we fitted multilevel logistic regression models to the presence/absence of YF reports by province in annual as well as seasonal models ., Annual models were fitted to the overall presence/absence of YF reports using annual mean covariates , while seasonal models were fitted to the presence/absence of YF reports in each calendar month , using monthly varying covariates within each province ., The seasonal models were fitted in a multilevel framework , where time points were nested in locations ., In order to facilitate the comparison of model predictions between the seasonal and annual models we converted the predictions of the regression model quantifying the monthly probability of YF reports to annual values using, pyear=1− ( ( 1−p1 ) ( 1−p2 ) … ( 1−p12 ) ) ,, 3, where p1 refers to January , p2 to February etc ., We refer to these models as compound seasonal models ., Models were fitted using every possible combination of environmental covariates , yielding 15 different variations for the annual and seasonal models each ., Models were scored based on Akaike Information Criterion ( AIC ) and the predictions from any model with an AIC less than 5 higher than the best performing model , as defined as the model with the lowest AIC value , were combined using Akaike weights 42 ., This is achieved by computing the differences in AIC ,, Δi=AICi−min ( AIC ) ,, 4, which was then used to obtain an estimate of the relative likelihood of model i , in proportion to the other models k = 1 … K included in the combined estimate through ,, wi=exp{−12Δi}∑k=1Kexp{−12Δk} ., 5, This provided model specific weights , wi , which we applied to the predicted values of the specific model , pi , to generate a single set of Akaike weighted predictions , pA ,, pA=∑k=1Kwkpk ., 6, For each set of covariates we assessed model fit using ROC ( receiver operating characteristic ) curves and the AUC ( area under the curve ) 42 ., The ROC is a graphical plot which illustrates diagnostic ability of a test , with the AUC providing a numerical value of this ability ., An AUC value of 0 . 5 indicates the diagnostic ability is no greater than random and a value of 1 as completely predictive ., The validity of predictions was ascertained through leave-one-out cross validation 43 , where we divided the dataset by randomly assigning countries to one of five non-overlapping subsets , then fitted the models to the dataset omitting each of the subsets in turn to generate out-of-sample predictions for the omitted subset ., This was repeated 10 times , resulting in 10 different allocations of provinces into subsets ., For each province , the average was taken across the 10 realisations ., Out-of-sample predictions for each subset were combined to generate a full set of out-of-sample predictions and its AUC compared with the full model ., All calculations and analyses were conducted in R version 3 . 2 . 5 ., A linear model fit to temperature dependent mortality data provided by Yang et al . , ( 2009 ) 35 predicts a mortality rate of 0 . 190 at 0°C which falls to a low of 0 . 027 at 25°C , before rising to 0 . 085 at 40°C ., A model fit to data provided by Christophers ( 1960 ) 36 at higher temperatures shows a rapid increase from 0 . 380 at 41°C to 1 . 000 at 47°C ( Fig 1A ) ., A linear model fit to temperature dependent biting rate , from Martens ( 1998 ) 37 , rises from 0 at 0°C to 4 . 5 at 47°C ., We identified 167 unique province-months with YF reports for which the month of the report was known and could be identified at the province level , occurring in 105 unique provinces ( Fig 2A ) ., Model predictions of the probability of a YF report across Africa from the annual model reproduced the geographical distribution of YF reports well , with an AUC of 0 . 83 ( 95% CI 0 . 80; 0 . 87 ) ( Fig 2B , see also fig in S3 Text for ROC curves of best performing models ) ., The YF report probabilities range from 0 for much of Southern and Northern Africa , to above 0 . 70 in parts of the Democratic Republic of the Congo ( DRC ) , Sierra Leone , Ghana and Côte dIvoire ., East Africa has notably low report probabilities except in Sudan and South Sudan ., The spatial distribution of the mean annual covariates is shown in Fig 2C–2F ., The areas with the highest mean annual temperature suitability index were found to be the Sahara and parts of the Horn of Africa ( Fig 2E ) ., Notably high index scores were found across West and parts of Central Africa , within the YF endemic zone 3 ., The temperature suitability is lower throughout Northern and Southern Africa , as well as regions of higher elevation in East and Central Africa ., Temperatures encountered in the Sahara appear highly suitable for transmission but the lack of rainfall is a limiting factor for transmission ( Fig 2D ) ., The largest values of the mean annual interaction of temperature suitability and rainfall ( Fig 2F ) are clustered around West and Central Africa , roughly approximating the regions where the burden of YF is the highest ., North and South Africa have negligible levels throughout , and East Africa shows marginally raised levels in limited locations ., These patterns are similar to the spatial distribution of the EVI ( Fig 2C ) , although the interaction of temperature suitability and rainfall shows a stronger concentration in the highly endemic zone in West and Central Africa than the EVI ., Fitting multilevel logistic regression models to time independent covariates ( surveillance quality and population sizes ) and monthly ( environmental ) covariate data produced seasonally variable predictions for the probability of YF reports across Africa with an AUC value of 0 . 81 ( 95% CI 0 . 79–0 . 84 ) ., The seasonality of YF reports , seasonal model predictions , the temperature suitability index , rainfall and EVI , differ between regions across Africa ( Fig 3 and S1 Movie ) ., The number of YF reports differs considerably between regions , but predominantly occur in September and October , with a secondary peak in spring/early summer across all regions , suggesting that there are two periods of heightened transmission annually ., Seasonal variation in the EVI is low , with the largest differences being observed in the Sahel ., This covariate has a significant effect on model fit ( Fig in S2 Text ) ., In YF endemic areas the temperature suitability index is fairly stable throughout the year while it drops to very low values in the respective winters in Northern and Southern Africa ., Rainfall and EVI patterns are broadly similar to each other , with EVI typically lagging behind rainfall by approximately a month ., These specific patterns differ across the continent with a single peak in the Sahel and West Africa , but with a bimodal pattern in Central and East Africa ., The interaction of temperature suitability and rainfall mirrors the EVI throughout much of Africa , albeit at a lower magnitude in all but the Sahel ., The probability of a report varies throughout the year , with an oscillating pattern rolling northwards across the continent from March where it peaks in the Sahel during September , then returns southwards , resulting in biannual peaks of risk in West Africa with the risk extending further east during the main peak in October ( Fig 3 , Fig 4 and S1 Movie ) ., The probability of YF reports , as predicted by the seasonal model varies geographically , with very low predictions in Northern , Southern and East Africa ., The probability of a YF report in the Sahel is minimal throughout most of the year but rises sharply from around June to a peak in September , while in West Africa there is a substantial YF report probability throughout the year , apart from January to March where it is slightly lower ., The seasonal variation of the probability of a YF report in Central Africa is synchronous to that in West Africa , although the amplitude of the variation is much reduced ., As for the annual model ( Fig 2 ) , the probability of a report is relatively high in the Democratic Republic of the Congo due to the large population size in these provinces ., For a given individual risk of infection , the probability of an report occurring is larger in a large underlying population , so this heightened probability reflects the aggregation level of populations more than the transmission intensity ., Aggregating the seasonal model predictions to a compound seasonal model gives very similar results to the annual model both in AUC and predictions , despite the magnitude of predictions being lower in the compound seasonal model ( see S4 Text for further details ) ., In both the annual and the seasonal models the EVI was a significant coefficient in all models included in the weighted model , and is a significant predictor of model fit ., For the weighted annual model , all contributing models also included the temperature suitability index , and for the weighted seasonal model , the interaction of the temperature suitability index and rainfall was included in all contributing models ( S2 Text ) ., Leave-one-out cross validation of the annual , compound seasonal and seasonal model assessed the predictive performance of our models and found the AUC values of out-of-sample predictions were not significantly different from the AUC values based on full model predictions , suggesting the models were not affected by over-fitting ( S5 Text ) ., In addition to this , each of the 10 out-of-sample realisations had overlapping confidence intervals indicating the individual predictions were not significantly different ., In this study we developed a temperature suitability index for YFV transmission based on a mechanistic model of mosquito life history and YFV replication within the mosquito ., We fitted multilevel logistic regression models to datasets of YF reports using the temperature suitability index as well as other relevant climatic variables to estimate monthly and annual probabilities of YF reports across mainland Africa at the province level ., By using rainfall , the temperature suitability index , their interaction evaluated on a monthly timescale to capture synchronicity , the EVI ( enhanced vegetation index ) , and factors related to case detection and human populations , we have explained a large amount of the geographic and temporal variation in the distribution of YF reports ., While further variables associated with arbovirus transmission could potentially increase the specificity and sensitivity of the model , by focusing on a few well documented variables we serve to enhance the mechanistic understanding while still managing to explain a large amount of geographic and temporal heterogeneity ., The model fit , as quantified by the AUC , is not significantly different between the compound seasonal and annual models despite the monthly model capturing seasonal trends in transmission in addition to the geographic heterogeneity without overfitting ( S5 Text ) ., The results show that areas with high annual probabilities of a report do not always possess the same risk throughout the year ., The 167 YF reports utilised were selected based on the availability of information on the location at the province level and the month during which the outbreak started ., These reports rely on the accurate identification of YF cases , which can be problematic due to the presence of asymptomatic infections and diseases with similar symptoms 4 , 45 , 46 ., In order to account for the inconsistencies between countries’ identification of cases , country-specific surveillance qualities were calculated using the reporting rate of suspected cases of YF for countries contributing to the YFSD and a constant incidence of syndrome “fever and jaundice” ., For the remaining countries we assigned the mean surveillance quality value in the database to account for the level of under-reporting 3 ., While the incidence of syndrome “fever and jaundice” is unlikely to be constant across the same region , the majority of cases highlighted by this will be caused by hepatitis and as such will not have the same variability geographically as a vector-borne disease such as YF ., Additionally , due to the timespan investigated ( 1971–2015 ) errors may be introduced in disease reporting by shifting political boundaries ., While important to note , this is unlikely to result in substantial issues with geolocation of cases , particularly as for any cases where more detailed geographical information was available we mapped these using a consistent set of modern political boundaries ., The highest values of the temperature suitability index were found in areas with the highest temperatures , such as the Sahara , Sahel and parts of Southern Somalia ( Fig 2E ) ., While the temperature suitability index in these regions is highly suitable for the transmission of YF , the regions’ low levels of rainfall ( Fig 2D ) coupled with mosquitoes’ dependence on water for breeding lead to an inhospitable environment for mosquito reproduction ., The necessity for both optimal temperatures and rainfall is depicted through the oscillating peaks in the probability of YF reports which closely follow the interaction of temperature suitability and rainfall ( S1B and S1F Movie ) ., The absence of this interaction offers an explanation for the lowered report predictions in Angola and Uganda , despite a historical presence of cases ., Additionally , with respect to Angola and Uganda , it is important to note that by using covariates averaged over a period of years ( 2003–2006 ) we are capturing the typical pattern of seasonality , but neglecting inter-annual variability ., This results in the de-emphasising of abnormal weather patterns and climate cycles such as El Niño Southern Oscillation , which have been found to affect vector-borne disease transmission 20 , 47 ., While the effect of temperature induced mortality on the temperature suitability index may be exaggerated in our parameterisation , temperatures in our dataset have a maximum of 38°C and so this does not affect the calculations when applied to our data ., Additionally , we do not capture well the effect of low temperatures ( <10°C ) on mortality , but at these temperatures the extreme value of the EIP will dominate , resulting in a low temperature suitability index ., The interaction of the temperature suitability index and rainfall was significant in all models contributing to the combined seasonal model , lending credence to the idea that this interaction of rainfall and temperature is more important than either factor alone ., However , the EVI was found in all the best-fitting annual and seasonal models ( S2 Text ) ., As continent wide covariates the temperature suitability and rainfall interaction is highly correlated with the EVI , and ranges from medium to high correlations at the regional level ( S1 Text ) ., Coupled with its high predictability , this suggests the EVI may account for the interaction of rainfall and the temperature suitability index , while providing additional information not captured by either , potentially due to the EVI quantifying a more complex relationship between temperature and environmental suitability than what is captured by their direct interaction ., One explanation for this may be the EVI offering a more suitable proxy for the presence of standing water ., While precipitation heavily influences the availability of standing water , ground permeability and the topographic redistribution of water additionally affect the presence of standing water , and so mosquito breeding sites ., To more accurately predict the occurrence of YF , the availability of standing water and its interaction with temperature , rather than levels of precipitation , should be considered ., To further increase accuracy , this could take into account anthropomorphic water storage and its influence on sustaining mosquito populations in the absence of naturally maintained water sources 48 ., However , despite the worse performance of the interaction of temperature suitability and rainfall compared to the EVI as a predictor of YF reports , the former offers a mechanistic insight into the spatio-temporal variability and therefore enhances our understanding of the factors influencing YFV transmission intensity ., The temperature suitability index ( Eq 2 ) assumes the inclusion of all necessary temperature dependent variables ., However there are mechanisms that may be temperature dependent that we have not included due to the lack of suitable data , such as the vector to host ratio 11 , 49 and the tendency to take multiple blood meals per gonotrophic cycle 50 ., Regardless , though our temperature suitability index may only quantify part of the temperature dependence of transmission , we believe that it offers an adequate parameterisation ., While this model has only investigated the role of Ae ., aegypti as a vector , the principal vector of urban YF , within Africa there are over 20 species of Aedes that can transmit YF as well as members of Eretmapodites 51 , 52 ., Though there are likely to be some differences in climate suitability and vector competence between species , the underlying mechanisms are likely similar and any differences observed will be differences in model parameterisation ., Therefore despite the model being calibrated for Ae ., aegypti , it provides a good generalised estimate for all mosquito vectors , with the possibility of easy refinement for other species given relevant data ., Furthermore , the framework could also be adapted to describe the seasonality of further mosquito-borne diseases , with the fewest adaptations necessary for other viruses transmitted by Ae ., aegypti ., Despite climatic factors integral role in vector-borne disease transmission , they do not operate in a vacuum ., Socioeconomic conditions such as informal housing and lack of sanitation greatly influence the presence of A . aegypti 53 , creating conditions ideal for disease transmission ., The importance of taking these factors into account is highlighted by our models’ low predictions in Luanda , where the 2015–2016 outbreak started , and suggests that the inclusion of measures of socioeconomic status may improve predictions of the geospatial occurrence of YF ., In contrast to earlier work on the distribution of burden of YF in Africa by Garske et al . , ( 2014 ) 3 , this body of work presents an initial simple framework for explaining YF report occurrence through a limited number of covariates , as well as exploring the role seasonal variation in these has on reports of YF ., We believe that the work presented here is a valuable contribution in its present form as the limited number of covariates allows us to focus on enhancing the understanding of underlying processes u

Introduction, Methods, Results, Discussion

Yellow fever virus ( YFV ) is a vector-borne flavivirus endemic to Africa and Latin America ., Ninety per cent of the global burden occurs in Africa where it is primarily transmitted by Aedes spp , with Aedes aegypti the main vector for urban yellow fever ( YF ) ., Mosquito life cycle and viral replication in the mosquito are heavily dependent on climate , particularly temperature and rainfall ., We aimed to assess whether seasonal variations in climatic factors are associated with the seasonality of YF reports ., We constructed a temperature suitability index for YFV transmission , capturing the temperature dependence of mosquito behaviour and viral replication within the mosquito ., We then fitted a series of multilevel logistic regression models to a dataset of YF reports across Africa , considering location and seasonality of occurrence for seasonal models , against the temperature suitability index , rainfall and the Enhanced Vegetation Index ( EVI ) as covariates alongside further demographic indicators ., Model fit was assessed by the Area Under the Curve ( AUC ) , and models were ranked by Akaike’s Information Criterion which was used to weight model outputs to create combined model predictions ., The seasonal model accurately captured both the geographic and temporal heterogeneities in YF transmission ( AUC = 0 . 81 ) , and did not perform significantly worse than the annual model which only captured the geographic distribution ., The interaction between temperature suitability and rainfall accounted for much of the occurrence of YF , which offers a statistical explanation for the spatio-temporal variability in transmission ., The description of seasonality offers an explanation for heterogeneities in the West-East YF burden across Africa ., Annual climatic variables may indicate a transmission suitability not always reflected in seasonal interactions ., This finding , in conjunction with forecasted data , could highlight areas of increased transmission and provide insights into the occurrence of large outbreaks , such as those seen in Angola , the Democratic Republic of the Congo and Brazil .

In this article , we describe the development of a model to quantify the seasonal dynamics of yellow fever virus ( YFV ) transmission across Africa ., YFV is a flavivirus transmitted , within Africa , primarily by Aedes spp where it causes an estimated 78 , 000 deaths a year despite the presence of a safe and effective vaccine ., The importance of sufficient vaccination , made difficult by a global shortage , has been highlighted by recent large scale , devastating , outbreaks in Angola , the Democratic Republic of the Congo and Brazil ., Here we describe a novel way of parameterising the effect of temperature on YFV transmission and implement statistical models to predict both the geographic and temporal heterogeneities in transmissions , while demonstrating their robustness in comparison to models simply predicting geographic distribution ., We believe this quantification of seasonality could lead to more precise applications of vaccination campaigns and vector-control programmes ., In turn this would help maximise their impact , especially vital with limited resources , and could contribute to lessening the risk of large scale outbreaks ., Not only this , but the methods described here could be applied to other Aedes-borne diseases and as such provide a useful tool in understanding , and combatting , several other important diseases such as dengue and zika .

death rates, invertebrates, medicine and health sciences, population dynamics, geographical locations, vector-borne diseases, animals, seasons, population biology, insect vectors, africa, infectious diseases, aedes aegypti, disease vectors, insects, arthropoda, people and places, yellow fever, population metrics, seasonal variations, mosquitoes, eukaryota, earth sciences, biology and life sciences, species interactions, viral diseases, organisms, geographic distribution

journal.pcbi.1000661

Automatic Assignment of EC Numbers

With the several thousand proteins found in each organism a highly developed hierarchical and consistent classification scheme is absolutely essential for a comparison of metabolic capacities of the organisms ., Unfortunately such a system exists only for the enzymes and not for the other protein classes but for the enzymes the classification scheme allows an immediate access or the enzyme functional properties including catalysed reaction , substrate specificity , etc ., In this respect a quick comparative assessment of enzymatic pathways between organisms is possible even when the enzymes in the different organisms have totally different sequences as long as they belong to the same EC-class ., A well reconstructed metabolic network provides a unified platform to integrate all the biological and medical information on genes , enzymes , metabolites , drugs and drug targets for a system level study of the relationship between metabolism and disease ., Therefore an accurate representation of biochemical and metabolic networks by mathematical models is one of the major goals of integrative systems biology ., Metabolic networks have been constructed for a number of genomes 1 , 2 ., An example for the reconstruction process of a metabolic network are schematically shown in Figure 1 ., It is essential to integrate information from different databases to get a more complete enzyme list for the reconstruction ., The main databases to be taken into account to provide a complete cross-link between genes and their corresponding enzymes are NCBI EntrezGene 3 , Ensembl 4 , KEGG 5 , MetaCyc 6 and BRENDA 7 ., The second step of the reconstruction procedure is to fill the gaps resulting from the first step based on information from literature ., This step is very time-consuming and it would be therefore highly desirable to make the first step an automatic and reliable procedure ., One of the problems is the different substrate specificity of enzymes in different organisms a fact that cannot be really accounted for by any classification system 8 ., A further problem is the wide-spread use of incomplete EC numbers such as 1 . - ., - ., - ( e . g . in UNIPROT entry AK1C3_HUMAN ) ., This often occurs because an enzymatic function is inferred from the existence of a certain pair of metabolites or only experimentally shown from a cell extract without a full characterisation of the enzyme with biochemical methods , which is the requirement for the assignment of EC-numbers by the IUBMB Nomenclature Committee 9 ., For example , in the UniProt database there are more than 800 proteins annotated with an incomplete EC number 10 ., Applications like drug design , ligand docking , or systems biology require the EC number classification to be correct , consistent , and accurate ., For these reasons the automatic assignment of EC numbers to enzymatic reactions is a current issue in bioinformatics and requires specific chemical knowledge , therefore just a few approaches have been published to handle the assignment problem ., The Kyoto Encyclopedia of Genes and Genomes ( KEGG ) developed a tool for computational assignment of EC numbers published by Kotera et al . 11 ., In this approach each reaction formula is decomposed by manual work into sets of corresponding substrate and product molecules , which are called reactant pairs ., In the second step every reactant pair is analysed by the structure comparison method SIMCOMP developed by Hattori et al . 12 ., Another approach proposed by Körner et al . 13 and Apostolakis et al . 14 considers reaction energetics to predict reaction sites ., Lationa et al . 15 introduced an EC number classification method based on self-organizing maps ., This approach allows to assign EC numbers at the sub-subclass levels for reactions with accuracies of 70% ., One of the authors being the current chairman of the IUBMB nomenclature committee we felt the need to develop a system that allows for a highly reliable classification system that can help to identify the sub-subclass of any given enzyme-catalyzed reaction , allow a quick assignment of new reactions and additionally serve in a retrospective quality control of existing EC-numbers ., With ca ., 4000 existing EC-numbers this can certainly not be done by hand ., In this article we present an efficient and reliable strategy for the automatic classification of enzyme-catalysed biochemical reactions based on the chemical structure of the involved substrates and products ., With one of the authors being the present chairman of the NC-IUBMB it is planned to use this and related tools to identify and remove errors and inconsistencies in the current EC-system and to optimise the system in a transparent and stable way ., We plan to develop a tool that assign EC sub-subclasses to new reactions , access to which will be provided to the scientific community in the Internet’ ., We used 3 , 788 different enzyme-catalysed reactions from an in-house-developed Database named BiReDa ( Biochemical Reaction Database ) ., The database held exclusively error-free MDL/MOL files as well as stoichiometrically and stereochemically correct reaction data from the BRENDA Database 7 and the KEGG LIGAND database 5 , which have been corrected manually or automatically , if required ., The key idea of this approach is to reproduce the classification system given by the IUBMB as closely as possible and not to create new classification rules ., The underlying procedure is divided into two steps:

Introduction, Results/Discussion, Materials And Methods

A wide range of research areas in molecular biology and medical biochemistry require a reliable enzyme classification system , e . g . , drug design , metabolic network reconstruction and system biology ., When research scientists in the above mentioned areas wish to unambiguously refer to an enzyme and its function , the EC number introduced by the Nomenclature Committee of the International Union of Biochemistry and Molecular Biology ( IUBMB ) is used ., However , each and every one of these applications is critically dependent upon the consistency and reliability of the underlying data for success ., We have developed tools for the validation of the EC number classification scheme ., In this paper , we present validated data of 3788 enzymatic reactions including 229 sub-subclasses of the EC classification system ., Over 80% agreement was found between our assignment and the EC classification ., For 61 ( i . e . , only 2 . 5% ) reactions we found that their assignment was inconsistent with the rules of the nomenclature committee; they have to be transferred to other sub-subclasses ., We demonstrate that our validation results can be used to initiate corrections and improvements to the EC number classification scheme .

The fundamental understanding of metabolism in organisms which can only be achieved by integrated studies on their biology using a systems biology approach will aid in the design of future metabolic engineering strategies ., Metabolic network reconstruction provides insight into the molecular mechanisms of a particular organism ., An annotated genome containing the specific metabolic genes found in a particular organism can be used to reconstruct its metabolic network ., The correlation between the genome and metabolism is made by searching gene databases or by searching protein databases with a known EC number in order to find the associated gene ., The success of the search process is critically dependent upon the consistency and reliability of the underlying data ., Therefore we have developed tools which can be used to identify wrong or inconsistent classification of enzymes and help to remove them from the relevant search databases .

biochemistry, computer science/systems and control theory, biochemistry/bioinformatics, computational biology/metabolic networks, computational biology

journal.pntd.0001330

The Long Term Effect of Current and New Interventions on the New Case Detection of Leprosy: A Modeling Study

The global new case detection rate of leprosy has dropped considerably during last century , but with approximately 250 , 000 new cases detected annually , leprosy is far from being eradicated 1 ., Currently , the primary strategy for controlling leprosy is case detection and treatment with multidrug therapy ( MDT ) ., Although new interventions are under development , their potential impact on disease control is unknown ., Recent clinical trials have indicated that a single chemoprophylactic dose of rifampicin given to individuals in contact with newly diagnosed leprosy patients could protect these contacts against leprosy disease 2 ., The results with a single dose of rifampicin are very comparable to trials with dapsone that were conducted in the pre-MDT era ., A meta-analysis showed that the combined results from the randomized controlled trials favored chemoprophylaxis to placebo with 2–4 years of follow-up ( relative risk 0 . 59 , 95% ( CI ) 0 . 50–0 . 70 ) 3 ., The advantage of a single dose of rifampicin is that it is only given once , while dapsone prophylaxis is given for at least 2 years ., Furthermore , new tests are under development for identifying subclinical infections 4 ., Other recent developments however , are cause for concern ., For example , the integration of leprosy control activities into general health care programs has in many countries led to the cessation of active case finding and contact tracing ., Consequently , diagnosis is delayed and patients are therefore infectious for a longer period causing more people in contact with patients to become infected ., Another concern is that a new vaccine may replace the current Bacillus Calmette-Guérin ( BCG ) tuberculosis vaccine , which is given to infants to prevent tuberculosis ( TB ) , but which also protects against leprosy 5 ., An update on progress describing new TB vaccine candidates that are currently entering clinical trials has recently been published 6 ., Most are pre-exposure vaccines and will most likely prevent TB disease ., Such vaccines are intended either to replace BCG ( recombinant live vaccines ) or to be given after BCG prime as boosters ( protein adjuvant formulations or recombinant viral carriers ) ., New and more specific TB vaccines may not induce cross-immunity to the bacterium responsible for leprosy , Mycobacterium leprae 6 , 7 ., It was recently pleaded that new candidate vaccines must be developed taking both diseases into account , and that the current TB candidate vaccines should be assessed for their potential to protect against leprosy as well as against TB 8 ., Therefore , the effect of new leprosy interventions strategies should be tested in the context of other related developments , such as possible changes to BCG ., Although the short-term effectiveness of new interventions can be assessed in trials , extrapolation to long-term effectiveness in the general population is difficult , due to the complex impact on transmission dynamics ., Hence , dynamic simulation models are necessary to assess the possible impact of different intervention strategies on future trends in the new case detection rate of leprosy ., We have developed a microsimulation model that simulates the transmission and control of leprosy ( the SIMCOLEP model ) , taking into account the population structure of households 9 ., The model has been quantified by data from northwest Bangladesh in 2003 ., Very detailed data were available for that year from a large randomised controlled trial of chemoprophylaxis with single dose rifampicin ( the COLEP study ) that was being conducted at the time 2 , 10–13 ., This is an area with a well-organized leprosy control program and with a decreasing trend in new case detection since the mid-1990′s ., Regardless , the current case detection rate remains one of the highest in Bangladesh , 2–3 per 10 , 000 population ., We applied our model to the situation in this area as a starting point for exploring the potential impact of seven different intervention strategies on the detection of new cases of leprosy over a 50-year period ., For the COLEP trial ( ISRCTN 61223447 ) , on which the data of this modeling study is largely based , ethical clearance was obtained from the Ethical Review Committee of the Bangladesh Medical Research Council in Dhaka ( ref . no . BMRC/ERC/2001–2004/799 ) ., All subjects were informed verbally in their own language and invited to participate ., Written consent was requested from each adult ., For children consent from a parent or guardian was given ., The microsimulation model simulates the life history of fictitious individuals 9 ., These individuals are members of a household that is formed , changes , and dissolves during the simulation ., Individual household movement occurs during adolescence and after marriage ., Some married couples start living in the household of the parents-in-law , and will form their own separate household after on average 12 years ., The life span of individuals is drawn from a life-table at birth; the number of newborn individuals maintains the simulated population growth rate equal to the observed population growth rate; newly born individuals are placed into the household of their mothers; and mothers are drawn from the population of married women and weighted with an age-dependent fertility function ., An individual that is susceptible to leprosy is defined as an individual that developed leprosy sometime during their lifetime , after acquiring the infection ., The large majority ( say 80–95% ) of the population is assumed not to be susceptible to leprosy 14–16 ., The remaining 5–20% of the population is susceptible ., For these individuals , it is assumed that 80% undergoes a self-healing infection and is never infectious to other individuals , that is 20% will become chronically infected and infectious 16 ., The mechanisms underlying leprosy susceptibility are currently unknown 9 ., Therefore , the model used six hypothetical mechanisms: Random ( no mechanism , but each individual has a fixed probability of being susceptible ) ; Household susceptibility ( all susceptibles live in a fraction of households , within these susceptible households a fraction of inhabitants is susceptible ) ; Dominant ( susceptibility is inherited by a dominant gene ) ; Recessive ( susceptibility is inherited by a recessive gene ) ; Household & dominant ( 50% of susceptibility is determined by the Household and 50% by a dominant gene ) ; Household & recessive inheritance ( 50% of susceptibility is determined by the Household and 50% by a recessive ) ., As described in a previous paper 9 , the model was unable to identify one single mechanism that could best explain the observed data ., However , for Random it turned out that 20% susceptibles provided the best fit , whereas this was 10% for the other mechanisms ., For Household this 10% was established by assuming on average 25% of the households contain on average 40% susceptible individuals ., The quantification of the model is based on the leprosy situation in 2003 and the control program of the last decades in the Nilphamari and Rangpur districts of Bangladesh 9 ., This control program consisted of passive case detection , with in 2003 an average detection delay of 2 years , treatment with MDT , and active tracing of people in contact with patients ., Contacts are examined annually for three consecutive years ., In this area , BCG vaccination was routinely given to newborn infants ., Since the introduction of the BCG vaccination in 1974 , the coverage had gradually expanded to 80% in 1990 and remained at that level in 2003 17 ., BCG had a protective effect of 60% 18 ., For a full and detailed description of the model , we refer to our previous paper 9 ., In the study we considered seven potential intervention scenarios for the future control of leprosy ., The baseline scenario was the current leprosy control program in the Bangladesh study area , as described above ., The other scenarios were modifications of the baseline control program ., These other six scenarios were:, 1 ) no contact tracing;, 2 ) with a single chemoprophylactic dose of rifampicin , which cured 50% of subclinical cases , for each individual in contact with a leprosy patient 2;, 3 ) with diagnosis of subclinical cases with a sensitivity of 70% 19 followed by effective treatment;, 4 ) with all newly born infants in the population receiving a new ( hypothetical ) tuberculosis vaccine that is ineffective against leprosy instead of BCG ( no BCG ) ;, 5 ) with the combination of no BCG and chemoprophylaxis; and, 6 ) with the combination of no BCG and early diagnosis with effective treatment ., In our intervention scenarios , contact tracing , chemoprophylactic treatment and early diagnosis were performed only on household members ., Contact tracing was repeated three times in three consecutive years with a 10% probability of loss to follow-up and a 90% of symptomatic cases being detected ., Early diagnosis was performed in the same schedule as the contact tracing with three consecutive visits to the household ., Chemoprophylactic treatment was given only once after examination , in which 90% of symptomatic cases will be detected ., The simulation of interventions was started based on the quantification of 2003 , because a detailed data set 16 was available from the COLEP study conducted during that period ., The Bangladesh districts at the time when the COLEP study took place can be seen as fairly representative for other areas in the Indian subcontinent with regard to demography , socio-economic condition , cultural tradition and the organization of the health system , including the leprosy control program ., The prevalence rate of leprosy at the time was well above the WHO elimination target of 1 per 10 , 000 population , which was also the case in many areas in India around the year 2000 ., Table 1 shows the predicted new case detection rates at 25 years after the initiation of the interventions ., Under the baseline control program , the different mechanisms that determined susceptibility showed up to three-fold differences in the predicted number of cases per 100 , 000 people ., In Figure 1 , the trends in the new case detection rates over 50 years are shown for all seven interventions ., All susceptibility mechanisms give qualitatively comparable trends ., When the intervention scenarios were ordered after 50 years by the amount of reduction in new case detection rates , the order was as good as identical for all mechanisms; i . e . early diagnosis lowest; then no BCG & early diagnosis; then chemoprophylaxis; then baseline; then no BCG & chemoprophylaxis together with no contact tracing; and finally no BCG had the highest new case detection rate ., Both the cessation of contact tracing and the replacement of BCG vaccine by a tuberculosis vaccine ineffective for leprosy ( no BCG ) would have detrimental effects on the rate of decline in leprosy ( Figure 2 ) ., Twenty-five years after introduction of the ineffective vaccine ( no BCG ) , the new case detection of leprosy was approximately 1 . 5 times higher than the baseline ( Table 1 ) ., The cessation of contact tracing was predicted to have a smaller impact , with a marked drop in detection of new leprosy cases during the first few years ., This sudden drop was due to the reduced number of examinations of people in contact with patients; thus , these cases would not be detected until later , through passive detection ( self-reporting ) ., Both chemoprophylaxis and early diagnosis were predicted to have substantial effects on the new case detection of leprosy ( Figure 2 ) ., With no BCG , chemoprophylaxis would partially compensate for the predicted increase in new case detection rates ., Furthermore , early diagnosis was predicted to more than compensate for the adverse effects of a leprosy-ineffective tuberculosis vaccine , and reduce the rate of new case detection compared to the baseline ., The effects were more promising with the ongoing presence of the BCG vaccine ., Under those conditions , at 25 years after the introduction of chemoprophylaxis , the new case detection rate was predicted to be 25% lower than baseline control ., Moreover , with the introduction of early diagnosis , the new case detection rate was predicted to halve the baseline incidence after 25 years ( Table 1 ) ., Early diagnosis of infection allows the detection of subclinical cases , of which part would be detected later or never at all ., These subclinical cases are added to the number of detected cases ., This is seen in the results of this intervention ., The introduction of early diagnosis would increase the total number of detected cases in the first 18 years , simply because of the detection of previously undetectable subclinical cases ., Over time however , the total number of new cases ( subclinical and clinical ) would finally drop below the number detected in the baseline control program ( Figure 3 ) ., In Figure 3 , we show that the new cases detected under the chemoprophylaxis intervention strategy drop immediately below the level of the baseline control program ., The additional effect of chemoprophylaxis is that additional new infections are prevented on top of the cure of subclinical infections ., These additional prevented infections are due to a shorter infectious period of the cured subclinical infections ., To illustrate this effect we show in Figure 3 the clinical and the subclinical cases that were cured by the chemoprophylactic intervention ., During the first 10 years , this total number of newly detected cases plus cured cases is equal to the number of newly detected cases under the baseline control program , but afterwards the number of cases plus cured subclinical cases in the chemoprophylaxis intervention group drops under the baseline control program , indicating the prevention of new infections ., We showed that the leprosy incidence would be reduced substantially by good BCG vaccine coverage and the combined strategies of contact tracing , early diagnosis , and treatment of infection and/or chemoprophylaxis among household contacts ., To effectively interrupt the transmission of M . leprae , it is crucial to continue developing immuno- and chemoprophylaxis strategies and an effective test for diagnosing subclinical infections .

Introduction, Methods, Results, Discussion

Although the number of newly detected leprosy cases has decreased globally , a quarter of a million new cases are detected annually and eradication remains far away ., Current options for leprosy prevention are contact tracing and BCG vaccination of infants ., Future options may include chemoprophylaxis and early diagnosis of subclinical infections ., This study compared the predicted trends in leprosy case detection of future intervention strategies ., Seven leprosy intervention scenarios were investigated with a microsimulation model ( SIMCOLEP ) to predict future leprosy trends ., The baseline scenario consisted of passive case detection , multidrug therapy , contact tracing , and BCG vaccination of infants ., The other six scenarios were modifications of the baseline , as follows: no contact tracing; with chemoprophylaxis; with early diagnosis of subclinical infections; replacement of the BCG vaccine with a new tuberculosis vaccine ineffective against Mycobacterium leprae ( “no BCG” ) ; no BCG with chemoprophylaxis; and no BCG with early diagnosis ., Without contact tracing , the model predicted an initial drop in the new case detection rate due to a delay in detecting clinical cases among contacts ., Eventually , this scenario would lead to new case detection rates higher than the baseline program ., Both chemoprophylaxis and early diagnosis would prevent new cases due to a reduction of the infectious period of subclinical cases by detection and cure of these cases ., Also , replacing BCG would increase the new case detection rate of leprosy , but this effect could be offset with either chemoprophylaxis or early diagnosis ., This study showed that the leprosy incidence would be reduced substantially by good BCG vaccine coverage and the combined strategies of contact tracing , early diagnosis , and treatment of infection and/or chemoprophylaxis among household contacts ., To effectively interrupt the transmission of M . leprae , it is crucial to continue developing immuno- and chemoprophylaxis strategies and an effective test for diagnosing subclinical infections .

Leprosy is a contagious disease that will remain prevalent , despite the declining number of patients worldwide over the last century ., With approximately 250 , 000 new cases detected annually , leprosy is far from being eradicated ., Leprosy can be treated with drugs after disease detection ., Some cases can be prevented with a tuberculosis vaccine ( BCG ) that cross-reacts with the bacterium responsible for leprosy , but this vaccine might be replaced in the future ., Furthermore , preventive drugs can reduce the number of new cases among people in contact with infectious patients , but this strategy has not yet become established in common practice ., Also , a new test is under development for the detection of infections before the appearance of symptoms ., In this study , we used a computer model to assess the effectiveness of seven possible leprosy control activities ., Our results showed that the decline in incidence of leprosy would slow down or halt with the introduction of a new tuberculosis vaccine that is ineffective against leprosy ., However , this effect could be offset by the implementation of effective tests for early diagnosis or the routine administration of preventative drugs to contacts of patients .

immunizations, medicine, infectious diseases, public health and epidemiology, epidemiology, infectious disease epidemiology, neglected tropical diseases, leprosy, infectious disease control, infectious disease modeling, public health

journal.pgen.1004595

The Groucho Co-repressor Is Primarily Recruited to Local Target Sites in Active Chromatin to Attenuate Transcription

Understanding how transcription factors regulate gene expression is essential for determining how genetically identical cells adopt different fates during animal development ., The expression of key genes involved with cell fate determination is often controlled by spatially restricted localization or activity of transcriptional repressors ., Many repressors do not have intrinsic repressive activity but recruit co-factors that inhibit productive transcription ., The Groucho/Transducin-Like Enhancer of split ( Gro/TLE ) family of co-repressors are conserved across metazoa and include a single ortholog in Drosophila ( Gro ) , and four orthologs in humans ( TLE1-4 ) and mouse ( Gro-related-gene: Grg1-4 ) ( reviewed in 1–4 ) ., Gro family proteins do not bind DNA directly , but are recruited to target genes by DNA-binding transcription factors ., Gro was first found as a co-factor for Hairy and the related Enhancer of split basic helix loop helix proteins E ( spl ) -bHLHs and Deadpan ( Dpn ) proteins during neurogenesis , segmentation , and sex differentiation in Drosophila 5 ., Subsequently , Gro family proteins have been identified as co-repressors for many other transcription factor families including Runx , Nkx , LEF1/Tcf , Pax , Six , Fox and c-Myc ( reviewed in 1 , 6 ) ., Recruiting partners for Gro/TLE proteins include transcription factors that are effectors of signaling pathways that determine cell fate including Notch and Wnt ., Thus , Gro family proteins have roles in a variety of biological processes including osteogenesis , somitogenesis , haematopoesis , and stem cell maintenance and proliferation ., Furthermore , human TLE proteins have been implicated in a variety of cancers including breast cancer , leukemia and lymphoma ( reviewed in 1 , 7 ) ., The primary structure of Gro/TLE proteins includes five distinguishable regions , of which the most highly conserved are the N-terminal glutamine-rich Q domain and the C-terminal WD-repeat domain 8 , 9 ., Sequences within the Q domain are predicted to form two coiled-coil motifs that facilitate oligomerization of Gro molecules in vitro 9–11 and also mediate interactions with some repressors 7 , 12 , 13 ., The WD-repeat domain has been shown by X-ray crystallography to form a β-propeller 14 , 15 , which binds many different transcription factors , including those containing the conserved “eh1” and WRPW and related peptide motifs 15 ., One model for Gro repression is that upon recruitment to a target site by a DNA binding transcription factor , Gro oligomerizes along the DNA and recruits factors that modify chromatin to inhibit transcription from promoters that may be over 1 kb from the initial recruitment site 9 , 16 ., This model is sometimes referred to as the “spreading model” and is based on the observations that oligomerization via the Q domain is required for Gro family proteins to repress reporter gene transcription in Drosophila S2 cells and in overexpression assays in the fly 9 , 11 , and that Gro interacts with a histone deacetylase ( HDAC1 , referred to as Rpd3 in Drosophila; 17 ) ., Recent support for this model comes from the observations that when a LexA-Hairy fusion protein recruits Gro to a reporter gene in flies , Gro recruitment is spread across 2–3 kb of the gene and is associated with Rpd3 recruitment and reduced histone acetylation 18 ., Gro-mediated repression of the fushi tarazu ( ftz ) gene by ectopic expression of Hairy induces histone deacetylation for several kilobases around ftz 19 ., Furthermore , the presence of histone deacetylase inhibitors or decreasing the dose of Rpd3 , lessen the defects caused by overexpressing Gro in wing imaginal discs in Drosophila 20 ., However , Gro repression is only partially dependent on Rpd3 , indicating that other modes of repression by Gro are important in vivo 20 , 21 ., Analysis of an endogenous Drosophila mutation revealed that oligomerization is not always required for the co-repressor function of Gro ., groMB12 is a single base pair substitution in the translation initiator ATG codon ( ATG-ATA ) that leads to an N-terminal truncation , deleting much of the Q-domain 3 ., MB12 protein does not oligomerize in vitro and is expressed at <5% normal levels in early embryos ., Nevertheless , groMB12 is not a null: maternal mutant embryos have intermediate segmentation phenotypes and retain more body mass than the null , indicating that MB12 retains some co-repressor activity ., The groMB12 mutation has differential effects on the expression of target genes in vivo ., For example , repression of the tailless ( tll ) gene by the Capicua-Gro complex is relatively normal in groMB12 embryos while repression of snail by Huckebein-Gro fails ., Thus , there are differential requirements for oligomerization via the Q domain during Gro-mediated repression ., In this study we have used chromatin immunoprecipitation followed by high throughput sequencing analysis ( ChIP-seq ) to profile the genome-wide recruitment of wild-type and non-oligomerizing Gro at high resolution in single cell types using Drosophila cell culture ., In addition , we have focused on Gro recruitment at a known target locus E ( spl ) mβ-HLH to establish a model for Gro function as a co-repressor ., To profile genome-wide Gro binding in Kc167 cells , we performed ChIP-seq using a previously validated anti-Gro antibody 22 ., We chose Kc167 cells as they had been characterized extensively for genome-wide transcription factor binding , chromatin modifications and gene expression by Filion et al . , 23 and the modENCODE project 24 ., Use of a single cell type avoided the complications of interpreting data derived from multiple cell types ( e . g . embryo collections ) where peaks may represent binding to overlapping or adjacent regulatory elements used at different times or by specific cell types ., Gro binding sites were determined by the maximum per cent overlap of called peaks in two independent biological samples ( see Materials and Methods for further details ) ., This analysis yielded 1912 peaks of endogenous Gro binding ( Figure 1A ) ., Depletion of Gro from Kc167 cells using RNAi against the 3′-untranslated region of the endogenous gro transcript led to a dramatic reduction of the number of significant peaks , demonstrating that ChIP with the anti-Gro antibody reflects bona fide Gro binding ( Figure 1B ) ., As subsequent experiments would require the expression of a mutated variant of Gro , we generated a wild-type Gro tagged with GFP ( Gro-GFP ) , tested its recruitment ( using an anti-GFP antibody ) in Kc167 cells depleted of endogenous Gro , and compared replicate samples as above ( Figure 1C ) ., To compare binding between the endogenous and GFP-tagged Gro , replicate samples were normalized together with the input , and the mean log fold change ( FC ) for each condition plotted ., The results were highly similar to the endogenous Gro ( Figure 1D ) and we therefore generated a “superset” of high confidence bound regions in Kc167 cells by selecting the 1376 peaks common to all datasets ( Table S1 ) ., We first examined the breadth of peaks bound by Gro in Kc167 cells to determine if Gro is recruited to discrete sites or spreads along the DNA - or if both types of recruitment occur but are target dependent ., The model that Gro spreads along chromatin ( via Q domain oligomerization ) to act as a long-range repressor predicts that Gro peaks would be typically greater than 1 kilobase wide and range to several kilobases 6 , 9 , 16 , 18 ., Previous studies of genome-wide Gro recruitment have either lacked the resolution to examine this due to the methodology used ( DamID; 23 ) or because they were performed using a highly mixed population of cells ( 0–12 hour embryos; 22 ) ., Our superset of high confidence ChIP-seq peaks of Gro in Kc167 cells typically span less than 1 kb ( Figure 2A ) with a mean width of 831 bp and a median width of 708 bp ( Table S1 ) ., Less than 3% ( 36 peaks ) of Gro bound regions extend beyond 2 kb , with the largest being 2922 bp ( in the region of Rh5 ) ., Peaks exclusive to individual replicates of Gro ChIP-seq tended to be narrower than those peaks found in the high confidence superset ( Figure S1 ) , indicating that selection of the superset did not exclude broad peaks found in individual replicates ., 33% of Gro peaks in the superset overlapped regions of the genome bound by Gro-Dam in Kc167 cells ( DamID data from 23 ) ( Figure S2A ) ., This is comparable to the overlap observed for ChIP-seq and DamID peaks of GAGA factor GAF; encoded by Trithorax-like ( Trl ) ( Figure S2B ) ., The conditions used during Gro-Dam analysis may have allowed the detection of broader , lower affinity Gro complexes on the chromatin that were potentially disrupted by the sonication regime necessary for Gro and Gro-GFP ChIP-seq ., However , the Gro-Dam peaks that did not overlap with peaks in our ChIP-seq replicates tended to be narrower than those which overlapped with Gro ChIP-seq peaks ( Figure S3 ) ., This indicates that the Gro-GFP ChIP-seq analysis was not biased against detecting broad Gro peaks ., We also compared the profile of Gro peak widths with those of other transcriptional regulators in Kc167 cells for which ChIP-seq data was currently available ., Gro peaks were broader than those produced by GAF , but were narrower than Tramtrack ( Ttk ) , Kruppel ( Kr ) , Zn finger homeodomain 1 ( Zfh1 ) and C-terminal Binding Protein ( CtBP ) ChIP-seq peaks in Kc167 cells ( Figure S4 ) ., Peaks from Hairy and Suppressor of Hairless Su ( H ) , proteins known to recruit Gro , were found over a broad range of sizes up to 5000 bp ., More generally , the dimensions we observe for Gro peaks correspond to peak widths observed from ChIP-seq experiments profiling “point sources” rather than “broad sources” 25 ., Our data demonstrate that Gro binding is not typically spread over multi-kilobase regions of the genome , while the conditions and analysis we used did not exclude the recovery of ≥2 kb peaks ., However , several genomic regions contain clusters of discrete Gro peaks that are spread across several kilobases ( Table S1 and Figure 2B , D ) that could be interpreted as single broad peaks using techniques and analysis with lower resolution ., Gro peaks commonly overlap annotated transcription start sites in Kc167 cells , although peaks are also found upstream of and inside genes ( Figure 2C ) ., One region that contains a cluster of Gro bound sites is the Enhancer of Split Complex E ( spl ) -C ( Figure 2D ) ., Gro has previously been shown to form a complex with Hairless ( H ) and Su ( H ) , contributing to the repression of target genes in the absence of Notch signaling 26 , 27 ., Su ( H ) represses Notch target gene expression ( including E ( spl ) -C genes ) in the absence of Notch signaling in Kc167 cells 28 ., We therefore assessed whether there was a relationship between the Gro and Su ( H ) bound regions within the E ( spl ) -C ., The Gro peaks overlapped Su ( H ) peaks close to E ( spl ) mβ-HLH and E ( spl ) m3-HLH ( Figure 2D ) ., The expression of E ( spl ) mβ-HLH and E ( spl ) m3-HLH was increased in Kc167 cells treated with Gro RNAi ( Figure 2E , Table S2 ) ., To test if depletion of Gro is sufficient to induce gene expression of repressed targets , we compared gene expression by RNA-seq of untreated and gro RNAi Kc167 cells ., There were very few genes differentially expressed genes and when looking at the whole transcriptome , we did not observe a general induction of genes ( e . g . at below statistical significance ) closely associated with ChIP-seq peaks in RNA-seq analysis ( Table S2 , Figure S5 ) , although the expression of two high confidence target genes within the E ( spl ) -C is upregulated when Gro is depleted by RNAi ., Gro is recruited as a co-factor by many different DNA-binding transcription factors in addition to Su ( H ) , thus Gro peaks are not expected to contain one consensus DNA binding sequence ., In agreement with this , no single consensus motif was found in the high confidence Gro peaks ( Figure 2F ) ., Instead , binding motifs for several different transcription factors expressed in Kc167 cells 29 with unrelated consensus recognition sequences were enriched in Gro peaks 30 ., These included binding motifs for known partners of Gro , including Hairy and Brinker ( Brk ) ., In addition , motifs for GAF and Mothers against dpp ( Mad ) , which have not previously been identified as Gro partners , were also enriched in Gro bound regions ., Gene Ontology analysis revealed that the terms over-represented in the genes nearest Gro binding sites in Kc167 cells included “cell morphogenesis” , “imaginal disc development” and “neuron differentiation”, ( Figure 2G ) ., These terms are consistent with Gros characterized biological role as a transcriptional co-repressor of developmentally regulated pathways , giving support to our ChIP-seq analysis representing bona fide Gro recruitment ., To determine if the features of Gro recruitment we observe in Kc167 cells are common to other cell types , we performed ChIP-seq to profile Gro binding in S2 cells ., Both Kc167 and S2 cell cultures are derived from late embryonic cells and have properties related to plasmatocytes , but they express distinct profiles of genes 31 ., The quality and consistency of the peaks derived from S2 cells were less reproducible between replicates and endogenous versus Gro-GFP ChIP experiments , probably due to the variable aneuploidy observed within S2 cell populations 31 ., However , by comparing the replicates with the most reads from ChIP using anti-Gro and ChIP using anti-GFP, ( to Gro-GFP ), we identified 1242 high confidence peaks in S2 cells, ( Figure 3A , Table S3 ) ., 519 of these peaks overlap the superset of high confidence peaks in Kc167 cells, ( Figure 3B ) , indicating that the genome-wide profile of Gro recruitment has a cell type specific component ., The peaks in S2 cells mapped to a similar profile of genomic features to those in Kc167 cells , although fewer overlapped the start of annotated transcripts, ( approximately 25% in S2 cells compared to 40% in Kc167; Figure 3C ) ., The high confidence peaks in S2 cells have an average peak width of 503 bp and median width of 425 bp ., The widest peak in S2 cells was 2301 bp , and there were just 4 peaks over 2 kb in breadth, ( Figure 3D ) ., Thus as in Kc167 cells , we did not observe Gro binding over broad domains of the genome in S2 cells ., In common with Kc167 cells , Gro peaks in S2 cells were enriched for GAF , Mad , Brk and Hairy binding sites , but also for l ( 3 ) neo38 motifs, ( Figure 3E ) ., Gene Ontology analysis indicated that the Gro peaks in S2 cells were associated with transcripts linked to developmental processes including “imaginal disc development” , “cell motion” , and “neuron differentiation”, ( Figure 3F ) ., We also tested if depletion of Gro is sufficient to induce gene expression of repressed targets in S2 cells ., Similar to Kc167 cells , the depletion of Gro from S2 cells by RNAi treatment resulted in very few differentially expressed genes and did not lead to general upregulation of Gro target genes, ( Table S4 , Figure S6 ) ., To examine the contribution of oligomerization via the Q-domain to the pattern of Gro recruitment , we used ChIP-seq to compare the binding profiles of a non-oligomerizing variant of Gro tagged with GFP, ( GroL38D , L87D-GFP; 11 ), with Gro-GFP in Kc167 cells depleted of endogenous Gro via RNAi ., The positions of the peaks of GroL38D , L87D-GFP showed a high degree of correlation with Gro-GFP peaks, ( Figure S7 ) ., Furthermore , blocking oligomerization of Gro did not decrease the average width of the peaks of Gro recruitment in Kc167 cells, ( Figure 4A , B ) ., Indeed , the average width of peaks bound by GroL38D , L87D-GFP was slightly higher than endogenous Gro and Gro-GFP, ( Figure 4B ) ., The width of the broadest Gro peak in Kc167 cells, ( at the Rh5 locus ), was not affected by blocking oligomerization and peaks bound by GroL38D , L87D-GFP at the E ( spl ) mβ-HLH locus closely resembled those bound by Gro-GFP, ( Figure 4C ) ., We saw no significant changes in the expression of genes bound by GroL38D , L87D-GFP with respect to those bound by Gro-GFP by RNA-seq analysis, ( Table S5 , Figure S8 ) ., Previous experiments demonstrating that the GroL38D , L87D variant is unable to repress transcription of a reporter gene were performed in S2 cells 11 ., Thus we repeated the ChIP-seq experiments comparing recruitment and activity of Gro-GFP and GroL38D , L87D-GFP in S2 cells ., The results were largely consistent with those obtained using Kc167 cells ., Gro-GFP and GroL38D , L87D-GFP exhibited highly similar binding profiles and peak widths in S2 cells, ( Figure 4A ) ., Furthermore , as in Kc167 cells , we observed no significant changes in the expression of genes bound by GroL38D , L87D-GFP with respect to those bound by Gro-GFP by RNA-seq analysis in S2 cells, ( Table S6 , Figure S8 ) ., To determine if the pattern of Gro binding in discrete peaks was conserved across evolution , we performed meta-analysis on published ChIP-seq data generated by using an antibody to the human Gro ortholog TLE3 in MCF7 cells 32 ., The average peak width for TLE3 was not significantly different to that of Gro in Kc167 cells , indicating that it is recruited in a similar manner to Gro and does not typically spread across broad chromatin domains, ( Figure 4B ) ., Gro has previously been shown to physically and genetically interact with the histone deacetylase Rpd3 in Drosophila , although Gro acts independently of Rpd3 in some contexts 17 , 18 , 20 , 21 , 33 ., Consistent with these observations , we found that 59% of our superset of Gro peaks overlapped with Rpd3 peaks in Kc167 cells, ( Figure 5A , Rpd3 peaks from modENCODE ChIP-chip data 24 ) ., Overexpression of Gro correlates with decreased acetylation of histones H3 and H4 around Gro-repressed targets , and phenotypes due to overexpression of Gro in the fly are partially rescued by histone deacetylase inhibitors 18–20 ., We observed that the peaks in our Gro superset are associated with sites that are depleted of acetylated histones , although histones in the regions adjacent to Gro binding are frequently acetylated, ( Figure 5B–G ) ., For example , the gene body of E ( spl ) mβ-HLH contains acetylated histones H3 and H4 , but the levels are lower at sites where Gro binds around the gene, ( Figure 5B ) ., To determine whether Gro induces changes in the acetylation status of histones around Gro target genes we profiled the acetylation status of H3 and H4 in wild-type and Gro depleted Kc167 cells ., Knockdown of Gro did not result in any significant changes in H3 or H4 acetylation profiles, ( Figure 5B–F ) ., There was no significant effect on histone acetylation around the E ( spl ) mβ-HLH gene , which undergoes increased transcription when Gro is depleted, ( Figures 5B , S9 ) ., Thus we found no evidence that depletion of Gro directly influences levels of H3 and H4 acetylation at Gro target sites in Kc167 cells ., Rpd3 has been implicated in the deacetylation of H3K27ac , a chromatin modification that is enriched at active enhancers and promoters in Drosophila embryos 34 , 35 ., Meta-analysis of H3K27ac ChIP-seq data in Kc167 cells 35 reveals that H3K27ac is excluded at Gro peaks, ( Figure 5G ) ., The lack of histone acetylation detected at Gro binding sites may have resulted from these regions being nucleosome-free ., However , we observe that Gro peaks are enriched for H3K4me3, ( H3K4me3 data from 35 ) , especially when Gro is bound at TSSs, ( Figure 5H ) ., Promoters are generally marked with high levels of H3K4me3 regardless of their transcriptional state 36 ., This overlap indicates that Gro is recruited to sites where there are nucleosomes present that may be modified ., Integrative analysis of the binding profiles of 53 DamID tagged chromatin associated factors in Kc167 cells produced a model in which the Drosophila genome contains five principal chromatin types 23; “Red”, ( active , developmentally regulated ) , “Yellow”, ( active , housekeeping ) , “Blue”, ( repressed , by Polycomb Group complexes ), “Green” ( repressed , classic heterochromatin ) , and “Black” ( highly repressed ) ., In agreement with 23 ( who used Gro-DamID to map Gro binding ) , we found Gro ChIP-seq peaks were most highly enriched in Red chromatin ( Figure 6A ) , which is associated with factors linked to active , developmentally regulated gene expression ., Gro binding appears to be excluded to some extent from the Black and Green types of repressed chromatin ., Furthermore , Gro peaks were found in regions associated with DNase I hypersensitivity ( Figure 6B ) , indicating that they lie in open chromatin where the turnover rate of nucleosomes is high 37 ., Although Gro may act as a “long range” repressor over distances of greater than 1 kb from the target promoter ( reviewed in 38 ) , we found that almost 40% of Gro peaks overlapped with transcription start sites ( TSSs ) in Kc167 cells ( Figure 2C ) ., Indeed , high resolution mapping revealed that the summits of Gro peaks most frequently map immediately downstream ( 25–50 bp ) of the TSS ( Figure 6C ) suggesting that Gro often acts on TSSs from a very short range ., However , the level of recruitment of Gro to different locations around genes was comparable ( Figure 6F ) ., Since Gro primarily bound annotated TSSs in Kc167 cells , one potential mechanism through which Gro could mediate repression would be to block RNAP II recruitment to TSSs ., We used ChIP-seq to profile RNAP II binding to determine if RNAP II is excluded from TSSs bound by Gro ., We found that the majority of Gro peaks found at TSSs overlap RNAP II peaks in Kc167 cells , indicating that Gro does not mediate repression by simply blocking RNAP II recruitment ( Figure 6D ) ., We observed that peaks of Gro binding that were not localized to TSSs did not show an association with RNAP II recruitment ( Figure 6D ) ., We detected transcripts in RNA-seq experiments from genes where Gro was bound at either the TSS or inside the gene ( Figure 6E ) indicating that these genes were not completely silenced ., Since Gro binding at TSSs does not exclude RNAP II recruitment , we attempted to establish if Gro affected the productivity of RNAP II ., One way Gro could attenuate transcription would be to promote promoter proximal RNAP II pausing ( reviewed in 39–42 ) ., Regulation of RNAP II release at the early elongation checkpoint is a major form of transcriptional regulation at genes directing anterior-posterior ( AP ) and dorsal-ventral ( DV ) patterning in the early Drosophila embryo , which include many known targets of Gro repression 42–44 ., To determine if Gro peaks were enriched at the start of transcripts that exhibit RNAP II pausing , the pause ratio of all transcripts was determined by establishing the ratio of total RNAP II at the TSS to that within the gene body ., Almost 50% of transcripts where Gro is bound at the TSS had a very high pause ratio ( in the top 10% of all transcripts; Figures 7A , S10 ) ., Furthermore , 82% of Gro peaks located at TSSs overlapped peaks of GAF binding ( Figure 7B ) ., GAF has previously been linked to promoter proximal pausing at many genes in Drosophila 45 , 46 ., The analysis therefore suggests that Gro is enriched at TSSs where there is promoter proximal pausing of RNAP II ., We did not detect any significant global effects on RNAP II pausing in cells depleted of Gro by RNAi ., However , we observed decreased RNAP II pausing at the E ( spl ) mβ-HLH locus , which is a high confidence target of Gro repression in Kc167 cells ( Figure 7C , D ) ., Gro was first described as a “long-range” co-repressor that could inhibit transcriptional initiation of reporter genes while bound to a distant ( >1 kb away ) enhancer element 47 ., However , the model that Gro spreads over multi-kilobase domains to repress transcription was derived from experimental approaches that lacked the resolution to determine if Gro was bound in continuous or clustered peaks around genes ., For example , Martinez and Arnosti 18 used ChIP and subsequent qPCR at sites spaced ≥1 kb apart around their single target gene to test the spreading model ., The Gro detected at the promoter and at 1 kb , 2 kb and 4 kb upstream of their target gene may have been derived from distinct , discrete peaks of Gro binding ., We observe that clusters of Gro peaks across the genome are common ( Figure 2B ) ., One example of this occurs at the E ( spl ) mβ-HLH locus where distinct Gro peaks lie less than 2 kb apart , either side of the coding region ( Figure 2D ) ., It seems most likely that these are distinct peaks , as they lie over distinct Su ( H ) peaks and are separated by peaks of histone H3 and H4 acetylation ( Figures 5B , S7 ) ., By selecting our superset of high confidence peaks common to all datasets for endogenous Gro and Gro-GFP , we may have excluded some “real” peaks from our general analysis ., However , the properties of the peaks excluded from the superset did not differ significantly from the peaks in the superset ., In general , peaks that were unique to one replicate were narrower than those included in the superset , further supporting the argument that our conditions and analyses were not biased against recovering broad peaks ( Figure S1 ) ., 33% of our high confidence Gro ChIP-seq peaks overlapped previously published Gro DamID peaks ., This overlap is relatively low , however , a comparable level of overlap ( 34% ) is observed between GAF ChIP-seq and GAF DamID peaks ( Figure S2 ) ., The Dam domain was fused to the C-terminal domain of Gro 48 , which is highly structured and interacts with many classes of transcription factor 15 ., Thus , the fusion of the Dam domain to the C-terminal of Gro may have interfered with Gro recruitment to the genome and excluded sites that we could detect with ChIP-seq ., Consistent with Martinez and Arnosti 18 , we were unable to obtain reproducible ChIP samples for Gro without the use of a two-step crosslinking method ., This may reflect that Gro is not directly recruited to chromatin , but rather via intermediate sequence specific DNA binding transcription factors ., Use of two cross-linking agents meant that relatively long sonication was required to generate DNA fragments of a suitable size for sequencing ( Materials and Methods ) ., Extended sonication may disrupt indirect chromatin interactions and select only for high affinity binding sites 49 ., However we recovered peaks with widths up to 2 . 9 kb from Kc167 cells ( Table S1 , Figure 4C ) indicating that the sonication regime was not inhibiting the recovery of broad peaks per se ., Furthermore , previously published Gro-Dam peaks that overlapped our ChIP-seq peaks tended to be broader than those that did not ( Figure S3 ) , indicating that our analysis was not biased against detecting any broad low affinity Gro peaks ., While we do observe some peaks of Gro binding in intergenic regions that may be associated with enhancer elements that are more than 1 kb from the nearest annotated TSS , our data support a model in which Gro is recruited locally by transcription factors and does not spread along the chromatin by oligomerization when it acts on a distant target promoter ., Thus , it is most likely that Gro recruited to distant regulatory elements is brought into the proximity of target promoters by “looping” of the DNA ., It is well established that chromatin looping can facilitate gene activation by bringing factors bound at intergenic enhancers into contact with the transcription machinery 50 , 51 and also facilitate repression by distant regulatory elements 52 ., Future studies using chromatin capture techniques in wild-type and Gro depleted cells will determine if Gro contributes to the formation and stability of chromatin loops from distant cis-regulatory elements to target promoters ., The RNA-seq experiments did not reveal a general upregulation of genes closely associated with Gro ChIP-seq peaks in cells treated with gro RNAi in either Kc167 or S2 cells ( Figures S5 , S6 , Tables S2 , S4 ) ., Indeed treatment with gro RNAi led to very few significant changes in gene expression ., Similarly , we did not observe widespread Gro-related changes to histone acetylation status or RNAP II recruitment or pausing ., We only observed highly significant changes to gene expression and RNAP II recruitment at a single known Gro target , E ( spl ) mβ-HLH ., It is possible that loss of Gro may have led to increased variability in target gene expression , and the average expression values from many cells in our two biological replicates is unlikely to be sufficient to show any change in variability ., However , genome-wide loss of Gro from its targets may not facilitate recruitment of activating factors in the absence of other changes in the nuclear environment ( e . g de novo expression of transcription factors in response to cell-cell signaling ) ., In addition , the residual Gro in these cells may be sufficient to maintain repression of most target genes ( Figure S5 , S7C ) ., The use of gronull cells made by newly available genome engineering techniques 53 may resolve this in the future - if gronull cells are viable ., Previous overexpression studies in S2 cells and in the fly indicate that oligomerization affects how Gro acts in cells 9 , 11 ., For example , ectopic expression of wild-type Gro leads to ectopic repression of the vgQ-lacZ reporter gene whereas overexpression of the non-oligomerizing GroL38D , L87D variant has no detectable effect on vgQ-lacZ expression 11 ., We do not observe dramatic differences in the breadth or location of Gro peaks with a variant that does not oligomerize ( L38D , L87D-GFP ) , lending support to the alternative models that it is the efficiency of Gro recruitment or overall structure of the co-repressor complex that is compromised in the presence of non-oligomerizing variants 9 ., We observe an apparent reduction in the amount of L38D , L87D-GFP binding with respect to Gro-GFP at the Rh5 locus ( Figure 4C ) although this effect is not observed at E ( spl ) mβ-HLH ., This indicates that the level of Gro binding may be dependent on oligomerization at a subset of targets ., Genetic evidence indicates that gro is not expressed in vast surplus to requirement as many genetic interactions can be detected with gro heterozygotes ., For example , multiple gro mutations were isolated in screens for dominant suppressors of roDom 54 and ectopic Hairy expression in the eye 55 ., Our results are generally consistent with those from previous studies that identified an association of Gro with hypoacetylated histones H3 and H4 17 , 20 , 21 ., However , we did not detect significant changes in the histone acetylation status of histones H3 and H4 at Gro target sites when we reduced Gro levels in Kc167 cells ., We cannot formally rule out that the residual Gro left in cells treated with RNAi against gro is sufficient to maintain histones in a hypoacetylated state or that there are subtle changes to acetylation levels that cannot be accurately detected by ChIP-seq methods ., Furthermore , loss of repression and gene activation are separable processes and depletion of Gro did not facilitate the recruitment and activity of histone acetylases at levels that we could detect ., Recent studies have revealed that regulation of promoter proximal pausing by RNAP II is a major point of control of the expression of many genes that respond to developmental and environmental cues ., Paused polymerase is highly enriched at genes in stimulus-responsive pathways 56 and in genes involved with patterning the axes in the early Drosophila embryo 44 ., Strikingly , Gro has critical functions regulating gene expression in stimulus-responsive pathways ( e . g . Notch and Wnt signaling ) and both AP and DV patte

Introduction, Results, Discussion, Materials and Methods

Gene expression is regulated by the complex interaction between transcriptional activators and repressors , which function in part by recruiting histone-modifying enzymes to control accessibility of DNA to RNA polymerase ., The evolutionarily conserved family of Groucho/Transducin-Like Enhancer of split ( Gro/TLE ) proteins act as co-repressors for numerous transcription factors ., Gro/TLE proteins act in several key pathways during development ( including Notch and Wnt signaling ) , and are implicated in the pathogenesis of several human cancers ., Gro/TLE proteins form oligomers and it has been proposed that their ability to exert long-range repression on target genes involves oligomerization over broad regions of chromatin ., However , analysis of an endogenous gro mutation in Drosophila revealed that oligomerization of Gro is not always obligatory for repression in vivo ., We have used chromatin immunoprecipitation followed by DNA sequencing ( ChIP-seq ) to profile Gro recruitment in two Drosophila cell lines ., We find that Gro predominantly binds at discrete peaks ( <1 kilobase ) ., We also demonstrate that blocking Gro oligomerization does not reduce peak width as would be expected if Gro oligomerization induced spreading along the chromatin from the site of recruitment ., Gro recruitment is enriched in “active” chromatin containing developmentally regulated genes ., However , Gro binding is associated with local regions containing hypoacetylated histones H3 and H4 , which is indicative of chromatin that is not fully open for efficient transcription ., We also find that peaks of Gro binding frequently overlap the transcription start sites of expressed genes that exhibit strong RNA polymerase pausing and that depletion of Gro leads to release of polymerase pausing and increased transcription at a bona fide target gene ., Our results demonstrate that Gro is recruited to local sites by transcription factors to attenuate rather than silence gene expression by promoting histone deacetylation and polymerase pausing .

Repression by transcription factors plays a central role in gene regulation ., The Groucho/Transducin-Like Enhancer of split ( Gro/TLE ) family of co-repressors interacts with many different transcription factors and has many essential roles during animal development ., Groucho/TLE proteins form oligomers that are necessary for target gene repression in some contexts ., We have profiled the genome-wide recruitment of the founding member of this family , Groucho ( from Drosophila ) to gain insight into how and where it binds with respect to target genes and to identify factors associated with its binding ., We find that Groucho binds in discrete peaks , frequently at transcription start sites , and that blocking Groucho from forming oligomers does not significantly change the pattern of Groucho recruitment ., Although Groucho acts as a repressor , Groucho binding is enriched in chromatin that is permissive for transcription , and we find that it acts to attenuate rather than completely silence target gene expression ., Thus , Groucho does not act as an “on/off” switch on target gene expression , but rather as a “mute” button .

sequencing techniques, genome expression analysis, invertebrates, gene regulation, dna-binding proteins, animals, next-generation sequencing, genome analysis, transcription factors, molecular biology techniques, drosophila, proteins, gene expression, molecular biology, insects, arthropoda, chromatin immunoprecipitation, biochemistry, high throughput sequencing, transcriptome analysis, genetics, biology and life sciences, genomics, computational biology, organisms

journal.pgen.1005975

Gifsy-1 Prophage IsrK with Dual Function as Small and Messenger RNA Modulates Vital Bacterial Machineries

The first systematic searches for bacterial sRNA were based on bio-computational identification of conserved genes in intergenic regions 1 ., The subsequent characterization of these conserved sRNAs identified them as important players in many adaptive and physiological responses ., These conserved core-genome encoded regulatory sRNAs comprise many antisense RNAs of which a subset are cis-encoded whereas the majority acts on trans-encoded target mRNAs by limited base complementarity 2 ., Most trans-acting base-pairing sRNAs of enteric bacteria require the RNA chaperone protein Hfq for both intracellular stability and for efficient annealing to target mRNAs 3 ., However , the chromosomes of these bacteria are mosaics , composed of conserved collinear regions interspersed with unique genetic islands that were acquired horizontally via once-mobile genetic elements ., Therefore , the early searches based on sequence conservation generally disregarded unique horizontally acquired sRNAs ., However , subsequent global cDNA cloning , comparative genomic based expression screens , Hfq-bound and global transcriptomic screens detected short RNA species in non-conserved horizontally acquired regions as well as highly abundant short RNA species from the UTRs of protein-coding genes 4–8 ., The function of these non-conserved sRNAs remains enigmatic yet promising , as they may inform new regulatory principles ., Members of the genus Salmonella carry numerous genomic islands that were acquired by horizontal transfer of phages , plasmids and transposons ., These islands carry fitness and virulence genes that are integral to Salmonella pathogenicity , enabling the bacteria to adapt to different niches , invade intestinal cells and multiply within cells of the immune response 9 ., In a previous study , we screened the horizontally acquired genomic islands of Salmonella typhimurium for non-conserved sRNA genes ., Our analysis led to identification of 19 unique island-encoded sRNAs including the Gifsy-1 prophage encoded IsrK RNA 6 ., The chromosome of Salmonella is lysogenic for a number of phages including Gifsy-1 and Gifsy-2 , both of which carry genes implicated in Salmonella virulence 9 , 10 ., Bacteriophage Gifsy-2 carries the sodCI gene encoding a periplasmic superoxide ( Cu , Zn ) -dismutase , with a proposed role in Salmonella defense against killing by macrophages as well as a number of genes encoding type III secreted effectors 11–13 ., Gifsy-1 carries multiple virulent factors such as gipA , which is involved in bacterial colonization of small intestine 14 , 15 ., To coordinate expression between core and island genes , bacteria often recruit sRNAs 16 , 17 ., For example , InvR from the major SPI-1 island represses the production of the core-encoded major outer membrane protein OmpD 18 ., IsrE , the paralogue of RyhB , represents another example of a cross talk between genes of core and islands 6 ., The island-encoded sRNA IsrE is regulated by Fur , a core-encoded repressor in response to iron-deplete conditions and contributes to control of the core-encoded iron regulon 6 , 19 ., Here we show that IsrK of Gifsy-1 prophage controls the expression of its genetic locus leading to growth arrest of Salmonella by acting as small and messenger RNA , in an Hfq-independent manner ., The growth inhibition is caused by an increase in expression of a Q-like antiterminator protein ( here denoted AntQ ) that is encoded on the same locus ., AntQ belongs to Q proteins’ family of lambdoid phages ., Bacteriophage λ Q protein is an operon specific transcription anti-termination factor required for expression of the phage late genes ., Q protein joins the elongation complex at early stages of transcription and enables RNA polymerase to read-through the terminator located upstream of the phage late genes 20–23 ., To join the elongation complex , Q interacts with a specific DNA sequence element as well as with RNA polymerase that is paused during early elongation ., The binding of Q alters the functional properties of the transcription elongation complex interfering with termination signals 24 , 25 ., We find that although Q proteins are known to bind specific sites within phages , the function of the Gifsy-1 Q-like protein AntQ is not limited to Salmonella or phage DNA ., By contrast AntQ promotes transcription elongation of core genome transcripts resulting in growth arrest and ultimately cell death ., The gene encoding IsrK sRNA is located within Gifsy-1 prophage ., Upstream of IsrK is SL2579 encoding a Q-like anti-terminator protein ( here denoted AntQ ) ., The transcription start-site of antQ was mapped 1 , 600 bases upstream of antQ 8 ., Downstream of isrK we noticed a putative ORF of 45 amino acids , followed by SL2578 encoding a predicted anti-repressor-like protein ( denoted AnrP ) and the previously identified sRNA encoding gene , isrJ ( Fig 1A ) 6 ., To examine whether transcription of the downstream gene anrP is linked to IsrK , we constructed isrK-orf45-anrP-lacZ transcriptional fusions with and without the isrK promoter ., The assays showed that IsrK promoter directs transcription of the downstream gene anrP ., Interestingly , the corresponding translation fusion demonstrated that although the operon is transcribed , there is no translation of anrP mRNA ( Table 1A ) ., We noticed that plasmid borne constitutive expression of isrK is toxic ., Salmonella cells fail to yield any colonies when transformed with plasmids expressing isrK , constitutively ( Fig 1B ) ., To further investigate the toxic phenotype , the isrK gene was cloned under the inducible PBAD promoter and we followed bacterial growth in wild type and a strain deficient of the chromosomal isrK promoter ., High levels of IsrK expressed in trans result in growth arrest of wild type Salmonella ( Fig 1C ) ., The growth inhibition is not observed when the chromosomal isrK promoter is deleted , suggesting that the genetic regulation leading to toxicity requires expression of the isrK locus in both cis and trans ., In addition , we examined whether the IsrK-dependent toxicity involves lysis of the host by Gifsy-1 induction ., To this end , we deleted Gifsy-1 genes SL2575 and SL2576 encoding proteins of phage lysis and phage lysozyme superfamily , respectively ., Both mutant strains failed to yield any colonies when transformed with plasmids expressing isrK constitutively , indicating that the growth arrest phenotype does not involve Gifsy-1 induction and lysis of the host ( S1 Fig ) ., Deletion mapping at the isrK locus to identify the cause of toxicity demonstrated that strains deleted for sequences upstream ( antQ ) or downstream ( anrP ) of isrK showed no growth inhibition , forming normal size colonies when transformed with a plasmid expressing constitutive high levels of IsrK , indicating involvement of both genes ( S2A Fig ) ., To define whether any of the above-mentioned genetic elements is toxic when expressed alone , in the absence of IsrK , we transformed strains deleted of the locus including the lysis genes up to antQ with plasmids expressing antQ or anrP from Ptac promoter under the control of the lacI repressor ., Whereas cells transformed with plasmids expressing anrP formed regular colonies ( S2B Fig ) , cells expressing antQ fail to grow , indicating that AntQ is sufficient for toxicity ., The growth arrest data indicated that toxicity involves IsrK RNA present in cis and expressed in trans , hence , we monitored the effect of IsrK expressed in trans on transcription and translation at the isrK locus ., A northern blot probed with an isrK specific primer , detected a long transcript of ~ 900 nucleotides when isrK was expressed in trans ( S3A Fig ) ., Probing the northern blot with an isrJ specific primer demonstrated that the long transcript encompasses anrP , suggesting that transcription starting at the isrK promoter reads-through the isrK Rho-independent transcription termination signal downstream into orf45 and anrP ., Analysis of shorter RNA species using isrK specific primer supports expression of the plasmid encoded isrK gene ( S3B Fig , lanes 4–9 ) , as well as chromosomally encoded short isrK form , indicating that transcription starting at the isrK promoter produces the IsrK sRNA as well as isrK operon mRNA ( S3B Fig , lanes 1–3 ) ., In addition , a lacZ-transcription fusion starting at the isrK promoter ( PisrK-isrK-orf45-anrP-lacZ ) showed that isrK expressed in trans increased the levels of the long transcript by ~8 fold ., The anrP translation fusion also showed that isrK expressed in trans activates translation of anrP by 140 fold ( Table 1B ) ., Together these data demonstrate that RNA polymerase partially reads-through the isrK Rho-independent transcription termination signal downstream into orf45-anrP and that IsrK sRNA acting in trans causes a slight increase in the levels of the downstream polycistronic mRNA and activates anrP translation ., Gifsy anti-repressor proteins bind to and inactivate the lysogenic repressor , thereby leading to transcription of phage operons 26 ., To learn about the correlation between increased levels of IsrK sRNA , the anti-repressor protein AnrP and the anti-terminator AntQ , we monitored antQ mRNA levels upon expression of isrK and anrP , using quantitative Real-Time PCR ., This analysis showed that in trans expression of isrK resulted in increased antQ mRNA levels ., Similarly , in trans expression of anrP led to higher antQ transcript levels ( Fig 2A and 2B ) ., To learn about the activity of the anti-repressor protein AnrP , we measured RNA levels of SL2581 , the second gene of antQ operon ., Similarly to antQ , SL2581 levels increased upon expression of isrK or anrP , indicating that AnrP activates transcription of the antQ operon most likely through activated transcription activity ( S4 Fig ) ., Together , the results indicate that IsrK activates expression of anrP , which in turn leads to AntQ synthesis ., Furthermore , we monitored Gifsy-1 prophage induction upon expression of isrK and anrP ( S5 Fig ) ., Phage plating on a susceptible strain demonstrates that Gifsy-1 phage induction by IsrK requires an intact isrK locus , whereas Gifsy-1 induction by AnrP is independent of isrK locus ., These results further support the regulatory cascade we present for the isrK locus and the biological relevance of this locus to phage development ., We also observed oxidative stress dependent general phage induction ( Materials and Methods and S5C Fig ) , upon which the levels of IsrK sRNA increase during the first minutes of exposure to hydrogen peroxide while antQ and SL2581 mRNA levels increase gradually ( S3D Fig and S6 Fig ) ., To visualize the translation pattern of the downstream operon including orf45 and anrP , we integrated the coding sequence of the sequential peptide affinity ( SPA ) tag 27 into orf45 and anrP to generate C-terminal fusion proteins , in two separated strains ., The western blot showed that IsrK expressed in trans increases translation of both orf45 and anrP ( Fig 3 ) ., orf45 carries two in frame initiation codons and a stop codon overlapping the initiation codon of anrP ( AUGA ) ., Nucleotide and amino acid conservation analysis demonstrated that the nucleotide sequence of orf45 is conserved among the enterobacteria ( S7 Fig ) , whereas the amino acid sequence of orf45 varies ( S8 Fig ) ., The proximity of orf45 to anrP prompted us to examine their potential translation coupling ., By mutating the initiation codons we found that translation starting at the first initiation codon leads to translation of anrP ., Moreover , insertion of a stop codon proximal to the translation initiation site reduced anrP translation activation by IsrK ( Table 1C ) ., Together , our data indicate that isrK expressed in trans and orf45 translation are required to stimulate anrP translation ., To investigate the mechanism of the translational regulation of isrK-orf45-anrP by IsrK , we induced random mutations at this locus using PisrK-orf45-anrP-lacZ translation fusion plasmid and screened for high-level expression mutants in the absence of in trans isrK ., Two high-level expression mutants were found to carry mutations within isrK ( G28A and G31A ) ., Structural prediction analysis using RNA fold program ( http://rna . tbi . univie . ac . at/ ) show that the wild type transcript ( 1–180 nt ) forms one conformation ( A ) having a ΔG of -85 . 66 kcal/mol , whereas the mutated RNA forms an alternative conformation ( B ) with a predicted ΔG value of -84 . 21 kcal/mol ( Fig 4A and 4B ) ., Functional studies of translation fusions carrying mutations G28A or G31A showed that G28A and G31A , which are predicted to form the alternative structure B , increase the basal level of AnrP translation ( Table 1D ) , indicating that structure B is translationally active , whereas structure A is translationally silent ., To visualize the two isoforms and to learn about their ratios in the wild type RNA , we examined the RNAs on nondenaturing polyacrylamide gels ., The native gels demonstrate that wild type RNA is found almost exclusively in one structure , while G31A and G28A mutant RNAs display two conformers of which one resembles the wild type conformation and the other represents the alternative structure B ( Fig 4C ) ., In the inactive structure formed by wild type RNA ( A ) , the middle part of IsrK ( in purple ) base pairs with ~ 30 nt long sequence of orf45 ( in blue ) forming helix b-I ( Fig 4A and 4B ) ., In this structure the ribosome-binding site of orf45 forms hairpin d-I ., In the alternative structure ( B ) , the middle part of IsrK forms an alternative hairpin ( b-II ) , whereas the RBS of orf45 forms a new helix by pairing with its 3’-end ( d-II ) ., Mutations G28A and G31A are likely to destabilize structure A by disrupting the middle helix , but are predicted to have no effect on structure B . To examine base pairing , we modified helix b opposite to G28A and G31A to carry the corresponding complementary mutations C162U and C159U , respectively and when combined , would restore formation of the helix ., RNA mutants carrying G28A/C162U or G31A/C159U exhibit one conformation , the same as wild type RNA ( Fig 4A , 4B and 4C ) , indicating that G28A basepairing with C159U and G31A/C162U basepairing form structure A . Functional studies of translation fusions carrying C162U and C159U showed that like mutations G28A or G31A , C162U and C159U mutants exhibited a high basal level of anrP translation ( Table 1D ) ., The basal levels decrease when these mutations are combined with the corresponding complementary mutations further indicating that structure A is translationally inactive , whereas structure B is translationally active ., Because mutations C162U and C159U affect the stability of structure B in addition to A , they exhibit a higher basal level of translation than that observed for the opposite mutations ( see below ) ., To affirm the differences between the two structures , we constructed mutations G114A and the corresponding complementary mutation C175U , both predicted to destabilize structure B with no effect on structure A . Given that wild type RNA is found almost exclusively in conformation A , these mutations only a mildly affected translation of anrP ( S1 Table ) ., Fig 4A shows that in structure A , the middle part of the cis-encoded IsrK ( in purple ) binds a long sequence of orf45 ( in blue ) ., Given the complementarity between cis-encoded IsrK and orf45 and the influence of in trans expression of isrK on downstream translation , we explored the functional and structural consequences of IsrK binding to structures A and B , in trans ., In binding to structure A , IsrK is predicted to compete with its own sequence for the binding of the middle helix ( Fig 5 ) ., Binding of structure B by IsrK is predicted to destabilize the helix d-II that sequesters the RBS of orf45 ( Fig 5 ) ., Mutational analysis supported binding of in trans IsrK to the cis-encoded isrK-orf45-anrP target mRNA ., Mutations C159U and C162U are predicted to affect base paring with IsrK by replacing CG pairs with UG pairs ( Fig 5 ) ., Functional studies of translation fusions of C159U and C162U mutant RNAs demonstrate that wild type IsrK weakly affected translation of anrP , indicating that stable binding of IsrK in trans is important for anrP translation activation and that destabilization of the helix formed between in trans IsrK and cis-encoded orf45 abrogates anrP translation control by IsrK ( Table 1D ) ., Similarly , mutation G173A is predicted to affect base paring with wild type IsrK by replacing a GC pair with an AC pair ( Fig 5 ) ., Translation fusions studies of orf45 carrying G173A mutation demonstrated that wild type IsrK RNA is less effective in activation of anrP translation than an IsrK mutant carrying the corresponding complementary mutation C18U ( Table 1E ) ., Likewise , mutation C175U replaces CG pair with UG pair and anrP translation activation by wild type IsrK is less productive ., Moreover , because of imperfect basepairing , IsrK activation of G173A is lower than that of C175U mutant ( S1 Table ) ., The effect of IsrK acting in trans is visible in native gels ., Incubation of wild type RNA with IsrK at 37°C results in minimal binding of structure A to IsrK ( S8 Fig ) ., In accordance , binding of structure A of isrK-orf45G31A RNA by IsrK is indistinct as opposed to binding of the structure B of this RNA mutant ( S8 Fig ) ., Pre-incubation of the target RNAs at 70°C to denature the structures , prior to their incubation with IsrK , facilitates binding by IsrK ., Under these conditions , IsrK binds structure A , characteristic of wild type RNA and both structure A and B that are characteristic of isrK-orf45G31A ( Fig 6A ) ., No binding can be detected when isrK-orf45 wild type RNA is incubated with isrKG31A mutant further confirming that this mutant is inactive as supported by our cultivation experiments ( Fig 6B , S10 Fig ) ., Since IsrK binds orf45 adjacent to its RBS , detection of 30S binding in real time ( toe printing ) is inconceivable ., However , 30S binding at the RBS would protect the neighboring upstream and downstream sequences ., To probe the accessibility of the sequence surrounding the RBS , we used dimethyl sulfate ( DMS ) , which methylates unpaired adenosine and cytidine residues at N1 and N3 positions , respectively ., Samples were incubated with and without 30S and/or IsrK prior to the addition of DMS ., The modified sites were detected by primer extension after phenol extraction ., A few nucleotides that surround the RBS of orf45 are susceptible to DMS modification in the presence of 30S ribosomes or IsrK ( Fig 7 lanes 2 , 4 ) ., The same nucleotides are protected from DMS in presence of both IsrK and 30S ( Fig 7 lane 6 ) , indicating that wild type IsrK facilitates 30S binding to the RBS of orf45 ., 30S protection from DMS decreases much in the presence of isrKG31A mutant that is unable to bind orf45 ( S11 Fig ) ., On the one hand , IsrK is part of the target isrK-orf45-anrP mRNA , and as such mutations in IsrK affect the structure this mRNA forms ., For instance , G28A disrupts helix b in structure A , shifting the equilibrium towards the translationally active structure B . Therefore , mRNA carrying isrKG28A-orf45-anrP exhibits a high basal level of anrP translation ., On the other hand , the short IsrK acts in trans to destabilize the cis-encoded translationally inactive target mRNA leading to anrP translation ., Therefore , one mutation in isrK gene is predicted to yield two different phenotypes ., We examined the effect of isrKG28A in cis and in trans and found that whereas , cis-encoded isrKG28A-orf45-anrP exhibits a high basal level of anrP translation; isrKG28A acting in trans is unable to activate anrP translation ( Table 1F ) ., In accordance , high levels of isrKG28A or isrKG31A expressed from the PBAD promoter have no effect on growth of wild type Salmonella ( S10 Fig ) ., We have shown that toxicity involves expression control of anrP by IsrK that in turn induces transcription of the anti-terminator protein AntQ ( Fig 2A and 2B ) ., Considering the origin of antQ , i . e . , Gifsy-1 prophage , we examined whether its toxicity is specific to Salmonella and/or phage genes ., Accordingly , the influence of high levels of AntQ was investigated in two E . coli K-12 strains; wild type ( MG1655 ) and MDS42 that is deleted of all genetic islands including prophages and insertion elements 28 ., antQ expression repressed growth and decreased survival of both strains ., At 40 minutes of induction , survival of wild type and MDS42 decreased by ~15 and ~30 fold , respectively , indicating that toxicity of the Q-like anti-terminator protein is not specific to Salmonella or phage DNA ( S12 Fig ) ., Moreover , the results suggest that AntQ protein has natural recognition sites within the core genome of these strains ., Bacteriophages Q antiterminator proteins interfere with transcription termination by binding specified sites at promoter regions and forming a persistent complex with RNA polymerase ., This complex of RNA polymerase and Q protein can bypass terminators 24 ., We examined changes in protein expression pattern of Salmonella upon exposure to AntQ using one-dimensional SDS-PAGE ., Gel areas showing differences in the pattern of proteins because of AntQ were isolated and subjected to mass spectrometry ( S13 Fig ) ., Two proteins , whose expression increased , were selected for further analysis because of their score and annotation; the transcription termination factor Rho and the DEAD-box-containing ATP-dependent RNA helicase SrmB 29–32 ., We suspected that expression of rho and srmB increased in response to transcription elongation related stress ., Thus , we examined whether co-expression of antQ with these genes would abolish the toxic effect of AntQ ., Survival assays show that co-expressing rho with antQ help to rescue cells from AntQ-mediated toxicity ., After 40 minutes of induction , survival of cells expressing antQ alone dropped to ~8% of their original amount , whereas cells expressing both antQ and rho managed to maintain a high CFU count , indicating that Rho can halt the toxicity inflicted by AntQ ., Likewise , survival assays in which srmB and antQ were co-expressed demonstrated that SrmB prevented AntQ toxicity ( Fig 8A and 8B ) ., Together the results show that proteins that harbor RNA helicase activity impede the toxic effects of transcription anti-termination ., Unregulated transcription elongation increases formation of DNA-RNA hybrids upstream of RNA polymerase ( R-loops ) ., The resulting R-loops may initiate DNA replication independently of oriC , leading to DNA damage 33 , 34 ., RNase H is an evolutionary conserved helicase that resolves R loops , thus protecting genomic DNA from breaks 30 , 35 ., Survival rates of cells co-expressing antQ and rnhA encoding RNase H were 10 fold higher than those expressing antQ alone , suggesting that the toxic effects of AntQ result from DNA damage due to the creation of R-loops ( Fig 8C ) ., It is well documented that exposure of bacteria to detrimental stressful conditions impairing protein synthesis or causing DNA damage , triggers genome condensation 36 ., We visualized the effect of antQ expression on chromatin morphology by florescence microscopy ., Images of Salmonella expressing a control plasmid display , as excepted , chromatin spread over the entire cytoplasm ., In contrast , images of Salmonella expressing antQ reveal condensed chromatin morphology ., Likewise , exposure of Salmonella cells to nalidixic acid ( NA ) , a pleiotropic drug that inflicts diverse DNA lesions ( nicks , gaps , and DSBs ) 36 resulted in genome condensation ( Fig 9 ) ., Furthermore , E . coli cells expressing antQ or exposed to NA exhibit genome condensation , similarly to Salmonella , indicating that antQ toxicity is mechanistically conserved ., The E . coli UvrD protein is a DNA helicase/translocase that functions in methyl-directed mismatch repair ( MMR ) nucleotide excision repair ( NER ) and more broadly in genome integrity maintenance 37 ., Recent studies in E . coli have shown that UvrD can act as an accessory replicative helicase that resolves conflicts between the replisome and transcription complexes 37–39 ., Using uvrD-yfp 40 , we demonstrate in vivo localization of the fluorescently tagged uvrD to the nucleoid upon expression of antQ and upon exposure to DNA damaging agents ( Fig 10 ) ., Together our data show that the function of the phage antiterminator protein is wide-ranging causing changes in bacterial chromatin morphology ., In this study we show that a subset of the IsrK sRNA transcripts reads through its transcription terminator to form a translationally inactive bi-cistronic mRNA ., Concomitantly , short IsrK RNA acts in trans , interacting with the inactive transcript to promote formation of a translationally active structure , in which orf45 translation leads to anrP translation by translational coupling ( model Fig 11 ) ., In bacteria , translational coupling provides a mechanism to coordinate expression of multiple proteins with adjacent or overlapping coding sequences ., Ribosomes terminating translation of upstream ORF dissociate and re-initiate translation at the downstream RBS 41 , 42 ., Re-initiation is enabled due to ribosomes elongating along the upstream ORF to unfold mRNA structures that sequester the downstream ribosome-binding site ., Such an example is PhrS sRNA that activates translation of pqsR mRNA by interaction with a sequence sequestering the RBS of an ORF upstream of pqsR 43 ., We find that inserting a stop codon proximal to the translation initiation site of orf45 reduces IsrK-controlled anrP translation activation , indicating that translational coupling is necessary for AnrP synthesis ., Given that the structural changes caused by IsrK and/or ribosome binding at the orf45 RBS do not seem to involve structural changes in the RBS of anrP , we suggest that the translational coupling between orf45 and anrP requires ribosome elongation from the RBS of the orf45 downstream to anrP ., A structural homolog of IsrK is SeqA RNA of P4-like phages 44 ., In the lysogenic state P4 prevents expression of its own replication genes by premature transcription termination ., The factor responsible for efficient termination is CI RNA that is generated by processing of a primary untranslated transcript ., CI RNA acting as an antisense RNA leads to transcription termination by pairing with two complementary sequences , seqA and seqC located upstream and downstream of CI , respectively 45–47 ., In Salmonella , transcriptome analysis revealed the existence of a stable non-coding RNA species downstream of IsrK ( STnc1160 ) 8 ., Our RNA analysis detected STnc1160 in wild type cells but not in a strain deleted of the isrK promoter , suggesting that STnc1160 is generated by processing of the readthrough transcript initiating from the isrK promoter ( S3 Fig ) ., It is possible that STnc1160 similarly to CI modulates transcription termination at IsrK Rho-independent terminator ., Since IsrK and STnc1160 share complementary sequences , IsrK binding of STnc1160 renders it inactive as a termination factor leading to transcription readthrough and thus to grow arrest ., However , in experiments of co-expression of isrK and orf45 ( STnc1160 ) in which STnc1160 was constitutively expressed , IsrK mediated growth arrest was even more pronounced ( S14 Fig ) ., In S3B Fig , we present a northern blot showing the levels of chromosomally and plasmid encoded isrK ., It is interesting to note that in addition to short IsrK , our analysis revealed a stable transcript ( isrK-orf45’ ) that is generated by processing of the long polycistronic transcript ., isrK-orf45’ species is observed upon expression of plasmid-encoded isrK in wild type cells ( lane 6 ) , as well as upon expression of chromosomally-encoded isrK ( see lane 1–3 ) , indicating that the pattern detected with high level expression is valid with chromosomally-encoded isrK ., In addition , the results demonstrating that Gifsy-1 phage induction by IsrK requires an intact isrK locus , whereas Gifsy-1 induction by AnrP is independent of isrK locus , further substantiate the regulatory cascade we present for the isrK locus and signify the biological relevance of this locus to phage activation ., Moreover , we show that wild type cells grown in minimal medium to stationary phase exhibit prophage induction , whereas isrK promoter deletion mutant ( ΔPisrK::frt ) fails to produce phages ( S15 Fig ) ., These findings indicate that IsrK is an important player in initiating prophage induction ., Concerning the conditions inducing isrK expression , we find that IsrK levels increase at stationary phase and under low Mg2+ conditions ( S15 Fig ) ., Salmonella global transcriptome analysis carried out by Kröger et al 8 shows that IsrK levels increase during conditions such as low Fe2+ shock , oxygen shock and growth in InSPI2 medium ., In addition they find that the levels of the transcript encoding orf45 resulting by transcription elongation through the isrK transcription terminator ( STnc1160 ) increase during low Fe2+ shock , InSPI2 and late stationary phase ., Together , the data indicate that isrK short and long forms are produced under a variety of environmental conditions ., The majority of the sRNA genes are encoded within intergenic regions acting in trans to control expression of physically unlinked target genes ., However , it is now increasingly appreciated that in addition to intergenic regions , many sRNAs originate from the 5’ or 3’ regions of coding mRNAs ., Such examples are 3’ UTR derived sRNAs that are generated either by internal processing of the related mRNA , as in the case of RybD or produced as a primary transcript like MicL and DapZ 7 , 48 ., Generated from within protein coding loci , these sRNAs act in trans controlling expression of unlinked target mRNAs ., Likewise , SreA and SreB originate from 5’ UTRs of two S-adenosylmethionine ( SAM ) riboswitches , and base pair with the unlinked prfA mRNA to repress translation 49 ., In Staphylococcus aureus , SprA1AS is transcribed from the strand opposite to SprA1 target mRNA encoding pepA1 ORF ., The antisense RNA SprA1AS acts in trans by base pairing with the 5’ domain of SprA1 to repress pepA1 translation by occluding its RBS 50 ., Somewhat different is the archaeal RNA regulator , sRNA162 ., sRNA162 masks the RBS of MM2441 by binding MM2440-MM2441 mRNA internally 51 ., Biochemical studies demonstrated that in addition to in trans binding of MM2441 RBS , encoded opposite of MM2442 , the 5’-end of sRNA162 targets the 5’-untranslated region of the cis-encoded MM2442 mRNA ., However , the regulatory outcome of this interaction is as yet unknown ., The mechanism of expression regulation of the isrK locus is unique , representing the first example of an RNA that acts as a small RNA on its own mRNA ., On the one hand , IsrK is part of a translationally inactive target mRNA , whereas on the other hand the short RNA species acts in trans to enable translation of the target mRNA ., Therefore one mutation within IsrK RNA yields two different phenotypes; when located in the long target mRNA it increases translation whereas the short mutant IsrK RNA can no longer activate translation ., Increasing evidences indicate that in prokaryotes and eukaryotes , common transcription-replication encounters lead to blockage of replication that is often accompanied by DNA damage and genome instability ., In bacteria , because replication and transcription proceed simultaneously on the same template DNA , yet DNA replication forks move 10 to 30 times faster than do RNA polymerases , both co-directional , and head-on collisions appear to be unavoidable 34 ., Transcription-replication conflicts may also result from stalled transcription elongation complexes , as they form stable barriers to the replication machinery ., These complexes increase the production and/or the length of DNA-RNA hybrid str

Introduction, Results, Discussion, Materials and Methods

While an increasing number of conserved small regulatory RNAs ( sRNAs ) are known to function in general bacterial physiology , the roles and modes of action of sRNAs from horizontally acquired genomic regions remain little understood ., The IsrK sRNA of Gifsy-1 prophage of Salmonella belongs to the latter class ., This regulatory RNA exists in two isoforms ., The first forms , when a portion of transcripts originating from isrK promoter reads-through the IsrK transcription-terminator producing a translationally inactive mRNA target ., Acting in trans , the second isoform , short IsrK RNA , binds the inactive transcript rendering it translationally active ., By switching on translation of the first isoform , short IsrK indirectly activates the production of AntQ , an antiterminator protein located upstream of isrK ., Expression of antQ globally interferes with transcription termination resulting in bacterial growth arrest and ultimately cell death ., Escherichia coli and Salmonella cells expressing AntQ display condensed chromatin morphology and localization of UvrD to the nucleoid ., The toxic phenotype of AntQ can be rescued by co-expression of the transcription termination factor , Rho , or RNase H , which protects genomic DNA from breaks by resolving R-loops ., We propose that AntQ causes conflicts between transcription and replication machineries and thus promotes DNA damage ., The isrK locus represents a unique example of an island-encoded sRNA that exerts a highly complex regulatory mechanism to tune the expression of a toxic protein .

As the function of conserved core-genome-encoded small RNAs ( sRNA ) reflects the basic lifestyle of bacteria , the function of non-conserved island-encoded sRNAs remains enigmatic ., The island-encoded sRNA IsrK belongs to Gifsy-1 prophage of Salmonella ., Here , we report a complex mechanism in which the IsrK RNA functions as both sRNA and mRNA to control the production of the toxic AntQ protein ., The isrK promoter directs the synthesis of two distinct RNA species: a full-length translationally inactive target mRNA and the correctly terminated , shorter IsrK sRNA ., IsrK sRNA binds the full-length inactive mRNA producing an antiterminator protein , AntQ , which interferes with transcription termination ., Expression of antQ results in bacterial growth arrest and ultimately cell death ., Fluorescence microscopy of E . coli and Salmonella expressing antQ revealed condensed chromatin morphology as observed upon exposure to DNA-damaging agents ., We propose that expression of the phage antiterminator protein results in conflicts between transcription and replication machineries and thus facilitates DNA damage ., In summary , the RNA regulator IsrK presents a new regulatory principle in which a horizontally acquired sRNA controls genome integrity .

medicine and health sciences, nucleic acid synthesis, pathology and laboratory medicine, bacteriophages, pathogens, messenger rna, microbiology, dna transcription, viruses, bacterial diseases, enterobacteriaceae, molecular biology techniques, rna synthesis, bacteria, bacterial pathogens, chemical synthesis, research and analysis methods, infectious diseases, artificial gene amplification and extension, medical microbiology, gene expression, microbial pathogens, denaturation, molecular biology, salmonella, biosynthetic techniques, biochemistry, rna, nucleic acids, polymerase chain reaction, protein translation, genetics, biology and life sciences, rna denaturation, organisms

journal.pgen.1000023

Redundant Function of REV-ERBα and β and Non-Essential Role for Bmal1 Cycling in Transcriptional Regulation of Intracellular Circadian Rhythms

Circadian rhythms in physiology and behavior are regulated by endogenous circadian clocks ., All the molecular clocks so far described in multicellular organisms constitute negative feedback loops in which protein products of clock genes inhibit transcription of their own genes 1 ., In mammals , the central pacemaker in the suprachiasmatic nuclei ( SCN ) integrates light-dark cycle input and coordinates oscillators in peripheral tissues 2 ., Like the SCN , peripheral tissues also contain cell-autonomous circadian oscillators ., The current cellular clock model comprises a core feedback loop consisting of PER and CRY repressors and BMAL1 and CLOCK activators 1 , 3 ., In the core loop , BMAL1/CLOCK heterodimers directly bind to E-box enhancer elements present in Per ( Per1 and Per2 ) and Cry ( Cry1 and Cry2 ) genes and activate their transcription; PER and CRY proteins in turn repress their own transcription through direct interactions with BMAL1/CLOCK ., The mammalian clock has been shown to contain additional interlocking loops ., In particular , the ROR/REV/Bmal1 feedback loop consists of the RORs ( RORa , b and c ) and REV-ERBs ( REV-ERBα and β ) , members of a subfamily of orphan nuclear receptors 4 , whose expression is directly regulated by the core loop 5–8 ., To drive rhythmic expression of Bmal1 , REV-ERBα represses Bmal1 transcription by directly binding to the ROR elements ( ROREs ) in the Bmal1 promoter 5 , 9; RORa and RORb , on the other hand , act as positive drivers to activate Bmal1 expression in the SCN 6 , 9–11 ., The roles of REV-ERBβ and RORc in clock function have not been addressed ., An analogous set of interlocking loops has been described in the Drosophila circadian clock 7 , 12 , 13 ., The dPER/dTIM repressors and dCLK/dCYC activators constitute the core feedback loop ., In the interlocked dClk feedback loop , the bZIP transcription factors dPDP1 and dVRI , which are directly controlled by the core loop , activate and repress dClk transcription , respectively ., However , unlike the requirement for cyclic expression of dPer and dTim mRNAs , it was shown that dClk mRNA cycling is not necessary for molecular and behavioral rhythms in flies 14–16 ., The dClk loop function in flies could not be precisely defined genetically , however , because mutants deficient in dVri and/or dPdp are developmentally lethal 12 , 16 ., The role of the ROR/REV/Bmal1 loop in mammals has not been precisely addressed either , due to functional redundancy of the RORs and REV-ERBs and pleiotropic effects of gene deletions ., As deletion of Rev-erbα , Rora or Rorb results in a broader distribution of circadian period lengths , it was suggested that the ROR/REV/Bmal1 loop serves as a stabilizing mechanism 5–7 , 10 ., However , Ror mutant mice exhibit potentially confounding non-circadian phenotypes ., Rorb−/− mice display reproductive deficits and a severe postnatal retinal degeneration 10 ., Rora knockout ( Rora−/− ) and mutant staggerer ( Rorasg/sg ) mice display cerebellar ataxia and are mostly infertile 6 , 11 , 17 , 18 ., Importantly , period dispersion is not unique to animals deficient in Ror or Rev-erb function; Per1−/− , Per2−/− and Clockm/m mice also display less persistent circadian behavior and larger variability of periods 19–22 ., Thus , circadian abnormalities in these mice measured using behavioral outputs may not faithfully reflect intracellular clock function ., Finally , functional redundancy cannot be addressed genetically at the behavioral level because compound knockout animals have gross defects ., The drawbacks of behavioral analysis can be circumvented by studies using cell-based clock models ., Strategically , molecular mechanisms required for rhythmicity are best studied at the cellular level using long-term recordings to assess persistence of circadian rhythmicity 23 ., In this study , by taking advantage of a cell-based experimental model and real-time bioluminescence monitoring of gene expression , we first define the roles of RORc and REV-ERBβ in peripheral clock function , and then extend our studies to include all the RORs and REV-ERBs and their respective contributions to circadian rhythms of Bmal1 expression ., Furthermore , we show that the REV-ERBs are necessary for Bmal1 rhythm while the RORs are dispensable , indicating that the REV-ERBs play a more prominent role than the RORs in the transcriptional circuitry of the clockwork ., Importantly , however , rhythmic Bmal1 mRNA and protein expression is not required for the basic operation of the core clock ., These results are in line with the observation that constitutive Bmal1 expression was able to rescue circadian behavioral rhythms in Bmal1−/− mice 24 ., We suggest that the major role of the ROR/REV/Bmal1 loop is to provide additional phase modulation for establishing transcriptional output networks ., We first examined expression of the Rors and Rev-erbs in various tissues ( Figure S1A , S1B , S1C ) ., In contrast to the ubiquitous expression of Bmal1 , Rev-erbα and Rev-erbβ in all the tissues examined , expression and rhythmicity of the RORs are more restrictive ., Rora expression is ubiquitous , but its circadian cycling is restricted to SCN ., Rorb is expressed in the SCN , hypothalamus , cerebral cortex and retina , but not in the liver ., Conversely , Rorc is rhythmically expressed in the liver , but not detected in the SCN or other brain regions ., Expression patterns of the Ror genes in the lung were similar to those in the liver ( data not shown ) ., The tissue-specific expression patterns of the RORs are consistent with previous reports 6 , 25–28 ., In this study , we extensively used fibroblasts derived from mice as a cell-based clock model ., Of the three Rors , only Rora is highly expressed in mouse fibroblasts , but no distinct mRNA rhythm was detected ( Figure S1A ) ; Rorb and Rorc were not detected in fibroblasts ( Figure S1A and S1B ) ., Differential tissue distribution and rhythmicity of the Rors suggests that they may have different functions in clock mechanisms ., Rora and Rorb have been characterized as clock components , functioning to regulate Bmal1 expression in the SCN ( Figure S1C ) 6 , 10 , 11 , 25 ., However , since RORc is not expressed in the SCN , it should not affect function of the SCN pacemaker , which drives circadian locomotor behavior ., We tested this hypothesis in a mouse line deficient in Rorc function ., Deletion of Rorc results in reduced survival of thymocytes and abnormal lymphoid organ development , but Bcl-xL transgene ( Bcl-xLTg ) expression restored most aspects of normal thymocyte development and significantly improved animal survival 29 ., Compared to Bcl-xLTg control ( period length τ\u200a=\u200a23 . 42 hr±0 . 08 , n\u200a=\u200a5 ) , Rorc−/−:Bcl-xLTg mice displayed normal circadian wheel-running activity under constant darkness ( τ\u200a=\u200a23 . 34 hr±0 . 2 , n\u200a=\u200a8 ) ., These mice also showed a normal response to a light pulse at CT16 ( Figure S1D ) ., We further examined the dynamics of molecular rhythms in the SCN and showed that SCN explants from Rorc−/−:Bcl-xLTg mice also displayed similar mPer2Luc bioluminescence rhythms to control mice ( data not shown ) ., Thus , consistent with the absence of Rorc gene expression in the SCN , these results confirm that RORc plays no role in SCN pacemaker function ., Based on the ability of RORc to activate a Bmal1-Luc reporter in vitro and its strong rhythmic expression in many peripheral tissues including the liver and lung 6 , 9 , 25 , we hypothesized that RORc , like RORa and RORb in the SCN , may play an important role as an activator of Bmal1 in peripheral oscillators ., We tested this hypothesis by analyzing Bmal1 expression in the mouse liver ., In Bcl-xLTg control mice , Bmal1 expression peaked at CT24 ( Figure 1A ) ., In contrast , Bmal1 expression at CT28 , CT44 and CT48 in the liver of Rorc−/−:Bcl-xLTg mice was significantly reduced , compared to those of Bcl-xLTg siblings ( Figure 1A ) ., These results suggest that RORc activates Bmal1 transcription in the positive arm of the ROR/REV/Bmal1 loop , functioning to maintain normal amplitude of Bmal1 rhythmicity ., Although Bmal1 peak expression levels are reduced in the absence of RORc , Bmal1 mRNA still retains a rhythm with fairly high amplitude , indicative of functional redundancy from RORa and/or contributions from the REV-ERBs ., RORc also regulates transcription of Cry1 , Clock and Npas2 , all of which are considered RORE-containing genes 30 , 31 , and their mRNAs were also reduced during peaking hours ( Figure 1A ) ., Despite the blunted rhythm amplitudes for Bmal1 , Clock , Npas2 and Cry1 , cyclic expression of Per2 and Dbp , however , was not dramatically affected by Rorc deletion , similar to observations in Rev-erbα−/− mice 5 ., As the SCN clock functions normally in the absence of Rorc , we assessed the effect of Rorc deletion on peripheral clock function in tissue-autonomous preparations in which confounding influences from the SCN are eliminated ., Tissue explants of the lung from Bcl-xLTg control mice displayed persistent mPer2Luc rhythms ( τ\u200a=\u200a24 . 00 hr±0 . 33 , n\u200a=\u200a4 ) ., Rorc−/−:Bcl-xLTg lung explants exhibited rhythmic mPer2Luc expression with comparable period lengths to controls ( τ\u200a=\u200a24 . 15±0 . 49 , n\u200a=\u200a4 ) ( Figure 1B ) ., Rorc−/−:Bcl-xLTg liver explants also displayed persistent mPer2Luc rhythms ( τ\u200a=\u200a22 . 59 hr±1 . 54 , n\u200a=\u200a5 ) , similar to controls ( τ\u200a=\u200a22 . 22 hr±0 . 71 , n\u200a=\u200a3 ) ., Surprisingly , no significant differences in circadian amplitude or damping rate were observed between controls and Rorc−/− mice ., The normal bioluminescence rhythms are consistent with unaltered molecular phenotypes of Per2 expression ( Figure 1A ) ., Moreover , we observed normal rhythms in fibroblasts , in which Rorc expression is not detectable ( data not shown ) , further confirming results from liver and explants ., In fibroblasts , over-expression of Rorc did not affect Bmal1 rhythms ( data not shown ) ., These results demonstrate that RORc does not play an essential role in maintaining circadian oscillation and suggest that a high-amplitude Bmal1 rhythm may not be critically required for basic clock operation , similar to phenotypes observed for Rev-erbα deficiency 5 ., So far , data suggest a functional redundancy among RORa , RORb and RORc ., In the liver and fibroblasts of both Rorasg/sg 6 , 11 and Rorc−/− mice , Bmal1 peak expression is reduced , but the mRNA rhythm is largely retained and Per2 oscillation is not altered ., Although Rora does not show strong rhythmicity in the liver , its expression alone could partially complement the loss of Rorc ., To study the ROR redundancy genetically , a mouse line deficient in both Rora and Rorc would represent an ideal reagent ., However , such a line is extremely difficult to obtain because Rorasg/sg mutant mice display cerebellar ataxia and mostly infertile 18 and Rorc−/− mice also have strongly abnormal phenotypes 29 ., Therefore , we decided to address the ROR redundancy using Rorasg/sg fibroblasts ., Because Rorb and Rorc are also not expressed in Rorasg/sg fibroblasts as determined by Q-PCR ( data not shown ) , thus excluding the possibility of a compensation mechanism , the positive arm of the ROR/REV/Bmal1 loop is essentially missing in cells lacking Rora function ., To monitor the function of the core loop and the ROR/REV/Bmal1 loop in parallel , we generated two lentivirus-mediated circadian reporters , pLV6-Per2-dLuc and pLV6-Bmal1-dLuc , designed to report the E-box and RORE-regulated rhythms , respectively ., As expected , WT cells displayed persistent Bmal1-dLuc rhythms ( τ\u200a=\u200a24 . 44±1 . 55 hr , n\u200a=\u200a17 culture dishes from 2 independent cell lines ) ., Importantly , Rorasg/sg fibroblasts also displayed rhythmic Bmal1-dLuc oscillations ( τ\u200a=\u200a24 . 34±0 . 95 hr , n\u200a=\u200a30 from 3 lines ) , comparable to WT cells ( Figure 1C ) ., Not surprisingly , these cells also exhibited Per2-dLuc rhythms similar to those of WT cells ( Figure 1C ) ., Our results demonstrate that the ROR activators contribute to Bmal1 rhythm amplitude , but are clearly not required for Bmal1 rhythmicity and core clock function in fibroblasts ., Next , we examined the consequence of disrupting the negative arm of the ROR/REV/Bmal1 loop ., Bmal1 expression is significantly higher in the liver 5 and fibroblasts of Rev-erbα−/− mice than in WT ( data not shown ) ., Given the abnormal Bmal1 expression in the liver and fibroblasts , we expected that deletion of Rev-erbα would dramatically compromise the Bmal1 rhythm , as previously suggested from mRNA analysis 5 ., Surprisingly , however , Rev-erbα−/− fibroblasts displayed rhythmic Bmal1-dLuc expression ( Figure 2A ) ., The period lengths for Rev-erbα−/− fibroblasts harboring Per2-dLuc reporter were determined to be 26 . 59±0 . 29 hr ( n\u200a=\u200a10 ) for cell line-1 and 24 . 25±0 . 72 hr ( n\u200a=\u200a10 ) for cell line-2 , and the corresponding WT fibroblasts exhibited a periodicity of 24 . 88±1 . 48 hr ( n\u200a=\u200a7 ) ., Thus , as expected , real-time longitudinal bioluminescence recording reveals the dynamics of gene expression , while mRNA profiling lacks temporal resolution and is generally more subject to noise ., Given the apparent redundant contribution from Rev-erbβ , Bmal1 rhythms in the liver and lung of Rev-erbα−/− mice are also likely to be rhythmic , similar to that observed in fibroblasts ., We assessed any redundant contribution from Rev-erbβ using small hairpin RNAs ( shRNA ) ., We designed and tested nine shRNA constructs against different regions of the Rev-erbβ gene , and three of them ( shRNA-β1 , β2 and β3 ) were found to be functional in efficiently knocking down Rev-erbβ expression ( Figure 2C ) ., We introduced Rev-erbβ-shRNA constructs into WT fibroblasts harboring Bmal1-dLuc reporter ., Knockdown of Rev-erbβ resulted in higher Bmal1 mRNA expression , with shRNA-β1 being the most potent ( Figure 2C ) ; these cells displayed rhythmic Bmal1-dLuc expression ( Figure 2A ) , similar to effects of Rev-erbα-knockout ., Thus , Rev-erbα and Rev-erbβ are functionally redundant and disruption of either one alone is not sufficient to disrupt Bmal1 rhythms ., To disrupt the function of REV-ERBα and β simultaneously , Rev-erbβ-shRNA constructs were stably introduced into Rev-erbα−/− fibroblasts harboring the Bmal1-dLuc reporter to obtain Rev-erbα−/−:Rev-erbβ-shRNA:Bmal1-dLuc cell lines ., In striking contrast to rhythmic Bmal1-dLuc expression in Rev-erbα-knockout or Rev-erbβ-knockdown fibroblasts , cells deficient in both Rev-erbα and β function displayed significantly higher levels but largely arrhythmic Bmal1-dLuc expression ( Figure 2B ) ., For cell line-2 , 15/18 dishes of Rev-erbα−/− cells expressing control shRNA displayed rhythmic Bmal1-dLuc expression ( FFT spectral amplitude\u200a=\u200a0 . 80±0 . 08 , n\u200a=\u200a15 ) , but only 6/19 of Rev-erbα−/−:Rev-erbβ-shRNA-β1 showed any rhythms , and those that were rhythmic showed significantly lower spectral amplitude ( FFT spectral amplitude\u200a=\u200a0 . 50±0 . 08 , n\u200a=\u200a6 ) ., The weak rhythms may likely result from residual levels of REV-ERBβ expression in these knockdown cells ., Similar results were observed in cell line-1 ( data not shown ) ., These results demonstrate that the REV-ERBα and β are required for rhythmic Bmal1 expression in fibroblasts ., The finding that cells lacking ROR function retain Bmal1-dLuc rhythms whereas those deficient in REV-ERB function are arrhythmic , suggests that the REV-ERB repressors play more prominent roles than the ROR activators in the ROR/REV/Bmal1 loop ., Given that the Bmal1-dLuc reporter is rhythmic in Rev-erbα−/− fibroblasts , it is not surprising to observe that the Per2-dLuc reporter was also rhythmic ( Figure 2D ) ., However , it was not known whether disrupting both Rev-erbα and β would affect the core feedback loop function ., We thus introduced Rev-erbβ-shRNA constructs into Rev-erbα−/−:Per2-dLuc fibroblasts and demonstrated that Rev-erbα−/−:Rev-erbβ-shRNA cells also displayed rhythmic patterns of Per2-dLuc expression ( τ\u200a=\u200a25 . 99±0 . 40 hr , n\u200a=\u200a7 for cell line-1; τ\u200a=\u200a25 . 12±0 . 60 hr , n\u200a=\u200a22 for cell line-2 ) , similar to cells expressing control shRNA ( τ\u200a=\u200a26 . 48±0 . 27 hr , n\u200a=\u200a7 for cell line-1; τ\u200a=\u200a25 . 31±0 . 52 hr , n\u200a=\u200a23 for cell line-2 ) ( Figure 2D ) ., We also examined effects of Rev-erbβ-knockdown on the expression of other clock genes ., In shRNA control cells , peaks of Rev-erbβ and Per2 mRNAs ( CT40–48 ) were almost anti-phasic to Bmal1 ( CT32–36 ) ., Bmal1 mRNA was effectively de-repressed , especially at CT46–52 when Bmal1 was at its nadir in control cells ( Figure 2C ) ., Consistent with rhythmic Per2-dLuc bioluminescence expression , the Per2 mRNA expression pattern was essentially the same in Rev-erbα−/− cells expressing control shRNA and in those expressing shRNA against Rev-erbβ ., Given that Cry1 is under combinatorial regulation by both BMAL1/CLOCK and REV-ERBs 5 , 30 , 31 , we expected that disruption of REV-ERB function would alter the Cry1 expression pattern ., Indeed , compared to WT cells , Cry1 mRNA levels were higher in Rev-erbα−/− fibroblasts ( data not shown ) , and even higher in Rev-erbα−/−:Rev-erbβ-shRNA fibroblasts ( Figure 2C ) , all consistent with REV-ERB proteins being repressors ., Although interference with the REV-ERBs clearly disrupted the Bmal1 rhythm , it did not seem to substantially alter the rhythm of Cry1 mRNA ., Cry1 mRNA remained to be rhythmic , reaching its nadir at CT36–40 and peaking at CT46–50 , illustrating the resilience of the intracellular clock mechanism ., It is possible that , even though the Bmal1 rhythm is abolished , the residual level of REV-ERBβ in the cells was sufficient for combinatorial regulation of Cry1 ., It is also possible that other unknown mechanisms contribute to Cry1 regulation ., This ambiguity can be resolved in future studies by examining cells completely deficient in both Rev-erbα and β function ., Nevertheless , our results suggest that REV-ERBα and β are required for rhythmic expression of Bmal1 , but REV-ERB function and the Bmal1 rhythm are not required for normal oscillations of Per and Cry ., To further test the role of RORE-mediated Bmal1 regulation , we eliminated all influences of the RORs and REV-ERBs on Bmal1 expression in cell-based genetic complementation experiments ., Fibroblasts derived from Bmal1−/−:mPer2Luc mice displayed arrhythmic patterns of bioluminescence expression , demonstrating that Bmal1 is an essential clock component for cellular rhythmicity in fibroblasts ( Figure 3A ) ., We asked whether constitutively expressed BMAL1 in Bmal1−/− fibroblasts could restore circadian rhythmicity ., This approach precludes residual REV-ERBβ function from shRNA knockdown and circumvents any off-target effects ., To manipulate Bmal1 expression , we used three promoters: Bmal1 ( WT ) contains a 526-bp DNA fragment from the Bmal1 promoter encompassing ROREs , Bmal1 ( Mut ) is identical to Bmal1 ( WT ) except that the RORE sites are mutated to prevent ROR/REV-ERB from binding , and UbC is a commonly used constitutive promoter from the UbC gene ., We showed that WT fibroblasts transduced with a lentiviral Bmal1 ( WT ) -dLuc reporter displayed rhythmic bioluminescence expression , but Bmal1 ( Mut ) or UbC promoters did not confer rhythmicity in these cells ( Figure 3B ) ., We next determined the ability of the promoters to regulate the expression of Bmal1 ., In lieu of Western blot analysis of BMAL1 , we monitored the bioluminescence expression of BMAL1::LUC fusion protein ., We demonstrated that BMAL1::LUC cycled only when it is driven by Bmal1 ( WT ) , and that UbC and Bmal1 ( Mut ) promoters did not confer rhythmic fusion protein expression ( Figure 3C ) ., Thus , BMAL1 protein itself does not cycle in the absence of a RORE-containing circadian promoter ., To carry out genetic complementation , we generated a lentiviral expression vector Bmal1 ( WT ) -Bmal1·Flag , in which Bmal1 cDNA is under the control of WT Bmal1 promoter ., When this construct was introduced into Bmal1−/−:mPer2luc fibroblasts , circadian rhythmicity was restored ( τ\u200a=\u200a22 . 02±0 . 68 hr , n\u200a=\u200a25 cultured dishes ) ( Figure 3D ) , but not in cells expressing a Bmal1 ( WT ) -GFP control construct ( data not shown ) ., Importantly , non-cyclically expressed BMAL1 under the control of either UbC or Bmal1 ( Mut ) also effectively restored circadian mPer2Luc rhythmicity in Bmal1−/− fibroblasts ( τ\u200a=\u200a22 . 08±0 . 46 hr , n\u200a=\u200a20 for UbC-Bmal1; τ\u200a=\u200a22 . 61±0 . 60 hr , n\u200a=\u200a27 for Bmal1 ( Mut ) -Bmal1 ) ( Figure 3D ) ., Taken together , these results demonstrate that rhythmic expression of BMAL1 protein is not essential for the basic functioning of the intracellular clock ., These results provide the cellular basis for the finding that constitutive Bmal1 expression was able to rescue circadian behavioral rhythms in Bmal1−/− mice 24 ., The Rorc gene has at least two E-boxes within the promoter region , and its circadian expression pattern is similar to Cry1 in the liver ., In vitro studies suggest that Rorc transcription is regulated by BMAL1/CLOCK 31 ., To verify the in vitro results , we demonstrated that , similar to the expression patterns of other BMAL1/CLOCK-regulated clock components , the Rorc mRNA rhythm was abolished in the Bmal1−/− mouse liver , confirming that Rorc is regulated by the core loop ( Figure 4A ) ., Interestingly , however , we observed that mRNA levels of Rorc as well as Cry1 are clearly elevated rather than reduced in Bmal1−/− liver ., This was surprising at first given that BMAL1 is a known activator of Cry1 and Rorc expression ., However , it should not be so surprising given the complexity of transcriptional circuitry of the clock ., Similarly , higher Cry1 mRNA levels were also reported previously in Bmal1−/− , Clockm/m and Clock−/− mice 32 , 33 ., A recent in silico study showed that Cry1 and Rorc genes contain two types of circadian regulatory elements , the E-box and the RORE 31 ., In vitro and in vivo evidence also supports the presence of RORE sites within the Cry1 gene 5 , 30 ., In the absence of E-box regulation , factors acting through the RORE , namely the RORs and REV-ERBs , are likely to govern Cry1 and Rorc transcription ., In line with this notion , Clock mRNA is also higher in Bmal1−/− liver ( Figure 4A ) , and Bmal1 mRNA is higher in Clock−/− mouse liver 33 ., A recent study proposed dual activator and repressor functions of BMAL1/CLOCK , in which its repressor function explains the elevated Cry1 expression in the absence of Bmal1 32 ., However , that study did not take into consideration Cry1 gene regulation through the ROREs ., In both WT and Bmal1−/− mouse liver , there exists a strong inverse correlation between Rev-erbα and Cry1/Rorc mRNA levels: when Rev-erbα is high , Cry1/Rorc is low , and vice versa ( Figure 4A ) ., Similar expression patterns were also observed in fibroblasts ( Figures 1B and 5B ) and in Rev-erbα−/− mice 5 , and suggested from in silico and in intro studies 30 , 31 ., Thus , the elevated Rorc and Cry1 expression in the absence of Bmal1 may be regulated primarily by the REV-ERBs rather than the repressor function of BMAL1 ., We therefore sought to experimentally demonstrate this notion ., We hypothesized that over-expression of Rev-erbα in Bmal1−/− cells would bring down the expression levels of Cry1 and Rorc ., Because Cry1 and Rorc genes are regulated similarly but Rorc is not expressed in fibroblasts , we focused our analysis on the Cry1 gene in this cell type ., To test this idea , we introduced Rev-erbα into Bmal1−/− cells by lentivirus-mediated delivery and obtained a Bmal1−/−:Rev-erbα-OX fibroblast cell line ., Indeed , over-expressed REV-ERBα in Bmal1−/− cells efficiently repressed Cry1 mRNA to levels similar to those in WT cells ( Figure 4C ) ., Taken together , we provide direct in vivo genetic and molecular evidence to support the notion that Cry1 and Rorc are regulated not only by BMAL1/CLOCK but also directly by the REV-ERBs ( Figure 5A ) , which is the underlying molecular mechanism for elevated Cry1 expression in Bmal1−/− cells ., Interestingly , the mRNA levels of other clock genes in the liver of Bmal1−/− mice are also very different ( Figure 4A ) : Dbp and Rev-erbα expression is dramatically reduced , and Per1 and Per2 are expressed at constant intermediate levels , consistent with sustained mPer2Luc expression in Bmal1−/− cells ( Figure 3A ) , whereas Rorc , Cry1 , Clock and E4bp4 are clearly de-repressed ., Based on mRNA expression patterns in both WT and Bmal1−/− cells ( Figure 4A ) , we suggest the following transcriptional regulatory scheme for clock gene expression ( Figure 5A ) : Dbp and Rev-erbα are activated primarily by BMAL1/CLOCK via the E-boxes , and that this E-box-mediated circadian regulation is essentially eliminated in the absence of BMAL1 ( and thus PER/CRY-mediated repression via the E-box is no longer relevant ) ., Per1 and Per2 are activated by BMAL1/CLOCK and other non-circadian mechanisms , accounting for the intermediate mRNA levels of Per1 and Per2 in Bmal1−/− mice ., Rorc and Cry1 are regulated not only by BMAL1/CLOCK but also by RORs/REV-ERBs via the RORE ., Bmal1 , Clock and E4bp4 are regulated by RORs/REV-ERBs via the RORE ., The different regulatory mechanisms offer mechanistic explanations for distinct phases of clock gene expression rhythms observed in vivo ( Figure 5A ) ., Dbp is controlled by BMAL1/CLOCK via the E-box , while E4bp4 is primarily regulated via RORE , explaining why the E4bp4 rhythm is in phase with Bmal1 and Clock , but is antiphasic to Dbp ., Rev-erbα and Rorc are both activated by BMAL1/CLOCK , but Rorc is also repressed by REV-ERBs , explaining how Rorc mRNA accumulation is phase-delayed compared to that of Rev-erbα ., Additional regulation of Per1 and Per2 by non-circadian factors ( and possibly also by E4bp4 ) may cause a phase-delay compared to Dbp and Rev-erbα ., In summary , our data provides novel mechanistic insights into how the genes in the clock circuitry are regulated in vivo 31 ., The RORs appear to regulate the amplitude of target gene expression , while the REV-ERBs regulate the rhythmic expression of Bmal1 and also participate in combinatorial regulation of Cry1 ., As these regulatory mechanisms are not required for basic clock function , we suggest that the ROR/REV/Bmal1 loop and its constituents provide additional opportunities to control time-specific expression of output genes in local clock physiology , especially in peripheral tissues ., In this context , the differential tissue expression patterns of the RORs also provide additional opportunities for tissue-specific local circadian biology ( Figure 5B ) ., The combinatorial regulatory mechanism provides a novel strategy for identifying and validating target genes of the RORs and REV-ERBs , as well as differentiating RORE-containing genes from those containing both RORE and E-boxes ( Figure 4A ) ., Here we examined several of the genes that exhibit phases similar to Bmal1 or Cry1 in the liver and contain potential RORE sequences 25 , 26 ., For example , mRNAs of heat-shock protein 60 ( Hsp60 ) , arginine vasopressin receptor 1A ( Avp-V1a ) and Apoc3 were reduced in the liver of Rorc−/− mice , especially at peak time ( CT40–48 ) , reflecting reduction of RORE-mediated activation , but their mRNA levels were up-regulated in Bmal1−/− mice at CT28–36 , corresponding to the trough time in WT , reflecting loss of E-box-mediated REV-ERB expression with subsequent relief of RORE-mediated repression ., Tubulin beta 5 ( Tubb5 ) and peptidyl-prolyl cis-trans isomerase FK506 binding protein 4 ( Fkbp4 ) also exhibited significantly higher mRNA levels at CT28–36 in Bmal1−/− mice , but their expression levels were not affected in Rorc−/− mice ., Thus , cyclic RORE-mediated activation and/or repression may modulate expression patterns of specific target genes involved in important biological processes in a tissue-specific manner ., Previous studies using mice deficient in Rora , Rorb or Rev-erbα function strongly suggested functional redundancy among the ROR and REV-ERB family members 5 , 6 , 10 , 11 ., Mutation of Rora was shown to reduce Bmal1 mRNA amplitude both in the SCN 6 and in fibroblasts 11 , and Rev-erbα deletion resulted in much higher levels of Bmal1 transcription 5 , but Bmal1 rhythms were still retained despite either deficiency ., While null mutations in core clock genes typically lead to severe impairment of clock function ( see below ) , deficiencies in clock components within the ROR/REV/Bmal1 loop only produce modest clock phenotypes 5–7 , 10 , 11 ., The ROR/REV/Bmal1 loop is thus thought to provide a “stabilizing” function ., However , mice deficient in core clock components ( e . g . , Per1−/− , Per2−/− or Clockm/m mice ) also similarly show less precise or less persistent circadian rhythms 19–22 ., In this study , we investigated the redundancy of functions among the ROR and REV-ERB family members and clarified their roles in regulating Bmal1 expression ., To circumvent pleiotropic effects of gene deletion , we directly tested this “stabilization function” hypothesis in cell-autonomous clock models by perturbing the BMAL1 rhythm ., We demonstrated that cells with Rev-erbα-knockout or Rev-erbβ-knockdown still rhythmically express Bmal1 ., The Bmal1-dLuc rhythm could be abolished only when both Rev-erbα and β were disrupted ( Figure 2B ) ., Thus , REV-ERBα and REV-ERBβ are required for Bmal1 rhythmicity , and they are functionally redundant ., In contrast , the RORs are not required for Bmal1 rhythmicity ( Figure 1 ) ., Thus , the REV-ERBs play a more prominent role than the RORs in regulating the rhythmic expression of Bmal1 ., The current models for mouse and fly circadian clocks indicate that the process of evolution has produced a genetic circuitry substantially more complex than a simple transcriptional feedback scheme ., Presumably , robustness is a key feature of circadian control that is likely to be under selective pressure , as it would underlie the adaptive significance of a particular physiological rhythm ., Robustness is the ability of a system to maintain essential properties despite internal noise and external perturbations , a property which is prevalent in biological control circuits 34 ., From a circadian clock perspective , the key measures of robustness are precision ( period stability over time ) , persistence ( how long a given clock system sustains rhythm amplitude without a resetting signal ) , and accuracy ( period consistency of cells , tissues , or organisms ) ., It should be noted , however , that period variation and alteration may be an indicator of robustness , not necessarily instability ., Mechanisms contributing to the robustness of the clock system include additional interlocking loops , gene redundancy , maintenance of amplitude , and intercellular coupling ., In contrast to the proposed “stabilizing” role of the ROR/REV/Bmal1 loop , we found that Per2-dLuc expression is rhythmic even in cells deficient in both REV-ERBα and β function ( Figure 2D ) or expressing constitutive BMAL1 protein ( Figure 3D ) ., This provides unambiguous evidence from cell-autonomous preparations that Bmal1 mRNA and protein rhythms are not essential for the basic operation of the intracellular clock ., In accord with our findings , constitutively expressed Bmal1 in the SCN of Bmal1−/− mice was able to rescue circadian behavioral rhythmicity 24 ., Using real-time bioluminescence imaging to monitor Per2 gene expression in tissues and cells from mutant mice 23 , we recently found that both Per1 and Per2 are required for sustained cell-autonomous rhythms in individual cells ., Impo

Introduction, Results, Discussion, Materials and Methods

The mammalian circadian clockwork is composed of a core PER/CRY feedback loop and additional interlocking loops ., In particular , the ROR/REV/Bmal1 loop , consisting of ROR activators and REV-ERB repressors that regulate Bmal1 expression , is thought to “stabilize” core clock function ., However , due to functional redundancy and pleiotropic effects of gene deletions , the role of the ROR/REV/Bmal1 loop has not been accurately defined ., In this study , we examined cell-autonomous circadian oscillations using combined gene knockout and RNA interference and demonstrated that REV-ERBα and β are functionally redundant and are required for rhythmic Bmal1 expression ., In contrast , the RORs contribute to Bmal1 amplitude but are dispensable for Bmal1 rhythm ., We provide direct in vivo genetic evidence that the REV-ERBs also participate in combinatorial regulation of Cry1 and Rorc expression , leading to their phase-delay relative to Rev-erbα ., Thus , the REV-ERBs play a more prominent role than the RORs in the basic clock mechanism ., The cellular genetic approach permitted testing of the robustness of the intracellular core clock function ., We showed that cells deficient in both REV-ERBα and β function , or those expressing constitutive BMAL1 , were still able to generate and maintain normal Per2 rhythmicity ., Our findings thus underscore the resilience of the intracellular clock mechanism and provide important insights into the transcriptional topologies underlying the circadian clock ., Since REV-ERB function and Bmal1 mRNA/protein cycling are not necessary for basic clock function , we propose that the major role of the ROR/REV/Bmal1 loop and its constituents is to control rhythmic transcription of clock output genes .

Circadian clocks in plants , fungi , insects , and mammals all share a common transcriptional network architecture ., At the cellular level , the mammalian clockwork consists of a core Per/Cry negative feedback loop and additional interlocking loops ., We wished to address experimentally the contribution of the interlocking Bmal1 loop to clock function in mammals ., Because behavioral rhythms do not always reflect cell-autonomous phenotypes and are subject to pleiotropic effects , we employed cell-based genetic approaches and monitored rhythms longitudinally using bioluminescent reporters of clock gene expression ., We showed that REV-ERB repressors play a more prominent role than ROR activators in regulating the Bmal1 rhythm ., However , significant rhythmicity remains even with constitutive expression of Bmal1 , pointing to the resilience of the core loop to perturbations of the Bmal1 loop ., We conclude that while the interlocking loop contributes to fine-tuning of the core loop , its primary function is to provide discrete waveforms of clock gene expression for control of local physiology ., This study has important general implications not only for circadian biology across species , but also for the emerging field of systems biology that seeks to understand complex interactions in genetic networks .

neuroscience/behavioral neuroscience, genetics and genomics/gene function, cell biology/gene expression

journal.pbio.2002860

Genome-wide identification of bacterial plant colonization genes

Plant health is intimately influenced by a diverse community of microorganisms inhabiting the root surface ( rhizoplane ) and endophytic compartment 1 ., This root microbiome is recruited from surrounding soil communities 2–4 and is thought to be modulated by host plant immune function , root exudate-mediated signaling and metabolic compatibility , as well as intermicrobial interactions within the rhizosphere 5–7 ., These interactions , especially during the initial colonization period , are critical for establishment of a root-associated bacterial community that is distinct from that of the surrounding soil ., Extensive studies of plant pathogens have established the role of plant genetic factors , including immune phytohormone pathways , in controlling the ability of bacteria to colonize plants 5 , 8–10 ., Although there is increasing recognition that root microbiomes , in particular plant growth-promoting rhizobacteria ( PGPR ) , may be harnessed to improve plant fitness in agricultural applications , progress toward this goal requires a more thorough understanding of the bacterial genetic factors contributing to root colonization and fitness in the root microbiome 11 ., Root-associated bacterial communities have been defined for several plants , including A . thaliana , using culture-independent 16S rRNA sequencing strategies 3 , 4 ., Bacterial communities across diverse plant species show similar dominant representation of Proteobacteria , Actinobacteria , and Bacteroidetes phyla 1 ., The Pseudomonadaceae ( within the Proteobacteria phylum ) , in particular , comprise many genera capable of plant association , with the best studied examples ( e . g . , Pseudomonas fluorescens and P . syringae ) being commensals or pathogens , respectively 12 , 13 ., Many other isolates within the Pseudomonadaceae family are characterized as PGPR , which can enhance plant growth and viability through beneficial immune stimulation ( induced systemic resistance ISR ) 14 , by improvement of soil nutrient acquisition , or by directly triggering plant growth pathways through phytohormone production 1 , 15 ., Additionally , P . fluorescens spp ., have been shown to actively protect crops from a variety of fungal pathogens 16 ., P . simiae WCS417r was originally characterized as a biocontrol isolate on wheat 17 ., This strain was originally characterized as a member of the P . fluorescens group but was reclassified based on its genome sequence homology to the P . simiae-type strain 18 and is a well-studied example of a PGPR 18 ., WCS417r displays other PGPR activities , including ISR induction , siderophore production , lateral root growth stimulation , and activation of auxin signaling pathways 19 ., Importantly , WCS417r can colonize the roots of many plant species including Arabidopsis 20 ., These features make colonization of Arabidopsis roots by WCS417r an ideal system for identifying generalized bacterial colonization traits ., Conventional , nonsaturation screens of transposon mutagenesis libraries of P . fluorescens and P . putida strains led to the identification of genes required for root and rhizosphere colonization 12 , 21 ., To enable the generation of a comprehensive genome-wide map of root colonization genes , we used randomly barcoded transposon mutagenesis sequencing ( RB-TnSeq ) , a barcode-enabled extension of transposon mutagenesis coupled to high-throughput sequencing ( transposon mutagenesis sequencing TnSeq ) 22 that allows for the generation of reusable libraries of unique , mapped , and barcoded insertion mutant strains 23 , 24 ., We adopted RB-TnSeq to construct a genome-wide map of P . simiae WCS417r plant-association factors in an in vivo screen using Arabidopsis as the host plant ., This screen revealed mutations in 115 genes that have a negative impact on the ability of P . simiae WCS417r to colonize roots ., In addition to genes linked to well-known colonization traits such as motility and carbon metabolism , our mutant screen revealed additional , previously uncharacterized genes ., Our screen also identified 243 genes , the loss of function of which enhances colonization fitness ., Many of the genes identified in each class are clustered into predicted operons ., Integration of the genome-wide colonization data with RB-TnSeq phenotypes from more than 90 different in vitro growth conditions 23 highlighted motility , stress response , amino acid metabolism , as well as potentially unknown pathways as being functionally important for root/bacterial interactions ., To enable the generation of a genome-wide map of genes required for plant root colonization in P . simiae , we used RB-TnSeq with a mariner transposon to create a saturation mutagenesis library of P . simiae WCS417r 22–25 ., We selected WCS417r based on its plant growth promoting potential , its ease of transformation at high efficiency , and its tractability for lab manipulation ., By high-throughput sequence analysis of barcoded insertion mutants , we identified and mapped 110 , 142 unique transposon insertion sites , distributed throughout the genome 18 at an average of approximately 18 insertions per 1 , 000 bp ( S1 Fig , S1 Data ) ., Most insertions ( 59 . 5% ) mapped to a gene body , with 84% of genes harboring at least 1 insertion event ( median insertions per gene: 9; S1 Fig ) ., Of the remaining 827 genes with no insertion mutant detected , nearly half shared significant homology ( Materials and methods ) to genes known to be essential in other species ( 385; 55 . 6% of such genes in the WCS417r genome; S1 Fig ) , suggesting that insertions in these genes are lethal in P . simiae ., Furthermore , 146 of the untargeted genes contained fewer than 3 potential thymine-adenine dinucleotide mariner transposon insertion sites , representing 59% of such genes in the WCS417r genome ., Thus , our library includes null mutations in the vast majority of nonessential genes in the P . simiae WCS417r genome , supporting its utility for large-scale genetic screening for various phenotypes ., To determine which genes are necessary for root association , we designed a competitive colonization screen in which P . simiae mutant strains migrate from support medium through a porous nylon filter toward the root system of Arabidopsis seedlings , where they can attach and propagate ( Fig 1 and S2 Fig ) ., After a colonization period and removal of loosely adhering bacteria , root-associated bacteria were isolated as a combined rhizoplane and endophytic sample ., Controls required for data analysis included a “no root initial” ( NRI ) sample ( i . e . , an empty nylon filter incubated on a plate containing the mutant library , harvested the same day as the library was inoculated ) and a “no root final” ( NRF ) sample ( i . e . , a filter incubated on the plate in the absence of plants for a full week ) ., In total , we analyzed samples and controls harvested from approximately 15 , 000 seedlings ., We sequenced approximately 181 , 300 unique barcodes from each pooled root sample ( see Materials and methods ) , corresponding to approximately 240 colonization events per individual root ., Of all barcode sequence reads , we mapped 70% to known barcoded insertion sites ., We used sequenced barcode read counts to quantify the representation of mutants in each sample and compared barcode frequencies across samples 24 ( Fig 1; Materials and methods ) ., After normalization of total counts across samples , we determined 3 separate derived fitness scores for each gene ., Each of these scores measures a different potential effect influencing microbial growth in these experimental conditions: a “mesh fitness score” comparing the NRF and NRI samples and thus measuring changes in the ability to growth on the nylon mesh alone; a “root + mesh fitness score” comparing the “root” and the NRI samples , which measures the overall ability to grow on the root and the nylon mesh; and a “root fitness score , ” comparing the root and NRF samples directly , which represents the “root + mesh fitness score” corrected for the “mesh fitness score” to quantify the ability to grow on the root after correction for mesh-related effects ., We used this root fitness score ( Materials and methods ) to identify mutant strains corresponding to 358 genes as significantly depleted or enriched in the root-associated sample , which included 115 colonization-depleted genes ( that , when mutated , results in reduced colonization ability ) and 243 colonization-enriched genes ( that , when mutated , increased colonization ability , S3 Fig ) ., We used the colonization fitness scores of individual genes to create a genome-wide map of the root colonization trait ( Fig 2 ) ., Genes significantly contributing to colonization fitness were distributed throughout the P . simiae genome , with many clustering together ( Fig 2 , Table 1 ) ., Strikingly , 45 of the 115 genes mutated in colonization-depleted strains are clustered into 8 predicted operons , each containing at least 3 genes that decrease colonization fitness when mutated ( Table 1 ) ., Similarly , 62 of the 243 genes mutated in colonization-enriched strains were located within 14 predicted operons containing at least 3 genes that significantly increased colonization fitness when mutated ( Table 1 ) ., Thus , 22 predicted operons contained 3 or more genes with significant fitness scores corresponding to enhanced or reduced colonization ability ., In 21 of these operons , all genes with significant fitness scores contributed to colonization in a consistent direction within the operon , with 14 operons exhibiting >50% of the constituent genes as significant ., We examined predicted functions of the identified colonization genes and operons based on clusters of orthologous groups ( COG ) of proteins annotations 26 , 27 ., Among colonization-depleted genes , motility was the most common COG category ( P < 1 . 88 e-20; hypergeometric test ) , followed by cell wall/membrane/envelope biogenesis ( P < 2 e-3 ) and carbohydrate transport and metabolism ( P < 2 . 43 e-2; Fig 2 ) , consistent with the known roles of motility , lipopolysaccharide production , and sugar metabolism in root colonization and activity12 ., Among colonization-enriched genes , common COG categories included amino acid metabolism and transport ( P < 1 . 38 e-2 ) , cell wall/membrane biogenesis ( P < 5 . 99 e-3 ) , and transcription ( P < 4 . 18 e-3; Fig 2 ) ., Taken together , our genome-wide colonization screen allowed for the simultaneous functional assessment of nearly all genes within the WCS417r genome for their contribution to plant colonization , and we identified a substantial number of genes and operons likely to be important for this process ., To evaluate the robustness of our screen , we isolated individual insertion mutant strains from sequence-informed WCS417r library arrays ( Materials and methods ) ., We selected 22 insertion mutant strains to validate ( using a single insertion mutant strain per gene ) covering a diversity of potentially interesting putative functions , with some representing operons containing multiple genes with significant fitness scores ( Table 1 ) , and others representing individual genes with a broad range of negative or positive fitness score effects ( S1 Data ) ., The selected mutants included 9 predicted to have compromised colonization fitness and 13 predicted to have increased colonization ability ( S1 Data ) ., We designed a competitive colonization screen in which individual mutants compete against a luminescent , but otherwise wild-type ( WT ) P . simiae WCS417r strain , and direct luminescence quantification of roots can be used to measure competitive fitness ( Fig 3 , Materials and methods ) ., We observed that 7 out of 9 colonization-depleted insertion mutants were out-competed by the luciferase-producing ( Lux+ ) strain ( Fig 3 ) ., Similarly , all 13 colonization-enriched insertion mutants competed either as well or better than the Lux+ strain at the root tip ( Fig 3 ) ., Overall , the direction of fitness change as assessed in luminescence-based competition assays was consistent with the direction predicted by RB-TnSeq in 20 of 22 cases ( 91% , chi-squared P < 0 . 00012 ) ., Selected insertion mutants were further validated using an analogous LacZ blue/white screening approach , as well as through colonization assays with corresponding loss-of-function mutants generated by targeted mutagenesis ( Materials and methods; S1 Data; S6 Fig ) ., Although the magnitude of estimated fitness changes for individual genes varied across validation methods , the results were largely consistent with luciferase-based validation screens and confirmed in particular the impact of mutations in predicted colonization-depleted genes on colonization fitness ( Materials and methods , S6 Fig , S1 Data ) ., We also explored dynamic aspects of root colonization , as the overall fitness of root colonizers might change over the days it takes to establish colonies on the root ., To test whether fitness is static across a time course , we selected 4 predicted poor colonizers ( PS417_00160 , PS417_01955 , PS417_22145 , and PS417_22775 ) and 4 predicted enhanced colonizers ( PS417_08165 , PS417_21035 , PS417_03095 , and PS417_10720 ) to inoculate Arabidopsis seedlings in competition with the LuxABCDE expressing P . simiae strain as above but sampled at 1 , 3 , 5 , and 7 days after inoculation ., We measured the proportion of Luciferase-negative cells ( i . e . , mutant strain ) from each root sample ., Although all poor colonizers tended to grow more slowly once present on the roots , 1 ( PS417_22775 ) failed to colonize very early on ( S7 Fig ) ., Most predicted enhanced colonizers appeared to grow more quickly than their poor colonizing counterparts , especially on later days ( S7 Fig ) , although were still present in reduced numbers than expected , indicating a possible bias towards measuring luciferase-positive cells in this assay ., Together , these detailed validation efforts support that the RB-TnSeq method applied to plant-bacteria interactions robustly defines both pronounced as well as subtle colonization defects ., Many of the genes identified by our screen have no or at best vague annotations ., To explore the physiological functions of the identified colonization genes in more detail , we compared our data to RB-TnSeq results of the same insertion mutant library tested under 90 distinct in vitro conditions , including 48 conditions using a defined compound as a sole carbon source in otherwise minimal media , 11 conditions using a defined nitrogen source , 29 stress conditions , and 2 in vitro motility conditions ( inner and outer cuts of a soft agar motility assay ) 23 ., Although the complexity of individual phenotypes measured by these in vitro assays is considerably lower than that of root colonization processes , these assays are scalable and can thus be used to rapidly assess many metabolic or stress responsive functions ., Within the large dataset covering genome-wide fitness across 90 conditions , we specifically examined the in vitro phenotypes of mutations in all 115 colonization-depleted and 243 colonization-enriched genes ( Materials and methods , Fig 4 and S8–S10 Figs , S1 Data ) ., We developed a genome-wide map of microbial genes required for colonization of plant roots in a plant/microbial system ., Building on the successful application of RB-TnSeq for the large-scale assessment of in vitro phenotypes 23 , the present study demonstrates the utility of this experimental paradigm for studies of bacterial plant root colonization in vivo , thus applying it to a process that considerably exceeds in vitro assays in terms of complexity ., By using the colonization of Arabidopsis roots by the biocontrol bacterium , P . simiae WCS417r , as a model of colonization , we observed a substantial variety of genes conferring altered survivability to the bacteria when mutated , mirroring the complex nature of this interaction system ., One challenge of TnSeq assays in general , and TnSeq assays targeting colonization phenotypes in particular , is the reliance on a diverse population of insertion mutants in the colonized host after coincubation ., During Arabidopsis colonization by P . simiae , we found that only 100 to 1 , 000 independent colonization events occur per individual root , creating a potential bottleneck for downstream analysis ., We mitigated this effect by sampling large numbers of plants ( approximately 1 , 000 pooled seedlings per sample ) , resulting in the recovery of most constituent mutant strain barcodes in every pooled sample ., Additional confounding factors include environmental or community considerations , namely that survival of individual mutants on the plant support medium and filter prior to root colonization might be reduced , which need to be corrected with appropriate controls ( S3 Fig , Materials and methods ) ., Certain functional deficiencies , especially those associated with secreted or extracellular activity , of some colonization genes might be effectively rescued by a largely WT population for that function ., Furthermore , some mutants identified by RB-TnSeq showed quantitatively weaker or no significant phenotypes in validation screens , raising the possibility that their fitness is higher when they are rare members of a diverse mutant population , as opposed to validation experiments where these mutants represent 50% of the population ., Notwithstanding these limitations , the high-validation rate of colonization genes in secondary validation assays supports the robustness of our genome-wide map of root colonization ., Many genes with significant fitness scores clustered within operons , further reinforcing the validity of RB-TnSeq-derived results ., Indeed , colonization genes within the operons shown in Table 1 represent the majority of genes included within these operons ., Additionally , 98 colonization genes were not predicted to be part of an operon or were part of an operon of only 2 to 3 genes ., Some colonization genes occurred in operons in which only 1 or a small subset of genes showed significant fitness scores ., For these operons , it is possible that not every gene is required for the given function , or that individual enzymes are shared across alternative pathways ., These results , along with the observation that many of the genes identified from our screen are involved in processes well known to be vital to colonization of plants ( e . g . , motility , carbohydrate utilization ) are consistent with the notion that fitness scores from genome-wide colonization reflect valid , biologically relevant genes and pathways ., We also compared the list of genes significantly affecting colonization to known colonization genes based on a number of smaller-scale mutant screens in P . putida 21 and found that 20 out of 87 P . putida homologues with colonization data in our screen showed altered fitness ( S1 Data ) ., Although this limited overlap is expected due to the heterogeneous nature of the assays performed across multiple studies in P . putida , as well as known differences between organisms and hosts , they further strengthen the biological validity of the genome-wide colonization map generated in the present study ., We observed a surprisingly large number of genes with positive colonization fitness scores ( 243 positive versus 115 negative ) ., While most of these mutants showed quantitatively less pronounced phenotypes in luciferase-based screens than predicted by the initial RB-TnSeq scores , in almost all cases the direction of the effect was confirmed ( Fig 3 ) ., This observation , along with the propensity of colonization-enriched genes to colocate in operons , supports the conclusion that the predicted phenotypes for these genes are biologically relevant ., A large proportion of these genes encode proteins involved in amino acid transport and metabolism ( Figs 2 and 4 ) , suggesting that auxotrophy for certain amino acids confers a selective advantage for survival in the plant-associated environment rich with exuded amino acids and sugars ., Two of these genes ( PS417_01565 and PS417_21035 ) were assayed in our luciferase-based competitive colonization screen , and behaved as predicted by the RB-TnSeq data ., This poses an intriguing technological opportunity: engineering strains to be more dependent on their plant hosts may have the dual effect of improving colonization while simultaneously restricting the survivability of engineered strains outside the context of the root ., Lastly , we identified 44 genes that had vague or no annotation information ., A particularly noteworthy subset of these genes lie within operons with multiple mutants with colonization phenotypes , yet without clear functional annotation ., These may represent truly novel genes or pathways contributing to colonization of and survival on roots and the results from our in vivo and in vitro screens pave the way for their targeted functional and biochemical characterization ., In summary , the genome-wide map of plant colonization genes described in the present study highlights diverse metabolic and physiological functions that support or hinder plant-microbe association and points to novel functions mediating this process ., A . thaliana Col-0 seeds were surface-sterilized in 70% ethanol for 5 minutes , followed by 10% bleach plus 0 . 1% Triton-X100 for an additional 5 to 10 minutes ., Sterilized seeds were washed 5 times in sterile water , and stratified in the dark for 2 to 3 days at 4°C ., After stratification , 100 seeds were plated on a nylon mesh filter ( 100 micron pore size , cut to an area of approximately 8 cm2 B0043D1XRE Amazon . com Inc , Seattle , WA ) placed on top of plant growth media ( 0 . 5X Murashige and Skoog basal salts MSP01 , Caisson Laboratories , Smithfield , UT , 2 . 5 mM MES M3671 , Sigma-Aldrich , St . Louis , MO , 0 . 6% phytagel P8169 , Sigma-Aldrich , pH 5 . 7 ) in a 10 cm square petri dish ., Seedlings were grown upright in a Percival incubator ( CU-36L5 , Geneva Scientific , Williams Bay , WI ) for 7 days prior to treatment ., A cultured isolate of P . simiae ( WCS417r ) was obtained from Dr . Corné Pieterse ( Utrecht University ) ., The barcoded insertion library for this strain was generated by transposon mutagenesis with a barcoded mariner transposon library , followed by TnSeq mapping and barcode association , as previously described 23 , 24 ., Glycerol stocks of this library were used for subsequent experiments , stored in 1 mL aliquots containing approximately 4 x 10^8 cells/mL ., On average , this represents greater than 1 , 000-fold excess of each individual strain and should avoid any filtration or passage effects associated with recovery from glycerol stocks ., A LuxABCDE-expressing transformant ( WCS417r:Lux+ ) was generated by inserting an IPTG-inducible expression cassette using a mariner transposase system , such that Luciferase expression could be visualized following IPTG induction ., The insertion site of the LuxABCDE transgene was determined to be at approximately position 1628942 ., WCS417r , WCS417r:Lux+ cultures were grown in LB Lennox media at 28°C in a shaking incubator at 200 rpm ., The insertion mutant library and single insertion mutants were grown in LB Lennox supplemented with 100 μg/mL kanamycin at 28°C in a shaking incubator at 200 rpm ., LacZ-expressing WCS417r was created using a previously engineered miniCTX-lacZ vector driven by the Vibrio cholera lacZ promoter36 ., Briefly , lacZ was transferred to the neutral phage attachment site ( attB ) of WCS417r via biparental mating using Escherichia coli SM10 and selected on LB plates containing 75 μg/mL tetracycline ., Each colonization experiment was comprised of 5 replicates of each sample type ., Three colonization experiments were performed at the DOE Joint Genome Institute ( sets A , B , and C ) ., For a single colonization experiment , a glycerol stock containing the transposon library was inoculated in 50 mL fresh LB and grown for approximately 6 hours until the culture reached the midlog phase ( OD between 0 . 2 and 0 . 6 ) ., Cells were then harvested by centrifugation ( 3 , 000 g for 3 minutes ) and washed 3 times by resuspending in 1 mL of 0 . 5X MS media and pelleting the cells ., After washing , the cells were resuspended in 1 mL , 0 . 5X MS , and the OD of the resuspension was calculated by spectrophotometer ( using a 1:10 dilution ) ., Cells were then normalized to OD 0 . 5 , and 50 μl ( corresponding to approximately 1 . 0 x 10^7 cells ) was spread onto 0 . 5X MS phytagel ( 0 . 6% ) plates using sterile glass beads ., Seven-day-old Arabidopsis seedlings grown on a sterile nylon mesh filter ( 110 μm pore size ) laid on top of 0 . 5X MS phytagel ( 0 . 6% ) plates were transferred by lifting and replacing the filter onto the inoculated plates ., Five aliquots of the OD 0 . 5 culture were saved as an input ( IPT ) culture for each experimental replicate ( set ) ., Ten plates were inoculated with bacteria and exposed to a nylon mesh filter without seedlings ., Five such filters ( NRI ) were allowed to contact the bacteria plate for 1 hour before being used to inoculate 50 mL LB + Kanamycin ( 100 μg/L ) overnight ., The remaining 5 filters ( NRF ) along with the plate/filters containing Col-0 seedlings were incubated vertically in a Percival growth chamber for 7 days under short-day ( 8 hour light/16 hour dark ) conditions ( 22°C ) ., Following cocultivation , the NRF filters were used to inoculate a 50 mL LB Lennox + Kanamycin ( 100 μg/L ) culture , and grown overnight ., Seedlings on plates containing bacteria were then cut just below the root/shoot junction , and the isolated roots were placed into 10 mL , 0 . 5X MS liquid ., Ten plates of roots were pooled into a single sample ., The pooled roots were vortexed for 15 seconds to wash loosely adhered cells from the surface of the roots , and the buffer removed ., The washing procedure was repeated 5 more times ( total 6 washes ) ., The washed roots were then cut into thirds , placed into 2 mL eppendorf tubes with 2 metal beads and 200 μl , 0 . 5X MS liquid ., The roots were ground in a TissueLyser bead mill for 2 cycles of 5-minute grinds at 30 Hz ( Qiagen , Hilden , Germany ) , inverting the tubes between cycles ., Ground roots ( rhizoplane + endophytic compartment; RPL ) were used to inoculate 50 mL LB Lennox + Kanamycin ( 100 μg/L ) cultures overnight ., Two mL samples from all overnight cultures ( IPT , NRI , NRF , RPL ) were harvested after 12 to 16 hours of growth , pelleted , and stored at −80°C prior to DNA extraction ., DNA from frozen pellets was isolated using the DNeasy Blood and Tissue kit ( Qiagen , Hilden , Germany ) according to the manufacturer’s instructions ., DNA was quantified with a Qubit fluorometer ( Thermo Scientific , Raleigh , NC ) according to the manufacturer’s instructions and normalized to 10 ng/μl ., Samples with ( DNA ) less than 10 ng/μl were not diluted ., Twenty microliters of the normalized ( or undiluted , in the case of low concentration samples ) DNA was used as template in a PCR using primers flanking the transposon barcode region , each containing an Illumina adapter and multiplexing index sequence ( BarSeq ) 23 , 24 ., PCR was performed using Q5 DNA polymerase with Q5 GC enhancer ( New England Biolabs , Ipswich , MA ) for 25 cycles of 30 seconds at 98°C , 30 seconds at 55°C , and 30 seconds at 72°C , followed by a final extension at 72°C for 5 minutes ., Following PCR , 10 μl of each reaction was pooled into 3 sets of 25 ( sets A , B , and C; see S1 Data ) amplicon libraries , corresponding to each experimental set ( see previous section ) ., Three pooled libraries ( representing sets A , B , and C ) were then purified using the DNA Clean & Concentrate Kit ( Zymo , Irvine , CA ) according to the manufacturer’s instructions ., Each set was sequenced on its own lane on an Illumina HiSeq 2500 machine using the 1T paired-end , 2 x 101 cycle protocol , producing an average of 3 to 8 million reads per sample ., We used barcode sequencing to quantify the representation of mutants in each sample and compared barcode frequencies across samples 24 ., Raw sequence reads were initially processed by looking for the 6 nt adapter sequences on either side of a 20 nt random barcode ., Reads with exactly 20 nt barcode sequences , no mismatches between mate-pair barcodes , and high-quality scores ( Q > 30 ) from each of the 60 libraries ( 15 IPT , 15 NRI , 15 NRF , and 15 RPL ) were then saved into filtered fastq files and used as input into the BarSeqR pipeline 24 ., For this analysis , the NRI samples were set as “Time0” controls , with all samples normalized to these samples ., The NRI samples were used as the normalization controls to factor out any amplification effects caused by overnight culturing ., To assess the saturation of our sampling method , we quantified the number of barcodes ( and genes mutated ) in all 60 samples ., On average , for each sample , we recovered >80% of the insertion mutants that we had mapping information for ., We also quantified the number of barcodes and genes with mutations when samples were considered together ( combining the unique barcode and gene count represented by any of the samples within a given sample type , e . g . , NRF or RPL ) ., With replication , our recovery rates approach saturation for each of the 4 sample types ( S11 Fig ) ., The BarSeqR scripts report per-gene fitness scores ( normalized log-ratio ) and t-like test statistics indicating the relative effect size and significance between each sample and the average of the Time0 controls , respectively , for each of the 60 samples , including the Time0 controls ( NRI samples ) 24 ., After normalization of total counts across samples , we determined 3 separate derived fitness scores for each gene ., Each of these scores measures a different potential effect influencing microbial growth in these experimental conditions: A mesh fitness score comparing the NRF and NRI samples and thus measuring changes in the ability to growth on the nylon mesh alone; a root + mesh fitness score comparing the RPL and the NRI samples , which measures the overall ability to grow on the root and the nylon mesh; and a root fitness score , comparing the RPL and NRF samples directly , which represents the root + mesh fitness score corrected for the mesh fitness score to quantify the ability to grow on the root after correction for mesh-related effects ., To classify genes based on their mesh phenotype , we compared fitness scores from the various sample types and computed 3 derived fitness scores , looking for significant differences based on an empirical P value corresponding to an FDR of 0 . 05 ( Student t test ) and an effect size ( absolute difference between the means ) of > 0 . 5: a root + mesh fitness score ( comparing RPL to NRI; P < 0 . 014 ) , a mesh fitness score ( comparing NRF to NRI; P < 0 . 01 ) and a root fitness score ( comparing RPL to NRF; P < 0 . 013 ) ., Considering that weak colonization fitness scores may not be biologically meaningful , we chose a threshold effect size cutoff of 0 . 5 , which eliminated nearly half of the genes that were significant based on P value alone ., We binned these genes into 2 main groups ( S3 Fig ) ., Genes in group 1 ( gray , S3 Fig; 149 genes ) had significant root + mesh fitness scores , but the quantitative magnitude of this effect was largely explained by changes to the ability to survive on the nylon mesh alone ( mesh fitness score ) ., Consequently , genes in group 1 were considered low-confidence candidate genes for root colonization , despite their significant root fitness scores ., Genes in group 2 ( blue and cyan , S3 Fig; 358 genes ) exhibited significant root fitness scores of at least moderate effect sizes ( root fitness score absolute value >0 . 5 ) ., A subset of these ( group 2b cyan , S3 Fig; 75 genes ) additionally did not exhibit significant root + mesh fitness scores” ., Mutations in these genes likely confer altered fitness on mesh , but this phenotype was at least partially compensated for b

Introduction, Results, Discussion, Materials and methods

Diverse soil-resident bacteria can contribute to plant growth and health , but the molecular mechanisms enabling them to effectively colonize their plant hosts remain poorly understood ., We used randomly barcoded transposon mutagenesis sequencing ( RB-TnSeq ) in Pseudomonas simiae , a model root-colonizing bacterium , to establish a genome-wide map of bacterial genes required for colonization of the Arabidopsis thaliana root system ., We identified 115 genes ( 2% of all P . simiae genes ) with functions that are required for maximal competitive colonization of the root system ., Among the genes we identified were some with obvious colonization-related roles in motility and carbon metabolism , as well as 44 other genes that had no or vague functional predictions ., Independent validation assays of individual genes confirmed colonization functions for 20 of 22 ( 91% ) cases tested ., To further characterize genes identified by our screen , we compared the functional contributions of P . simiae genes to growth in 90 distinct in vitro conditions by RB-TnSeq , highlighting specific metabolic functions associated with root colonization genes ., Our analysis of bacterial genes by sequence-driven saturation mutagenesis revealed a genome-wide map of the genetic determinants of plant root colonization and offers a starting point for targeted improvement of the colonization capabilities of plant-beneficial microbes .

Plants fix carbon to create an abundance of sugars and amino acids , thus providing an enticing environment for microorganisms that reside in soil ., Once these microorganisms have colonized the root environment , they can dramatically influence plant growth and development ., We set out to identify a comprehensive set of microbial genes that control or influence root colonization , using a genome-wide transposon mutagenesis approach ( randomly barcoded transposon sequencing RB-TnSeq ) ., By using this method , we identified several hundred genes that , when mutated , affect the ability of the bacterium P . simiae to competitively colonize the root system of the model plant A . thaliana ., These included many genes purported to be involved in carbohydrate metabolism , cell wall biosynthesis , and motility , underscoring the notion that sugar metabolism , defense , and motility are all key features of a root-colonizing microbe ., We also identified several amino acid transport and metabolism genes with mutations that confer a fitness advantage in root colonization ., Lastly , we identified several genes with no known function that significantly alter root colonization ability when mutated ., These findings suggest novel engineering strategies to improve biological product development , and will facilitate the mechanistic exploration of the root colonization process .

methods and resources, transposon mutagenesis, brassica, operons, mutation, model organisms, materials science, experimental organism systems, dna, molecular biology techniques, mutagenesis and gene deletion techniques, seedlings, plants, macromolecules, materials by structure, arabidopsis thaliana, research and analysis methods, polymers, polymer chemistry, mutant strains, gene mapping, chemistry, nylons, molecular biology, biochemistry, eukaryota, plant and algal models, nucleic acids, genetic screens, gene identification and analysis, genetics, biology and life sciences, physical sciences, organisms

journal.pntd.0007349

Modelling the impact of a Schistosoma mansoni vaccine and mass drug administration to achieve morbidity control and transmission elimination

Schistosomiasis inflicts significant levels of human morbidity and mortality in regions of the world with endemic infection ., It is estimated that nearly 258 million people are infected worldwide with up to 700 million at risk of being infected , leading to an estimated 280000 deaths annually 1–3 ., Schistosomiasis is an intestinal or urogenital disease caused predominantly by infection with Schistosoma mansoni , S . japonicum or S . haematobium , and is one of the diseases included within the World Health Organization ( WHO ) 2020 goals for neglected tropical diseases ( NTD ) control ., Individuals become infected when cercariae ( larval forms of the parasitic worm ) , released by an intermediate host ( various freshwater snail species ) , penetrate the skin during contact with contaminated water 4 ., Control programmes are at present based on mass drug administration ( MDA ) using the drug praziquantel , and behaviour modification directed at reducing water contact and improvements in sanitation ., MDA has to be repeatedly used , since clearing infection does not result in acquired immunity and treated individuals can be re-infected ., Age-related water contact behaviour results in most infection residing in school-aged children ( SAC; 5–14 years of age ) , since age intensity of infection profiles are convex in shape ., Treatment is therefore specifically focused on this age group ., At present , pre-school aged children ( pre-SAC ) are not eligible for treatment with praziquantel 5 due to the absence of clinical data on the drug effects and safety in the very young ., In the coming years a new formulation of praziquantel may be approved for very young children 6 ., In areas of high transmission , WHO guidelines also recommend treatment of adults at risk 1 , 7 ., By 2020 , WHO aims to increase coverage in areas of endemic infection such that 75% of SAC at risk will be regularly treated 2 , but progress to date in reaching this target has been poor in many regions ., Currently WHO recommends using prevalence of infection in SAC to determine how often to treat in a given endemic area 1 ., The recommended treatment strategy for schistosome infection is dependent upon whether the community has a low ( < 10% ) , moderate ( 10–50% ) or high ( ≥ 50% ) prevalence at baseline before the implementation of MDA ., The strategy for low-risk communities is to treat all SAC twice during their primary schooling age , generally once every three years , and supply praziquantel in local health centres to treat suspected cases ., For moderate-risk communities , the recommendation is to treat all SAC and at-risk adults once every two years ., For high-risk communities , the recommended approach is to treat all SAC and at-risk adults once a year ., At present in national NTD control programmes , schistosomiasis has one of the lowest levels of MDA coverage of all helminth diseases 8 , 9 ., Given that MDA needs to be administered to individuals frequently , and that it does not provide long-term protection against the infection in the absence of a strong acquired immunological response to infection , a vaccine is ideally needed for control in the longer term ., At present , there is no vaccine for use in humans that can protect against the schistosome infection ., However , recent experimental studies by Afzal Siddiqui and colleagues on a candidate vaccine against Schistosoma mansoni infection in a baboon animal model have produced some encouraging results ., In four independent , double-blinded studies , a Sm-p80-based vaccine exhibited potent prophylactic , anti-egg induced pathology and transmission-blocking efficacy against S . mansoni in the baboon ( Papio ursinus ) animal model 10 ., The vaccine reduced female worm establishment by 93 . 45% and significantly resolved the major clinical manifestations of hepatic/intestinal schistosomiasis by reducing the tissue-egg load by 91 . 35% ., A 40-fold decrease in faecal egg excretion by those few female parasites that established in the vaccinated animals , combined with a 79 . 21% reduction in hatching ability of eggs ( the release of viable miracidia ) , suggests the vaccine may have a high transmission blocking potential ., The study showed comprehensive evidence for the effectiveness of a Sm-p80-based vaccine for schistosomiasis and provided support for the need to move beyond animal models to human studies ., Based on the baboon experiments by Siddiqui and colleagues , and assuming efficacy would be similar in humans , published epidemiological analyses based on mathematical models have predicted that the Sm-p80-based vaccine could potentially block infection in areas of low and moderate transmission provided the duration of protection provided by the vaccine is 5 years or more 11 , 12 ., These models were simple in structure and built on a deterministic framework ., This study extends these analyses using an individual based stochastic model to look at the impact of a vaccine , with varying durations of protection , employed in different community-based vaccination programmes involving either vaccinating young children in a cohort-based approach or vaccinating the whole community across all age classes ) ., Analyses are also presented of the impact on transmission and the prevailing levels of infection using either vaccination alone , MDA alone ( the current most commonly used intervention to control morbidity ) and or using both in different combinations ., A description of the impact of MDA , alone on the prevalence and intensity of S . mansoni infection in various transmission settings , is covered in a series of recent publications , as is model structure , model assumptions and data sources for the key transmission and biological parameters 3 , 4 , 7 , 8 ., The focus in the present analyses is on the relative merits of vaccination versus MDA , alone or in combination , as a tool for the community control of the morbidity induced by S . mansoni and the likelihood of transmission elimination ., Past work on the impact of MDA on Schistosoma mansoni has employed a hybrid deterministic model ( with deterministic and stochastic components ) based on sets of partial-differential equations to describe changes in the mean worm burden M ( t , a ) , for host a over time t 13–15 ., Stylianou et al , developed an age independent deterministic model to explore the effect of community vaccination programmes 11 ., We extend this deterministic model and develop an individual-based stochastic model ( an earlier version is described in 4 ) , where an individual of age a can be in one of the two categories;, ( i ) unvaccinated group or, ( ii ) vaccinated group , denoted by Nu ( a , t ) and Nv ( a , t ) respectively ., We assume that the number of births is the same as the number of deaths ( constant size for the human host ) , hence the total population of age a , at time t is N ( a , t ) = Nu ( a , t ) +Nv ( a , t ) ., The unvaccinated and vaccinated host dynamics can be described by the following system of partial differential equations ( PDEs ) :, ∂Nu ( a , t ) ∂t+∂Nu ( a , t ) ∂a=−q ( a , t ) Nu ( a , t ) +ωNv ( a , t ) −μ ( a ) Nu ( a , t ), ( 1 ), ∂Nv ( a , t ) ∂t+∂Nv ( a , t ) ∂a=q ( a , t ) Nu ( a , t ) −ωNv ( a , t ) −μ ( a ) Nv ( a , t ), ( 2 ), Here q ( a , t ) is the fraction of the population of age a vaccinated at time t , ω=1durationofvaccineprotection is the vaccine decay rate and μ ( a ) is the host mortality rate ., The vaccine candidate is assumed to act on the following variables cf . Eqs ( 3 ) and ( 4 ) ;, ( i ) parasite establishment within the human host by reducing the rate of infection , β ,, ( ii ) parasite survival and growth within the human host , by reducing adult worm life expectancy , σ and, ( iii ) reducing the rate of egg production , λ , due to a reduced growth rate in humans ., We assume that the vaccine’s impact on worm death rate , eggs per gram ( EPG ) and age-specific contact rates are v1 , v2 and v3 respectively , where the values range from 0 to 1 . The total worm burden in the unvaccinated and vaccinated hosts are denoted by Mu and Mv and the changes in Mu and Mv , over time for host a are described by the following equations:, ∂Mu ( a , t ) ∂t+∂Mu ( a , t ) ∂a=Lβ ( a ) Nu ( a , t ) −q ( a , t ) Mu ( a , t ) +ωMv ( a , t ) − ( μ ( a ) +σ ) Mu ( a , t ), ( 3 ), ∂Mv ( a , t ) ∂t+∂Mv ( a , t ) ∂a=Lv3β ( a ) Nv ( a , t ) +q ( a , t ) Mu ( a , t ) −ωMv ( a , t ) − ( μ ( a ) +v1σ ) Mv ( a , t ), ( 4 ), Here L represents the concentration of the infectious material in the environment , namely , how each individual of age a , contributes to the pool of released eggs ., This is discussed in detail in 14 and 16 ., It is assumed that the rates of turn over for the miracidia , snail intermediate host and cercaria are much faster ( life expectancies days to weeks ) than the adult worm in the human host ( life expectancy 4–6 years ) , so the dynamics of these life cycle stages are collapsed into the equations for the adult worms in humans as detailed in Anderson & May 15 ., The total worm burden in the population is given by the sum of the total worm burden in the unvaccinated and vaccinated hosts ., If we denote the total worm burden in the population as the sum of the total worm burden in the unvaccinated and vaccinated hosts by M ( a , t ) ¯=Mu ( a , t ) +Mv ( a , t ) and add Eqs ( 3 ) and ( 4 ) together we obtain the following ,, ∂M ( a , t ) ¯∂t+∂M ( a , t ) ¯∂a=Lβ ( a ) Nu ( a , t ) +Lv3β ( a ) Nv ( a , t ) −σM ( a , t ) ¯−μ ( a ) M ( a , t ) ¯, ( 5 ), In Eq ( 5 ) we have assumed v1 = 1 . We can express M ( a , t ) ¯ in terms of the mean worm burden , M ( a , t ) , as M ( a , t ) ¯=N ( a , t ) M ( a , t ) ., Then we obtain;, ∂M ( t , a ) ∂t+∂M ( t , a ) ∂a=Lv3β ( a ) Nv ( a , t ) +Lβ ( a ) Nu ( a , t ) N ( a , t ) −σM ( a , t ), ( 6 ), The egg output ( from the vaccinated and unvaccinated populations ) is given by, E=ψL¯∫a=0∞{Nu ( a , t ) F ( Mu ( a , t ) Nu ( a , t ) ;λ ) +Nv ( a , t ) F ( Mv ( a , t ) Nv ( a , t ) ;v2λ ) }ρ ( a ) da, ( 7 ), given, dLdt=E−μ2L, ( 8 ), where the death rate is that of infected snails ., In the above equation ψ describes the flow of the infectious material into the reservoir while the function F ( M ( a , t ) ; λ ) generates the egg output as a function of mean worm burden and ρ ( a ) represents the age-specific relative contribution of infectious stages to the environmental reservoir ., In our simulations we assume the host contribution to the reservoir to be the same as the age-specific contact rates , β ( a ) ., This model has a full age structure for the human host where the outputs are grouped into three age categories , pre-SAC ( 0–4 years of age ) , SAC ( 5–14 years of age ) and adults ( 15+ years of age ) ., We use these age groupings based on WHO definitions of treatment groups 1–3 to calculate the necessary coverage levels ( MDA or vaccination ) for each category in order to interrupt transmission ., This is typically defined as the overall R0 <1 in infectious disease epidemiology , but as shown by Anderson and May 14 , the system of equations defined above has three possible equilibria; namely , a stable endemic state , an unstable boundary ( transmission breakpoint ) and a stable state of parasite extinction ., This model is hybrid in the sense that assumes a negative binomial form for the distribution of parasite numbers per host with a fixed aggregation parameter k , density dependent fecundity , and assumed monogamous sexual reproduction among worms ., The mean expected behavior of the individual based stochastic model is identical to the predictions of a deterministic version of the model ., However , an individual-based stochastic model permits the examination of the probability distribution of a given event occurring , such as transmission elimination , in a defined period of time during which control measures are applied ., Autopsy data show that worms tend to aggregate more in some individuals than in others , due to poorly understood factors such as environmental , social , host genetic or immunological effects 17 ., Epidemiological studies also show that those heavily infected are predisposed to this state 18 ., To take account of such effects in our model , individuals in each age category are assigned a contact rate drawn from a gamma distribution with shape parameter α , which , via compounding across individual distributions , leads to a negative binomial distribution of worms within the total host population ., It is important to note that the aggregation parameter , k , within the stochastic model , fluctuates in value over time , as a result of changes in the mean worm burden ., In the deterministic model k is held fixed in value ., The stochastic model more accurately mirrors observed patterns where k tends to decrease in value as prevalence declines under the impact of control measures 19 ., The egg contribution to the infectious reservoir depends on the age-specific contact rate for each individual and is governed by a deterministic formulation ., Treatment events are predetermined , they occur at time tj and the time step to the next treatment event is randomly drawn from an exponential distribution ., The rate parameter for this distribution is given by the overall rate that any event happens ., Which event occurs is drawn at random , on the basis of the relative magnitude of each individual event relative to the combined rate of all events ., Table 1 provides a description of these rates ., In this paper we consider 15 years of MDA and vaccination administration ., Most of the parameter values used in this paper are taken from within the ranges found in the literature ( Table 2 ) ., However , the data for the age-specific contact rates of hosts within the infectious reservoir ( β ) and age-specific contribution of hosts to the reservoir are unknown ., They are estimated by using MCMC method in parameter estimation from age intensity and prevalence curves as described in references detailed in the text and Table 2 . Precise details of the model fitting procedure are described in previous publications 4 , 14 , 15 , 17 ., In the numerical evaluations of the model’s behavior ( stochastic simulations ) , we follow the WHO guidelines for the implementation of MDA ., Starting with an untreated population , we administrate MDA over a 15-year period with coverage levels and treatment intervals based on the baseline prevalence ., For low baseline prevalence in SAC , we treat once every 3 years; for moderate baseline prevalence in SAC , we treat once every 2 years and for high baseline prevalence in SAC , we treat once a year ., The intensity of transmission is determined by R0 ( the basic reproductive number ) which varies for different baseline settings ., When MDA alone is used as the treatment strategy , we simulate the following treatment strategies:, ( i ) the WHO recommended treatment coverage of 75% SAC only;, ( ii ) 60% of SAC only;, ( iii ) 40% of SAC only and, ( iv ) 85% of SAC and 40% of adults ., In this paper we consider an ideal case-perfect vaccine , meaning that the rate of infection and the rate of egg production are essentially reduced by 100% , which is comparable to the efficacy of the Sm-p80 vaccine in the baboon model ., This efficacy considers the prevention of worm establishment , the fecundity falling dramatically in those few worms that establish , and the inability of eggs from these worms to hatch and release viable miracidia ., Vaccination is given annually to the children with the pre-specified age of administration , and the coverage levels depend on the age group that is treated and the duration of vaccine protection ., In various experimental settings Sm-p80 has demonstrated robust antibody titres in baboons for up to 5–8 years 10 suggesting a reasonably long duration of protection ., In this paper we simulate scenarios where, ( i ) the vaccine gives a 5 year duration of protection ( from 10 ) and, ( ii ) an ideal scenario where the vaccine gives a 20 years of protection which is longer than the duration of treatment ( 15 years ) ., It should be noted here that the same results will be obtained for vaccines with a duration of protection longer than 20 years as we are only calculating the probability of achieving the WHO goals within 15 years of initiating vaccination ., Also , it should be noted that the vaccine decay rate is given by 1/ ( duration of protection ) ., Duration of vaccine protection has a direct impact on the vaccine administration schedule and the coverage levels required to have a significant impact ., Here we consider the epidemiology of schistosome infections and the human host age-groups contributing most to parasite transmission ., The aim is to cover children from ages 5–15 by vaccinating children in cohorts ., We also analyze control strategies where the vaccine is given to younger children in their first year of life ., The schistosomiasis vaccine will very likely be administered in conjunction with other vaccines already present in traditional immunization programmes ( HPV , DTP ) ., Therefore , the achievable coverage will typically match that achieved for one of the other co-administered vaccines ., Vaccination coverage in the first year of life ranges between 85% and 91% at global level and reduces significantly in the following years ( Table 3 ) ., The coverage levels for school age children vary between 60% and 70% and for out of school individuals this range is 40%-50% 24–27 ., Based on these coverage levels , for a vaccine that provides a 20-year protection against schistosomiasis , we vaccinate at age 1 ( early start ) or age 5 ( school start ) , with coverage levels of 85% and 60% respectively ., For a vaccine that provides a 5-year duration of protection against infection , to ensure continuous protection , we vaccinate either at ages 1 , 6 and 11 with coverage levels 85% , 60% and 70% respectively , or at ages 5 , 10 and 15 with coverage levels 60% , 70% and 45% respectively ., In this case ( 5-year duration of protection ) we have a 3-dose schedule of vaccination , similar to the HPV administration schedule ., We consider MDA and vaccination , alone or in combination , as control strategies , where treatment is delivered at random at each round within the population with a given coverage ., In other words , we do not consider individual compliance to treatment 19 in these analyses and just assume the individuals treated or vaccinated are chosen at random at each round ., At the end of the treatment period , we calculate the probability of reaching WHO morbidity and elimination as a public health problem goal , by evaluating the fraction of SAC heavy-intensity infection prevalence ( ≤5% heavy-intensity infection in SAC for the morbidity goal and ≤1% heavy-intensity infection in SAC for the elimination as a public health problem goal ) ., In our results we include the prevalence of infection ( population having egg count threshold > 0 ) and prevalence of heavy-intensity infections ( population having egg count threshold > 16 ) ., The probability of reaching the 5% and 1% WHO goals are calculated as the fraction of repetitions that reach the target , by averaging across 300 simulations ( to ascertain the mean expectation of the stochastic model ) ., A summary of the treatment strategies is presented in Fig 1 ., First MDA alone is examined as the treatment strategy , using the WHO targets for treatment of 75% coverage for SAC ., The results are presented in Fig 2 and Table 4 ., Model simulations ( based on the parameter values listed in Table 2 ) suggest that for low prevalence regions , the 5% morbidity goal in SAC can be achieved within 5 years of treatment , while the elimination as a public health problem goal in the total population can be achieved within 10 years of treatment ., Similarly , for moderate-prevalence regions , the 5% morbidity goal in SAC can be achieved within 5 years of treatment , whereas the 1% elimination as a public health problem goal can be achieved within 15 year of MDA treatment ., Again , both goals will be achieved within 15 years with a probability of unity ., In high transmission regions , we can achieve the SAC 5% morbidity goal in 85% of the simulations ., However , the 1% elimination as a public health problem goal in such high transmission ( large R0 values ) settings can be achieved in 35% of our simulations ., In these settings , increasing the SAC coverage to > 75% and/or include other age bands in the treatment is highly desirable ., In low to moderate transmission settings , using the recommended target coverage of 75% for SAC , the SAC 5% morbidity goal can be achieved within 5 years of MDA ., Given the difficulties countries with endemic infection are experiencing in achieving this level of coverage , SAC coverages between 40% and 60% were also examined to explore if it is still possible to achieve the WHO goals with 15 years of MDA treatment ., The impact of MDA decreases as SAC coverage declines as indicated in Table 4 ., The SAC 5% morbidity goal can be achieved within 5 years at 60% SAC coverage ( in low to moderate settings ) ., However , for the <1% heavy infection in the total population goal ( = elimination as a public health problem ) to be achieved within 15 years the probabilities of achieving this are 90% and 70% , respectively , in low and moderate transmission regions ., Lowering the SAC coverage to 40% is predicted to achieve the WHO goals in low transmission settings ., However , in moderate transmission settings , the SAC 5% morbidity goal can be achieved within 15 years of treatment with probability of 0 . 9 , but the 1% elimination as a public health problem goal is only achieved with probability 0 . 4 in that time ., These results highlight the importance of using different MDA coverage levels in different transmission settings , as opposed to following the recommended 75% SAC coverage for all transmission levels ., In stochastic ( and deterministic ) models ( and in the real world ) there is always a chance that the prevalence of infection will bounce back after control measures cease since in some simulation runs the breakpoint in transmission is not crossed ., It is therefore important to analyze the probability of true elimination ( also known as ‘transmission interruption’ ) which results in the prevalence within the whole community in which control measures are introduced going to zero ., As in previous studies 28 it is assumed that if the overall prevalence is less than 1% it is almost certain that transmission interruption has been achieved ., We find that treating only 75% of SAC cannot interrupt transmission ( see Fig 2A , 2C and 2E ) , since the reservoir of untreated people in the adult age classes is able to seed the whole population once control ceases at year 15 ., As discussed earlier , in high transmission settings it is necessary to treat both SAC and adults ., Here we present the simulation results for the scenario 85% of SAC and 40% of adults are annually treated with MDA ., These results are summarized in Fig 3 which shows that with this approach the WHO goals can be achieved , although the probability of complete elimination by year 15 is still low ( <0 . 3 ) ., Longer durations of treatment and/or more frequent treatment are required to increase this probability ., In this section , the effects of both vaccination coverage , and the average duration of protection provided by the vaccine , are examined ., It should be noted that , based on the animal model results , we assume the vaccine is 100% efficacious ., In the previous two sections it is shown that the WHO 5% morbidity control goals can be achieved in low to moderate transmission settings if either MDA alone or vaccination alone are administrated in endemic regions ., However , these goals , particularly the 1% elimination as a public health problem goal , are unlikely to be achieved in high transmission settings ., Whether it is beneficial to combine both treatments together is examined in this section ., In practice , this is a likely scenario since MDA will remain the main control options for many years to come ( possibly 10 to 15 years ) even if Phase I , II and III trials in humans of the new vaccine go smoothly ., The simulation results suggest that giving MDA to 75% of SAC and administrating vaccination with a wide range of coverage levels ( see Figs 6 and 7 , Tables 7 and 8 ) , can reach the 1% elimination as a public health problem goal in high settings with a probability of nearly 0 . 55 and 0 . 82 for vaccines with durations of protection of 20 and 5 years , respectively ., The 5% SAC morbidity goal is achieved in all transmission settings ., Therefore , a vaccine that provides 5 years of protection and covers three age groups , can achieve the WHO 5% morbidity control and 1% elimination as a public health problem goals ., However , for a vaccine that provides 20 years protection we need to increase MDA and vaccination coverage levels , or include other age categories in the vaccination programme , to increase the probability of achieving elimination as a public health problem ( <1% ) in high transmission settings ., However , do note that the short duration vaccine must be delivered to multiple age groups ., Over 15 years an individual may need three vaccinations ( or 3 short courses of vaccination ) to maintain protection ., As such costs and delivery may be important issues with a short duration of protection vaccine ., The results presented in this paper are very sensitive to the values of certain parameters ., The two most important are the negative binomial aggregation parameter k and the magnitude of transmission before control measures are initiated ( the magnitude of R0 ) ., Using k = 0 . 24 , λ = 0 . 24 in low transmission settings , the model cannot support endemic parasite populations when R0 is low ., As a result , the model typically cannot reproduce endemic prevalences less than about 49% ., The two possible causes are:, ( i ) Diagnostic; due to poor sensitivity in the standard diagnostic test , measured prevalences may be much lower than the real values and, ( ii ) model transmission structure; transmission may be confined to specific age groups as elimination is approached , giving a low community-level prevalence ., To manage this limitation , we use k = 0 . 04 value for low transmission setting and k = 0 . 24 for moderate to high transmission settings ., We have chosen the extreme baseline prevalences ( just below 10% for low transmission settings and just below 50% for moderate transmission settings ) ., For these values there is a high probability to achieve the WHO goals and hence lowering the baseline prevalence does not alter the outcome ., For a baseline prevalence between 50% and 58% ( high transmission settings ) we obtain qualitatively similar results with the ones produced in moderate settings ., Therefore , for high transmission settings , we consider endemic regions with a baseline prevalence of around 62% ( R0 = 3 . 5 ) which is a realistic upper bound of prevalence for S . mansoni in most endemic regions 29 , 30 ., In this study , we have used parameter values fitted to data collected in Iietune village in Kenya ( refer to Table 2 ) , but the same model and analysis can be used for other endemic regions ., We should note here , that if the age-related contact rates and death rates are similar to the ones we have used , the results will be similar ., If the prevalence of intensity is higher ( lower ) in SAC , the probability of achieving the WHO goals will be lower ( higher ) in these regions ., These results are based on data for S . mansoni , but the analysis can be easily extended to S . haematobium ., A possible key parameter in the analysis and not included in our study is the buildup of acquired immunity ., To date , there aren’t enough evidences to show the presence of immunity in S . mansoni and we have assumed that the shape of age-intensity of infection is influenced only by rate of exposure to infection ., It will be of great importance , in the future , to extend our model so that we can explore the effect of acquired immunity on morbidity ., Currently schistosome control strategies suggested by WHO and widely implemented in endemic regions include mass drug administration of school aged children and adults in high transmission settings ., The primary goal is morbidity prevention in SAC or morbidity elimination in populations in areas of endemic infection ., Snail control , snail habitat alterations and improving water , sanitation and hygiene ( WASH ) are also recommended ( there is little information on their efficacy ) , but MDA is the major route for morbidity control at present ., In this paper , we have extended the individual based stochastic age structured model developed by Anderson and colleagues , which is constructed on the template of an age structured deterministic model 13–15 where its predictions have been validated using observed infection trends under defined levels of MDA in a number of field settings 31 ., We specifically extend past work to include the effect of a vaccine on parasite establishment ., The aim has been to explore the impact a vaccine with an efficacy of 100% might have on control efforts to attain the WHO goals for morbidity control in SAC and morbidity elimination in the total population ( but not infection ) ., Different treatment and vaccination strategies have been considered in numerical analyses; namely: MDA alone , vaccination alone , or MDA plus vaccination combined ., Analyses are conducted for three different transmission settings as defined by WHO on the basis of prevalence; low ( <10% baseline prevalence among SAC ) , moderate ( 10–50% baseline prevalence among SAC ) and high ( ≥50% baseline prevalence among SAC ) settings ., These transmission conditions at baseline are determined by the magnitude of R0 , and , concomitantly , by the overall prevalence of infection and the average intensity of infection in defined community ., We find that the optimal strategy to control or eliminate morbidity depends on the transmission setting , vaccine coverage level achieved , the duration of vaccine protection and the timeline of vaccination in different age groupings of the human host ., In low prevalence settings , MDA alone or vaccination alone , with different levels of protection , can achieve the WHO goals with a probability of close to unity ., Furthermore , our results show that treating just 40% of SAC with MDA alone can achieve the morbidity control goal and potentially elimination as a public health problem goal ., This is an encouraging prediction considering the difficulties endemic regions are having in achieving the WHO recommended treatment coverage for SAC at 75% ., In moderate prevalence settings , treating 60% of MDA can achieve the morbidity goal with probability of unity and possibly the elimination as a public health problem goal with probability of 0 . 7 ., Increasing the SAC coverage to 75% increases the probability of elimination to 0 . 96 ., Vaccination with a duration of protection of 5 years can achieve the morbidity control goal within 5 years of treatment and elimination as a public health problem goal within 15 years ., However , a vaccine with a longer duration of protection ( 20 years ) achieves the morbidity goal with a probability of near unity , but the probability of elimination as a public health problem goal decreases to nearly 0 . 55 ., In high transmission settings , we obtain the following outcomes:, ( i ) the WHO recommended MDA treatment coverage for SAC at 75% can achieve the morbidity control goal with a probability of 0 . 85 , but there is only a 0 . 35 chance that we can achieve the elimination as a public health problem goal ., ( ii ) Vaccinating 85% of 1-year olds with a vaccine that provides 20 years of protection , can achieve the morbidity control goal with probability of 0 . 61 , but it is very unlikely that the elimination as a public health problem goal will be achieved ., ( iii )

Introduction, Methods, Results, Sensitivity analysis and model limitations, Discussion

Mass drug administration ( MDA ) is , and has been , the principal method for the control of the schistosome helminths ., Using MDA only is unlikely to eliminate the infection in areas of high transmission and the implementation of other measures such as reduced water contact improved hygiene and sanitation are required ., Ideally a vaccine is needed to ensure long term benefits and eliminate the need for repeated drug treatment since infection does not seem to induce lasting protective immunity ., Currently , a candidate vaccine is under trial in a baboon animal model , and very encouraging results have been reported ., In this paper , we develop an individual-based stochastic model to evaluate the effect of a vaccine with similar properties in humans to those recorded in baboons in achieving the World Health Organization ( WHO ) goals of morbidity control and elimination as a public health problem in populations living in a variety of transmission settings ., MDA and vaccination assuming different durations of protection and coverage levels , alone or in combination , are examined as treatment strategies to reach the WHO goals of the elimination of morbidity and mortality in the coming decade ., We find that the efficacy of a vaccine as an adjunct or main control tool will depend critically on a number of factors including the average duration of protection it provides , vaccine efficacy and the baseline prevalence prior to immunization ., In low prevalence settings , simulations suggest that the WHO goals can be achieved for all treatment strategies ., In moderate prevalence settings , a vaccine that provides 5 years of protection , can achieve both goals within 15 years of treatment ., In high prevalence settings , by vaccinating at age 1 , 6 and 11 we can achieve the morbidity control with a probability of nearly 0 . 89 but we cannot achieve elimination as a public health problem goal ., A combined vaccination and MDA treatment plan has the greatest chance of achieving the WHO goals in the shorter term .

Nearly 258 million people are infected worldwide by schistosome parasites ., The World Health Organization ( WHO ) has set control guidelines to combat the morbidity and mortality induced by infection , defined by reaching ≤5% and ≤1% prevalence of heavy-intensity infections in school-aged children ( SAC ) , respectively ., Mass drug administration ( MDA ) is the major route for morbidity control and elimination ., However , MDA does not provide long-term protection against schistosome parasites and frequent drug administration is therefore required to control morbidity ., Infection does not induce lasting acquired immunity to reinfection ., Drug resistance is another issue with MDA which , if it arises , could possibly make drug treatment ineffective over time as drug-resistant genes in the parasite population increase in frequency ., A vaccine is ideally needed to both reduce the possibility of reinfection and to achieve transmission elimination within a feasible time frame ., Based on the recent results obtained for a new candidate vaccine in the baboon animal model , we employ an individual-based stochastic model to assess the impact of a vaccine with an efficacy of 100% when applied in endemic regions with different intensities of transmission ., Simulations suggest that the probability of achieving morbidity control and elimination as a public health problem depends on the duration of protection provided by vaccination , the age categories of the human host population vaccinated , and the coverage levels achieved ., In order to achieve elimination as a public health problem , model simulations suggest that combining vaccination ( with 5 years of protection ) with MDA ( treating 75% of school-aged children , 5–14 years of age ) is the best option , particularly in high transmission settings .

schistosoma, invertebrates, schistosoma mansoni, medicine and health sciences, helminths, immunology, vertebrates, parasitic diseases, animals, mammals, health care, vaccines, preventive medicine, age groups, primates, infectious disease control, vaccination and immunization, morbidity, old world monkeys, public and occupational health, infectious diseases, monkeys, baboons, people and places, health statistics, eukaryota, biology and life sciences, population groupings, amniotes, organisms

journal.pntd.0000545

Spatial Evaluation and Modeling of Dengue Seroprevalence and Vector Density in Rio de Janeiro, Brazil

Dengue is a mosquito-borne viral infection , considered a major public health problem in many tropical regions of the world , including Brazil 1 , 2 ., Aedes aegypti is the most important dengue vector worldwide 3–5 and the only known vector in Brazil 6 ., Dengue infection can manifest itself as clinically unapparent , an undifferentiated febrile illness , classic dengue fever ( DF ) , or dengue hemorrhagic fever ( DHF ) ., Prevalence of dengue is highest in tropical areas of Asia and the Americas , with 50–100 million estimated cases of dengue fever and 250 , 000–500 , 000 cases of dengue hemorrhagic fever occurring annually worldwide as explosive outbreaks in urban areas 7 , 8 ., In Brazil , three dengue virus serotypes ( DENV ) have been introduced through Rio de Janeiro in the past three decades: DENV-1 in 1986 9 , DENV-2 in 1990 10 , and DENV-3 in 2000 11 ., Figure 1 shows the time series of dengue cases in Rio de Janeiro State from 2000 to 2008 12 ., The introduction of DENV-3 in the state of Rio de Janeiro led to severe epidemics in 2002 with the largest number of cases ( 288 , 245 notified ) , with 1 , 831 DHF cases and 91 deaths , corresponding to 1 , 735 reported cases per 100 , 000 inhabitants 13 , and a case-fatality ratio of 3 . 15∶10 , 000 ., Eight years later , in 2007–2008 , during the current study , Rio de Janeiro ( and Brazil ) experienced the most severe dengue epidemics ever reported in terms of morbidity and mortality 14 ., During this period , 322 , 371 cases and 240 deaths were registered , with 100 deaths due to DHF/dengue shock syndrome ( DSS ) and 140 due to other dengue-related complications 12 ., That represented a case-fatality rate of 9 . 4∶10 , 000 ., Contrasting with the previous epidemics , the 2008 epidemic , essentially caused by DENV-2 , was characterized by a higher incidence of severe cases in children ., In fact , 36% of deaths reported occurred in individuals ≤15 years old 12 , 15 ., Rio de Janeiro presents highly favorable conditions for transmission of dengue 13 , as shown by serological cross-sectional surveys carried out after the arrival of DENV-1 and DENV-2 ., In 1987 , after the first wave , 45 . 5% of schoolchildren were positive for DENV-1 haemagglutination inhibition antibodies ( HAI ) 16 ., HAI antibody persists for a long period , but is highly cross-reactive 3 ., In the neighbor city of Niterói , 55% of schoolchildren were positive in 1988 , and 66% in 1992 ( after the arrival of DENV-2 ) 17 , 18 ., In Paracambi , another neighbor city , 29 . 2% schoolchildren were positive in 1997 19 ., Dengue surveillance and control in large urban areas with high levels of dengue transmission pose important challenges ., Clinical surveillance is impaired by the high proportion of asymptomatic infections 20 , 21 , 22 , and mosquito surveillance is very time and resource consuming ., Moreover , despite the theoretical association between vector abundance and risk of transmission , the quantitative nature of this relationship is poorly known 23 ., Understanding the epidemiology of this disease requires studies that integrate epidemiological and entomological data 19 , 21 , 24 , 25 ., The main objective of this study is to model the spatial patterns of seroprevalence in three neighborhoods with different socioeconomic profiles in Rio de Janeiro ., As blood sampling coincided with the peak of dengue transmission , we were also able to identify recent dengue infections and visually relate them to Aedes aegypti spatial distribution abundance ., We analyzed individual and spatial factors associated with seroprevalence using Generalized Additive Model ( GAM ) ., Surveys were performed in three neighborhoods of Rio de Janeiro city: Higienópolis , Tubiacanga , and Palmares , which differ in human population density , sanitation , vegetation cover , and history of dengue ( Fig . 2 ) ., Since neighborhoods were large and heterogeneous , we restricted the survey to an area of approximate 0 . 25 km2 in each one 26 ., The serological surveys were carried out in July-November 2007 and February-April 2008 , the latter coinciding with the 2008 high transmission period 12 ., The study areas had been under entomological surveillance since September 2006 ( see Mosquitoes surveillance section ) 26 ., The entomological surveillance consisted of weekly collections of Ae ., aegypti eggs and adults using traps located in 80 households per site ., All householders participating in the entomological surveillance were invited to participate in the serological surveys ., Only 72 out of 240 householders agreed to participate ( 13 in Higienópolis , 31 in Tubiacanga and 28 in Palmares ) ., To increase the sample , we invited additional residents from nearby houses , reaching a total of 171 participating households ( 19 in Higienópolis , 93 in Tubiacanga and 59 in Palmares ) , with 337 individuals ( 44 in Higienópolis , 162 in Tubiacanga and 131 in Palmares ) ., Since previous studies reported lower seropositive rates in the younger age classes 16 , 18 , 19 , we concentrated our sample effort in the age group of 1–20 years old to increase the chance of detecting seroconversion events 30 ., However , due to problems related to participant refusal , particularly for small children in the urban area , older people were included as well , to increase the sample size ., The range and median age in the sample is presented in Table 1 ., A questionnaire was applied to each enrolled individual , with questions regarding sex , age , education level , yellow fever ( YF ) vaccination status , clinical symptoms of dengue-like disease and past dengue episodes ., The location of each household was determined by a hand-held , 12 channel global positioning system ( Garmin ) , which accurate to 15 m ., Recent dengue infection was defined by the detection of DENV IgM antibodies in any sample ( first or second sample ) within the last 6 weeks or so ., Seroprevalence was defined by detection of DENV IgG antibodies in the first sample ( July–November/2007 ) ., Seroconversion was defined only for the paired samples – negative in the first sample and positive in the second one – considering both IgM and IgG ., Primary infection was defined as a negative IgG in the first sample with positive IgM in the second and secondary infection when DENV IgG antibodies were detected in the first sample ., Individuals with DENV IgM antibodies were considered asymptomatic cases when clinical definition of dengue – high fever , accompanied by at least two of the associated symptoms: headache , myalgia , arthralgia retro-orbital pain and rash – was not met 31 ., A blood sample ( 5 mL ) was collected from all participants during the household visit , stored at −20°C and processed within 12 hours ., Sera were tested for DENV- reactive IgM and IgG immunoglobulin by using PANBIO dengue IgM capture and dengue IgG indirect Elisa ( Brisbane , Australia ) ., Viral RNA for the nested RT-PCR and real-time RT-PCR assays was extracted from 140 µL of serum samples by the QIAamp Viral RNA Mini Kit ( QIAGEN , Valencia , CA ) , according to the manufacturers instructions ., RNA was eluted in 60 µL of buffer ( AVE ) and stored at −70°C ., For the quantitative TaqMan assay , a 10-fold-dilution series containing a known amount of target viral RNA ( 107 RNA copies/mL ) was used for RNA extraction ., The nested RT-PCR protocol for DENV detection and typing was performed on serum samples , which tested DENV IgM positive according to 32 ., One-step real-time RT-PCR assays were performed in the ABI Prism 7000 Sequence Detection System ( Applied Biosystems , Foster City , CA ) in all IgM positive samples ., Briefly , samples were assayed in a 25 µl reaction mixture containing 5 µl of extracted RNA , 1 µl of 40X Multiscribe enzyme plus RNAse inhibitor , 12 . 5 µl TaqMan 2X Universal PCR Master Mix ( Applied Biosystems , Foster City , CA ) and 300 nM of each specific primer and fluorogenic probe ., Positive and negative controls were included ., To detect specific DENV1-2 , primer and probe sequences were obtained from 33 ., To detect specific DENV-3 , primer and probe sequences were obtained from 34 ., The TaqMan probe was labeled at the 5′ end with the 5-carboxyfluorescein ( FAM ) reporter dye and at the 3′ end with 6-carboxy-N , N , N′ , N′-tetramethylrhodamine ( TAMRA ) quencher fluorophore ., The number of viral RNA copies detected was calculated by generating a standard curve from 10-fold-dilutions of DENV-3 RNA , isolated from a known amount of local virus propagated in Aedes albopictus C6/36 cells 13 , the titer of which was determine by plaque assay ., The same model of DENV-3 standard curve was applied to build DENV-1 and DENV-2 curves ., Quantitative interpretation of the results obtained was performed by interpolation from the standard curve included in each independent run for each serotypes ., Entomological surveillance was carried out with two types of traps for ovipositing females , egg traps and adult traps ., Egg traps are black plastic containers , filled with 300 ml of a 10% hay infusion , and a wooden paddle held on the wall for oviposition 26 , 35 , 36 , 37 ., Adult traps ( version 1 . 0 , Ecovec Ltd ) consists of a matte black container ( 16 cm high×11 cm diameter ) with approximately 280 ml of water and a removable sticky card ., A synthetic oviposition attractant was used to attract gravid female mosquitoes 38 ., Surveillance was conducted weekly from September , the 6th 2006 to March , 24th 2008 in the three study areas , encompassing two wet-hot seasons and one dry-cool season ., In each study area , 40 adult traps and 40 egg traps were installed in a random sample of premises 26 ., Two infestation indexes were calculated: mean adult density ( MAD = number of trapped female Ae ., aegypti/number of adult traps and mean egg density ( MED = number of collected eggs/number egg traps ) ., Details on the entomological methods and results are described in 26 ., To evaluate potential heterogeneities in the spatial distribution of mosquito abundance during the serological surveys , we aggregated the weekly entomological collections over time , from April/2007 to March/2008 , into a single index ., Recent dengue infections are plotted on this vector abundance map to inspect for possible associations ., Breteau Index ( number of Ae . aegypti-positive containers per 100 houses ) measured in March , June , August , November of 2007 and January and April of 2008 in each study area was also obtained from Public Health Office of Rio de Janeiro city ., The number of recent dengue infections was very small , and consequently , not statistically modeled ( descriptive data in Table 1 ) ., To compare and possibly to advance further investigations , the coordinates of negative and positive ( in any sample ) DENV IgM antibodies were mapped over the aggregated distribution of adult mosquito abundance ., The technique to build the interpolated surface is presented in the section below ., To compare seroprevalence among the areas we standardized the proportion of positive samples ( direct method ) using the total number of samples in all areas ., Seroprevalence data was analyzed using a Generalized Additive Model ( GAM ) : a statistical model that extends the generalized linear models to include non-parametric smoothing terms ., In the generalized linear model , the response variable belongs to the exponential family , and its mean value is related to the linear predictors through a link function ., The canonical link function for binomial response , such as positive or negative sera , is the logit link ., To evaluate possible non-linearity of the age effect on the outcome we used a smooth-spline and plotted the predicted against the observed value ., The spatial distribution was modeled using a bi-dimensional smooth function 39 ., The complete model thus included a set of directly observed covariates and a function – in our case , a thin plate spline – applied on the geographical coordinates of each household , as depicted in the equation below: is the response variable , are the slope coefficients of the model , so is the adjusted odds ratio , are the explanatory variables at the individual and household levels , the function is a smooth function of geographic co-ordinates and are the residuals ., All covariates with a p-value ≤0 . 10 in the univariate analysis were included in the multivariate model ., The approach used to analyze the spatial distribution started with a model with just the smooth function of the coordinates ., Then explanatory variables were included successively until the final adjusted model was obtained ., Contour lines at p-value ≤0 . 05 were drawn on the maps to identify areas with significantly higher ( red lines ) and lower risk ( blue lines ) than the overall mean ., In the case of the mosquito interpolation surface , the adults counts were the outcome variable and the smoothed geographic coordinates of the adult traps were the independent variables ., All statistical analyses were performed using the statistical software R 2 . 8 . 1 40 , with library mcgv 41 ., Ethical clearance was obtained from the Ethical Committee in Research ( CEP 365/07 ) from the Oswaldo Cruz Foundation , Ministry of Health , Brazil ., Written consent to participate in the two surveys was obtained from each participant and in case of minor , from their legal guardians ., All administrative areas containing the studied neighborhoods had a history of dengue cases recorded by the local public health authorities 42 ., Figure 3 shows the time series of reported dengue cases from Public Health Office of Rio de Janeiro city , with a clear peak between December/2007 and April/2008 , during the present study ., In 2008 , the attack rates were: 45 . 94/‰ in Higienópolis , 35 . 17 in Galeão area ( where the neighborhood of Tubiacanga is located ) and 19 . 68 in Vargem Pequena area ( where the suburban slum of Palmares is located ) ., Aedes aegypti abundance was consistently high throughout the year in the urban and suburban sites ( Higienópolis and Tubiacanga ) , and low in the suburban slum ( Palmares ) ., The largest increase in notified dengue fever cases began in December/2007 and apparently was not preceded by an increase in vector density as measured by our study ., The mosquito indices ( MAD and MED ) time series fluctuated over the time ., An increase in summer is clear in both suburban areas , but not in the urban area ., The bars at the bottom of the picture , showing the number of recent dengue infections relative to the number of collected blood samples , coincide with the high peak of the 2008 epidemic ., The Breteau index ranged from 4 . 20 to 11 . 32 in Higienópolis , 4 . 10 to 20 . 51 in Tubiacanga and 3 . 30 to 15 . 38 in Palmares ., Table 1 shows the results of the serological surveys ., From 337 individuals , 247 provided paired serum samples ( 73 . 3% ) ( Higienópolis: paired/unpaired =\u200a28/16; Tubiacanga =\u200a117/45; Palmares =\u200a102/29 ) ., Age of participants ranged from 1 to 79 years , with an average of 16 . 9 ., There were 156 ( 46 . 3% ) males and 181 ( 53 . 7% ) females ., For education level , 29 ( 8 . 6% ) were illiterate , 241 ( 71 . 5% ) reported elementary school , 56 ( 16 . 6% ) high school , and 11 ( 3 . 3% ) college ., Only 6 . 2% of the study subjects reported vaccination against yellow fever and 16% reported a previous history of dengue ., The combination of four methods provided diagnostic confirmation of dengue infection as follows: previous exposure to dengue ( IgG ) in the first survey detected in 199 ( 61 . 0% ) out of the 326 individuals ., Recent dengue infection ( IgM ) was detected in 30 individuals ( 4 in Higienópolis , 7 in Tubiacanga , and 19 in Palmares ) , which were subjected to nested RT-PCR and real-time RT-PCR ( Table 1 ) ., DENV-RNA was detected in 5 individuals ( 4 DENV-2 and 1 DENV-3 ) , by Nested RT-PCR and Real Time RT-PCR ( TaqMan ) ., Adopting quantitative real-time RT-PCR , we examined levels of DENV-RNA ., The results revealed low viral RNA , ranging from 1 to 45 RNA copies/mL ., Dengue seroprevalence varied between the study areas ., The age standardized proportions were 60 . 26% in Higienópolis , 56 . 07% in Tubiacanga and 77 . 44% in Palmares ( Table 1 , Fig . 4 ) ., In Higienópolis , the urban area , participation in the study was the lowest in all age groups , and the largest number of samples was in the interval of 5 to 9 years old ., Frequency of seropositive samples increased with age ( Fig . 4 ) ., In Tubiacanga a non-linear relationship between age and seroprevalence was observed , with a plateau at about 15 year old ( Fig . 5 ) ., In the other two areas , the relationship between seroprevalence and age was linear and significant ., Due to the non-linearity observed in Tubiacanga , we categorized the variable age , using cut points at 10 and 20 years old , to analyze the effect of age on seroprevalence in the multivariate models ., The variable sex was significant only in Tubiacanga , while self-reported past dengue was a predictor of seropositivity in Tubiacanga and Palmares ., Yellow fever vaccination was not statistically associated with dengue seropositivity in any study area ( Table 2 ) ., Prevalence smooth maps , with darker gray colors indicating higher odds ratio ( OR ) , are shown in Figure 6 ., In Higienópolis , the urban area , the spatial distribution of seroprevalence showed a linear North-South trend , with the highest odds ratios three times larger than the average value ., However , no location in this area presented statistically significant differences in OR ., Tubiacanga , the suburban area , presented similar variation in spatial odds ratio , with a high OR 3 . 0 region in the middle of the map , and this variation in chance significant ( depicted by the red line in the map ) ., In Palmares , the suburban slum , we observed the highest differences in seroprevalence distribution , with significantly high risk patch with OR =\u200a56 on the Northeast , where the main access to the community is located ., Towards the South , a protective spatial effect is evident , and an area with a protective effect was observed , located close to a forested area ., The OR maps resulting from the models adjusting for individual covariates ( sex and age ) presented a very similar pattern , and therefore are not shown ., Figure 7 shows maps of adult Ae ., aegypti abundance ., Dots indicate the location of surveyed households with and without cases of recent dengue infection ., Darker shades of gray indicate higher levels of mosquito abundance , measured in terms of relative risk ( RR ) ., Visual inspection , the only possible analysis due to the small number of recent dengue infections , suggests no evidence of a coincident pattern ., In the urban area , Higienópolis , mosquito RR varied from ca 0 . 25 to 4 . 5 , with a significantly high mosquito density area ( depicted in red in the map ) ., Only one of the four new infections is located inside or close to this area ., In the suburban area , Tubiacanga , spatial variability in mosquito density was smaller , with RR going up to 3 ., Recent dengue infections are spread evenly over the entire area , just two in seven located inside a mosquito hotspot ., Palmares , the suburban slum , showed the smallest variation in the vector density – with mosquitoes homogeneously covering the whole area , and recent dengue infections are also homogeneously distributed over the region , without any detectable pattern ., High dengue virus activity in Brazil during the past 20 years is evidenced by the large number of reported cases , in almost all states 13 , 21 , 22 ., Rio de Janeiro , located in the Southeast Region of Brazil , is one of the most densely populated cities and has always been an important entry point for dengue viruses into the country 13 , 43 , 44 ( Fig . 3 ) ., In 2008 , DENV-2 was the predominant serotype 12 , 42 ., In the current study , we confirmed the co-circulation of DENV-2 and DENV-3 serotypes in 5 individuals ( 4 DENV-2 and 1 DENV-3 ) , by molecular methods , DENV-2 serotype invaded Rio de Janeiro 19 years before this study 10 , when it caused an epidemic that resulted in about 100 . 000 notified cases ., The 2008 DENV-2 epidemic struck a population were most children had no previous contact with this serotype , while most in the 10–20 years old group probably had experienced previous infections with either DENV-2 or DENV-3 ., Our results confirm this epidemiological scenario , with a high predominance of recent infections in children under 15 years old ( 18/30 ) ., Although the number of recent dengue infections was small , we decided to present the data because it is rare to have any recent infection data in population surveys ., The epidemic that occurred during our field work presented the largest number of severe cases in children 12 , 42 ., However , in our data , only 23 . 3% of infections were symptomatic , suggesting that even during such severe epidemic , silent circulation of the virus is highly prevalent 20 , 21 , 45 , 46 ., A consequence of high frequency of asymptomatic infections is that measures of notified cases greatly underestimate the true incidence of infection and difficult the identification of high risk transmission areas within cities 47 ., We observed events of recent dengue infection in residences located in areas with low mosquito densities , suggesting that infection took place out of the residence , either in other premises – school , for instance – or outdoors , ( where children in these neighborhoods stay most of daytime , when Aedes mosquitoes are more active ) ., However , the lack of coherence between household mosquito counts and recent dengue infection should be further investigated in future work , by comparing the current data with infected Ae ., aegypti information 48 ., In parallel , information on human population movement patterns could also bring further insight on dengue fever transmission dynamics and the main places of transmission , eventually serving to build an early warning system for dengue outbreaks ., Entomological surveillance is of great importance for early detection of transmission risk and for directing vector control measures ., However , in Brazil , vector surveillance using Premise and Breteau indices correlates poorly with dengue incidence 49 , 50 , 51 , and moderately with the rate of epidemic growth 25 ., In Puerto Rico a study 52 to investigate the relationship between serological and epidemiological surveys and mosquito density showed that none of the household characteristics evaluated was significantly associated with recent dengue infection , except the number of female Ae ., aegypti per person ., In Colombia , the only entomological factor related to dengue infection in humans was the pooled infection rate of mosquitoes ., It would be helpful to discover the threshold of mosquito density that would trigger an epidemic 51 , 53 ., Epidemiological studies have identified statistical risk factors for human infection or diseases 54 , 55 , 56 ., Statistical models can bridge the gaps between landscape ecology , vector biology and human epidemiology , providing a sound approach to understanding risk and planning for control in heterogeneous environments , especially when the models are based on the ecology of the local vector populations 55–58 ., Additionally , understanding the space and time distribution of risk for mosquito-borne infections is an important step in planning and implementing effective infection control measures 59 , 60 ., This is because space and time are two important dimensions in describing epidemic dynamics and risk distribution 61 ., Our results point to larger spatial heterogeneity in dengue seroprevalence in the most isolated areas – Tubiacanga and Palmares ., In Tubiacanga , seroprevalence concentrated in the area with more intense commercial activity , schools and the main bus station ., In Palmares , seroprevalence was concentrated in the slum entrance , also an area of high commercial activity and human movement ., We hypothesize that such isolated populations are too small to maintain the dengue virus endemically and that the observed seroprevalence maps are the result of multiple viral introductions through the last 20 years , always through the same entrance ., Such spatial clustering of dengue has being reported in the literature 45 , 46 , and supports the hypothesis that mosquito-borne disease incidence is highly focal 46 , 62 ., On the other hand , a spatial pattern was not observed in Higienópolis , a neighborhood with multiple accesses and surrounded by slums with high population density ., These results highlight the important role on dengue transmission , of public spaces where human movement is intense , possibly more important than the households ., Further characterization of human movement patterns should provide additional information in the understanding of dengue transmission dynamics 63 ., Some authors have suggested that people rather than mosquitoes rapidly move dengue virus within and among communities 64 , 65 ., The present study is consistent with this information ., Our results must be considered in the context of the limitations of the serological survey ., First , the small number of recent dengue infections precluded a more adequate modeling of incidence versus mosquito density associations ., Second , the age distribution , particularly in Higienópolis , was not comparable to the other areas ., Third , households in the entomological and serological surveys did not match exactly what may precluded the identification of association between mosquito abundance and risk of infection ., This study contributes to a better understanding of the dynamics of dengue in Rio de Janeiro by assessing the relationship between dengue seroprevalence , recent dengue infection , and vector density ., In conclusion , the variation in spatial seroprevalence patterns inside the neighborhoods , with significantly higher risk patches close to the areas with the greatest human movement , suggests that humans may be responsible for virus inflow to small neighborhoods in Rio de Janeiro ., Surveillance guidelines should be further discussed , considering these findings , specially the spatial patterns for both human and mosquito populations .

Introduction, Methods, Results, Discussion

Rio de Janeiro , Brazil , experienced a severe dengue fever epidemic in 2008 ., This was the worst epidemic ever , characterized by a sharp increase in case-fatality rate , mainly among younger individuals ., A combination of factors , such as climate , mosquito abundance , buildup of the susceptible population , or viral evolution , could explain the severity of this epidemic ., The main objective of this study is to model the spatial patterns of dengue seroprevalence in three neighborhoods with different socioeconomic profiles in Rio de Janeiro ., As blood sampling coincided with the peak of dengue transmission , we were also able to identify recent dengue infections and visually relate them to Aedes aegypti spatial distribution abundance ., We analyzed individual and spatial factors associated with seroprevalence using Generalized Additive Model ( GAM ) ., Three neighborhoods were investigated: a central urban neighborhood , and two isolated areas characterized as a slum and a suburban area ., Weekly mosquito collections started in September 2006 and continued until March 2008 ., In each study area , 40 adult traps and 40 egg traps were installed in a random sample of premises , and two infestation indexes calculated: mean adult density and mean egg density ., Sera from individuals living in the three neighborhoods were collected before the 2008 epidemic ( July through November 2007 ) and during the epidemic ( February through April 2008 ) ., Sera were tested for DENV-reactive IgM , IgG , Nested RT-PCR , and Real Time RT-PCR ., From the before–after epidemics paired data , we described seroprevalence , recent dengue infections ( asymptomatic or not ) , and seroconversion ., Recent dengue infection varied from 1 . 3% to 14 . 1% among study areas ., The highest IgM seropositivity occurred in the slum , where mosquito abundance was the lowest , but household conditions were the best for promoting contact between hosts and vectors ., By fitting spatial GAM we found dengue seroprevalence hotspots located at the entrances of the two isolated communities , which are commercial activity areas with high human movement ., No association between recent dengue infection and households high mosquito abundance was observed in this sample ., This study contributes to better understanding the dynamics of dengue in Rio de Janeiro by assessing the relationship between dengue seroprevalence , recent dengue infection , and vector density ., In conclusion , the variation in spatial seroprevalence patterns inside the neighborhoods , with significantly higher risk patches close to the areas with large human movement , suggests that humans may be responsible for virus inflow to small neighborhoods in Rio de Janeiro ., Surveillance guidelines should be further discussed , considering these findings , particularly the spatial patterns for both human and mosquito populations .

Dengue is a major public health problem in many tropical regions of the world , including Brazil , where Aedes aegypti is the main vector ., We present a household study that combines data on dengue fever seroprevalence , recent dengue infection , and vector density , in three neighborhoods of Rio de Janeiro , Brazil , during its most devastating dengue epidemic to date ., This integrated entomological–serological survey showed evidence of silent transmission even during a severe epidemic ., Also , past exposure to dengue virus was highly associated with age and living in areas of high movement of individuals and social/commercial activity ., No association was observed between household infestation index and risk of dengue infection in these areas ., Our findings are discussed in the light of current theories regarding transmission thresholds and relative role of mosquitoes and humans as vectors of dengue viruses .

ecology, public health and epidemiology, virology

journal.pntd.0000311

Dermal-Type Macrophages Expressing CD209/DC-SIGN Show Inherent Resistance to Dengue Virus Growth

Dengue is probably the most important mosquito-transmitted viral disease of humans worldwide ., It is caused by dengue virus ( DV ) , which exists as four serotypes ( DV1-4 ) and circulates in an endemic-epidemic mode in most tropical and sub-tropical territories ., Transmission of DV to humans occurs when an infected mosquito probes for blood vessels and during a blood meal , through injection of infectious saliva into the human dermis ., As a member of the Flaviviridae family , DV infection involves virus uptake into endosomal vesicles that undergo acidification ., The low pH induces structural alterations in the envelope ( E ) protein that lead to membrane fusion and the release of the nucleocapsid into the cytoplasm 1 ., After uncoating , the RNA genome is translated to initiate virus replication ., It has been proposed that non-neutralizing antibodies raised against one DV serotype may enhance infection by a heterotypic serotype 2 ., This may explain why secondary infections are often associated with the more severe forms of dengue fever ( hemorrhagic fever with or without shock ) ., Much research on DV relies on relevant human cell culture models due to the difficulty of establishing appropriate animal models ., Progress has been made by showing that DV E protein recognizes the C-type lectin CD209 and its homologue L-SIGN and that expression of either of these lectins is sufficient to render cells permissive to DV grown in mosquito cells 3 , 4 ., Recently , the mannose receptor ( MR ) has also been shown to mediate DV binding and infection 5 ., Dendritic cells ( DC ) , generated from monocytes in the presence of GM-CSF and IL-4 , express CD209 , L-SIGN and the MR and are highly susceptible to DV infection 3 , 4 , 6 ., These monocyte-derived DC are thought to be representative of dermal DC ( dDC ) , yet there is increasing evidence that CD209 is not expressed by dDC but primarily by dermal macrophages ( dMφ ) 7–9 ., This underscores the importance of dMφ in early infection events and raises the question of whether dMφ are permissive for productive DV infection ., Studies of these cells have been hampered by the lack of suitable isolation techniques from human skin and culture methods to generate the cells from monocytic precursors ., Here , we confirmed that human dMφ express CD209 and showed that they bind DV E protein ., Based on the finding that dMφ stained for intracellular IL-10 , we developed a method to generate the cells from monocytes in the presence of IL-10 ., The monocyte-derived dMφ bound E protein and acquired DV in intracellular vesicles , but were resistant to viral replication ., The inability of DV to grow in these dermal-type Mφ was attributable to accumulation of internalized virus particles into poorly-acidified phagosomes ., These findings advance our understanding of the host innate resistance to DV at the early stages of infection and have implications for other pathogens recognizing CD209 ., Before blood and tissue samples were collected for the study , all healthy donors and patients gave written informed consent in agreement with the Helsinki Declaration and French legislation ., A prospective IRB approval was not obtained since there was no need as specified by French law of the health protection act when employing healthy material destined for disposal or one-time biomedical research ., A retrospective IRB approval was given ., Fresh skin ( about 50 cm2 ) was obtained from patients undergoing breast reduction surgery or abdominoplasty ., The skin was trypsinized to peel off the epidermis and the remaining dermis was processed as described elsewhere 10 with the modification that only collagenase type I ( 1 mg/ml , Invitrogen ) was used for 18 h at 37°C ., The resulting cell suspension was pipetted and serially filtered through 100 µm and 70 µm cell strainers ( BD Biosciences ) to remove undigested tissue fragments and to obtain a homogeneous cell suspension ., A DNA fragment containing the DV3 genomic region ( Swiss-Prot accession number P27915 ) coding for the prM-E protein ( 1674 nt in total , including all of prM and the E ectodomain , ending at codon 392 of E , at the end of domain III ) was amplified by PCR with forward primer 5′TTATGCATATTACTGGCCGTCGTGGCC and reverse primer 5′CTCGCCCGCAGACATGGCCTTATCGTCATCGTCGGGCCCCTTCCTGTACCA-GTTGATTTT and inserted into the plasmid pT352 ., This is a shuttle vector containing selection markers for yeast and E . coli , as well as a metallotheionein-inducible expression cassette for Drosophila cells ., In the construct , called pT352/DV3 sE-GFP , the DV prM-E sequence is in-frame with the Drosophila BiP signal peptide , which directs the recombinant protein to the secretory pathway ., Drosophila S2 cells ( Invitrogen ) were co-transfected with pT352/DV3 sE-GFP and a vector conferring resistance to blasticidine , using the effectene transfection reagent ( Qiagen ) ., The selected cells were adapted to serum-free growth medium and grown to high density before induction with CuSO4 at 500 µM ., The supernatant was collected 10 days later and concentrated using a flow concentration system with a 10 KDa-cutoff membrane ( Vivascience ) , and DV3 sE-GFP was purified by affinity chromatography using a Steptactin column ., The eluate was concentrated and further purified by size-exclusion chromatography , using a Superdex 200 10/300 column ( GE Healthcare ) with 0 . 5 M NaCl and 50 mM Tris ( pH 8 . 0 ) ., Purified DV3 sE was concentrated to 10 g/liter in Vivaspin ultrafiltration spin columns ( Sartorius ) ., Dermal cells were collected 48 h after culturing in complete medium , RPMI medium supplemented 10% fetal calf serum ( FCS ) and antibiotics ( Invitrogen ) , and 3×105 cells were incubated with 1 , 2 , 4 or 8 µg recombinant DV3 sE-eGFP fusion protein in 0 . 1 ml complete medium at 37°C for 30 min ., The cells were then washed twice with complete medium and incubated with anti-CD14-APC , anti-CD1a-PE and anti-HLA-DR-PerCP mAb ( BD Biosciences ) in PBS/2% FCS for 15 min ., Following 3 washes , the cells were fixed in 0 . 4% formaldehyde and analyzed by flow cytometry ( FACS Calibur , BD Biosciences ) ., The relative MFI for 3 donors was determined in triplicate after gating for CD1a+HLA-DR+ or CD14+HLA-DR+ cells using the following formula: ( MFI ( FL1 ) protein sE-eGFP – MFI ( FL1 ) no protein sE-eGFP ) /MFI ( FL1 ) no protein sE-eGFP ., To determine CD209 expression , 3×105 cells were incubated with anti-CD209-PerCPCy5 . 5 ( clone DCN46 , BD Biosciences ) , anti-CD14-APC , anti-CD1a-PE and anti-HLA-DR-PerCP mAb in PBS/2% FCS for 15 min and , after washing , fixed and analyzed by flow cytometry ., Formaldehyde-fixed , paraffin sections were rehydrated and antigen was retrieved in citrate buffer pH 6 at 97°C for 45 min ., Biotin was blocked using the avidin-biotin blocking kit ( Vector Inc . ) , and sections were saturated in 5% human serum at room temperature for 40 min ., The following primary Abs were used: goat-anti IL-10 ( 1∶75 dilution , R&D Systems ) , mouse anti-CD209 ( 2 µg/ml , R&D Systems ) , mouse anti-CD1a ( Immunotech ) and mouse anti-CD14 ( 1∶40 dilution , Novocastra ) ., The secondary Ab ( Jackson ) were: biotin-conjugated donkey anti-goat followed by streptavidin-Alexa 488 ( Molecular Probes-Invitrogen ) and F ( ab ) 2 rabbit anti-mouse followed by Cy3-conjugated donkey anti-rabbit ., Sections were observed by confocal microscopy ( LSM510 Zeiss ) ., Monocytes were isolated from 200 ml of adult human peripheral blood using negative-depletion beads ( Dynal-Invitrogen ) or by counterflow centrifugal elutriation ., To obtain MDdMφ , 3×106 monocytes were cultured for 5 days in 5 ml of complete medium containing 10 ng/ml M-CSF ( R&D Systems ) , 20 ng/ml IL-10 ( Immunotools ) and 20 ng/ml GM-CSF ( Schering-Plough ) with refreshment of GM-CSF ( 10 ng/ml ) and IL-10 ( 10 ng/ml ) at day 3 ., For MDDC , 3×106 monocytes were cultured for 5 days in 5 ml of complete medium containing 50 ng/ml GM-CSF and 10 ng/ml IL-4 ( Schering-Plough ) with readdition of cytokines at day 3 ., Non-adherent cells were harvested ., Expression of markers was measured by FACS using specific antibodies and their corresponding isotype controls ., To assay for DV3 sE protein binding , cells were pre-incubated for 10 min in complete medium in the absence or presence of 5 mM EDTA before adding 3 µg DV3 sE-eGFP protein ., After 30 min at 37°C , the cells were washed three times in complete medium and analyzed by flow cytometry ., 5×105 MDdMφ and MDDC were exposed to DV serotype 1 ( strain FGA/NA d1d ) 11 , serotype 2 ( strain 16681 ) , or serotype 3 ( strain PaH 881 , isolated in 1988 in Thailand ) in RPMI medium supplemented with 0 . 2% bovine serum albumin for 2 h ., Viral growth was determined at 40 h post-infection ., Virus titration was performed as previously described 3 ., Infectivity titers were expressed as focus forming unit ( FFU ) on mosquito AP61 cell line ( DV1 and DV3 ) or plaque forming unit ( PFU ) on mammalian BHK cell line ( DV2 ) ., Different titering assays were performed to independently confirm our findings , despite the fact both methods may not be equivalent ., The limit of titer determination was fixed at 103 , below which viral production was considered non-significant ., For FACS analysis , infected cells were fixed and labeled for intracellular viral antigens with antiserum raised in mice that had received intracerebral DV injection 3 ., IFN-α released from DV1-infected MDdMφ and MDDC was measured by ELISA ( R&D Systems ) ., To observe live DV internalization by MDDC and MDdMφ , the cells were exposed to DV1 at an MOI of 100 at 4°C for 30 min or at 37°C for 1 h and fixed in 2 . 5% glutaraldehyde ., Cells were postfixed in osmium tetroxide , dehydrated in ethanol containing 1% uranyl acetate , treated with propylene oxide and embedded in resin ( Durcupan ACM , Fluka ) ., Ultrathin sections were stained with lead citrate and examined by transmission electron microscopy ( TEM ) ( Hitachi H600 ) ., Images were acquired using a CCD camera ( Hamamatsu ) ., To visualize DV3 sE-eGFP internalization and endosomal acidification , cells were incubated with 10 µM LysoSensor Blue DND-167 ( Molecular Probes-Invitrogen ) for 30 min at 37°C ., Protein sE-eGFP was added at a concentration of 3 µg/ml , and cells were viewed after different incubation times using a confocal microscope ( LSM510 , Zeiss ) ., The blue color emitted by the LysoSensor dye was digitally converted into red ., For TEM , cells were fixed in 2% paraformaldehyde and 0 . 2% glutaraldehyde ., Cells were embedded in 1% agarose , permeabilized with 0 . 2% saponin and saturated with 2% BSA before incubation with 5 µg/ml polyclonal rabbit anti-GFP antibody ( Rockland ) ., The antibody was visualized by pre-embedding labeling using a goat anti-rabbit IgG conjugated to 0 . 8 nm gold particles , according to manufacturers instructions ( Aurion ) ., Cells were fixed in 1% glutaraldehyde , and gold particles were enhanced using a silver kit ( HQ silver , Nanoprobes ) ., Cells were then treated and observed as above ., We wished to determine whether human dMφ are targets of DV infection ., To this end , healthy human skin from patients undergoing plastic surgery was processed to obtain a dermal cell suspension ., The cells were then cultured without additional cytokines for 48 h to allow re-expression of cell surface markers , such as CD1a and CD209 , lost during the collagenase treatment ( data not shown ) ., Binding of DV3 E protein to dermal cells was assessed by flow cytometry after staining with CD14 and CD1a-specific antibodies ., CD14 is expressed by dMφ and CD1a by dDC 7–9 ., To detect E protein binding , the soluble form of DV3 E protein ( sE ) was fused to the reporter protein eGFP and purified from a Drosophila expression system ., As shown in Figure 1A , CD1a+ dDC showed only a limited capacity to interact with DV3 sE protein , whereas CD14+ dMφ readily bound the protein ., This is corroborated by the distinct expression of CD209 by dMφ ( Fig . 1A ) , whereas dDC expressed little , if any , CD209 ( data not shown ) ., Increasing amounts of DV3 sE protein were added to the dermal cell suspension to test if dDC bound the protein at higher concentrations ., Figure 1B shows that even at high concentrations , there was little binding of DV3 sE protein to dDC , whereas it bound to dMφ in a dose-dependent fashion ., These findings identify dMφ as potential key cellular targets of DV ., To address the question of whether dMφ are infected by DV and whether they are permissive for viral production , we established cell culture conditions to generate dermal-type Mφ from monocytes ., We observed on human skin tissue sections that dMφ expressing CD14 or CD209 , but not the CD1a+ dDC , stained for IL-10 ( Fig . 2A ) ., When purified human monocytes were cultured in M-CSF and increasing concentrations of IL-10 , the cells expressed CD14 and CD209 in an IL-10 dose-dependent manner ( Fig . 2B ) ., Similar to DC 12 , the addition of GM-CSF increased CD209 levels ( Fig . 2B ) , so that a homogeneous CD14+CD209+ cell population could be obtained with CD209 expression nearly identical to that of DC derived from monocytes in the presence of GM-CSF and IL-4 ( Figure S1A ) ., Western blotting of cell lysates confirmed the presence of CD209 as a major band of 49 kDa in both cell-types 13 ( Figure S1B ) ., The Mφ expressed coagulation factor XIIIa and CD163 , two other cell surface markers of dMφ 14 ( Fig . 2C ) ., The Mφ and the DC were both able to bind eGFP-tagged DV3 sE protein , which was inhibited by EDTA ( Fig . 2C ) ., This distinguishes the monocyte-derived DC from dDC ., Upon activation by lipopolysaccharide ( LPS ) , the Mφ rapidly released IL-10 , whereas DC or monocytes produced little of this cytokine ( Figure S1C ) ., Monocyte-derived dMφ ( MDdMφ ) and monocyte-derived DC ( MDDC ) were analyzed for DV infection using low-passage DV1 and DV3 strains grown in mosquito cells 3 as well as the prototype DV2 strain 16681 15 ., The cells were exposed to DV1 at a multiplicity of infection ( MOI ) of 1 for 2 h , washed , and then cultured for 40 h ., As shown in Figure 3A , intracellular viral antigen was clearly detected in MDDC by flow cytometry , whereas no specific immuno-labeling was observed in MDdMφ ., An analysis of DV replication in these cells infected at an MOI of 1 ( DV1 and DV3 ) or 2 ( DV2 ) showed that MDDC were highly permissive to productive infection ( ∼105 FFU/ml or PFU/ml ) ( Fig . 3B ) ; in contrast , progeny virus production was undetectable in DV-infected MDdMφ ( <103 FFU/ml or PFU/ml ) ., Consistent with this finding , no IFN-α was produced by DV-infected MDdMφ , even at an MOI of 10 , whereas MDDC readily released IFN-α when infected with DV at an MOI of 1 or 10 16 ( Fig . 3C ) ., To verify that MDdMφ acquired the virus , both myeloid cell-types were exposed to high DV input ( MOI of 100 ) and electron microscopy analysis was performed after 30 min at 4°C and after 1 h at 37°C ( Fig . 3D ) ., Cell surface-bound ( at 4°C ) and endosomal vesicle-associated virus particles ( at 37°C ) were clearly detected in both cell-types ., Thus , internalization of DV can occur in MDdMφ but does not result in productive infection ., In an effort to define the molecular basis of the inability of DV to grow in MDdMφ ., we asked whether internalized DV was sequestered in a manner that hampers productive infection , using DV3 sE-eGFP fusion protein ., To monitor DV3 sE protein internalization in MDdMφ and MDDC , the cells were incubated with pH-sensitive LysoSensor dye and analyzed by confocal microscopy ( Fig . 4 ) ., This dye accumulates in acidic organelles , where its fluorescence emission is highest ., After 5 min at 37°C , DV3 sE protein was observed in vesicle-like structures in both cell-types ., By 30 min and 60 min , DV3 sE protein dispersed to acidified perinuclear lysosomes in MDDC ., In marked contrast , when MDdMφ were examined at these time-points , a large fraction of internalized DV3 sE protein was excluded from the acidic compartment and remained in non-acidic , large endosomes ., Electron microscopy analysis using a colloidal gold-conjugated antibody to GFP demonstrated that DV3 sE protein accumulated in large phagosomes in MDdMφ , located close to the plasma membrane ( Fig . 5 ) ., On the other hand , at 30 min , in MDDC , DV3 sE protein was mostly found in small perinuclear vesicles in the environment of the endoplasmic reticulum ., Taken together , these data suggest that the inability of DV to productively infect MDdMφ is due to accumulation of virus particles in immature endosomal vesicles whose pH does not allow efficient viral-cell membrane fusion and subsequent virus uncoating ., In the present study , we demonstrated for the first time the interaction of dMφ with DV3 sE glycoprotein , which correlates with the expression of the DV attachment receptor CD209 ., Dermal DC displayed only a limited capacity to interact with DV3 sE protein and expressed little CD209 ., In accordance with these findings , in situ immuno-labeling of human skin section revealed CD209 expression by dMφ but little on DC 7–9 ., Both cell types carry the MR 7 , which also recognizes DV E protein 5 ., Due to the nature of our binding assay , the dermal cells with the highest affinity for DV3 sE protein would acquire the most DV3 sE protein , suggesting that dDC may capture the recombinant envelope protein when physically isolated from dMφ ., In the skin , the abundance , the location and the co-expression of CD209 , L-SIGN and MR are likely to determine the nature of the DV-capturing immune cell ., Based on the observations that dMφ stained for intracellular IL-10 in situ and that IL-10 is produced by dMφ ex vivo 17 , 18 , we tested the effect of IL-10 on the formation of dMφ from monocytes ., By combining IL-10 , M-CSF and GM-CSF , a homogenous cell population was obtained which carried CD209 and other markers characteristic of dMφ , rapidly produced IL-10 in response to LPS or other toll-like receptor ligands ( data not shown ) , and bound DV3 sE protein ., Like MDDC , the MDdMφ were capable of internalizing live DV but , distinct from MDDC , they displayed an inherent resistance to viral growth ., In contrast to DV3 sE protein found in acidified compartments in MDDC , we observed that DV3 sE protein accumulated in non-acidified phagosomes in MDdMφ ., The DC vesicles containing DV3 sE protein or live virus were bell-shaped or tubular , whereas they were round , larger and close to the plasma membrane in the Mφ ., To our knowledge , this identifies MDdMφ as the first innate immune cell capable of protecting the human host from DV infection and virus propagation ., From this data , we propose that dMφ can act to trap infecting virions in a fusion-incompetent endosomal environment and thus to prevent DV spread to dDC at the anatomical site of the mosquito bite ., We cannot formally exclude the possibility that downstream delays in the viral life cycle contribute to the inability of DV to replicate in MDdMφ , but the finding that West Nile virus productively infects these cells ( data not shown ) indicates that they are not generally refractory to flavivirus growth ., IL-10 , required for CD209 expression and blockage of endosome acidification , is likely to be produced by the dMφ themselves , constitutively , or in response to stimuli such as UV-light 17 ., In this context , a key question is whether mosquito salivary proteins , co-injected with the infectious virus , would also trigger IL-10 production by dMφ or , on the contrary , provoke an inflammatory response ., Inflammatory cytokines of the Th2 T-helper cell type , IL-4 and IL-13 , may be responsible for the formation of CD209+MR+ DC , which are permissive for DV infection and viral progeny production 3–6 ., Alternatively , the presence of anti-DV non-neutralizing antibodies raised against a heterotypic DV serotype may render dDC susceptible to DV infection at the site of the mosquito bite ., The abundance and strategic position of the Mφ in the dermis is consistent with their function as first defense barrier against pathogens by isolating and eliminating them and thus avoiding unnecessary immune activation ., However , other pathogens that recognize C-type lectins , such as mycobacteria , may exploit these cells to escape immune attack ., Accumulating CD209+ Mφ in leprosy skin lesions have been associated with mycobacterial persistence 19 ., Important questions to address in future are whether DV is eliminated in MDdMφ , whether infected MDdMφ gradually release DV , as shown for the foot-and-mouth disease virus and pulmonary Mφ 20 , and whether rapid DV growth can occur when the Mα convert to DC ., Improved knowledge of the molecular mechanisms for suppressing pathogen growth in MDdMφ will provide new insight into the crucial role of dMφ in protective immunity to infectious agents at the skin level .

Introduction, Materials and Methods, Results, Discussion

An important question in dengue pathogenesis is the identity of immune cells involved in the control of dengue virus infection at the site of the mosquito bite ., There is evidence that infection of immature myeloid dendritic cells plays a crucial role in dengue pathogenesis and that the interaction of the viral envelope E glycoprotein with CD209/DC-SIGN is a key element for their productive infection ., Dermal macrophages express CD209 , yet little is known about their role in dengue virus infection ., Here , we showed that dermal macrophages bound recombinant envelope E glycoprotein fused to green fluorescent protein ., Because dermal macrophages stain for IL-10 in situ , we generated dermal-type macrophages from monocytes in the presence of IL-10 to study their infection by dengue virus ., The macrophages were able to internalize the virus , but progeny virus production was undetectable in the infected cells ., In addition , no IFN-α was produced in response to the virus ., The inability of dengue virus to grow in the macrophages was attributable to accumulation of internalized virus particles into poorly-acidified phagosomes ., Aborting infection by viral sequestration in early phagosomes would present a novel means to curb infection of enveloped virus and may constitute a prime defense system to prevent dengue virus spread shortly after the bite of the infected mosquito .

Mosquito-transmitted pathogens are a major challenge to humans due to ever-increasing distribution of the vector worldwide ., Dengue virus causes morbidity and mortality , and no anti-viral treatment or vaccine are currently available ., The virus is injected into the skin when an infected mosquito probes for blood ., Among the skin immunocytes , dendritic cells and macrophages are equipped with pathogen-sensing receptors ., Our work has shown that dermal macrophages bind the dengue virus envelope protein ., We demonstrate that monocyte-derived dermal macrophages are resistant to infection and present evidence that this is due to sequestration of the virus into fusion-incompetent intracellular vesicles ., This identifies skin macrophages as the first innate immune cell potentially capable of protecting the human host from infection by dengue virus shortly after a mosquito bite ., These findings have important implications for better understanding the early infection events of dengue virus and of other skin-penetrating pathogens .

infectious diseases/neglected tropical diseases, immunology/immune response, dermatology/skin infections, cell biology/membranes and sorting, cell biology/microbial growth and development, immunology/innate immunity, infectious diseases/viral infections, virology/emerging viral diseases, infectious diseases/skin infections, immunology/immunity to infections, virology/host invasion and cell entry